GENÉTICA

MICROBIANA

MMCR 7003 – 1
Segundo Semestre
2022-2023
Gabriel Trueba

Nomenclature of Microorganisms
Binary names, consisting of a generic name and a specific epi-
thet (e.g., Escherichia coli), must be used for all microorgan-
isms. Names of categories at or above the genus level may be
used alone, but specific and subspecific epithets may not. A
specific epithet must be preceded by a generic name, written
out in full the first time it is used in a paper. Thereafter, the
generic name should be abbreviated to the initial capital letter
(e.g., E. coli), provided there can be no confusion with other
genera used in the paper. Names of all taxa (kingdoms, phyla,
classes, orders, families, genera, species, and subspecies) are
printed in italics and should be italicized in the manuscript;
strain designations and numbers are not. Vernacular (com-
mon) names should be in lowercase roman type (e.g., strepto-
coccus, brucella). For Salmonella, genus, species, and subspe-
cies names should be rendered in standard form: Salmonella
enterica at first use, S. enterica thereafter; Salmonella enterica
subsp. arizonae at first use, S. enterica subsp. arizonae thereaf-
ter. Names of serovars should be in roman type with the first
letter capitalized: Salmonella enterica serovar Typhimurium.
After the first use, the serovar may also be given without a
species name: Salmonella Typhimurium, S. Typhimurium, or
Salmonella serovar Typhimurium. For other information re-
garding serovar designations, see Antigenic Formulae of the
Sal-monella Serovars, 9th ed. (P. A. D. Grimont and F.-X.
Weill, WHO Collaborating Centre for Reference and
Research on Salmonella, Institut Pasteur, Paris, France, 2007;
see http: //www.pasteur.fr/ip/portal/action/
WebdriveActionEvent/oid/ 01s-000036-089). For a summary
of the current standards for Salmonella nomenclature and the
Kaufmann-White criteria, see the article by Brenner et al. (J.
Clin. Microbiol. 38:2465–2467, 2000), the opinion of the
Judicial Commission of the International Committee on
Systematics of Prokaryotes (Int. J. Syst. Evol. Microbiol.
55:519 –520, 2005), and the article by Tindall et al. (Int. J.
Syst. Evol. Microbiol. 55:521–524, 2005). The spelling of
bacterial names should follow the Approved Lists of Bacterial
Names (Amended) & Index of the Bacterial and Yeast
Nomenclatural Changes (V. B. D. Skerman et al., ed.,
American Society for Microbiology, Washington, DC, 1989)
and the validation lists and notification lists published in the
International Journal of Systematic and Evolutionary Microbiol-
ogy (formerly the International Journal of Systematic Bacteriol-
ogy) since January 1989. In addition, two sites on the World
Wide Web list current approved bacterial names: Prokaryotic
No-menclature Up-to-Date (http://www.dsmz.de/bacterial-
diversity /prokaryotic-nomenclature-up-to-date.html) and List
of Pro-karyotic Names with Standing in Nomenclature
(http://www .bacterio.net/). If there is reason to use a name
that does not have standing in nomenclature, the name
should be enclosed in quotation marks in the title and at its
first use in the abstract
August 2014, Instructions to Authors
Instructions to Authors
and the text and an appropriate statement concerning the no-
menclatural status of the name should be made in the text.
“Candidatus” species should always be set in quotation marks.
Microorganisms, viruses, and plasmids should be given des-
ignations consisting of letters and serial numbers. It is gener-
ally advisable to include a worker’s initials or a descriptive sym-
bol of locale or laboratory, etc., in the designation. Each new
strain, mutant, isolate, or derivative should be given a new
(serial) designation. This designation should be distinct from
those of the genotype and phenotype, and genotypic and phe-
notypic symbols should not be included. Plasmids are named
with a lowercase “p” followed by the designation in uppercase
letters and numbers. To avoid the use of the same designation
as that of a widely used strain or plasmid, check the designation
against a publication database such as Medline.
Genetic Nomenclature
To facilitate accurate communication, it is important that
standard genetic nomenclature be used whenever possible
and that deviations or proposals for new naming systems be
endorsed by an appropriate authoritative body. Review
and/or publication of submitted manuscripts that contain new
or nonstandard nomenclature may be delayed by the editor or
the Journals Department so that they may be reviewed.
Bacteria. The genetic properties of bacteria are described in
terms of phenotypes and genotypes. The phenotype describes
the observable properties of an organism. The genotype refers
to the genetic constitution of an organism, usually in reference
to some standard wild type. The guidelines that follow are
based on the recommendations of Demerec et al. (Genetics
54:61–76, 1966).
(i) Phenotypic designations must be used when mutant loci
have not been identified or mapped. They can also be used to
identify the protein product of a gene, e.g., the OmpA protein.
Phenotypic designations generally consist of three-letter sym-
bols; these are not italicized, and the first letter of the symbol is
capitalized. It is preferable to use Roman or Arabic numerals
(instead of letters) to identify a series of related phenotypes.
Thus, nucleic acid polymerase mutants might be designated
Pol1, Pol2, and Pol3, etc. Wild-type characteristics can be des-
ignated with a superscript plus (Pol⫹), and, when necessary for
clarity, negative superscripts (Pol⫺) can be used to designate
mutant characteristics. Lowercase superscript letters may be
used to further delineate phenotypes (e.g., Strr for streptomy-
cin resistance). Phenotypic designations should be defined.
(ii) Genotypic designations are also indicated by three-letter
locus symbols. In contrast to phenotypic designations, these
are lowercase italic (e.g., ara his rps). If several loci govern
related functions, these are distinguished by italicized capital
letters following the locus symbol (e.g., araA araB araC). Pro-
moter, terminator, and operator sites should be indicated as
described by Bachmann and Low (Microbiol. Rev. 44:1–56,
1980), e.g., lacZp, lacAt, and lacZo.
(iii) Wild-type alleles are indicated with a superscript plus
(ara⫹ his⫹). A superscript minus is not used to indicate a mu-
tant locus; thus, one refers to an ara mutant rather than an
ara⫺ strain.
(iv) Mutation sites are designated by placing serial isolation
Journal of Bacteriology jb.asm.org 15
Instructions to Authors
numbers (allele numbers) after the locus symbol (e.g., araA1
araA2). If only a single such locus exists or if it is not known in
which of several related loci the mutation has occurred, a hy-
phen is used instead of the capital letter (e.g., ara-23). It is
essential in papers reporting the isolation of new mutants that
allele numbers be given to the mutations. For Escherichia coli,
there is a registry of such numbers: the Coli Genetic Stock
Center (http://cgsc.biology.yale.edu/). For the genus Salmo-
nella, the registry is the Salmonella Genetic Stock Centre (http:
//people.ucalgary.ca/~kesander/). For the genus Bacillus, the
registry is the Bacillus Genetic Stock Center (http://www.bgsc
.org/).
(v) The use of superscripts with genotypes (other than ⫹ to
indicate wild-type alleles) should be avoided. Designations in-
dicating amber mutations (Am), temperature-sensitive muta-
tions (Ts), constitutive mutations (Con), cold-sensitive muta-
tions (Cs), production of a hybrid protein (Hyb), and other
important phenotypic properties should follow the allele num-
ber [e.g., araA230(Am) hisD21(Ts)]. All other such designa-
tions of phenotype must be defined at the first occurrence. If
superscripts must be used, they must be approved by the editor
and defined at the first occurrence in the text.
Subscripts may be used in two situations. Subscripts may be
used to distinguish between genes (having the same name)
from different organisms or strains; e.g., hisE. coli or hisK-12 for
the his gene of E. coli or strain K-12, respectively, may be used
to distinguish this gene from the his gene in another species or
strain. An abbreviation may also be used if it is explained. Sim-
ilarly, a subscript is also used to distinguish between genetic
elements that have the same name. For example, the promoters
of the gln operon can be designated glnAp1 and glnAp2. This
form departs slightly from that recommended by Bachmann
and Low (e.g., desC1p).
(vi) Deletions are indicated by the symbol ⌬ placed before
the deleted gene or region, e.g., ⌬trpA432, ⌬(aroP-aceE)419, or
⌬(hisQ-hisJo)1256. Similarly, other symbols can be used (with
appropriate definition). Thus, a fusion of the ara and lac oper-
ons can be shown as ⌽(ara-lac)95. Likewise, ⌽(araB⬘-
lacZ⫹)96 indicates that the fusion results in a truncated araB
gene fused to an intact lacZ gene, and ⌽(malE-lacZ)97(Hyb)
shows that a hybrid protein is synthesized. An inversion is
shown as IN(rrnD-rrnE)1. An insertion of an E. coli his gene
into plasmid pSC101 at zero kilobases (0 kb) is shown as
pSC101 ⍀(0kb::K-12hisB)4. An alternative designation of an
insertion can be used in simple cases, e.g., galT236::Tn5. The
number 236 refers to the locus of the insertion, and if the strain
carries an additional gal mutation, it is listed separately. Addi-
tional examples, which utilize a slightly different format, can be
found in the papers by Campbell et al. and Novick et al. cited
below. It is important in reporting the construction of strains
in which a mobile element was inserted and subsequently de-
leted that this fact be noted in the strain table. This can be done
by listing the genotype of the strain used as an intermediate in
a table footnote or by making a direct or parenthetical remark
in the genotype, e.g., (F⫺), ⌬Mu cts, or mal::⌬Mu cts::lac. In
setting parenthetical remarks within the genotype or dividing
the genotype into constituent elements, parentheses and
brackets are used without special meaning; brackets are used
outside parentheses. To indicate the presence of an episome,
16 August 2014, Instructions to Authors
parentheses (or brackets) are used (␭, F⫹). Reference to an
integrated episome is indicated as described for inserted ele-
ments, and an exogenote is shown as, for example, W3110/
F⬘8(gal⫹).
For information about the symbols in current use, consult
Berlyn (Microbiol. Mol. Biol. Rev. 62:814 –984, 1998) for E.
coli K-12, Sanderson and Roth (Microbiol. Rev. 52:485–532,
1988) for Salmonella serovar Typhimurium, Holloway et al.
(Microbiol. Rev. 43:73–102, 1979) for the genus Pseudomonas,
and Piggot and Hoch (Microbiol. Rev. 49:158 –179, 1985) for
Bacillus subtilis.
Conventions for naming genes. It is recommended that
(entirely) new genes be given names that are mnemonics of
their function, avoiding names that are already assigned and
earlier or alternative gene names, irrespective of the bacterium
for which such assignments have been made. Similarly, it is
recommended that, whenever possible, orthologous genes
present in different organisms receive the same name. When
homology is not apparent or the function of a new gene has not
been established, a provisional name may be given by one of
the following methods. (i) The gene may be named on the basis
of its map location in the style yaaA, analogous to the style used
for recording transposon insertions (zef) as discussed below. A
list of such names in use for E. coli has been published by Rudd
(Microbiol. Mol. Biol. Rev. 62:985–1019, 1998). (ii) A provi-
sional name may be given in the style described by Demerec et
al. (e.g., usg, gene upstream of folC). Such names should be
unique, and names such as orf or genX should not be used. For
reference, the Coli Genetic Stock Center’s database includes an
updated listing of E. coli gene names and gene products. It is
accessible on the Internet (http://cgsc.biology.yale.edu/index
.php). A list can also be found in the work of Riley (Microbiol.
Rev. 57:862–952, 1993). For the genes of other bacteria, con-
sult the references given above.
For prokaryotes, gene names should not begin with prefixes
indicating the genus and species from which the gene is derived
(for example, do not use EcmecA for the mecA gene from E.
coli). However, subscripts may be used where necessary to dis-
tinguish between genes from different organisms or strains as
described in section v of “Bacteria” above.
Locus tags. Locus tags are systematic, unique identifiers
that are assigned to each gene in GenBank. All genes men-
tioned in a manuscript should be traceable to their sequences
by the reader, and locus tags may be used for this purpose in
manuscripts to identify uncharacterized genes. In addition,
authors should check GenBank to make sure that they are us-
ing the correct, up-to-date format for locus tags (e.g., upper-
case versus lowercase letters and the presence or absence of an
underscore, etc.). Locus tag formats vary between different or-
ganisms and also may be updated for a given organism, so it is
important to check GenBank at the time of manuscript prep-
aration.
“Mutant” versus “mutation.” Keep in mind the distinc-
tion between a mutation (an alteration of the primary se-
quence of the genetic material) and a mutant (a strain carrying
one or more mutations). One may speak about the mapping of
Journal of Bacteriology jb.asm.org
 Instructions to Authors

a mutation, but one cannot map a mutant. Likewise, a mutant
has no genetic locus, only a phenotype.
“Homology” versus “similarity.” For use of terms that de-
scribe relationships between genes, consult the articles by
Theissen (Nature 415:741, 2002) and Fitch (Trends Genet. 16:
227–231, 2000). “Homology” implies a relationship between
genes that have a common evolutionary origin; partial homol-
ogy is not recognized. When sequence comparisons are dis-
cussed, it is more appropriate to use the term “percent se-
quence similarity” or “percent sequence identity,” as appropri-
ate.
Strain designations. Do not use the genotype as a name
(e.g., “subsequent use of leuC6 for transduction”). If a strain
designation has not been chosen, select an appropriate word
combination (e.g., “another strain containing the leuC6 muta-
tion”).
tors, and of Roberts et al. (Nucleic Acids Res. 31:1805–1812,
2003) for restriction enzymes, DNA methyltransferases, hom-
ing endonucleases, and their genes should be used. The no-
menclature for recombinant DNA molecules constructed in
vitro follows the nomenclature for insertions in general. DNA
inserted into recombinant DNA molecules should be de-
scribed by using the gene symbols and conventions for the
organism from which the DNA was obtained.
Tetracycline resistance determinants. The nomenclature
for tetracycline resistance determinants is based on the pro-
posal of Levy et al. (Antimicrob. Agents Chemother. 43:1523–
1524, 1999). The style for such determinants is, e.g., Tet B; the
space helps distinguish the determinant designation from that
for phenotypes and proteins (TetB). The above-referenced ar-
ticle also gives the correct format for genes, proteins, and de-
terminants in this family

Bacteriophages. The genetic nomenclature for phages dif-

fers from that for bacteria and tends to have separate conven-
tions for each phage. Genetic symbols may be one, two, or
three letters and are italicized. For example, a mutant strain of
␭ might be designated ␭ Aam11 int2 red114 cI857; this strain
carries mutations in genes cI, int, and red and an amber-sup-
pressible (am) mutation in gene A. Phenotypic symbols and
designations of gene products are not italicized (e.g., “the Spi
phenotype” or “Int protein”), and superscript plus and minus
symbols can be used to indicate wild-type and mutant pheno-
types, respectively. Host DNA insertions into phages should be
delineated by square brackets, and the genetic symbols and
designations for such inserted DNA should conform to those
used for the host genome. Lists of gene symbols for several
phages can be found in Genetic Maps, 6th ed. (S. J. O’Brien, ed.,
Cold Spring Harbor Laboratory, Cold Spring Harbor, NY,
1993). Relevant references for some of the more widely studied
phages are as follows: for phage ␭, Daniels et al. (p. 469 –515, in
R. W. Hendrix, J. W. Roberts, F. W. Stahl, and R. A. Weisberg,
ed., Lambda II, Cold Spring Harbor Laboratory, Cold
Spring Harbor, NY, 1983); for phage T4, Kutter et al. (p.
491–519, in J. D. Karam, ed., Molecular Biology of Bacterio-
phage T4, American Society for Microbiology, Washington,
DC, 1994); and for phage T7, Dunn and Studier (J. Mol.
Biol. 166:477–535, 1983).

Transposable elements, plasmids, and restriction en-

zymes. Nomenclature of transposable elements (insertion se-
quences, transposons, and phage Mu, etc.) should follow the
recommendations of Campbell et al. (Gene 5:197–206, 1979),
with the modifications given in section vi of “Bacteria,” above.
The Internet site where insertion sequences of eubacteria and
archaea are described and new sequences can be recorded is
http://www-is.biotoul.fr/.
The system of designating transposon insertions at sites
where there are no known loci, e.g., zef-123::Tn5, has been
described by Chumley et al. (Genetics 91:639 – 655, 1979). The
nomenclature recommendations of Novick et al. (Bacteriol.
Rev. 40:168 –189, 1976) for plasmids and plasmid-specified ac-
tivities, of Low (Bacteriol. Rev. 36:587– 607, 1972) for F⬘ fac-

August 2014, Instructions to Authors Journal of Bacteriology jb.asm.org 17


100 CHAPTER 4· 0 GENES IN POPU-LATIONS
rates (see Chapter 7), while 95% of the mammalian Y chromosome does not
recombine.
An important example of linkage disequilibrium at the molecular level
occurs in the genes which make the inajor histocompatibility complex (MHC),
a vital part of the vertebrate inunune system (the evolution of this gene complex
is discussed in detail in Chapter .7): Alleles at different MHC genes, called
haplotypes;· are óften inherited in combination; with . different haplotypes
found in different·populations. For example, the alleles AL B8 and DR3 show
significant.linkage disequilibriulTl in Caucasian populations. Unfortunately, this
haplotype has also been associated with a faster progression to AIDS in mv­
infected individuals.
4.2.2 Natural selection
Natural selectionis at the heart of the evolutionary process. Because of the
'struggle for existence' that faces organisms as they compete for resources,
those with genes that better adapt them to their environment have a greater
probability of surviVing this struggle, so that their favourable genes áre
preferentially passed on and will increase in frequency in the population.
Although natural selection is synonymous with the name of Charles Darwin,
it was .R.A. Fisher who <lid much to show the power of this process at the
gen�tic leveL
Natural -selection can act in a variety of ways and at different points during
an otganism's life-cycle. For example, genotypes rnay differ in their viability
of producing ·adult organisms from zygotes. La ter in the life-cycle, individuals
may·differ·in their·fecundity-the numbers of offspring they produce­
which:wil1 also lead to natural selection. However, before this takes place there
may also be-sexual selection in which organisms differ in their mating success.
·Finally, gametes may have different probabilities of achieving fertilisation
leading to a fom1 of gamétic selection.
, The simplest way of thinking about whether one organism is better adapted
than another is to-describe its fitness. In population genetic terms, fitness can
be défined as the ·capability-that any particular génotype has to survive and
reproduce. This is usually expressed in relative terms, such as whether the
heterozygote Aa has a higher fitness than the homozygotes AA and aa. Fitness
is also specifi.c to each environment, because a genotype which is beneficia! in
one location might be deleterious in another (and vice versa).
In molecular evolution, fim•esses are most often expressed _in terms of the
selection coefficient, · denoted · s, which is a measure of the reduction in
fitness comPared with the best genotype in the population: for instance, an s
of 0.01 rn:eansthat a genotype has a I % less chance ohurvival than the Qest
genotype-it is 99% as fit. Despite this very .simple notation, fitness is a
complex entity which may change through time, and is difficult to measure in
nature. Indeed, there ai'e very few • cases in which the differences between
n
�
101
alleles at a single locus make- enough difference to phenotype to have a
measurable effect on fitness.
In most cases mutations will be deleterious and so lower fimess. These
mutations will be removed fairly quickly from the population by purifying or
negative selection. At other times a new mutation might have a higher füness
than ali others in the population and so is able to multiply. This is usually
called positive selection in molecular evolution, and will lead to.the eventual
fixation of the selectively favoured allele (see below).. For example, it may
be that allele A has a higher fitness than. allele a. This m·eans .(assuming A is
dominan! to a) that selection will favour the AA andAa genotypes over the aa
genotype. In terms of selective coefficients, we can think of this as:
Genotype AA Aa ªª
Fitness 1 ¡ -s
On the other hand, if there is codorninance, so that the Aa het_erozygote l;J.as a
fitness intermediate to those of the two . homozygotes, then:
Genotyp� AA Aa ªª
Fitness l 1-s I -2s
As well as Iooking at a single Iocus, it is also possible to think of natural
selection acting on a phenotypic character which is under the control .of many
different loci, such as height or body weight. The genetics of these quantitative
characters is discussed more fully in section 4.3; 'fypically,: characters · of this
son have a bell-shaped normal disnibuti.on of phenotypes, with most ·indivi.du.als
somewhere near the middle (the mean) of the distrib.ution,.and-decreasing
numbers ar more extreme values (represented by a variance). Positive selection
on qua:f1:titative characters; more often referred to as directional selection,
will move the mean·of the distribution in one direction (-Fig. 4.3a).
Sorne of the best examples of positive sele'ction_ involve the eyolution..of
resistance to antibiotics, drugs and pesticides, such as that built up by mosquitoes
to the chernical DDT once used to control them. A spectacular example of this,
and.one that has been of. great practica! importance in recent years, is- the
resistan.ce developed by lllV-1 to the drug AZ�. AZT inhibits viral replication
by terminating transcription of the enzyme reverse transcriptase. In the absence
of AZT, the wild-type allele for reverse transcriptase was found to be between
0.4 and 2.3.% more fit than alleles adapte:d for dnig resistance (s = 0.004.-
0.023 ). Although the drug works extremely well for the first few inon�s, -and
greatly reduces .the amount of circulating virns, the benefit is only shorHived
and within six months or so the virus is able to evolve resistance through a
small number of mutations. The frequency of the wild-type allele then declines
rapidly. Current apprbaches to IllV therapy therefore involve combinat.ions
of ·drugs because the virus has a much smaller chance of developing multiple
resistance.
In other cases.natural s.election wilJ favour the hetero�ygote over either of
102 CHAPTER 4 GENES IN POPULATIONS 103

whkh takes exactly the samc fnrm as 1he equilíbrtum achieved under forward
(a)
and backward mutatiort pressure {eguatíon 4.7}.
Phenotypic
distributlon Directíonal The most ce[ebrated exampies of overdomlnance are the sickle-ce!l poJy•
in the _selec;tlon morphísm in human fi-haemogiobin and the alcohol dehydrogenase {Adh)
populatían
polymorphism in Drosophila (see Chapter 7). Overdominant selectíon is also
likely to be opcrating on the MHC beca use a heterozygote witb dlfíerent Jv1HC
molecules will recognise more parasites than a homozygote with only one
type of MHC molecule. The analogo1.1s process to overdominant selection for

(b)
, quanti1ative characters, in that genetk variafüm is actively mainudne<l, is
usually called stabilising selection and meaos that indlviduaJs in the mean
oí the dístributlon wm benefü over those at thc extremes (Fig. 4.3b).
St.i:bHí$lng The alternative to overdominance is where the hcterozygote has a hnver
selectíon Fig. 4.3 Differcnt types nf fii.ness rhan either hornozygote:
natural selectlon and how
they cffec1 the phenotypk Genotype AA Aa "ª
dl.stributicm of a dwracter Fitness 1 1-s
controllcd by many genes
(Le. a qunn:i!a:ive d aracter This is called underdominance and although homozygous genotypes are
with a nDnnal dis1rib111ion) favoured, it still preserves both the A anda allcles although this polymorphisrn
and fht' frequende o( alleles
will he unstable, wlth a change in aJlele frequency away from the equilibrium
iH a single lorus where each
Disruptive shaded elilpse represents a leading to the fixation of A or a, For quantitatíve cha.ractcrs an anafogous
se!ection population of individuals process is cafled disruptlve selection whereby individuais at the extremes
(cirdes:) with an allele of a of the distribution have a higher füness than thos.e with mean values (Fig.
pan.icu!ar colour-black ,,r 4.3c).
whitc. Thc open and shaded Underdominant selection appears to be controiling bill -size in the African
bells represen! the phe:notypic
dlsufüutiom before and after finch Pyrenestes ostrinus studie<l by Thomas Bates Smith. Bill sízes in this species
selection. respectively, and are either large or small. but not of intermedíate dimensions. The basis for this
the arrows signiiy the differem:e appears to be the hardness of the seeds eaten by the birds: duríng
dircct!on of sele.::tion. the reproductive season both the large• and srnaU-biHed morphs preíer soft
seeds but durlng the dry season, when food abundance is low, the farge-billed
morph tends to leed on hará seeds while the smail-biUed morph expands
its diet to indude other foods. No seeds of íntermediate hardness are found 1n
the homozygotes. This is caUed overdominance (sorne-times heterozygous
the Jocafüies where these birds líve so that selection acts agalnst those wlth
advantage) and wc can think of it in tenns of fitnesses as follows:
intermedíate bill sizes.
Gen-0type AA Aa ª" Polymorphísms can also be rnaintained in populations through frequency-
Fitness ¡-s 1 1- t dependent selection (Fig, 4.4). Thjs occurs when the fitness of a gcnotype
depends on its frequency in the population. Por example, the genotype at
where t is simply the selection coeffident for the aa genotype. Overdominant
the !owest frequency may have thc highest fitness {negative frequency-
selection wH! produce high heterozygosíty (Le. over that expected under tbe
dependent selectiún). Thís may occur tri host-parasite sys'tems where a
Hardy-Weinberg eguilihrium) and beca use it preserves both the p and q alleles
parasite at low frequency is not recognlsed hy the host immune system and is
will maintain genetic varíation in populations, giving rise to a balanced
therefore able to increase in frequency, whereas a strain at higher frequency
polymorphism. The eqtiilibrium allele frequencies' {i.e. fa) whkh reflect
may elicit a stronger hnmune response agalnst it and so be more easily deared.
Lhis bá.fance depcnd only on the relative fitncsses of the three genotypes, so
This wíll cause an oscillation in the frequcncy of different strains as they evade
1hat:
or are recognised by the immune system, leading to a polymorphism in thc
population. In other circumstances t.he genotype vvith the híghcst frcqucncy
(4.9) may be íavoured (positive frequency..<fependent s.eiection), whicb wiU
104 CHAPTER 4 GENES 1N POPULAT!ONS 105

allele} will be after a generation of this seiective process. We can denote this
aa new frcquency of p as p', and can cakulate it by:

il, p'=_P_2 (4. 10)

J,__
_________
1l l-sq
;.M The change in allele frequency (tlp} can then be simply computed as p' - por
O 1 directly by:
Frequencyof allele A
'
Fig. 4.4 AUele and genotype frequendes under frequency•dependcnt selecrlon. Genotypes Ap=...!E'L (4.11)
AA and ha are suhjttt to n tlve frequency-dependent sele<tion whereas the fitness oí Aa 1 - s q "'
L frequency-independent. To the rlght of the po\m where tbe lines cn) s. the A allele- is It is worth m1.1:strating this with a numera eiample. Let's assume p arid·q i
favourt"d and ilu:rea.,i¡es in irequency althoilgh the fitness oí thé AA genotype decreases, To
thc left of this pohlt, tbc fre{}uency Of the A·anetc declines aud the a1lelé is favoured. the _population are initially 0.25 and O. 75, respectively, and the sclection
Althoúgh the relationship be1weCil íimess and írequency Is shown here as a stratght line, coeffícient against aa is at the top end of what we observed in the case oí AZT
in reallty it can take on any shape. Adapted Jrom Rid.ley (1996). resistance, say 0.025. Applying equation 4.11 to get Ap we discover that-

spq 2 0.025 X 0,25 X (0.75)'

prevent polymorphisms from heing established. Ftequency dependent selecrton
may also ocrur if a population occupies a heterogenem.rn envíronment,. so that
1- sq2 = 1-0.025 X (0.75)'
dlfferent alleles are favoured in different patches (mkro-environments) or at 0.0035
different times. =0.9854
Unfortunately, detecting natural selection at the molecular leve! is a =0.0035
notoriously difficult thing to do because its effects can be subtle and becaus.c which means· that p V\'ill have íncreased fróm 0.25 to 0.2535. Although thís
different evoltHionary processes can leave similar signatUres in sequences (see seems Jike a tiny íncrease ít has happened in a single generation, and represents
section 4,.5 and Chapter 7). An added complkation is that the agem of selection spectacular progress compared with what happe-ns under mutatíon alone. lf
may. no longer· be presen:t in the population. For instance, the chemokine we run this process for long enough:rhen the A allele wilI eventually achleve
receptor 5 (CC 5) gene ts an importan t. factor in aUowíng HIV-1 to infect cells fixation. However, the rate at which this occurs is not constant, befng fastest
called macr,ophages. About 10% of Caucasian people have a 32-bp deletion when lhere are many heterozygotes in the population and slówest ·When
(called A32} which inactiva tes this gene and those homozygous for this deletion there are mostly homozygotes (plug in sorne different values for p and q and
(about 1% of individua)s) have a greatly reduced susceptibility to BIV infection, see for .yoursdf}. The same will aiso -be true lf thfie •is codominance" Jn
whil thosc that an; heterózygous may become HIV infected but seem to which case the change in allele freqUencies takes: on ai1 S shaped· (s.ymmetrkal)
progress more slowly to AIDS. Whatis puzzling is that this mutation is at such form (Fig, 45). In su:m, we can see that natural selection operares most
hi.gh frequency in Caucasians {see Fig. 4.13) but; is very rare in African and efficiently when there are large amo-unts of genetic variation for it to Work
Asían populations, Because HJV is so recentan invader it cannot be the selective with, Furthermore, the efíidency o( natura} selection is also detennined by
agent responsible for the high frequency of this mntation, which raises the the size of the population, working best when it is large, When popuiation
posslbility that a past infectious disease may have acted as a strong selectíve
force favoming individuals 1-'Vith this deletion.
Despite chis complexity, it is still possible to show how natural selection
changes g ne frequendes \"ilith some simple mathematks. Taking the simplest Fig. 4.5· The change in al!ele
example, of positive selection favouring AA and Aa over aa, so that: frequendes under posltive
natural sele.:tion for a
Genotype AA Aa ªª oodominant advamageoll5
Fitness l l 1-,. A ailek that has entered
a population of na
it ís easy to see that frequenCf of the aa genotyJ}e, or q2 if we rccall our Hardy- homwygotes.. Frmn Griffiths
Weinberg fo1mula, will be. reduced by 1 - s in a given generatton, The next eta!. (1993), wilhpemtlsslon Number.of generatlom. - - - ►
thing we would like to do is calculate what p {the frequency of lhe favoured írom WH Freeman.
106 CHAPTER 4 GENES IN POPULATI'ONS 107

sizes are small, mutations are more under the control oI chance lhOcesses.
This is the subject of the nexr section.
límeto
füation
4.2.3 Genetic drift
Once a mutation has arisen in a popu1ation it can experlence one of two
evolutionary fates: it cat1 be fixed and so come to domínate the population, or •A!lelernpy
dies out
ít can be ]ost. \Vhích outcome a new allele faces ls not ahvays down to how because it
much betrer or worse it is compared to those alleles already present in the faih to
population. Instead it may simp-ly be down to dlance. For example, sorne reproduce
ali eles may reside in individuals who leave no offspring and even if an ir1dívidual
does produce offsprlng only a rnndom sample of their genes. those present in
the successfui egg or sperm, will be inherited. Because of bad luck, most gametes
wil.l not make it to the next generation. This means that alleles are in effect
randornly s.ampled in,every generation, and tbis random sampHng can change
Flg. 4.6 Fixation oI an allele
altete frequendes. This is caHed genetíc ddft. (bfack circle) by genclíc drift.
To understand how allele frequendes change wlth genetk drift we need to The shaded cltipse represems
devefop a stochastic model of evolution, fn which chance plays the leadjng a population observcd at
role. Thís differs from our prevíous discussions of mutation and selection wbere difiermt time points, The
we dealt with determinist:k models, which ignored random effects, First othcr ailde in the poJJUlatlon
(individuals. with whit¡;
imagine a diploid population of size N, so that there are a total of 2N allelic
drdes) dies out becaose ft
copícs of cach gene. Because of the random sampling of gamctes there is a fatls to reproduce. On average
chance that some of these alJeles will contrtbute no copies of themselves to it takes an allí'lí' 4iV Odg!nof
thc ncxt generation, whilst others will colltribute many. Sorne a.Beles wlll Henerations t) gí't to fixation. afte!e
.at1hough the probabUity of (generation 1)
therefore be Iost from the population ín each generation because the gamctes
carrying them have not been passed on. This will cause aUele frequendes to this happcniug is oµly l/2N.
fh.1ctuate slightly from generation to generation as sorne alleles are lost and
others reproduce. Ifthis stochastk variation in reproductive sucress continues advantage, most of whkh 'WiH also be lost by chance unless their benefü is
for long eno11gh tl1ere wilI eventually come a time when all 2N alleles wmbe substantiaL For example, R,A. Fisher cakulated that a mutant with a l %
descended from a slngie of our odginaI 2N alleles because all others have failed selecüve advantage only has about a '2% chance of being fixed (So a 9íP'lo
to reproduce at some stage along the way (Fig. 4.6}. This aHeie wíll then have chance of being Jost). Loss by genetic drift is espedally Bkely soon after thc
hecn ffxed by the process oí genetic drift alone, The probability oí fixation of mutation has appeared and so is still at low frequency.
an alleJe by random gcnetk drift is simply 1/(2N}, which is its frequency in tbe Usins a matbematical modcl it has been shown that H takes an average of
Jlopulation after it has arisen by mutation. Thls also means rhat the probability 4N generations for a neutral altele to get to fixation through genetic driit,
of fixatíon is greater when the population slze is small. because l/2N is larger, aJrhough the random naturc of the process rneans that tbere is a large vatiance
Not only will genetic drift work better in small populations, bul it will also a.round thh time. Mutations with a selective advantage will bC fixC'd qukker
be most important for those mmations which have no effect on phenotype, rhan neutral alleles (that is, in less than 4N generations) because indíviduals
and so are neithe advar.t..1geous nordeleterious compared to theirpredecessors. with the better gene "vlll produce more offspring than those with other alleles,
These mútations, whlch are free from the rigours of natural selecfüm, are Conversely, when natural selection is actively maintaining a polymorphism,
called neutral mutations and their evolution is díscussed in detaíl in Chapter .alleles can coexist in the popuíarJon wilhout any going to fixation far as long
7. Beca use an individual carrying a neutral rnutation will not beata selective as the selection p:ressure exists, so that they may survive for Ionger than 4N
advantage, the only way they can he fixed is through the chance action generations.
of genetic drift (tbe one exception-genetic hitchhiking-is discussed in During the time it takes for an allele ro be Jixed by genetic drift, thc
Chapter 7). However, genetíc drift wiH aiso affect new mutations wíth a selective population will exhibit polymorphísm at the locus in question. Obviously, the
 FEMS Microbiology Reviews 26 (2002) 355^374
www.fems-microbiology.org

The bacterial replication initiator DnaA. DnaA and oriC,

the bacterial mode to initiate DNA replication


Walter Messer
Max-Planck-Institut fu«r molekulare Genetik, Ihnestrasse 73, D-14195 Berlin-Dahlem, Germany

Received 2 June 2002; received in revised form 1 July 2002; accepted 2 July 2002

First published online 3 October 2002

Abstract

The initiation of replication is the central event in the bacterial cell cycle. Cells control the rate of DNA synthesis by modulating the
frequency with which new chains are initiated, like all macromolecular synthesis. The end of the replication cycle provides a checkpoint
that must be executed for cell division to occur. This review summarizes recent insight into the biochemistry, genetics and control of the
initiation of replication in bacteria, and the central role of the initiator protein DnaA.
, 2002 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.

Keywords : Protein domain ; DnaB helicase ; ATP-binding DnaA box; Plasmid

Contents

1. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
2. The di¡erent functions of DnaA protein, an overview . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
3. Steps in the initiation of E. coli DNA replication . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 356
4. Domains of DnaA and the homology between DnaA proteins . . . . . . . . . . . . . . . . . . . . . 357
5. Domain functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 358
6. The role of ATP binding on DnaA structure and function . . . . . . . . . . . . . . . . . . . . . . . . 359
7. DnaA and membranes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
8. DnaA boxes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
9. DnaA oligomerization and cooperativity. Rules for DnaA binding . . . . . . . . . . . . . . . . . . 360
10. DnaA-mediated origin unwinding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361
11. DnaA-mediated helicase loading: the DnaA primosome . . . . . . . . . . . . . . . . . . . . . . . . . . 362
12. Variations of the theme: DnaA and oriC in other bacteria . . . . . . . . . . . . . . . . . . . . . . . . 362
12.1. B. subtilis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 362
12.2. Streptomyces . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 363
12.3. T. thermophilus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364
12.4. H. pylori . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364
12.5. Synechocystis sp. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364
12.6. Caulobacter crescentus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 364
13. DnaA as a transcription factor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365
14. Joint action of DnaA and plasmid initiators in plasmid replication . . . . . . . . . . . . . . . . . . 365
14.1. P1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365
14.2. F . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365
14.3. RK2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365
14.4. R6K . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366
14.5. R1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366
14.6. pSC101 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366

* Tel. : +49 (30) 8173788; Fax: +49 (30) 8413 1385. E-mail address: messer@molgen.mpg.de (W. Messer).

0168-6445 / 02 / $22.00 , 2002 Federation of European Microbiological Societies. Published by Elsevier Science B.V. All rights reserved.
PII: S 0 1 6 8 - 6 4 4 5 ( 0 2 ) 0 0 1 2 7 - 4

FEMSRE 756 24-10-02 Cyaan Magenta Geel Zwart

356 W. Messer / FEMS Microbiology Reviews 26 (2002) 355^374

15. The special case of DnaA and V plasmids . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 366

16. How to limit initiation to once per generation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367
16.1. Sequestration . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367
16.2. Titration of DnaA to the datA locus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367
17. DnaA and oriC in the cell cycle, control of initiation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367
18. Perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368
19. Note added in proof . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368

Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 368

1. Introduction
The initiator protein DnaA is found in all eubacteria
analyzed so far. The name derives from the fact that
dnaA was the ¢rst mutant isolated that is a¡ected in
DNA replication. In fact, mutants from this series were
the ¢rst temperature-sensitive mutants in bacteria and es-
tablished the use of this type of conditional lethal muta-
tions for cellular functions [1,2]. The replicon hypothesis
by Jacob et al. [3] postulates two basic elements for the
initiation of replication : the initiator, a trans-acting sub-
stance, DnaA protein, and the cis-acting replicator, which
we now call replication origin, oriC. Most bacteria possess
a unique replication origin on the usually circular chromo-
some. Several aspects of the initiation of bacterial replica-
tion are covered in a number of recent reviews [4^7].
2. The di¡erent functions of DnaA protein, an overview
All functions of DnaA protein depend on its ability to
bind speci¢cally to an asymmetric 9-bp recognition se-
quence, the DnaA box: 5P-TTATNCACA. A replication
origin, oriC, usually consists of an array of several DnaA
boxes. Binding of DnaA to such an array, origin recogni-
tion, is the ¢rst step in the assembly of a specialized nu-
cleoprotein complex [8], the initiation complex. Structural
distortions of the DNA within such a complex are the
prerequisite for the second function of DnaA in the ini-
tiation process. It acts as a DnaA primosome. Protein^
protein interaction results in the loading of the replicative
helicase, DnaB in the case of Escherichia coli.
DnaA protein is also a transcription factor (for a review
see [9]). Binding to one or two DnaA boxes in a promoter
region may result in repression. Most importantly, the
dnaA gene itself is repressed by DnaA, it is autoregulated.
Other promoters are activated by DnaA binding. If DnaA
boxes are within a transcription unit, DnaA binding may
result in transcription termination.
3. Steps in the initiation of E. coli DNA replication
Most of our knowledge on DnaA and oriC is based on
E. coli. Especially the elegant work over many years from
the Kornberg laboratory using the DnaA- and oriC-de-
pendent in vitro replication system is the basis of our
understanding [10].
oriC of E. coli contains in 260 bp ¢ve DnaA boxes, and
at its left end an AT-rich region consisting of three 13-mer
repeats and the so-called AT cluster (see Fig. 1). In addi-
tion, there are binding sites for accessory proteins like IHF
and FIS, and control factors like IciA, Rob, H-NS whose
precise function is not known (see [5] and references there-
in). The importance of 11 GATC sites, recognition sequen-
ces for the Dam methyltransferase, will be discussed in
Section 16.
DnaA protein binds to its ¢ve binding sites in oriC as a
monomer, introducing a 40‡ bend at each site [11]. Only
DnaA complexed with ATP is active in initiation [12].
ATP-DnaA binds to additional sites, 6-mer ATP-DnaA
boxes with the sequence 5P-AGATCT or a close match
[13,14]. These sites are predominant in the AT-rich region
which is unwound, and subsequently stabilized due to spe-
ci¢c binding of ATP-DnaA to single-stranded ATP-DnaA
boxes. In this sequence of interactions binding to 9-mer
DnaA boxes in oriC is a high-a⁄nity interaction (KD =
1 nM), binding to 6-mer ATP-DnaA boxes in double-
stranded DNA is a low-a⁄nity interaction that requires
cooperativity among the proteins, and binding to 6-mer
ATP-DnaA boxes in single-stranded DNA is again a
high-a⁄nity interaction [14]. A threshold level of DnaA
protein is required for the conversion of the initial to the
open complex. It is presumably because of these graded
a⁄nities and the cooperativity between the proteins that a
precise timing of initiation in the cell cycle is possible. The
net result is unwinding of the AT-rich origin region [15^
17]. FIS protein has a negative e¡ect on the reaction
[18,19], HU and IHF proteins [20,21], a high ATP concen-
tration ( s 2 mM), high temperature (38‡C) [22], and tran-
scriptional activation enhance unwinding [21,23^25]. The
replicatively active complex contains 20^30 DnaA mono-
mers, as determined by electron microscopy [26,27]. Direct
measurement by surface plasmon resonance gives a stoi-
chiometry of 18 DnaA monomers per oriC in the initial
complex [7]. The open complex with six more binding sites
in the single-stranded AT-rich region thus must contain 24
DnaA monomers, corroborating the electron microscopic

FEMSRE 756 24-10-02 Cyaan Magenta Geel Zwart

 W. Messer / FEMS Microbiology Reviews 26 (2002) 355^374
Fig. 1. Schematic replication origin of E. coli.
results. However, lower stoichiometries have also been re-
ported, and the in£uence of binding conditions on DnaA
stoichiometries has been emphasized [28].
The unwound region spans 28 bp without and 44^46 bp
with SSB present [17]. Since single-stranded DNA covered
with SSB is a poor substrate for DnaB helicase, it has to
be loaded with the help of DnaA. Two double hexamers of
DnaB and the helicase loader DnaC, one double hexamer
for each replication direction, are positioned by DnaA
into the loop [28,29]. DnaC leaves the complex immedi-
ately after or during loading, accompanied by ATP hydro-
lysis. This activates the helicase activity of DnaB [30]. The
two DnaB hexamers slide past each other in 5P-3P direc-
tion, and expand the bubble to about 65 nucleotides [29].
Now primase can enter the complex and synthesize two
leading strand primers. Synthesis of opposed and overlap-
ping leading strands has been postulated before in order to
ensure complete synthesis of oriC [31].
The sliding clamp, a ring-shaped dimer of the L-subunit
of DNA polymerase III, is loaded onto each primed tem-
plate by the clamp loader, the polymerase III Q complex
[32]. This activates the intrinsic ATPase activity of DnaA
[33,34], in cooperation with a protein with sequence ho-
mology to DnaA, Hda [35]. Inactivation of DnaA by ATP
hydrolysis prevents further initiations. The E. coli initia-
tion cycle is shown schematically in Fig. 2.
4. Domains of DnaA and the homology between DnaA
proteins
Sequence homology based on four species [36,37] and
later on 30 species [6] made it possible to subdivide DnaA
protein into four domains. These domains were later
shown to correspond to functional domains, with minor
correction of their borders [38,39]. The functional studies
were inspired by earlier attempts to combine sequence
comparisons with secondary structure predictions [40].
These results were recently corroborated and extended
by a thorough study that compares 104 di¡erent DnaA
proteins (C. Weigel and W. Messer, http://www.molgen.
mpg.de/Vmesser/). Despite considerable e¡ort no three-
dimensional structure of a DnaA protein is available.
FEMSRE 756 24-10-02
357
However, the high number of proteins to compare and
the possibility to relate the predicted structure to experi-
mentally determined structures of other members of this
protein family, the AAAþ family, make this prediction
pretty reliable. The AAAþ class of ATPases is a large
family of proteins with common sequence motifs. Many
of them are involved in the initiation of replication and
other functions of DNA metabolism throughout all king-
doms with a presumptive role in remodeling and loading
of proteins. DnaA and also DnaC are members of this
family [41].
Grouping of DnaA sequences from eubacteria on the
basis of similarity reproduced closely the phylogenetic re-
lations. DnaA domains are shown in Fig. 3. For domain 1
(86 residues) a consensus sequence is not very pronounced,
except for a few patches of conserved residues. Domain 2
is highly variable in sequence and length, between 6 and
247 (Streptomyces) amino acids. Therefore we assume it to
form a £exible region. The C-terminal two-thirds are quite
similar and make it possible to derive a reasonable con-
sensus sequence. Domain 3 contains an ATP-binding site,
i.e. Walker A and B motifs, and other motifs found in the
AAAþ protein family. Domain 4 is also well conserved
and lacks homologues in the archaea as well as in the
eukarya. It contains the DnaA signature, a de¢nition
used in the SWISS PROSITE online search facility,
which has been updated on the basis of the now known
104 DnaA sequences: [STPQ]-x(3)-[IL]-[GA]-x(2)-[FLIM]-
x(1,2)-[RK]-[DTSNER]-[HA]-[TSPA]-[TSVA]-[VILM] (C.
Weigel and W. Messer, http://www.molgen.mpg.de/
Vmesser/).
Secondary structure prediction was done by the PHD
server (http://www.embl-heidelberg.de/predictprotein/pre-
dictprotein.html) [42]. Results were virtually identical for
di¡erent prediction methods. The secondary structure pre-
diction showed a much higher conservation for the struc-
tural elements than for primary sequence. It is shown in
Fig. 3. Domain 3 shows an [KL]5 Rossman fold-type ATP-
ase motif [43] suggesting a tight structure which aligns
perfectly with the known structure of polymerase IIINP
[44]. This illustrates the reliability of the prediction. The
prediction for the DNA-binding domain 4 shows four
K-helices, the C-terminal long one is divided in two in
Cyaan Magenta Geel Zwart
358 W. Messer / FEMS Microbiology Reviews 26 (2002) 355^374

some DnaA proteins. K-Helices 12 and 13 (Fig. 3) with a

basic loop in between were suggested to represent the
DNA-binding motif [45].

5. Domain functions

The N-terminal domain 1 mediates protein^protein in-

Fig. 2. The initiation cycle of E. coli.
teractions, DnaA oligomerization and interaction with
DnaB protein. Several techniques were used to demon-
strate oligomerization. Domains 1 of E. coli [46] and of
Streptomyces lividans [47] were able to functionally replace
the oligomerization domain of VcI repressor. Physical in-
teraction of isolated domain 1 was demonstrated with a
solid phase binding assay in vitro, showing that dimeriza-
tion via domain 1 does not require other parts of DnaA
protein [46]. Cooperative binding of DnaA to two adja-
cent DnaA boxes depended on domain 1 [48]. Expression
of isolated domain 1 was inhibitory to initiation, as mea-
sured by suppression of the overinitiation phenotype of
dnaA219 (see Section 6) [46], showing that dimerization
via domain 1 is an essential step in initiation in vivo.
The larger part of domain 1, amino acids 24^86, is also
involved in the interaction of DnaA with DnaB, as shown
in vitro with a solid phase binding assay [49]. In in vivo
experiments domain 1, or amino acids 24^86, was required
for the loading of DnaB [49,50]. There is a second site for
DnaA^DnaB interaction in the N-terminal region of do-
main 3 (see below). In summary, physical interaction be-
tween DnaA and DnaB requires amino acids 24^86 of
DnaA, whereas DnaA dimerization requires the complete
domain 1, amino acids 1^86 ; in later experiments amino
acids 1^77 were found to be su⁄cient. Surprisingly, do-
mains 1 and 2 were found not to be required for DNA
unwinding of the AT-rich region in vitro [50]. However, it
has also been reported that mutants with altered residues
26 or 40 within domain 1 were unable to unwind the AT-
rich region [51]. Additional experiments are required to
clarify this point.
Domain 2 is very di¡erent in size in di¡erent bacteria, as
mentioned above. In E. coli DnaA all of domain 2, amino
acids 87^134, can be deleted without loss of function. A
spontaneous mutant, dnaA216, ‘DnaA light’ lacking resi-
dues 87^104, was functionally indistinguishable from wild-
type [38]. However, long domains 2, e.g. in Streptomyces,
have been conserved in evolution with respect to length
not sequence. Therefore the length of domain 2 may well
have an important function in certain conditions.
Domain 3 is organized as an open twisted KL-structure
[11] and contains the ATP-binding region of AAAþ ATP-
ase-type proteins, a P-loop or Walker A motif [41,43,52^
54]. The presence of an ATPase is shown by the Walker B
motif as well as by the isolation of mutants a¡ected in
ATPase activity, E204Q and R334H [55,56]. The impor-
tance of ATP binding and hydrolysis will be discussed in
Section 6.

FEMSRE 756 24-10-02 Cyaan Magenta Geel Zwart

 W. Messer / FEMS Microbiology Reviews 26 (2002) 355^374
Fig. 3. Secondary structure prediction for E. coli DnaA protein with sequence motifs and interaction domains.
In the N-terminal part of domain 3 is a region, amino
acids 130^148, that comprises a second interaction site
with DnaB [49,50,57,58]. DnaA thus contains two sites
for interaction with DnaB, amino acids 24^86, interacting
with DnaB amino acids 154^210 (C-terminal L Q~ fragment
[59]), and DnaA amino acids 130^148 interacting with the
DnaB N-terminus (K fragment) [49].
Domain 3 also contains a second region that promotes
oligomerization. This has been shown for Streptomyces
DnaA by kinetic analysis [48] and by gel retardation
that even showed the existence of mixed dimers containing
full-length and truncated DnaA [47]. Likewise, in E. coli
DnaA the presence of a region from the C-terminal part of
domain 3 gave a strongly reduced dissociation rate in sur-
face plasmon resonance analysis, suggesting cooperativity
between monomers that did not contain domain 1 [7]. This
same region is conserved in the NtrC family of transcrip-
tion factors [45].
The DNA-binding domain 4 is represented by the 94
C-terminal amino acids [45]. The DNA-binding motif
probably consists of K-helices 12 and 13 (see Fig. 3) with
a basic loop between the helices. In fact, mutation of some
basic amino acids in the loop abolishes DNA binding
[60,61]. A thorough mutational analysis of this domain
revealed that mutations with impaired binding were found
throughout domain 4 [60]. Mutations that a¡ect sequence
speci¢city are clustered in the beginning of helix 15 [60,62].
Suppressors of temperature-sensitive dnaX mutants [63]
also map to this speci¢city region. Binding of ATP to its
site in domain 3 modi¢es the sequence speci¢city of DnaA
protein [13] (see Section 9). In concert with this property,
bound ATP is located close to the DNA-binding domain
in the three-dimensional structure, as found by cross-link-
ing the Q-phosphate of ATP to Lys-415 in domain 4 [64].
FEMSRE 756 24-10-02
359
6. The role of ATP binding on DnaA structure and function
Binding of ATP is required for the unwinding of the
AT-rich region by DnaA [12]. In fact, this is the only
reaction that requires ATP-complexed DnaA, since heli-
case loading occurs in mutants that are de¢cient in ATP
binding [49]. ATP promotes an allosteric modi¢cation and
does not provide energy for the unwinding reaction since
non-hydrolyzable ATP analogues are equally e¡ective [12].
Several of the ‘classical’ dnaA mutants, e.g. dnaA5 and
dnaA46, carry a mutation A184V close to the ATP-bind-
ing site in the Walker A motif [65]. These mutants are
temperature-sensitive and do not bind ATP or ADP at
any temperature [66^68], but can be activated to ATP
binding by DnaK and GrpE proteins [68]. All A184V
mutants carry secondary mutations in dnaA [65,69]. How-
ever, when the A184V mutation is separated from the sec-
ondary mutations, it confers similar temperature sensitiv-
ity [68]. When A184V is present on multicopy plasmids,
such plasmids confer cold sensitivity to dnaAþ or dnaA46
hosts [65,68,70]. Intragenic suppressors of dnaA46 have
been isolated, dnaAcos [71] which carries two additional
mutations (Q156L and Y271H) [69], and dnaA219 with a
R342C mutation [46]. Both suppressor strains are cold-
sensitive due to DNA overinitiation at 30‡C [46,71]. The
dnaA219-carrying strain has been successfully used to
monitor for conditions that are detrimental to initiation
and thereby relieve the cold sensitivity and provide a pos-
itive selection for impaired initiation [46,49].
The active structure of ATP-complexed DnaA is sensed
by a region close to or in DNA-binding domain 4, since
ATP-DnaA acquires an additional sequence speci¢city.
Mutations that a¡ect ATP hydrolysis were found just up-
stream of domain 4, R334H [56], as well as at a position in
Cyaan Magenta Geel Zwart
360 W. Messer / FEMS Microbiology Reviews 26 (2002) 355^374
domain 3, E204Q, that is involved in hydrolysis in similar
ATPases [55]. In addition, physical proximity between
ATP bound in the P-loop to the DNA-binding region
was demonstrated by crosslinking the Q-phosphate posi-
tion of ATP with K415 in domain 4 (see Section 5) [64].
The intrinsic ATPase of DnaA is weak. As outlined in
Section 3 it can be stimulated and DnaA thereby inacti-
vated. Since DnaA protein with the A184V mutation does
not bind ATP (or ADP) we must assume that the sup-
pressing mutations confer an active conformation onto
the protein without nucleotide binding. Of course, such a
protein cannot be inactivated by ATP hydrolysis, which
may explain the phenotype of overinitiation [72,73].
The intrinsic ATPase activity of DnaA is stimulated at
the end of the initiation cycle. This reaction is promoted
by the loading of the sliding clamp to single-stranded
DNA, the dimeric L-subunit of DNA polymerase III ho-
loenzyme [33]. This reaction requires Hda as cofactor [35]
(see Section 3). Therefore, hda mutants have an overinitia-
tion phenotype similar to dnaAcos or dnaA219. The over-
all ratio of ADP to ATP in exponentially growing cultures
was 4:1. In synchronized cultures ATP-DnaA increased
prior to initiation and was hydrolyzed to ADP-DnaA dur-
ing initiation [74]. This also results in a £uctuation of
DnaA synthesis in the cell cycle since ATP-DnaA is the
active repressor for the dnaA promoter [13]. The cycling
between ATP- and ADP-bound states of initiation pro-
teins is apparently a ubiquitous regulatory principle [75].
It has been described that DnaA at low Mg2þ or nucle-
otide-free DnaA protein binds DNA unspeci¢cally
[28,76,77]. However, this is an artifact of the Sekimizu
protocol of DnaA preparation [78] which involves a dena-
turation step. Native DnaA protein is always complexed
with ATP or ADP.
7. DnaA and membranes
A large part of the DnaA protein in E. coli cells is found
in a membrane fraction [78,79]. An E. coli mutant, pgsA,
with a limited ability to synthesize acidic phospholipids
leads to arrest of cell growth [80]. This can be suppressed
by overproduction of DnaA proteins with mutations in
the membrane interaction domain described below [81].
Also, some dnaA mutants have an altered cell membrane
permeability [82]. There are thus a number of indications
that DnaA^membrane interaction is physiologically rele-
vant.
Acidic phospholipids decrease the a⁄nity of DnaA for
ATP or ADP, and therefore enhance the exchange of these
nucleotides and promote a ‘rejuvenation’ of inactive ADP-
DnaA [83^86]. Phospholipids also decrease the a⁄nity of
DnaA for oriC sequences [87]. The region of DnaA that
interacts with membrane was de¢ned by mutations in
K-helices 10 and 11 (see Fig. 3), amino acids 327^344
and 357^374 [88^91]. K-helix 11 (amino acids 357^374)
FEMSRE 756 24-10-02
was found to be indispensable for membrane-mediated
release of nucleotides [92], and a peptide between amino
acids 373^381, adjacent to K-helix 11 on the C-terminal
side, was found to bind to phospholipids [92]. So far, it is
not known whether the physiological relevance for DnaA^
membrane interaction lies in DnaA-mediated binding of
the initiation complex to the membrane or in the ATP-
ADP ‘rejuvenation’ reaction.
8. DnaA boxes
ATP-DnaA and ADP-DnaA bind with the same a⁄nity
(KD between 0.6 and 50 nM) to an asymmetric 9-mer
consensus sequence, the DnaA box 5P-TTA /T TNCACA,
as measured by gel retardation of oligonucleotides carry-
ing a single DnaA box [11] and by surface plasmon reso-
nance [60]. UDG footprinting shows that within this se-
quence T2, T4, T7P, and T9P are of particular importance.
In this technique speci¢c T residues are replaced by dU,
and their accessibility to uracil-DNA glycosylase is deter-
mined with and without DnaA [93]. Previously a more
relaxed consensus sequence which allowed 1^2-bp devia-
tion from this stringent consensus sequence was found
using DNase I footprinting or retention on nitrocellulose
¢lters [94^97]. When the ability of DnaA-DNA box com-
plexes to block transcribing RNA polymerase was used
to de¢ne DnaA boxes an even more relaxed consensus
sequence was found: 5P-(T,C)(T,C)(A,T,C)T(A,C)C(A,G)
(A,C,T)(A,C) [98]. The reason for the apparent discrep-
ancy are the rules for binding of E. coli DnaA protein
described below. A monomer of E. coli DnaA binds
only to the stringent DnaA box. Relaxed consensus se-
quences require the cooperation of two monomers, and
hence two DnaA boxes [7,13,99].
9. DnaA oligomerization and cooperativity. Rules for DnaA
binding
The dnaA promoter region was analyzed by DNase I
footprinting and surface plasmon resonance for the modes
of DnaA binding. The surprising observation was that
ATP-complexed DnaA, but not ADP-DnaA, has a second
sequence speci¢city for a 6-mer consensus sequence, the
ATP-DnaA box [13]. This is also true for the dnaA pro-
moter region of Vibrio harveyi [100]. DnaA binding to
‘non-canonical’ sites was observed before with V DNA
[101]. A summary of these sites and the ATP-DnaA box
consensus sequence is given in Table 1.
The rules for DnaA binding as derived from the dnaA
promoter region are as follows :
1. E. coli DnaA, both ATP-DnaA and ADP-DnaA, binds
to a single DnaA box as a monomer. But this box must
conform to the stringent consensus sequence 5P-TTA /
T TNCACA, henceforth called a ‘strong’ box.
Cyaan Magenta Geel Zwart
 W. Messer / FEMS Microbiology Reviews 26 (2002) 355^374
2. Both forms do not bind to single ‘weak’ boxes, i.e.
DnaA boxes with one or two mismatches to the con-
sensus sequence, unless cooperation occurs with a sec-
ond DnaA protein bound to a DnaA box close by.
3. ATP-DnaA recognizes in addition 6-mer ATP-DnaA
boxes with a consensus sequence that corresponds to
BglII recognition sites. These constitute weak boxes,
and therefore require an adjacent strong box for
DnaA binding.
4. ATP-DnaA also binds to single-stranded ATP-DnaA
boxes. This was observed for the unwound region in
E. coli oriC [14] and for a substrate with a double-
stranded strong box and adjacent single-stranded
DNA with an ATP-DnaA box [102,234]. In the latter
case the strong box was required for binding to the
single-stranded ATP-DnaA box, in oriC ATP-DnaA
could bind to single-stranded ATP-DnaA boxes with-
out the help of a strong box, presumably because of the
high number of such boxes in oriC. Single-stranded
ATP-DnaA boxes have an intermediate a⁄nity
(KD = 40 nM); strong double-stranded boxes have a
KD of about 1 nM, and weak double-stranded boxes,
including ATP-DnaA boxes, have a KD of around
400 nM [14].
As discussed in Section 5, there are two regions respon-
sible for oligomerization of DnaA monomers, domain 1,
amino acids 1^77, and a region in domain 3. This second
oligomerization region was localized in the C-terminal
part of domain 3 between amino acids 334 and 373 [7].
The domain 1 oligomerization domain can operate over a
distance of at least 150 bp [46,49], and does not even
require DnaA protein to be bound to DNA, the domain
3 oligomerization region requires a precise spacing of ad-
jacent DnaA boxes [7].
Rules for binding of DnaAs from other bacteria are
Table 1
E. coli ATP-DnaA boxes
DNA background ATP-DnaA boxes
E. coli dnaAp [13] AGAACT
AGATCT
AGTTTA
AGATTT
E. coli oriC [14] AGATCT
AGATCT
AGATCA
AGGATC
V [101] AGAACT
AGATCC
AGTCAT
AGTATT
V. harveyi dnaAp [100] TGATCG
AGATCG
AGACTG
AGATCT
AGATCA
Consensus AGatct
FEMSRE 756 24-10-02
361
similar but not identical, and therefore show the variabil-
ity of the binding reaction, see also Section 12. Bacillus
subtilis DnaA apparently follows similar rules, since a hy-
brid origin with the main part of oriC from E. coli and the
AT-rich region from B. subtilis can execute most initiation
reactions, primarily it can be unwound [103]. DnaA from
S. lividans has the same two interaction domains with
similar kinetic consequences [47,61,104]. However, oligo-
merization via domain 3 does not require binding to a
second DnaA box as in the case of E. coli, since mixed
oligomers made out of full-length and truncated DnaA
proteins, respectively, can bind to a single DnaA box
[47]. Streptomyces species have a high G/C content in their
DNA, and consequently the consensus sequence of the
Streptomyces DnaA box has a G or C at position 3
[105,106]. Although this is a strong box, binding a⁄nity
to the E. coli consensus sequence is still better, and
although binding to a single box of this kind is possible,
binding to two boxes is preferred [48]. It is not known
whether the ATP-complexed form of Streptomyces DnaA
recognizes a 6-mer box.
Thermus thermophilus is also a G/C-rich organism.
However, contrary to Streptomyces, the DnaA boxes in
the Thermus origin have the stringent E. coli consensus
sequence. However, single DnaA boxes do not bind
T. thermophilus DnaA (approx. KD s 1500 nM), and
only cooperativity allows reasonable binding with an ap-
parent KD of 30^60 nM. This cooperativity requires ATP-
DnaA, the a⁄nity of T. thermophilus oriC to ADP-DnaA
is about 10 times lower [107].
10. DnaA-mediated origin unwinding
The unwinding of the AT-rich region is the crucial step
in the initiation of most origins, prokaryotic and eukary-
otic ones. In the case of E. coli, the AT-rich region con-
tains six ATP-DnaA boxes adjacent to the 9-mer DnaA
box R1 (Fig. 4). The sequential binding of ATP-DnaA to
these sites has been determined using DNase I footprinting
and surface plasmon resonance [14]. Initial binding is to
DnaA box R1. This then serves as an anchor for the
cooperative binding of ATP-DnaA to the double-stranded
ATP-DnaA boxes. As outlined above, these boxes pro-
mote low-a⁄nity binding. Therefore cooperative interac-
tion is required. DnaA bends the DNA by 40‡ upon bind-
ing [11]. The topological stress in this complex unwinds
the DNA in the AT-rich region. ATP-DnaA then binds to
the single-stranded ATP-DnaA boxes, providing an initial
stabilization of the single-stranded state. This is again a
high-a⁄nity interaction. It seems to be a general feature in
the assembly of protein complexes at speci¢c sites and at
de¢ned times that a sequence of high-low-high a⁄nity is
followed. Cooperative binding to the double-stranded AT-
rich region is therefore presumably the limiting step in the
initiation reaction, followed by unwinding [14].
Cyaan Magenta Geel Zwart
362 W. Messer / FEMS Microbiology Reviews 26 (2002) 355^374
Initially 28 bp are unwound by DnaA protein alone,
and 44^52 when SSB is also present. This is true both
for E. coli and for B. subtilis oriC [17]. A hybrid origin
with the main part from oriC of E. coli and the AT-rich
region from B. subtilis could be unwound by E. coli DnaA,
and helicase was loaded into the bubble [103]. Therefore
we assume that the initiation process and the action of
DnaA are similar for E. coli and for B. subtilis. We also
conclude that the species speci¢city of the DnaA^oriC
interaction lies in the main part of oriC with its ¢ve
9-mer DnaA boxes.
11. DnaA-mediated helicase loading: the DnaA primosome
Two double hexamers of the replicative helicase DnaB
and the helicase loader protein DnaC, one for each repli-
cation direction, must be loaded into the unwound region.
Since single-stranded DNA covered with SSB is not a
substrate for DnaB, helicase must be positioned by
DnaA. As outlined in Section 5, DnaA and DnaB interact
physically. The DnaA N-terminus (amino acids 24^86)
interacts with the DnaB LQ fragment (amino acids 154^
210), and the DnaB N-terminus (K fragment) with domain
3 of DnaA (amino acids 130^148) [49,50,57,58]. In an
elegant study, O’Donnell and co-workers could show
that two DnaB6 ^DnaC6 complexes are actually introduced
into the bubble in a reaction that does not require helicase
activity [29]. This stoichiometry was subsequently con-
¢rmed [28]. The DnaB hexamer forms a ring around sin-
gle-stranded DNA [108,109]. The channel through the
helicase ring corresponds to about 20 nucleotides. The
bubble can therefore accommodate just two helicase rings.
A binding activity to single-stranded DNA induced in
DnaC upon interaction with DnaB has been described.
This seems to facilitate the transfer of DnaB to the tem-
plate [110]. DnaC dissociates from the complexes with
concomitant ATP hydrolysis resulting in two head-to-
head helicase hexamers. This activates DnaB as a helicase
[111]. The two helicase complexes move past each other in
5P-3P direction, and only after they moved s 65 nucleoti-
des can they interact with primase [29].
The strand in oriC to which the ¢rst helicase is posi-
tioned by DnaA was determined using an arti¢cial sub-
strate that consisted of a double-stranded DnaA box R1
of oriC and single-stranded DNA from the AT-rich region
in one strand and random single-stranded DNA in the
other strand. At limiting DnaA concentrations DnaB
was loaded to the lower strand [102,234]. This suggests
that the ¢rst helicase hexamer is positioned onto the lower
strand, as illustrated in Fig. 2. This is compatible with
results obtained by footprinting [29].
There is a very striking similarity between the region of
E. coli oriC where DnaB is loaded into the AT-rich bubble
and structures found in other systems where DnaB heli-
case is loaded with the help of DnaA. Even the distance
FEMSRE 756 24-10-02
Fig. 4. Mechanism of unwinding of the AT-rich region. ATP-DnaA is
shown in red, ATP-DnaA boxes are shown in blue, the high-a⁄nity box
R1 in magenta.
between the 3P-end of the DnaA box and the potentially
single-stranded DNA is similar in most cases. Such a
structure is found in several bacterial origins, e.g. B. sub-
tilis [17,112], Pseudomonas putida and aeruginosa [113,114],
Coxiella burnetti [115], T. thermophilus [107], and others.
This also holds for some plasmid origins like P1 [116] or
mini-F [117]. In an arti¢cial system, the ABC primosome,
a DnaA box is positioned in a hairpin structure within
single-stranded phage DNA [118]. In all cases, a single
double-stranded DnaA box is followed by a single-
stranded region to its 3P-side.
12. Variations of the theme : DnaA and oriC in other
bacteria
Di¡erent bacteria have replication origins with widely
di¡ering sizes, but all (except Synechocystis, see below)
contain several DnaA boxes and an AT-rich region (Fig.
5). For example, E. coli oriC consists of 260 bp and ¢ve
DnaA boxes, B. subtilis oriC contains three clusters of
DnaA boxes which are separated by the dnaA gene
[119]. T. thermophilus has symmetrically arranged 12 or
13 boxes (number 12 and 13 overlap) [107], and Strepto-
myces has 19 DnaA boxes in 600 bp [106,120].
12.1. B. subtilis
All three clusters of DnaA boxes are required for oriC
function. The intervening dnaA gene can be largely deleted
if DnaA is provided in trans. There are three AT-rich 16-
mers at the rpmH-proximal border of oriC, and an AT-
rich cluster of 27 bp at the dnaN side [121]. The unwinding
reaction occurs at this 27-bp AT cluster and surrounding
sequences [122]. DnaA proteins bound to the DnaA box
clusters interact and form loops that can be visualized in
the electron microscope. This is also seen with E. coli oriC,
but the physiological relevance is not clear [122]. The
Cyaan Magenta Geel Zwart
 W. Messer / FEMS Microbiology Reviews 26 (2002) 355^374
DnaA protein is biochemically very similar to that of
E. coli [123]. As described above, the structure of the re-
gion between the right-most DnaA box in Fig. 5 and the
AT cluster is very similar to the corresponding region in
oriC of E. coli, including the positions of ATP-DnaA
boxes, and the extent and positions of unwound nucleo-
tides are virtually identical between E. coli and B. subtilis
[17].
The location of oriC with DnaA box clusters upstream
of dnaA or between dnaA and dnaN is conserved in many
bacteria [124], e.g. Micrococcus luteus [36], Mycoplasma
capricolum [125], Spiroplasma citri [126], Mycobacterium
[127], Helicobacter pylori [128], and Streptomyces [106].
Likewise, many bacteria carry a ‘dnaA’ operon down-
stream, containing besides dnaA and dnaN, recF and
gyrB. It has therefore been suggested that this genomic
arrangement represents a primordial structure [104,113,
121,124]. However, there are also many exceptions with
dnaA and oriC located in di¡erent environments on the
genome [129,130].
A minichromosome replicating from oriC of B. subtilis
has been isolated [131]. However, distinct from minichro-
mosomes of E. coli, such plasmids have a strong incom-
patibility to the chromosomal origin, and consequently
their copy number is low. Incompatibility is also observed
when isolated DnaA box clusters from oriC are cloned in
high-copy-number vectors [112]. Presumably, the relaxed
copy number control of E. coli minichromosomes is due to
the Dam-SeqA system (see Section 16) that allows the cell
to discriminate between replicated and not yet replicated
origins. An organism like B. subtilis without such a meth-
ylation must control its replication initiation much more
stringently. An in vitro replication system for oriC plas-
mids has also been established for B. subtilis [132]. Expres-
Fig. 5. Replication origins of di¡erent bacteria. DnaA boxes are shown as red arrows, AT-rich regions are shown in yellow.
FEMSRE 756 24-10-02
363
sion of the dnaA gene is autoregulated, as in the case of
E. coli (see Section 13) [133].
One of the additional controls found in B. subtilis but
not in E. coli are two replication checkpoints 100^200 kb
downstream of oriC on either side. They are dependent on
the stringent response and on a RTP protein-binding site
of the type normally used for replication arrest at the
chromosomal terminus [134].
12.2. Streptomyces
Streptomyces species have large linear chromosomes
with a centrally located oriC. This oriC can be cloned as
a circular, autonomously replicating minichromosome,
which also has a low copy number [120]. Each oriC con-
tains 19 DnaA boxes whose location and orientation are
conserved [106]. Streptomyces oriC contains no extended
AT-rich region, but ¢ve short AT-rich stretches. Conse-
quently, unwound bases are interspersed with paired ones
(J. Majka and J. Zakrzewska-Czerwinska, personal com-
munication). Therefore it is unknown where and how the
replicative helicase is loaded.
Streptomyces DnaA protein is similar to that from
E. coli in many respects [7]. It can oligomerize via two
regions, one in domain 1 and one in domain 3 [47,48,
61,104]. The consensus sequence of Streptomyces DnaA
boxes shows a G or C at position 3 [105,106], similar to
M. luteus [36]. Both are organisms with a high G+C con-
tent. Such a strong Streptomyces DnaA box (5P-
TTGTCCACA) can bind a DnaA monomer, and di¡erent
from the situation in E. coli, also a dimer [47,61]. How-
ever, binding to a DnaA box with the E. coli consensus
sequence is still better (KD around 10 nM for TTGTCCA-
CA and around 3 nM for TTATCCACA) [48]. Weak
Cyaan Magenta Geel Zwart
364 W. Messer / FEMS Microbiology Reviews 26 (2002) 355^374
boxes with mismatches require cooperation between DnaA
proteins, and therefore require two boxes [48,135]. In an
experiment where optimal binding sites were selected by
evolution from random sequences, the optimal spacing
was 3 bp and the boxes were facing each other [48]. At
higher concentrations long-range interactions between
DnaA protein bound to distant DnaA boxes in oriC pro-
mote loop formation [47,104]. The promoter region of
dnaA contains two DnaA boxes with the same spacing
and arrangement as found in the selection from random
sequence [136]. Cooperative binding of DnaA to these
boxes results in repression, and thus in autoregulation of
the dnaA gene [135].
12.3. T. thermophilus
The T. thermophilus replication origin is also located
between the dnaA and dnaN genes. It has a quite unusual,
nearly symmetrical structure [107]. Six DnaA boxes in the
¢rst half of oriC point toward dnaA, six boxes in the
second half point toward the dnaN gene. Box number 13
overlaps with box number 12. Like Streptomyces, T. ther-
mophilus is an organism with a high G+C content in its
chromosome. Nevertheless, all DnaA boxes, except num-
bers 6 and 13, have the E. coli consensus sequence, 5P-
TTATCCACA. At the dnaN side is a 40-bp-long AT-
rich stretch.
The Thermus DnaA protein has the same overall struc-
ture as the other DnaA proteins. The binding properties,
however, are di¡erent. A single DnaA box is nearly inert
to DnaA. DnaA protein binding to oriC requires cooper-
ativity between monomers, and this cooperativity is en-
hanced when DnaA is in the ATP-complexed form (Table
2) [107]. There are no DnaA boxes in the promoter region
of the dnaA gene, and consequently, T. thermophilus dnaA
is not autoregulated [137].
A comparison of the two G+C-rich organisms, Strepto-
myces and T. thermophilus, shows that they follow di¡er-
ent strategies to adjust the properties of their DnaA pro-
teins to their origins which are both rich in DnaA boxes.
Streptomyces with a high-a⁄nity DnaA protein reduces
the quality of its DnaA boxes. Thermus, on the other
hand, compensates potentially high-a⁄nity DnaA boxes
with a DnaA protein with low binding a⁄nity.
Table 2
Apparent dissociation constants of T. thermophilus DnaA protein with
DnaA boxes from T. thermophilus oriC
DnaA box(es) [DnaA] at 50% of Rmax (nM)
ADP-DnaA ATP-DnaA
4 1300 1500
7 2700 2200
1^3 735 82
1^6 270 31
1^13 420 60
FEMSRE 756 24-10-02
12.4. H. pylori
H. pylori DnaA also has similar structural properties as
the other DnaA proteins [128]. The replication origin of
H. pylori has been de¢ned by its location close to DnaA,
by a strand switch in the G/C composition, and by ¢ve
DnaA boxes in a 180-bp stretch. These DnaA boxes, how-
ever, all have at least one mismatch to the consensus se-
quence. Despite this, the DNA-binding domain 4 of Heli-
cobacter DnaA can bind to each isolated DnaA box as a
monomer [128].
12.5. Synechocystis sp.
Cyanobacteria also have regular DnaA proteins
[129,138]. The Synechocystis DnaA protein binds to
DnaA boxes, as does its C-terminal domain 4 with a⁄n-
ities comparable to E. coli DnaA. The genome of Syn-
echocystis has been completely sequenced, and there is
no stretch with several DnaA boxes that could qualify as
a replication origin. In addition, the dnaA gene could be
deleted without loss of viability [139]. However, DnaA has
been conserved during evolution, and its transcription is
not autoregulated but light-dependent instead, and follows
the circadian rhythm of DNA synthesis [139]. The easiest
way to interpret these seemingly incompatible observa-
tions is to assume that normal replication initiation oper-
ates via DnaA and an origin that is even more di⁄cult to
detect than H. pylori oriC. In addition, Synechocystis must
have a very e⁄cient alternative initiation pathway, possi-
bly involving R-loops or D-loops as in the case of stable
replication in E. coli [140].
12.6. Caulobacter crescentus
This is a bacterium with a developmentally di¡erenti-
ated cell cycle. Asymmetric cell division results in a repli-
cating stalked cell and a non-replicating swarmer cell. Mo-
tile swarmer cells must di¡erentiate into sessile stalked
cells before they can replicate [141]. An essential control
system for the cell cycle in this bacterium is a response
regulator, CtrA (for cell cycle transcription regulator),
which is homologous to E. coli OmpR. The Caulobacter
replication origin, Cori, contains ¢ve DnaA boxes, of
which none has the E. coli consensus sequence, and ¢ve
binding sites for CtrA [142]. CtrA binds to these sites with
a 40-fold enhanced a⁄nity if it is phosphorylated by the
histidine kinase CckA [143]. CtrA is only present in
swarmer cells and represses among many other genes rep-
lication from Cori [144]. During di¡erentiation from
swarmer to stalked cells CtrA is degraded by regulated
proteolysis [145]. Four of the ¢ve binding sites at which
CtrA represses replication could, by AT-richness and by
their location to the left-most DnaA box of Cori, represent
potential sites of unwinding which have to be cleared be-
fore initiation. Whatever the precise mechanism, regula-
Cyaan Magenta Geel Zwart
 W. Messer / FEMS Microbiology Reviews 26 (2002) 355^374
tion of replication by a transcriptional regulator, that is a
response regulator which can be phosphorylated, is a very
unusual way to regulate replication in this unusual organ-
ism. However, initiation of chromosome replication is
strictly dependent on DnaA also in Caulobacter [146].
Like E. coli, Caulobacter has a DNA methyltransferase,
CcrM with the recognition sequence GANTC, that is cell
cycle-regulated. However, it is not clear whether it has a
similar role as Dam methyltransferase in E. coli [147].
13. DnaA as a transcription factor
DnaA boxes are found in the promoter regions of many
genes where they can mediate repression, transcriptional
activation, or transcription termination due to loop for-
mation between two DnaA boxes in a transcription unit
and long-range DnaA^DnaA interaction [99,148]. The
property of DnaA as a transcription factor has been cov-
ered in a review by Messer and Weigel [9]. Here I want to
limit myself to some aspects of the autoregulation of the
dnaA gene. E. coli dnaA is transcribed from two pro-
moters. There is a consensus DnaA box between them.
Inactivation of DnaA in temperature-sensitive dnaA mu-
tants results in derepression of dnaA promoters, whereas
overproduction of dnaA cloned in a plasmid represses
dnaA transcription [149^154]. However, a dnaA gene
with a mutation of the DnaA box (dnaA820) was still
subject to autoregulation [155]. The solution of this para-
dox is the presence of additional binding sites for DnaA
between the promoters. There is a second DnaA box with
a mismatch and four ATP-DnaA boxes. The dnaA pro-
moter region was the model sequence for derivation of the
rules for DnaA binding [13]. ADP-DnaA binds to the two
9-mer DnaA boxes, ATP-DnaA binds to all boxes as
shown by DNase I footprinting and surface plasmon res-
onance. The result is that ATP-DnaA is a much better
repressor for the dnaA promoter than ADP-DnaA [13].
The promoter region with the dnaA820 mutation is bound
about 60% as e⁄ciently by ATP-DnaA as the wild-type
promoter. There are additional controls for dnaA expres-
sion, e.g. methylation of GATC sites present in the region
(see Section 16), or Fis and IciA proteins. These are listed
in [9].
14. Joint action of DnaA and plasmid initiators in plasmid
replication
DnaA is involved in the initiation of replication of many
E. coli plasmids. These plasmids contain in their replica-
tion origins one or several DnaA boxes and an AT-rich
region for unwinding, and, in addition, several iterons,
repeated binding sites for a plasmid-encoded initiation
protein. Whereas bacterial initiation requires the single
initiator DnaA, initiation of plasmids with iterons requires
FEMSRE 756 24-10-02
365
dual initiator proteins ; DnaA works in concert with a
plasmid-encoded initiator protein [156]. DnaA can assist
in the unwinding reaction, in helicase loading, or it can
have a primary role in the structure of the initiation com-
plex, e.g. in pSC101. The dependence of the unwinding
reaction on DnaA protein is di¡erent for di¡erent plas-
mids. It is absolutely required for some, e.g. P1, where
DnaA can unwind the AT-rich region even alone,
although quite ine⁄ciently. DnaA is required, in concert
with the plasmid initiator, for unwinding in mini-F, RK2
and R6K. At the other end of the list is plasmid R1 which
requires DnaA in vitro, but not in vivo. The cooperation
between DnaA and plasmid initiator proteins extends to
the loading of DnaB helicase, at least in the cases that
have been analyzed.
14.1. P1
The origin of plasmid P1 contains two groups of DnaA
boxes at either end. The left box region is followed by an
AT-rich region containing GATC sites. These sites must
be methylated by Dam methyltransferase for e⁄cient ini-
tiation [157]. Adjacent to the right box region are ¢ve 19-
bp repeats, binding sites for the plasmid initiator protein
RepA. Only one of the two DnaA box regions is required.
Even one box is su⁄cient, provided it has the consensus
sequence for a strong box. Optimal replication needs both
DnaA box regions [158,159]. The phasing of DnaA boxes
relative to one another and the rest of the origin is impor-
tant [160]. DnaA is required for the unwinding of the AT-
rich region [159]. DnaA, together with HU protein, is
su⁄cient for unwinding. However, the reaction is much
more e⁄cient in the presence of RepA, as measured by
KMnO4 footprinting [116,160].
14.2. F
Plasmid mini-F possesses all necessary functions for rep-
lication of F plasmids. The replication origin (ori2) has
two DnaA boxes, followed by an AT-rich region and
four 19-bp direct repeats, iterons to which the plasmid-
encoded initiator protein RepE binds as monomer. Both
DnaA and RepE proteins alone, but together with HU
protein, are able to unwind mini-F. For e⁄cient unwind-
ing, as well as sequence speci¢city for the AT-rich region,
however, the concerted action of DnaA, RepE, and HU is
required [117].
14.3. RK2
The replication origin of the broad-host-range plasmid
RK2/RP4 (393 bp) carries four (lousy) DnaA boxes, ¢ve
17-bp iterons for binding of the plasmid initiator TrfA in
its monomeric form, and an AT-rich region consisting of
four 13-mer repeats that are similar to the 13-mers in
E. coli oriC. DnaA binds cooperatively to the four
Cyaan Magenta Geel Zwart
366 W. Messer / FEMS Microbiology Reviews 26 (2002) 355^374
DnaA boxes, in agreement with the rules for DnaA bind-
ing formulated above. DnaA box number 4 seems to serve
as an anchor point [161]. However, DnaA binding by itself
does not result in unwinding of the AT-rich region, di¡er-
ent from the plasmids discussed above [162]. TrfA, bound
to the iterons, unwinds the AT-rich region in the presence
of HU and ATP. This unwinding is enhanced by DnaA
[162]. DnaA is indispensable for the following step, the
delivery of the helicase [163]. However, DnaA cannot by
itself activate the DnaB helicase at the RK2 origin. Both
DnaA and TrfA are required for DnaB-induced template
unwinding [163]. Surprisingly, it has been found that the
DnaA^DnaB^DnaC complex is not assembled at the un-
wound AT-rich region but at the DnaA box region that is
about 200 bp away [164]. We must therefore assume a
physical interaction by DNA looping between the DnaA
box region with the bound complex and the unwound AT-
rich region, presumably with the help of TrfA.
14.4. R6K
The replication of plasmid R6K is initiated at either of
three origins, K, L, and Q. K and L origins require the
presence of ori Q in cis for function. Initiation from ori
K and ori Q requires both the plasmid-encoded initiator Z
and DnaA protein (reviewed in [165]. ori Q contains (in
this order) a DnaA box, an AT-rich region, and seven
iterons for binding of Z. Between the AT-rich region and
the iterons is a binding site for IHF. It has been suggested
that its function is to bend the origin and thereby bring
into contact DnaA and Z [166]. Unwinding of the AT-rich
region in ori Q depends on an interaction between DnaA
and the N-terminal part (amino acids 1^116) of Z [167].
The ATP-complexed form of DnaA is not required, since
a mutation in the ATP-binding site (K178A) of DnaA is
as e¡ective [167].
14.5. R1
DnaA protein is required for the in vitro replication of
plasmid R1, whereas it is dispensable for replication in
vivo [168,169]. Therefore, dnaA(Null) mutants can be in-
tegratively suppressed by R1, but not by F [170]. R1 rep-
lication initiation is di¡erent from the other plasmids dis-
cussed so far. Instead of iterons it contains two partially
palindromic 10-bp sequences at each end of a 100-bp re-
gion. This region is completely protected in DNase I foot-
printing experiments. At one end is a DnaA box, and at
the other end an AT-rich region [171]. A model has been
proposed in which a DNA loop is formed by binding and
interaction of two RepA molecules, probably dimers, to
the binding sites at the ends of ori R, and then the 100-bp
loop is ¢lled with more RepA. The AT-rich region is then
unwound in this higher-order complex [156,171]. Binding
of DnaA to its box is very ine⁄cient and requires DnaA^
RepA contacts [172,173]. Although DnaA is required for
FEMSRE 756 24-10-02
initiation in vitro, the DnaA box can be mutated [172].
Presumably, similar to loading of DnaA to a DnaA box
with mismatches, such a loading is also possible due to
interaction with RepA. The role of DnaA in R1 replica-
tion is, however, not clear.
14.6. pSC101
The origin of pSC101 contains a consensus DnaA box
followed by an AT-rich region, a binding site for IHF, and
¢ve iterons for RepA binding. Within the iterons is a weak
DnaA box with two mismatches to the consensus sequence
[174]. DnaA is required for unwinding of the AT-rich re-
gion and for helicase loading, in concert with RepA and
IHF [174,175]. However, since domains 2 and 3 of DnaA
are dispensable for pSC101 replication [49,176], the in-
volvement of DnaA must be indirect. It has been sug-
gested that DnaA ful¢lls a structural role in pSC101 ini-
tiation, together with IHF, by stabilizing a loop between
the DnaA box at the left end of the origin and the iteron
region [174]. Mutant DnaA protein consisting of domains
1 and 4 only might dimerize with domain 1, bind with
domain 4 to the left DnaA box and with domain 4 of
the second monomer to the iteron region. This might hap-
pen by binding to the weak DnaA box or by interaction
with RepA. In fact, interaction of DnaA protein domain 1
and of domain 4 with RepA has been demonstrated [176].
An alternative anchor at the iteron side of pSC101 ori
could be a second region of DnaA that can support
pSC101 replication [39]. DnaA domains 1 and 4 do not
need to be covalently linked. A leucine zipper added to
each domain allows a non-covalent association that is suf-
¢cient to provide a bridge between the left DnaA box and
the iteron region in the origin [176].
15. The special case of DnaA and V plasmids
Replication of bacteriophage V, and of plasmids derived
from phage V, depends on the phage encoded proteins O
and P (for review see [177]). O is a functional analogue of
DnaA, and P replaces DnaC during loading of DnaB heli-
case. The replication origin (ori V) is located within the O
gene. Transcription from the rightward V promoter pR is
required for expression of O and P, but also for transcrip-
tional activation of the origin (for review see [177]). Tran-
scriptional activation is required for early replication of
phage V, and for replication of V plasmids. Replication
of V plasmids depends on DnaA protein [178]. This was
surprising, since there is no DnaA box in the V origin
region. However, binding of DnaA to several 9-bp sequen-
ces that have similarity to DnaA boxes, as well as to 6-mer
ATP-DnaA boxes, has been demonstrated [101]. Since all
these boxes represent low-a⁄nity binding sites, coopera-
tive binding of DnaA is required. DnaA activates tran-
scription, i.e. transcriptional activation, from pR in vivo
Cyaan Magenta Geel Zwart
 W. Messer / FEMS Microbiology Reviews 26 (2002) 355^374
and in vitro [179,180]. A DnaA box located several base
pairs downstream of pR is particularly important for this
activation, since mutation of this box abolishes transcrip-
tion stimulation by DnaA [180]. It has been shown that
early V replication via the bidirectional a mode depends on
activation of pR by DnaA. It is predominantly unidirec-
tional in dnaA mutants [181]. Unidirectional a type of
replication precedes the late c type of replication. We
may therefore assume that in dnaAþ hosts a depletion of
the DnaA pool due to binding to DnaA boxes is the rea-
son for the switch from a to c replication [181]. The acti-
vation is delicately balanced, e.g. certain dnaA(Ts) mutants
are unable to activate pR even at permissive temperature
[182]. All these results show that the role of DnaA in
V plasmid replication is a role as transcription factor,
and not as a replisome organizer.
16. How to limit initiation to once per generation
All organisms have developed mechanisms that ensure
chromosomal replication occurs once and only once per
generation [183]. In an E. coli cell there are at least three
systems that prevent reinitiation of origins that have al-
ready initiated: (i) sequestration of oriC and of the dnaA
promoter region after replication and hence blocking of
their activities for a given time window; (ii) binding of
DnaA protein to a region close to oriC that provides a
sink for DnaA ; (iii) regulatory inactivation of ATP-DnaA
at the end of the initiation cycle. This has been discussed
in Sections 3 and 6.
16.1. Sequestration
The replication origin of E. coli contains an unusually
high number (11) of GATC sequences, recognition sites
for the Dam methyltransferase. These sites must be meth-
ylated for e⁄cient initiation [184,185]. Especially hemi-
methylated GATC sites, as they are present shortly after
replication, are detrimental to initiation [186]. Hemimethy-
lated GATC sites in oriC, and also in the promoter region
of dnaA, remain hemimethylated for about one-third of a
generation, whereas elsewhere on the chromosome they
become remethylated within about 1 min [187,188]. The
length of this eclipse period depends on the supply of Dam
methyltransferase [184,189,190]. Since in vitro initiation
can occur on unmethylated and hemimethylated origins
[184,191,192], it has been concluded that within cells these
regions are sequestered by binding to cellular factors. The
most prominent of these factors is SeqA [193,194]. Anoth-
er member of the membrane-bound sequestration complex
is SeqB [195].
SeqA has a high a⁄nity for hemimethylated oriC, some-
what less for fully methylated oriC, and does not bind
speci¢cally to unmethylated DNA [196,197]. It binds spe-
ci¢cally to two sites in oriC, one on each side of DnaA box
FEMSRE 756 24-10-02
367
R1. From there it seems to spread by cooperative binding
to adjacent regions, competing with DnaA for binding
[198^200]. This is likely to be the main reason for the
negative e¡ect of SeqA on initiation of replication. Also
sequestration of the dnaA promoter region a¡ects initia-
tion negatively [193], since methylation of GATC sites in
the promoter region is required for maximum promoter
activity [151,201]. Hence, shortly after initiation at the
time when the dnaA promoter is sequestered, there is a
transient suppression of dnaA expression [202]. Other in-
hibitory e¡ects could come from changes in oriC topology
made by SeqA protein [203,204].
Sequestration by Dam methylation is restricted to En-
terobacteriaceae. It gives E. coli and its relatives the op-
portunity to regulate initiation in a very relaxed way.
Since the cells are able to tell a new origin from an old
one they do not need to keep track of individual copies
and need not control the copy number of oriC tightly.
Consequently, the copy numbers of E. coli minichromo-
somes are around 10 per cell, and they show a very large
variation, and therefore a high loss rate [205,206]. Quite
contrary, the copy number of minichromosomes from
B. subtilis [131] or Streptomyces [120] is around 1 per
cell, and there is strong incompatibility between minichro-
mosomes and the chromosomal origin. As expected, ini-
tiation of origins that are not subject to methylation is
more tightly controlled.
16.2. Titration of DnaA to the datA locus
About 300 DnaA boxes, located around the E. coli
chromosome, are able to bind DnaA protein. Five chro-
mosomal regions show an especially high a⁄nity [207].
The locus with the highest DnaA-binding capacity is
datA, located close to oriC [208]. It can bind about eight
times more DnaA than the combined oriC-mioC region
with a similar number of DnaA boxes. This high capacity,
compared to oriC, might result from a missing competi-
tion with SeqA since there are only few GATC sites in
datA. So far, it has not been analyzed whether there are
ATP-DnaA boxes in datA or close to the other high-a⁄n-
ity sites that might be responsible for the high binding
capacity.
There is evidence for a direct involvement of datA in the
regulation of initiation. Deletion of datA results in over-
initiation [209]. Additional copies of datA on a plasmid
limit initiation or block it completely [209,210]. We can
therefore safely assume that datA is a sink for DnaA pro-
tein which regulates the availability of DnaA and hence
a¡ects initiation.
17. DnaA and oriC in the cell cycle, control of initiation
When E. coli grows in a rich medium it contains several
chromosomes. These initiate synchronously within a very
Cyaan Magenta Geel Zwart
368 W. Messer / FEMS Microbiology Reviews 26 (2002) 355^374
short time interval [183,211,212]. oriC present on mini-
chromosomes initiate in the same interval [213]. A number
of conditions upset this synchrony. These could either be
suboptimal conditions of initiation, e.g. in some dnaA(Ts)
mutants even at permissive temperature [214], or interfer-
ence with the sequestration apparatus, e.g. in dam and
seqA mutants [189,212].
Initiation occurs at a de¢ned time in the cell cycle. Don-
achie discovered that at the time of initiation cells have a
more or less constant volume per replication origin, called
the initiation volume [215]. Later it was discovered that
there is some variation in this initiation volume [216].
After initiation it takes a constant time to replicate the
chromosome, independent of growth rate as long as it is
6 60 min at 37‡C, and a constant time between termina-
tion of replication and cell division [217,218]. The fre-
quency of initiation thus determines the rate of DNA syn-
thesis, and indirectly the rate of cell division.
The DNA-bending proteins Fis and IHF have speci¢c
binding sites in oriC (Fig. 1) [219^224]. The occupation of
binding sites by these architectural proteins and by DnaA
has been determined in synchronized cells by in vivo foot-
printing. The result was that throughout most of the cell
cycle, DnaA was bound to DnaA boxes R1, R2, and R4,
and Fis was bound to its site, presumably inhibiting ini-
tiation. At the time of initiation Fis was lost from oriC,
DnaA bound to DnaA box R3, and IHF bound to its site
[225,226]. This results in bending of oriC and in a redis-
tribution of DnaA [227]. So far a correlation of these in
vivo results with the results obtained in vitro has not been
attempted.
The molecular basis of the initiation volume discussed
above apparently is the concentration of DnaA protein. It
has been shown that the initiation volume can be changed
by modulation of the DnaA level [228]. A hypothesis has
been formulated [229,230] that postulates that initially all
newly synthesized DnaA protein is bound to DnaA boxes
around the chromosome. These boxes increase in number
due to replication, and DnaA continues to be synthesized
until the number of DnaA molecules exceeds the number
of DnaA boxes. This is the moment of initiation. A neces-
sary condition for this hypothesis is that the binding event
that triggers initiation is of lower a⁄nity than binding to
the other boxes [229]. The switch to cooperative binding at
ATP-DnaA boxes (see Section 10) is ideally suited to ful¢ll
this condition.
In addition, it has been suggested that DnaA protein
molecules that become free due to initiation and replica-
tion are immediately used for initiation at adjacent origins
in the form of an initiation cascade or an avalanche, there-
by ensuring the highly synchronous initiation of all origins
in a cell [231].
Both these hypotheses have been formulated before the
discovery of inactivation of ATP-DnaA due to ATP hy-
drolysis [33], and before the drastic changes in the propor-
tion of ATP-DnaA before and after initiation had been
FEMSRE 756 24-10-02
measured [74]. This does not invalidate these hypotheses,
but they have to be re¢ned in view of the newer results.
18. Perspectives
DnaA and oriC and the replicative helicase are the ma-
jor elements in bacterial replication initiation. They have
been the topic for extensive research for close to 40 years,
but many questions are still unanswered. So far most of
the research has centered on E. coli, but this one-sided
view is slowly broadening. Following good tradition in
molecular biology, the initiation cycle in bacteria, in
E. coli, can serve as a paradigm for other more compli-
cated systems. Eukaryotic viruses like SV40, polyoma or
papilloma have origins and initiators that are basically
comparable to E. coli [232]. Yeast origins have an AT-
rich region and a binding site for the six-protein origin
recognition complex (ORC). ORC binds like DnaA to
double-stranded and to single-stranded DNA, and the
change between the binding speci¢cities is associated
with an ATP/ADP switch [75,233]. ORCs and correspond-
ing origins have been found in all metazoa that have been
looked at. The unwinding of double-stranded origin DNA
and the loading of helicase into this region is one of the
basic reactions in the cell cycle. The E. coli way of life
gives us a good example how to think about this process
and design experiments.
19. Note added in proof
A crystal structure of domains 3 and 4 of one DnaA
protein has recently been published [Erzberger, J.P., Pir-
ruccello, M.M. and Berger, J.M. (2002) The structure of
bacterial DnaA: implications for general mechanisms
underlying DNA replication initiation. EMBO J. 21,
4763^4773]. It extends and corroborates many of the pre-
dictions given in Section 4.
Acknowledgements
I thank Harald Seitz, Kirsten Skarstad, Grzegorz Wegr-
zyn, Christoph Weigel and Jolanta Zakrzewska-Czerwin-
ska for critically reading the manuscript. I gratefully ac-
knowledge support from the Fonds der Chemischen
Industrie.
References
[1] Kohiyama, M., Cousin, D., Ryter, A. and Jacob, F. (1966) Mutants
thermosensibles d’E. coli K-12. I. Isolement et characterisation ra-
pide. Ann. Inst. Pasteur 110, 465^486.
[2] Kohiyama, M. (1968) DNA synthesis in temperature sensitive mu-
tants in E. coli. Cold Spring Harbor Symp. Quant. Biol. 43, 317^324.
Cyaan Magenta Geel Zwart
 W. Messer / FEMS Microbiology Reviews 26 (2002) 355^374
[3] Jacob, F., Brenner, S. and Cuzin, F. (1963) On the regulation of
DNA replication in bacteria. Cold Spring Harbor Symp. Quant.
Biol. 28, 329^348.
[4] Skarstad, K. and Boye, E. (1994) The initiator protein DnaA: Evo-
lution, properties and function. Biochim. Biophys. Acta 1217, 111^
130.
[5] Messer, W. and Weigel, C. (1996) Initation of chromosome replica-
tion. In: Escherichia coli and Salmonella, Cellular and Molecular
Biology (Neidhardt, F.C., Curtiss III, R., Ingraham, J., Lin,
E.C.C., Low, K.B., Magasanik, B., Rezniko¡, W.S., Riley, M.,
Schaechter, M. and Umbarger, H.E., Eds.), pp. 1579^1601. ASM,
Washington, DC.
[6] Kaguni, J.M. (1997) Escherichia coli DnaA protein: the replication
initiator. Mol. Cells 7, 145^157.
[7] Messer, W., Blaesing, F., Jakimowicz, D., Majka, J., Nardmann, J.,
Schaper, S., Seitz, H., Speck, C., Weigel, C., Wegrzyn, G., Welzeck,
M. and Zakrzewska-Czerwinska, J. (2001) Bacterial replication ini-
tiator DnaA. Rules for DnaA binding and roles of DnaA in origin
unwinding and helicase loading. Biochimie 83, 1^9.
[8] Echols, H. (1990) Nucleoprotein structures initiating DNA replica-
tion, transcription, and site-speci¢c recombination. J. Biol. Chem.
265, 14697^14700.
[9] Messer, W. and Weigel, C. (1997) DnaA initiator ^ also a transcrip-
tion factor. Mol. Microbiol. 24, 1^6.
[10] Kornberg, A. and Baker, T.A. (1992) DNA Replication. W.H. Free-
man and Company, New York.
[11] Schaper, S. and Messer, W. (1995) Interaction of the initiator protein
DnaA of Escherichia coli with its DNA target. J. Biol. Chem. 270,
17622^17626.
[12] Sekimizu, K., Bramhill, D. and Kornberg, A. (1987) ATP activates
dnaA protein in initiating replication of plasmids bearing the origin
of the E. coli chromosome. Cell 50, 259^265.
[13] Speck, C., Weigel, C. and Messer, W. (1999) ATP- and ADP-DnaA
protein, a molecular switch in gene regulation. EMBO J. 18, 6169^
6176.
[14] Speck, C. and Messer, W. (2001) Mechanism of origin unwinding:
sequential binding of DnaA initiator protein to double-stranded and
single-stranded DNA in the AT-rich region of the replication origin.
EMBO J. 20, 1469^1476.
[15] Bramhill, D. and Kornberg, A. (1988) Duplex opening by dnaA
protein at novel sequences in initiation of replication at the origin
of the E. coli chromosome. Cell 52, 743^755.
[16] Gille, H. and Messer, W. (1991) Localized unwinding and structural
perturbations in the origin of replication, oriC, of Escherichia coli in
vitro and in vivo. EMBO J. 10, 1579^1584.
[17] Krause, M. and Messer, W. (1999) DnaA proteins of Escherichia coli
and Bacillus subtilis: coordinate actions with single-stranded DNA-
binding protein and interspecies inhibition during open complex for-
mation at the replication origins. Gene 228, 123^132.
[18] Hiasa, H. and Marians, K.J. (1994) Fis cannot support oriC DNA
replication in vitro. J. Biol. Chem. 269, 24999^25003.
[19] Wold, S., Crooke, E. and Skarstad, K. (1996) The Escherichia coli Fis
protein prevents initiation of DNA replication from oriC in vitro.
Nucleic Acids Res. 24, 3527^3532.
[20] Dixon, N.E. and Kornberg, A. (1984) Protein HU in the enzymatic
replication of the chromosomal origin of Escherichia coli. Proc. Natl.
Acad. Sci. USA 81, 424^428.
[21] Skarstad, K., Baker, T.A. and Kornberg, A. (1990) Strand separation
required for initiation of replication at the chromosomal origin of
E. coli is facilitated by a RNA-DNA hybrid. EMBO J. 9, 2341^2348.
[22] Sekimizu, K., Bramhill, D. and Kornberg, A. (1988) Sequential early
stages in the in vitro initiation of replication at the origin of the
Escherichia coli chromosome. J. Biol. Chem. 263, 7124^7130.
[23] Messer, W. (1972) Initiation of DNA replication in E. coli B/r. Chro-
nology of events and transcriptional control of initiation. J. Bacteriol.
112, 7^12.
FEMSRE 756 24-10-02 Cyaan Magenta Geel Zwart
369
[24] Lark, K.G. (1972) Evidence for direct involvement of RNA in the
initiation of DNA replication in E. coli 15T. J. Mol. Biol. 64, 47^60.
[25] Baker, T.A. and Kornberg, A. (1988) Transcriptional activation of
initiation of replication from the E. coli chromosomal origin: An
RNA-DNA hybrid near oriC. Cell 55, 113^123.
[26] Funnell, B.E., Baker, T.A. and Kornberg, A. (1987) In vitro assembly
of a prepriming complex at the origin of the Escherichia coli chro-
mosome. J. Biol. Chem. 262, 10327^10334.
[27] Crooke, E., Thresher, R., Hwang, D.S., Gri⁄th, J. and Kornberg, A.
(1993) Replicatively active complexes of DnaA protein and the Es-
cherichia coli chromosomal origin observed in the electron micro-
scope. J. Mol. Biol. 233, 16^24.
[28] Carr, K.M. and Kaguni, J.M. (2001) Stoichiometry of DnaA and
DnaB protein in initiation at the Escherichia coli chromosomal ori-
gin. J. Biol. Chem. 276, 44919^44925.
[29] Fang, L.H., Davey, M.J. and O’Donnell, M. (1999) Replisome as-
sembly at oriC, the replication origin of E. coli, reveals an explana-
tion for initiation sites outside an origin. Mol. Cell 4, 541^553.
[30] Wahle, E., Lasken, R.S. and Kornberg, A. (1989) The dnaB-dnaC
replication protein complex of Escherichia coli. I. Formation and
properties. J. Biol. Chem. 264, 2463^2468.
[31] Messer, W., Seufert, W., Schaefer, C., Gielow, A., Hartmann, H. and
Wende, M. (1988) Functions of the DnaA protein of Escherichia coli
in replication and transcription. Biochim. Biophys. Acta 951, 351^
358.
[32] Kelman, Z. and O’Donnell, M. (1995) DNA polymerase III holoen-
zyme: structure and function of a chromosomal replicating machine.
Annu. Rev. Biochem. 64, 171^200.
[33] Katayama, T., Kubota, T., Kurokawa, K., Crooke, E. and Sekimizu,
K. (1998) The initiator function of DnaA protein is negatively regu-
lated by the sliding clamp of the E. coli chromosomal replicase. Cell
94, 61^71.
[34] Katayama, T. and Sekimizu, K. (1999) Inactivation of Escherichia
coli DnaA protein by DNA polymerase III and negative regulations
for initiation of chromosomal replication. Biochimie 81, 835^840.
[35] Kato, J. and Katayama, T. (2001) Hda, a novel DnaA-related pro-
tein, regulates the replication cycle in Escherichia coli. EMBO J. 20,
4253^4262.
[36] Fujita, M.Q., Yoshikawa, H. and Ogasawara, N. (1990) Structure of
the dnaA region of Micrococcus luteus: Conservation and variations
among eubacteria. Gene 93, 73^78.
[37] Skovgaard, O. and Hansen, F.G. (1987) Comparison of dnaA nucle-
otide sequences of Escherichia coli, Salmonella typhimurium, and Ser-
ratia marcescens. J. Bacteriol. 169, 3976^3981.
[38] Messer, W., Blaesing, F., Majka, J., Nardmann, J., Schaper, S.,
Schmidt, A., Seitz, H., Speck, C., Tu«ngler, D., Wegrzyn, G., Weigel,
C., Welzeck, M. and Zakrzewska-Czerwinska, J. (1999) Functional
domains of DnaA proteins. Biochimie 81, 819^825.
[39] Sutton, M.D. and Kaguni, J.M. (1997) The Escherichia coli dnaA
gene: Four functional domains. J. Mol. Biol. 274, 546^561.
[40] Schaper, S. and Messer, W. (1997) Prediction of the structure of the
replication initiator protein DnaA. Proteins Struct. Funct. Genet. 28,
1^9.
[41] Neuwald, A.F., Aravind, L., Spouge, J.L. and Koonin, E.V. (1999)
AAA(+): A class of chaperone-like ATPases associated with the as-
sembly, operation, and disassembly of protein complexes. Genome
Res. 9, 27^43.
[42] Rost, B. and Sander, C. (1994) Combining evolutionary information
and neural networks to predict protein secondary structure. Proteins
19, 55^72.
[43] Rossmann, M.G., Moras, D. and Olsen, K.W. (1974) Chemical and
biological evolution of nucleotide-binding protein. Nature 250, 194^
199.
[44] Guenther, B., Onrust, R., Sali, A., O’Donnell, M. and Kuriyan, J.
(1997) Crystal structure of the deltaP subunit of the clamp-loader
complex of E. coli DNA polymerase III. Cell 91, 335^345.
370 W. Messer / FEMS Microbiology Reviews 26 (2002) 355^374
[45] Roth, A. and Messer, W. (1995) The DNA binding domain of the
initiator protein DnaA. EMBO J. 14, 2106^2111.
[46] Weigel, C., Schmidt, A., Seitz, H., Tuengler, D., Welzeck, M. and
Messer, W. (1999) The N-terminus promotes oligomerisation of the
Escherichia coli initiator protein DnaA. Mol. Microbiol. 34, 53^66.
[47] Jakimowicz, D., Majka, J., Konopa, G., Wegrzyn, G., Messer, W.,
Schrempf, H. and Zakrzewska-Czerwinska, J. (2000) Architecture of
the Streptomyces lividans DnaA protein-replication origin complexes.
J. Mol. Biol. 298, 351^364.
[48] Majka, J., Zakrzewska-Czerwinska, J. and Messer, W. (2001) Se-
quence recognition, cooperative interaction and dimerization of the
initiator protein DnaA of Streptomyces. J. Biol. Chem. 276, 6243^
6252.
[49] Seitz, H., Weigel, C. and Messer, W. (2000) The interaction domains
of the DnaA and DnaB replication proteins of Escherichia coli. Mol.
Microbiol. 37, 1270^1279.
[50] Sutton, M.D., Carr, K.M., Vicente, M. and Kaguni, J.M. (1998)
Escherichia coli DnaA protein. The N-terminal domain and loading
of DnaB helicase at the E. coli chromosomal origin. J. Biol. Chem.
273, 34255^34262.
[51] Mima, S., Yamagachi, Y., Kondo, T., Tsuchiya, T. and Mizushima,
T. (1999) Role of the amino-terminal region of the DnaA protein in
opening of the duplex DNA at the oriC region. FEMS Microbiol.
Lett. 176, 163^167.
[52] Walker, J.E., Saraste, M., Runswick, M.J. and Gay, N.J. (1982)
Distantly related sequences in the alpha and beta subunits of ATP
synthase, myosin, kinases and other ATP requiring enzymes and a
commom binding fold. EMBO J. 1, 945^951.
[53] Saraste, M., Sibbald, P.R. and Wittinghofer, A. (1990) The P-loop ^
a common motif in ATP- and GTP-binding proteins. Trends Bio-
chem. Sci. 15, 430^434.
[54] Koonin, E.V. (1993) A common set of conserved motifs in a vast
variety of putative nucleic acid-dependent ATPases including MCM
proteins involved in the initiation of eukaryotic DNA replication.
Nucleic Acids Res. 21, 2541^2547.
[55] Mizushima, T., Nishida, S., Kurokawa, K., Katayama, T., Miki, T.
and Sekimizu, K. (1997) Negative control of DNA replication by
hydrolysis of ATP bound to DnaA protein, the initiator of chromo-
somal DNA replication in Escherichia coli. EMBO J. 16, 3724^3730.
[56] Su’etsugu, M., Kawakami, H., Kurokawa, K., Kubota, T., Takata,
M. and Katayama, T. (2001) DNA replication-coupled inactivation
of DnaA protein in vitro: a role for DnaA arginine-334 of the
AAA(+) Box VIII motif in ATP hydrolysis. Mol. Microbiol. 40,
376^386.
[57] Marszalek, J. and Kaguni, J.M. (1994) DnaA protein directs the
binding of DnaB protein in initiation of DNA replication in Esche-
richia coli. J. Biol. Chem. 269, 4883^4890.
[58] Marszalek, J., Zhang, W.G., Hupp, T.R., Margulies, C., Carr, K.M.,
Cherry, S. and Kaguni, J.M. (1996) Domains of DnaA protein in-
volved in interaction with DnaB protein, and in unwinding the Es-
cherichia coli chromosomal origin. J. Biol. Chem. 271, 18535^18542.
[59] Biswas, E.E. and Biswas, S.B. (1999) Mechanism of DnaB helicase of
Escherichia coli : Structural domains involved in ATP hydrolysis,
DNA binding, and oligomerization. Biochemistry 38, 10919^10928.
[60] Blaesing, F., Weigel, C. and Messer, W. (2000) Analysis of the DNA
binding domain of Escherichia coli DnaA protein. Mol. Microbiol.
36, 557^569.
[61] Majka, J., Jakimowicz, D., Messer, W., Schrempf, H., Lisowski, M.
and Zakrzewska-Czerwinska, J. (1999) Interactions of the Streptomy-
ces lividans initiator protein DnaA with its target. Eur. J. Biochem.
260, 325^335.
[62] Sutton, M.D. and Kaguni, J.M. (1997) Threonine 435 of Escherichia
coli DnaA protein confers sequence-speci¢c DNA binding activity.
J. Biol. Chem. 272, 23017^23024.
[63] Blinkova, A., Gines-Candelaria, E., Ross, J.D. and Walker, J.R.
(2000) Suppression of a DnaX temperature-sensitive polymerization
FEMSRE 756 24-10-02 Cyaan Magenta Geel Zwart
defect by mutation in the initiation gene, dnaA, requires functional
oriC. Mol. Microbiol. 36, 913^925.
[64] Kubota, T., Ito, Y., Sekimizu, K., Tagaya, M. and Katayama, T.
(2001) DnaA protein Lys-415 is close to the ATP-binding site: ATP-
pyridoxal a⁄nity labeling. Biochem. Biophys. Res. Commun. 288,
1141^1148.
[65] Hansen, F.G., Koefoed, S. and Atlung, T. (1992) Cloning and nucle-
otide sequence determination of twelve mutant dnaA genes of Esche-
richia coli. Mol. Gen. Genet. 234, 14^21.
[66] Hwang, D.S. and Kaguni, J.M. (1988) Interaction of dnaA46 protein
with a stimulatory protein in replication from the Escherichia coli
chromosomal origin. J. Biol. Chem. 263, 10633^10640.
[67] Katayama, T. (1994) The mutant DnaAcos protein which overiniti-
ates replication of the Escherichia coli chromosome is inert to neg-
ative regulation for initiation. J. Biol. Chem. 269, 22075^22079.
[68] Carr, K.M. and Kaguni, J.M. (1996) The A184V missense mutation
of the dnaA5 and dnaA46 alleles confers a defect in ATP binding and
thermolability in initiation of Escherichia coli DNA replication. Mol.
Microbiol. 20, 1307^1318.
[69] Braun, R.E., O’Day, K. and Wright, A. (1987) Cloning and charac-
terization of dnaA(Cs), a mutation which leads to overinitiation of
DNA replication in Escherichia coli K-12. J. Bacteriol. 169, 3898^
3903.
[70] Nyborg, M., Atlung, T., Skovgaard, O. and Hansen, F.G. (2000)
Two types of cold sensitivity associated with the A184V change in
the DnaA protein. Mol. Microbiol. 35, 1202^1210.
[71] Kellenberger-Gujer, G., Podhajska, A.J. and Caro, L. (1978) A cold
sensitive dnaA mutant of E. coli which overinitiates chromosome
replication at low temperature. Mol. Gen. Genet. 162, 9^16.
[72] Katayama, T. and Kornberg, A. (1994) Hyperactive initiation of
chromosomal replication in vivo and in vitro by a mutant initiator
protein, DnaAcos, of Escherichia coli. J. Biol. Chem. 269, 12698^
12703.
[73] Katayama, T. and Crooke, E. (1995) DnaA protein is sensitive to a
soluble factor and is speci¢cally inactivated for initiation of in vitro
replication of the Escherichia coli minichromosome. J. Biol. Chem.
270, 9265^9271.
[74] Kurokawa, K., Nishida, S., Emoto, A., Sekimizu, K. and Katayama,
T. (1999) Replication cycle-coordinated change of the adenine nucle-
otide-bound forms of DnaA protein in Escherichia coli. EMBO J. 18,
6642^6652.
[75] Lee, D.G. and Bell, S.P. (2000) ATPase switches controlling DNA
replication initiation. Curr. Opin. Cell Biol. 12, 280^285.
[76] Hwang, D.S. and Kornberg, A. (1992) Opening of the replication
origin of Escherichia coli by DnaA protein with protein HU or
IHF. J. Biol. Chem. 267, 23083^23086.
[77] Makise, M., Tsuchiya, T. and Mizushima, T. (2002) Sequence-inde-
pendent DNA binding activity of DnaA protein, the initiator of chro-
mosomal DNA replication in Escherichia coli. J. Biochem. 131, 419^
425.
[78] Sekimizu, K., Yung, B.Y. and Kornberg, A. (1988) The dnaA protein
of Escherichia coli. Abundance, improved puri¢cation, and mem-
brane binding. J. Biol. Chem. 263, 7136^7140.
[79] Newman, G. and Crooke, E. (2000) DnaA, the initiator of Escheri-
chia coli chromosomal replication, is located at the cell membrane.
J. Bacteriol. 182, 2604^2610.
[80] Xia, W. and Dowhan, W. (1995) In vivo evidence for the involvement
of anionic phospholipids in initiation of DNA replication in Esche-
richia coli. Proc. Natl. Acad. Sci. USA 92, 783^787.
[81] Zheng, W.D., Li, Z.Y., Skarstad, K. and Crooke, E. (2001) Muta-
tions in DnaA protein suppress the growth arrest of acidic phospho-
lipid-de¢cient Escherichia coli cells. EMBO J. 20, 1164^1172.
[82] Wegrzyn, A., Wrobel, B. and Wegrzyn, G. (1999) Altered biological
properties of cell membranes in Escherichia coli dnaA and seqA mu-
tants. Mol. Gen. Genet. 261, 762^769.
[83] Yung, B.Y.M. and Kornberg, A. (1988) Membrane attachment acti-
 W. Messer / FEMS Microbiology Reviews 26 (2002) 355^374
vates dnaA protein, the initiation protein of chromosome replication
in Escherichia coli. Proc. Natl. Acad. Sci. USA 85, 7202^7205.
[84] Sekimizu, K. and Kornberg, A. (1988) Cardiolipin activation of
DnaA protein, the initiation protein in E. coli. J. Biol. Chem. 263,
7131^7135.
[85] Castuma, C.E., Crooke, E. and Kornberg, A. (1993) Fluid mem-
branes with acidic domains activate DnaA, the initiator protein of
replication in Escherichia coli. J. Biol. Chem. 268, 24665^24668.
[86] Mizushima, T., Ishikawa, Y., Obana, E., Hase, M., Kubota, T.,
Katayama, T., Kunitake, T., Watanabe, E. and Sekimizu, K.
(1996) In£uence of cluster formation of acidic phospholipids on
decrease in the a⁄nity for ATP of DnaA protein. J. Biol. Chem.
271, 3633^3638.
[87] Crooke, E., Castuma, C.E. and Kornberg, A. (1992) The chromo-
some origin of Escherichia coli stabilizes DnaA protein during reju-
venation by phospholipids. J. Biol. Chem. 267, 16779^16782.
[88] Hase, M., Yoshimi, T., Ishikawa, Y., Ohba, A., Guo, L., Mima, S.,
Makise, M., Yamaguchi, K., Tsuchiya, T. and Mizushima, T. (1998)
Site-directed mutational analysis for the membrane binding of
DnaA protein. Identi¢cation of amino acids involved in the func-
tionsl interaction between DnaA protein and acidic phospholipids.
J. Biol. Chem. 273, 28651^28656.
[89] Yamaguchi, Y., Hase, M., Makise, M., Mima, S., Yoshimi, T.,
Ishikawa, Y., Tsuchiya, T. and Mizushima, T. (1999) Involvement
of Arg-328, Arg-334 and Arg-342 of DnaA protein in the functional
interaction with acidic phospholipids. Biochem. J. 340, 433^438.
[90] Makise, M., Mima, S., Tsuchiya, T. and Mizushima, T. (2000)
Identi¢cation of amino acids involved in the functional interaction
between DnaA protein and acidic phospholipids. J. Biol. Chem. 275,
4513^4518.
[91] Makise, M., Mima, S., Tsuchiya, T. and Mizushima, T. (2001) Mo-
lecular mechanism for functional interaction between DnaA protein
and acidic phospholipids ^ Identi¢cation of important amino acids.
J. Biol. Chem. 276, 7450^7456.
[92] Garner, J. and Crooke, E. (1996) Membrane regulation of the chro-
mosomal replication activity of E. coli DnaA requires a discrete site
on the protein. EMBO J. 15, 3477^3485.
[93] Speck, C., Weigel, C. and Messer, W. (1997) From footprint to
toeprint : a close-up of the DnaA box, the binding site for the ini-
tiator protein DnaA of Escherichia coli. Nucleic Acids Res. 25,
3242^3247.
[94] Fuller, R.S. and Kornberg, A. (1983) Puri¢ed dnaA protein in ini-
tiation of replication at the Escherichia coli chromosomal origin of
replication. Proc. Natl. Acad. Sci. USA 80, 5817^5821.
[95] Fuller, R.S., Funnell, B.E. and Kornberg, A. (1984) The dnaA pro-
tein complex with the E. coli chromosomal origin (oriC) and other
sites. Cell 38, 889^900.
[96] Matsui, M., Oka, A., Takanami, M., Yasuda, S. and Hirota, Y.
(1985) Sites of dnaA protein-binding in the replication origin of
the E. coli K-12 chromosome. J. Mol. Biol. 184, 529^533.
[97] Holz, A., Schaefer, C., Gille, H., Jueterbock, W.-R. and Messer, W.
(1992) Mutations in the DnaA binding sites of the replication origin
of Escherichia coli. Mol. Gen. Genet. 233, 81^88.
[98] Schaefer, C. and Messer, W. (1991) DnaA protein/DNA interaction.
Modulation of the recognition sequence. Mol. Gen. Genet. 226, 34^
40.
[99] Konopa, G., Szalewska-Palasz, A., Schmidt, A., Srutkowska, S.,
Messer, W. and Wegrzyn, G. (1999) The presence of two DnaA-
binding sequences is required for e⁄cient interaction of the Esche-
richia coli DnaA protein with each particular weak DnaA box re-
gion. FEMS Microbiol. Lett. 174, 25^31.
[100] Berenstein, D., Olesen, K., Speck, C. and Skovgaard, O. (2002)
Genetic organization of the Vibrio harveyi dnaA gene region and
analysis of the function of the V. harveyi DnaA protein in Escheri-
chia coli. J. Bacteriol. 184, 2533^2538.
[101] Szalewska-Palasz, A., Weigel, C., Speck, C., Srutkowska, S., Kono-
FEMSRE 756 24-10-02 Cyaan Magenta Geel Zwart
371
pa, G., Lurz, R., Marszalek, J., Taylor, K., Messer, W. and Wegr-
zyn, G. (1998) Interaction of the Escherichia coli DnaA protein with
lambda phage DNA. Mol. Gen. Genet. 259, 679^688.
[102] Seitz, H. (2000) Interaktionen des E. coli Initiatorproteins DnaA mit
der replikativen Helikase DnaB. PhD Thesis, FU Berlin. http://
www.diss.fu-berlin.de/diss/index.html.
[103] Seitz, H., Welzeck, M. and Messer, W. (2001) A hybrid bacterial
replication origin. EMBO Rep. 2, 1003^1006.
[104] Zakrzewska-Czerwinska, J., Jakimowicz, D., Majka, J., Messer, W.
and Schrempf, H. (2000) Initiation of the Streptomyces chromosome
replication. Antonie van Leeuwenhoek 78, 211^221.
[105] Zakrzewska-Czerwinska, J., Majka, J. and Schrempf, H. (1995)
Minimal requirements of the Streptomyces lividans 66 oriC region
and its transcriptional and translational activities. J. Bacteriol. 177,
4765^4771.
[106] Jakimowicz, D., Majka, J., Messer, W., Speck, C., Fernandez, M.,
Martin, M.C., Sanchez, J., Schauwecker, F., Keller, U., Schrempf,
H. and Zakrzewska-Czerwinska, J. (1998) Structural elements of the
Streptomyces oriC region and their interactions with the DnaA pro-
tein. Microbiology 144, 1281^1290.
[107] Schaper, S., Nardmann, J., Lu«der, G., Lurz, R., Speck, C. and
Messer, W. (2000) Identi¢cation of the chromosomal replication
origin from Thermus thermophilus and its interaction with the rep-
lication initiator DnaA. J. Mol. Biol. 299, 655^665.
[108] San Martin, M.C., Stamford, N.P.J., Dammerova, N., Dixon, N.E.
and Carazo, J.M. (1995) A structural model for the Escherichia coli
DnaB helicase based on electron microscopy data. J. Struct. Biol.
114, 167^176.
[109] Egelman, H.H., Yu, X., Wild, R., Hingorani, M.M. and Patel, S.S.
(1995) Bacteriophage T7 helicase/primase proteins form rings
around single-stranded DNA that suggest a general structure for
hexameric helicases. Proc. Natl. Acad. Sci. USA 92, 3869^3873.
[110] Learn, B.A., Um, S.-J., Huang, L. and McMacken, R. (1997) Cryp-
tic single-stranded-DNA binding activities of the phage V P and
Escherichia coli DnaC replication initiation proteins facilitate the
transfer of E. coli DnaB helicase onto DNA. Proc. Natl. Acad.
Sci. USA 94, 1154^1159.
[111] Wahle, E., Lasken, R.S. and Kornberg, A. (1989) The dnaB-dnaC
replication protein complex of Escherichia coli. II. role of the com-
plex in mobilizing dnaB functions. J. Biol. Chem. 264, 2469^2475.
[112] Moriya, S., Fukuoka, T., Ogasawara, N. and Yoshikawa, H. (1988)
Regulation of initiation of the chromosomal replication by DnaA-
boxes in the origin of the Bacillus subtilis chromosome. EMBO J. 7,
2911^2917.
[113] Fujita, M.Q., Yoshikawa, H. and Ogasawara, N. (1989) Structure of
the dnaA region of Pseudomonas putida : Conservation among three
bacteria, Bacillus subtilis, Escherichia coli and P. putida. Mol. Gen.
Genet. 215, 381^387.
[114] Smith, D.W., Yee, T.W., Baird, C. and Krishnapillai, V. (1991)
Pseudomonad replication origins: A paradigm for bacterial origins.
Mol. Microbiol. 5, 2581^2587.
[115] Suhan, M., Chen, S.-Y., Thompson, H.A., Hoover, T.A., Hill, A.
and Williams, J.C. (1994) Cloning and characterization of an auton-
omous replication sequence from Coxiella burnetii. J. Bacteriol. 176,
5233^5243.
[116] Park, K., Mukhopadhyay, S. and Chattoraj, D. (1998) Require-
ments for and regulation of origin opening of plasmid P1. J. Biol.
Chem. 273, 24906^24911.
[117] Kawasaki, Y., Matsunaga, F., Kano, Y., Yura, T. and Wada, C.
(1996) The localized melting of mini-F by the combined action of
the mini-F initiator protein (RepE) and HU and DnaA of Escheri-
chia coli. Mol. Gen. Genet. 253, 42^49.
[118] Masai, H., Nomura, N. and Arai, K. (1990) The ABC-primosome.
A novel priming system employing dnaA, dnaB, dnaC, and primase
on a hairpin containing a dnaA box sequence. J. Biol. Chem. 265,
15134^15144.
372 W. Messer / FEMS Microbiology Reviews 26 (2002) 355^374
[119] Ogasawara, N., Moriya, S., von Meyenburg, K., Hansen, F.G. and
Yoshikawa, H. (1985) Conservation of genes and their organization
in the chromosomal replication origin region of Bacillus subtilis and
Escherichia coli. EMBO J. 4, 3345^3350.
[120] Zakrzewska-Czerwinska, J. and Schrempf, H. (1992) Characteriza-
tion of an autonomously replicating region from the Streptomyces
lividans chromosome. J. Bacteriol. 174, 2688^2693.
[121] Yoshikawa, H. and Wake, R.G. (1993) Initiation and termination of
chromosome replication. In: Bacillus subtilis and Other Gram-Pos-
itive Bacteria : Biochemistry, Physiology, and Molecular Genetics
(Sonenshein, A.L., Hoch, J.A. and Losick, R., Eds.), pp. 507^528.
ASM, Washington, DC.
[122] Krause, M., Lurz, R., Ru«ckert, B. and Messer, W. (1997) Com-
plexes at the replication origin of Bacillus subtilis with homologous
and heterologous DnaA protein. J. Mol. Biol. 274, 365^380.
[123] Fukuoka, T., Moriya, S., Yoshikawa, H. and Ogasawara, N. (1990)
Puri¢cation and characterization of an initiation protein for chro-
mosomal replication, DnaA, in Bacillus subtilis. J. Biochem. (Tokyo)
107, 732^739.
[124] Yoshikawa, H. and Ogasawara, N. (1991) Structure and function of
DnaA and the DnaA-box in eubacteria: Evolutionary relationships
of bacterial replication origins. Mol. Microbiol. 5, 2589^2597.
[125] Fujita, M.Q., Yoshikawa, H. and Ogasawara, N. (1992) Structure of
the dnaA and DnaA-box region in the Mycoplasma capricolum chro-
mosome: Conservation and variations in the course of evolution.
Gene 110, 17^23.
[126] Ye, F., Renaudin, J., Bove¤, J.-M. and Laigret, F. (1994) Cloning
and sequencing of the replication origin (oriC) of the Spiroplasma
citri chromosome and construction of autonomously replicating ar-
ti¢cial plasmids. Curr. Microbiol. 29, 23^29.
[127] Salazar, L., Fsihi, H., de Rossi, E., Riccardi, G., Rios, C., Cole, S.T.
and Taki¡, H.E. (1996) Organization of the origins of replication of
the chromosomes of Mycobacterium smegmatis, Mycobacterium le-
prae and Mycobacterium tuberculosis and isolation of a functional
origin from M. smegmatis. Mol. Microbiol. 20, 283^293.
[128] Zawilak, A., Cebrat, S., Mackiewicz, P., Krol-Hulewicz, A., Jaki-
mowicz, D., Messer, W., Gosciniak, G. and Zakrzewska-Czerwin-
ska, J. (2001) Identi¢cation of a putative chromosomal replication
origin from Helicobacter pylori and its interaction with the initiator
protein DnaA. Nucleic Acids Res. 29, 2251^2259.
[129] Richter, S., Hess, W.R., Krause, M. and Messer, W. (1998) Unique
organization of the dnaA region from Prochlorococcus marinus
CCMP1375, a marine cyanobacterium. Mol. Gen. Genet. 257,
534^541.
[130] Mushegian, A.R. and Koonin, E.V. (1996) Gene order is not con-
served in bacterial evolution. Trends Genet. Sci. 12, 289^290.
[131] Moriya, S., Atlung, T., Hansen, F.G., Yoshikawa, H. and Ogasa-
wara, N. (1992) Cloning of an autonomously replicating sequence
(ars) from the Bacillus subtilis chromosome. Mol. Microbiol. 6, 309^
315.
[132] Moriya, S., Firshein, W., Yoshikawa, H. and Ogasawara, N. (1994)
Replication of a Bacillus subtilis oriC plasmid in vitro. Mol. Micro-
biol. 12, 469^478.
[133] Ogura, Y., Imai, Y., Ogasawara, N. and Moriya, S. (2001) Auto-
regulation of the dnaA-dnaN operon and e¡ects of DnaA protein
levels on replication initiation in Bacillus subtilis. J. Bacteriol. 183,
3833^3841.
[134] Autret, S., Levine, A., Vannier, F., Fujita, Y. and Seror, S.J. (1999)
The replication checkpoint control in Bacillus subtilis : identi¢cation
of a novel RTP-binding sequence essential for the replication fork
arrest after induction of the stringent response. Mol. Microbiol. 31,
1665^1679.
[135] Jakimowicz, D., Majka, J., Lis, B., Konopa, G., Wegrzyn, G.,
Messer, W., Schrempf, H. and Zakrzewska-Czerwinska, J. (2000)
Structure and regulation of the dnaA promoter region in three
Streptomyces species. Mol. Gen. Genet. 262, 1093^1102.
[136] Zakrzewska-Czerwinska, J., Nardmann, J. and Schrempf, H. (1994)
FEMSRE 756 24-10-02 Cyaan Magenta Geel Zwart
Inducible transcription of the dnaA gene from Streptomyces lividans
66. Mol. Gen. Genet. 242, 440^447.
[137] Nardmann, J. and Messer, W. (2001) Identi¢cation and character-
ization of the dnaA upstream region of Thermus thermophilus. Gene
261, 299^303.
[138] Richter, S. and Messer, W. (1995) Genetic structure of the dnaA
region of the cyanobacterium Synechocystis sp. PCC6803. J. Bacte-
riol. 177, 4245^4251.
[139] Richter, S., Hagemann, M. and Messer, W. (1998) Transcriptional
analysis and mutation of a dnaA-like gene in Synechocystis sp.
PCC6803. J. Bacteriol. 180, 4946^4949.
[140] Kogoma, T. (1997) Stable DNA replication : interplay between
DNA replication, homologous recombination, and transcription.
Microbiol. Mol. Biol. Rev. 61, 212^238.
[141] Marczynski, G.T., Dingwall, A. and Shapiro, L. (1990) Plasmid and
chromosomal DNA replication and partitioning during the Caulo-
bacter crescentus cell cycle. J. Mol. Biol. 212, 709^722.
[142] Marczynski, G.T. and Shapiro, L. (1992) Cell-cycle control of a
cloned chromosomal origin of replication from Caulobacter cres-
centus. J. Mol. Biol. 226, 959^977.
[143] Siam, R. and Marczynski, G.T. (2000) Cell cycle regulator phos-
phorylation stimulates two distinct modes of binding at a chromo-
some rplication origin. EMBO J. 19, 1138^1147.
[144] Quon, K.C., Yang, B., Domian, I.J., Shapiro, L. and Marczynski,
G.T. (1998) Negative control of bacterial DNA replication by a cell
cycle regulatory protein that binds at the chromosome origin. Proc.
Natl. Acad. Sci. USA 95, 120^125.
[145] Domain, I.J., Quon, K.C. and Shapiro, L. (1997) Cell-type speci¢c
phosphorylation and proteolysis of a transcriptional regulator con-
trols the G1 -to-S transition in a bacterial cell cycle. Cell 90, 415^424.
[146] Gorbatyuk, B. and Marczynski, G.T. (2001) Physiological conse-
quences of blocked Caulobacter crescentus dnaA expression, an es-
sential DNA replication gene. Mol. Microbiol. 40, 485^497.
[147] Marczynski, G.T. (1999) Chromosome methylation and measure-
ment of faithful, once and only once per cell cycle chromosome
replication in Caulobacter crescentus. J. Bacteriol. 181, 1984^1993.
[148] Schaefer, C. and Messer, W. (1988) Termination of the E. coli asnC
transcript. The DnaA protein/dnaA box complex blocks transcribing
RNA polymerase. Gene 73, 347^354.
[149] Braun, R.E., O’Day, K. and Wright, A. (1985) Autoregulation of
the DNA replication gene dnaA in E. coli. Cell 40, 159^169.
[150] Atlung, T., Clausen, E. and Hansen, F.G. (1985) Autoregulation of
the dnaA gene of Escherichia coli. Mol. Gen. Genet. 200, 442^450.
[151] Ku«cherer, C., Lother, H., Ko«lling, R., Schauzu, M.A. and Messer,
W. (1986) Regulation of transcription of the chromosomal dnaA
gene of Escherichia coli. Mol. Gen. Genet. 205, 115^121.
[152] Wang, Q. and Kaguni, J.M. (1987) Transcriptional repression of the
dnaA gene of Escherichia coli by dnaA protein. Mol. Gen. Genet.
209, 518^525.
[153] Polaczek, P. and Wright, A. (1990) Regulation of expression of the
dnaA gene in Escherichia coli: Role of the two promoters and the
DnaA box. New Biol. 2, 574^582.
[154] Lee, Y.S. and Hwang, D.S. (1997) Occlusion of RNA polymerase by
oligomerization of DnaA protein over the dnaA promoter of Esche-
richia coli. J. Biol. Chem. 272, 83^88.
[155] Smith, R.W.P., McAteer, S. and Masters, M. (1997) Autoregulation
of the Escherichia coli replication initiator protein, DnaA, is indi-
rect. Mol. Microbiol. 23, 1303^1315.
[156] Del Solar, G.H., Giraldo, R., Ruiz-Echevarr|¤a, M.J., Espinosa, M.
and D|¤az-Orejas, R. (1998) Replication and control of circular bac-
terial plasmids. Microbiol. Mol. Biol. Rev. 62, 434^464.
[157] Abeles, A.L. and Austin, S.J. (1988) P1 plasmid replication requires
Escherichia coli Dam-methylated DNA. Gene 74, 185^186.
[158] Abeles, A.L., Reaves, L.D. and Austin, S.J. (1990) A single dnaA
box is su⁄cient for initiation from the plasmid P1 origin. J. Bacte-
riol. 172, 4386^4391.
[159] Mukhopadhyay, G., Carr, K.M., Kaguni, J.M. and Chattoraj, D.K.
 W. Messer / FEMS Microbiology Reviews 26 (2002) 355^374
(1993) Open-complex formation by the host initiator, DnaA, at the
origin of P1 plasmid replication. EMBO J. 12, 4547^4554.
[160] Park, K. and Chattoraj, D.K. (2001) DnaA boxes in the P1 plasmid
origin : The e¡ect of their position on the directionality of replica-
tion and plasmid copy number. J. Mol. Biol. 310, 69^81.
[161] Doran, K.S., Helinski, D.B. and Konieczny, I. (1999) A critical
DnaA box directs the cooperative binding of the Escherichia coli
DnaA protein to the plasmid RK2 replication origin. J. Biol.
Chem. 274, 17918^17923.
[162] Konieczny, I., Doran, K.S., Helinski, D.R. and Blasina, A. (1997)
Role of TrfA and DnaA proteins in origin opening during initiation
of DNA replication of the broad host range plasmid RK2. J. Biol.
Chem. 272, 20173^20178.
[163] Konieczny, I. and Helinski, D.R. (1997) Helicase delivery and acti-
vation by DnaA and TrfA proteins during the initiation of replica-
tion of the broad host range plasmid RK2. J. Biol. Chem. 272,
33312^33318.
[164] Pacek, M., Konopa, G. and Konieczny, I. (2001) DnaA box sequen-
ces as the site for helicase delivery during plasmid RK2 replication
initiation in Escherichia coli. J. Biol. Chem. 276, 23639^23644.
[165] Filutowicz, M., Dellis, S., Levchenko, I., Urh, M., Wu, F. and
York, D. (1994) Regulation of replication in an iteron-containing
DNA molecule. Progr. Nucleic Acids Res. Mol. Biol. 68, 239^273.
[166] Miron, A., Mukherjee, S. and Bastia, D. (1992) Activation of dis-
tant replication origins in vivo by DNA looping as revealed by a
novel mutant form of an initiator protein defective in cooperativity
at a distance. EMBO J. 11, 1205^1216.
[167] Lu, Y.B., Datta, H.J. and Bastia, D. (1998) Mechanistic studies of
initiator-initiator interaction and replication initiation. EMBO J. 17,
5192^5200.
[168] Bernander, R., Dasgupta, S. and Nordstro«m, K. (1991) The E. coli
cell cycle and the plasmid R1 replication cycle in the absence of the
DnaA protein. Cell 64, 1145^1153.
[169] Tang, X.B., Womble, D.D. and Rownd, R.H. (1989) DnaA protein
is not essential for replication of IncFII plasmid NR1. J. Bacteriol.
171, 5290^5295.
[170] Hansen, E.B. and Yarmolinsky, M.B. (1986) Host participation in
plasmid maintenance: Dependence upon dnaA of replicons derived
from P1 and F. Proc. Natl. Acad. Sci. USA 83, 4423^4427.
[171] Giraldo, R. and Diaz, R. (1992) Di¡erential binding of wild-type
and a mutant RepA protein to oriR sequence suggests a model for
the initiation of plasmid R1 replication. J. Mol. Biol. 228, 787^802.
[172] Ortega-Jime¤nez, S., Giraldo-Sua¤rez, R., Ferna¤ndez-Tresguerres,
M.E., Berzal-Herranz, A. and D|¤az-Orejas, R. (1992) DnaA depen-
dent replication of plasmid R1 occurs in the presence of point mu-
tations that disrupt the dnaA box of oriR. Nucleic Acids Res. 20,
2547^2551.
[173] Masai, H. and Arai, K.-I. (1987) RepA and DnaA proteins are
required for initiation of R1 plasmid replication in vitro and interact
with the oriR sequence. Proc. Natl. Acad. Sci. USA 84, 4781^4785.
[174] Stenzel, T.T., MacAllister, T. and Bastia, D. (1991) Cooperativity at
a distance promoted by the combined action of two replication
initiator proteins and a DNA bending protein at the replication
origin of pSC101. Genes Dev. 5, 1453^1463.
[175] Datta, H.J., Khatri, G.S. and Bastia, D. (1999) Mechanism of re-
cruitment of DnaB helicase to the replication origin of the plasmid
pSC101. Proc. Natl. Acad. Sci. USA 96, 73^78.
[176] Sharma, R., Kachroo, A. and Bastia, D. (2001) Mechanistic aspects
of DnaA-RepA interaction as revealed by yeast forward and reverse
two-hybrid analysis. EMBO J. 20, 4577^4587.
[177] Taylor, K. and Wegrzyn, G. (1995) Replication of coliphage lambda
DNA. FEMS Microbiol. Rev. 17, 109^119.
[178] Kur, J., Gorska, I. and Taylor, K. (1987) Escherichia coli dnaA
initiation function is required for replication of plasmids derived
from coliphage lambda. Mol. Gen. Genet. 198, 203^210.
[179] Wegrzyn, G., Szalewska-Palasz, A., Wegrzyn, A., Obuchowski, M.
and Taylor, K. (1995) Transcriptional activation of the origin of
FEMSRE 756 24-10-02 Cyaan Magenta Geel Zwart
373
coliphage lambda DNA replication is regulated by the host DnaA
initiator function. Gene 154, 47^50.
[180] Szalewska-Palasz, A., Wegrzyn, A., Blaszczak, A., Taylor, K. and
Wegrzyn, G. (1998) DnaA-stimulated transcriptional activation of
orilambda: Escherichia coli RNA polymerase L subunit as a tran-
scriptional activator contact site. Proc. Natl. Acad. Sci. USA 95,
4241^4246.
[181] Baranska, S., Gabig, M., Wegrzyn, A., Konopa, G., Herman-Anto-
siewicz, A., Hernandez, P., Schvartzman, J.B., Helinski, D.R. and
Wegrzyn, G. (2001) Regulation of the switch from early to late
bacteriophage lambda DNA replication. Microbiology 147, 535^
547.
[182] Glinkowska, M., Konopa, G., Wegrzyn, A., Herman-Antosiewicz,
A., Weigel, C., Seitz, H., Messer, W. and Wegrzyn, G. (2001) The
double mechanism of incompatibility between V plasmids and Es-
cherichia coli dnaA(ts) host cells. Microbiology 147, 1923^1928.
[183] Boye, E., Lobner-Olesen, A. and Skarstad, K. (2000) Limiting DNA
replication to once and only once. EMBO Rep. 1, 479^483.
[184] Messer, W., Bellekes, U. and Lother, H. (1985) E¡ect of dam-meth-
ylation on the activity of the replication origin, oriC. EMBO J. 4,
1327^1332.
[185] Smith, D.W., Garland, A.M., Herman, G., Enns, R.E., Baker, T.A.
and Zyskind, J.W. (1985) Importance of state of methylation of oriC
GATC sites in initiation of DNA replication in Escherichia coli.
EMBO J. 4, 1319^1326.
[186] Russel, D.W. and Zinder, N.D. (1987) Hemimethylation prevents
DNA replication in E. coli. Cell 50, 1071^1079.
[187] Ogden, G.B., Pratt, M.J. and Schaechter, M. (1988) The replicative
origin of the E. coli chromosome binds to cell membranes only when
hemimethylated. Cell 54, 127^135.
[188] Campbell, J.L. and Kleckner, N. (1990) E. coli oriC and the dnaA
gene promoter are sequestered from dam methyltransferase follow-
ing the passage of the chromosomal replication fork. Cell 62, 967^
979.
[189] Boye, E. and Lobner-Olesen, A. (1990) The role of dam methyl-
transferase in the control of DNA replication in E. coli. Cell 62,
981^989.
[190] von Freiesleben, U., Krekling, M.A., Hansen, F.G. and Lobner-
Olesen, A. (2000) The eclipse period of Escherichia coli. EMBO J.
19, 6240^6248.
[191] Landoulsi, A., Hughes, P., Kern, R. and Kohiyama, M. (1989) dam
methylation and the initiation of DNA replication on oriC plasmids.
Mol. Gen. Genet. 216, 217^223.
[192] Boye, E. (1991) The hemimethylated replication origin of Escheri-
chia coli can be initiated in vitro. J. Bacteriol. 173, 4537^4539.
[193] Lu, M., Campbell, J.L., Boye, E. and Kleckner, N. (1994) SeqA : a
negative modulator of replication initiation in E. coli. Cell 77, 413^
426.
[194] von Freiesleben, U., Rasmussen, K.V. and Schaechter, M. (1994)
SeqA limits DnaA activity in replication from oriC in Escherichia
coli. Mol. Microbiol. 14, 763^772.
[195] Shakibai, N., Ishidate, K., Reshetnyak, E., Gunji, S., Kohiyama, M.
and Roth¢eld, L. (1998) High-a⁄nity binding of hemimethylated
oriC by Escherichia coli membranes is mediated by a multiprotein
system that includes SeqA and a newly identi¢ed factor, SeqB. Proc.
Natl. Acad. Sci. USA 95, 11117^11121.
[196] Slater, S., Wold, S., Lu, M., Boye, E., Skarstad, K. and Kleckner,
N. (1995) E. coli SeqA protein binds oriC in two di¡erent methyl-
modulated reactions appropriate to its roles in DNA replication
initiation and origin sequestration. Cell 82, 927^936.
[197] Brendler, T. and Austin, S. (1999) Binding of SeqA protein to DNA
requires interaction between two or more complexes bound to sep-
arate hemimethylated GATC sequences. EMBO J. 18, 2304^2310.
[198] Skarstad, K., Lueder, G., Lurz, R., Speck, C. and Messer, W. (2000)
The Escherichia coli SeqA protein binds speci¢cally and coopera-
tively to two sites in hemimethylated and fully methylated oriC.
Mol. Microbiol. 36, 1319^1326.
374 W. Messer / FEMS Microbiology Reviews 26 (2002) 355^374
[199] Skarstad, K., Torheim, N., Wold, S., Lurz, R., Messer, W., Fossum,
S. and Bach, T. (2001) The Escherichia coli SeqA protein binds
speci¢cally to two sites in fully and hemimethylated oriC and has
the capacity to inhibit DNA replication and a¡ect chromosome to-
pology. Biochimie 83, 49^51.
[200] Taghbalout, A., Landoulsi, A., Kern, R., Yamazoe, M., Hiraga, S.,
Holland, B., Kohiyama, M. and Malki, A. (2000) Competition be-
tween the replication initiator DnaA the sequestration factor SeqA
for binding to the hemimethylated chromosomal origin of E. coli in
vitro. Genes Cells 5, 873^884.
[201] Braun, R.E. and Wright, A. (1986) DNA methylation di¡erentially
enhances the expression of one of the two E. coli dnaA promoters in
vivo and in vitro. Mol. Gen. Genet. 202, 246^250.
[202] Theisen, P., Grimwade, J.E., Leonard, A.C., Bogan, J.A. and Helm-
stetter, C.E. (1993) Correlation of gene transcription with the time
of initiation of chromosome replication in Escherichia coli. Mol.
Microbiol. 10, 575^584.
[203] Torheim, N.K. and Skarstad, K. (1999) Escherichia coli SeqA pro-
tein a¡ects DNA topology and inhibits open complex formation at
oriC. EMBO J. 18, 4882^4888.
[204] Weitao, T., Nordstro«m, K. and Dasgupta, S. (2000) Escherichia coli
cell cycle control genes a¡ect chromosome superhelicity. EMBO
Rep. 1, 494^499.
[205] Lobner-Olesen, A. (1999) Distribution of minichromosomes in indi-
vidual Escherichia coli cells: implications for replication control.
EMBO J. 18, 1712^1721.
[206] Jensen, M.R., Lobner-Olesen, A. and Rasmussen, K.V. (1990) Es-
cherichia coli minichromosomes : Absence of copy number control,
and random segregation. J. Mol. Biol. 215, 257^265.
[207] Roth, A. and Messer, W. (1998) High-a⁄nity binding sites for the
initiator protein DnaA on the chromosome of Escherichia coli. Mol.
Microbiol. 28, 395^401.
[208] Kitagawa, R., Mitsuki, H., Okazaki, T. and Ogawa, T. (1996) A
novel DnaA protein-binding site at 94.7 min on the Escherichia coli
chromosome. Mol. Microbiol. 19, 1137^1147.
[209] Kitagawa, R., Ozaki, T., Moriya, S. and Ogawa, T. (1998) Negative
control of replication initiation by a novel chromosomal locus ex-
hibiting exceptional a⁄nity for Escherichia coli DnaA protein.
Genes Dev. 12, 3032^3043.
[210] Morigen, Boye, E., Skarstad, K. and Lobner-Olesen, A. (2001) Reg-
ulation of chromosomal replication by DnaA protein availability in
Escherichia coli : e¡ects of the datA region. Biochim. Biophys. Acta
1521, 73^80.
[211] Skarstad, K., Boye, E. and Steen, H.B. (1986) Timing of initiation
of chromosome replication in individual E. coli cells. EMBO J. 5,
1711^1717.
[212] Boye, E., Stokke, T., Kleckner, N. and Skarstad, K. (1996) Coor-
dinating DNA replication initiation with cell growth: Di¡erential
roles for DnaA and SeqA proteins. Proc. Natl. Acad. Sci. USA
93, 12206^12211.
[213] Helmstetter, C.E. and Leonard, A.C. (1987) Coordinate initiation of
chromosome and minichromosome replication in Escherichia coli.
J. Bacteriol. 169, 3489^3494.
[214] Skarstad, K., von Meyenburg, K., Hansen, F.G. and Boye, E.
(1988) Coordination of chromosome replication initiation in Esche-
richia coli: e¡ects of di¡erent dnaA alleles. J. Bacteriol. 170, 852^
858.
[215] Donachie, W.D. (1968) Relationship between cell size and time of
initiation of DNA replication. Nature 219, 1077^1079.
[216] Wold, S., Skarstad, K., Steen, H.B., Stokke, T. and Boye, E. (1994)
FEMSRE 756 24-10-02 Cyaan Magenta Geel Zwart
The initiation mass for DNA replication in Escherichia coli K-12 is
dependent on growth rate. EMBO J. 13, 2097^2102.
[217] Helmstetter, C.E., Cooper, S., Pierucci, D. and Revelas, E. (1968)
The bacterial life cycle. Cold Spring Harbor Symp. Quant. Biol. 33,
809^822.
[218] Donachie, W.D. (2001) Co-ordinate regulation of the Escherichia
coli cell cycle or The cloud of unknowing. Mol. Microbiol. 40,
779^785.
[219] Filutowicz, M., Ross, W., Wild, J. and Gourse, R.L. (1992) Involve-
ment of Fis protein in replication of the Escherichia coli chromo-
some. J. Bacteriol. 174, 398^407.
[220] Filutowicz, M. and Roll, J. (1990) The requirement of IHF protein
for extrachromosomal replication of the Escherichia coli oriC in a
mutant de¢cient in DNA polymerase I activity. New Biol. 2, 818^
827.
[221] Polaczek, P., Kwan, K., Liberles, D.A. and Campbell, J.L. (1997)
Role of architectrual elements in combinatorial regulation of initia-
tion of DNA replication in Escherichia coli. Mol. Microbiol. 26,
261^275.
[222] Polaczek, P. (1990) Bending of the origin of replication of E. coli by
binding of IHF at a speci¢c site. New Biol. 2, 265^271.
[223] Gille, H., Egan, J.B., Roth, A. and Messer, W. (1991) The FIS
protein binds and bends the origin of chromosomal DNA replica-
tion, oriC, of Escherichia coli. Nucleic Acids Res. 19, 4167^4172.
[224] Roth, A., Urmoneit, B. and Messer, W. (1994) Functions of histone-
like proteins in the initiation of DNA replication at oriC of Esche-
richia coli. Biochimie 76, 917^923.
[225] Samitt, C.E., Hansen, F.G., Miller, J.F. and Schaechter, M. (1989)
In vivo studies of DnaA binding to the origin of replication of Es-
cherichia coli. EMBO J. 8, 989^993.
[226] Cassler, M.R., Grimwade, J.E. and Leonard, A.C. (1995) Cell cycle-
speci¢c changes in nucleoprotein complexes at a chromosomal rep-
lication origin. EMBO J. 14, 5833^5841.
[227] Grimwade, J.E., Ryan, V.T. and Leonard, A.C. (2000) IHF redis-
tributes bound initiator protein, DnaA, on supercoiled oriC of Es-
cherichia coli. Mol. Microbiol. 35, 835^844.
[228] Lobner-Olesen, A., Skarstad, K., Hansen, F.G., von Meyenburg, K.
and Boye, E. (1989) The DnaA protein determines the initiation
mass of Escherichia coli K-12. Cell 57, 881^889.
[229] Hansen, F.G., Christensen, B.B. and Atlung, T. (1991) The initiator
titration model : Computer simulation of chromosome and mini-
chromosome control. Res. Microbiol. 142, 161^167.
[230] Christensen, B.B., Atlung, T. and Hansen, F.G. (1999) DnaA boxes
are important elements in setting the initiation mass of Escherichia
coli. J. Bacteriol. 181, 2683^2688.
[231] Lobner-Olesen, A., Hansen, F.G., Rasmussen, K.V., Martin, B. and
Kuempel, P.L. (1994) The initiation cascade for chromosome repli-
cation in wild-type and Dam methyltransferase de¢cient Escherichia
coli cells. EMBO J. 13, 1856^1862.
[232] DePamphilis, M.L. (1996) Origins of DNA replication. In: DNA
Replication in Eukaryotic Cells (DePamphilis, M.L., Ed.), pp. 45^
86. Cold Spring Harbor Laboratory Press, Cold Spring Harbor,
NY.
[233] Lee, D.G., Makhov, A.M., Klemm, R.D., Gri⁄th, J.D. and Bell,
S.P. (2000) Regulation of origin recognition complex conformation
and ATPase activity: di¡erential e¡ects of single-stranded and dou-
ble-stranded DNA binding. EMBO J. 19, 4774^4782.
[234] Weigel, W. and Seitz, H. (2002) Strand-speci¢c loading of DnaB
helicase by DnaA to a substrate mimicking unwound oriC. Mol.
Microbiol., in press.
Molecular Microbiology (2000) 37(3), 492±500
MicroReview
Plasmid copy number control: an ever-growing story
Gloria del Solar* and Manuel Espinosa
Centro de Investigaciones BioloÂgicas, CSIC, VelaÂzquez,
144, E-28006 Madrid, Spain.
Summary
Bacterial plasmids maintain their number of copies
by negative regulatory systems that adjust the rate of
replication per plasmid copy in response to fluctua-
tions in the copy number. Three general classes of
regulatory mechanisms have been studied in depth,
namely those that involve directly repeated
sequences (iterons), those that use only antisense
RNAs and those that use a mechanism involving an
antisense RNA in combination with a protein. The
first class of control mechanism will not be discussed
here. Within the second class (the most `classical'
one), exciting insights have been obtained on the
molecular basis of the inhibition mechanism that
prevents the formation of a long-range RNA structure
(pseudoknot), which is an example of an elegant
solution reached by some replicons to control their
copy number. Among the third class, it is possible to
distinguish between (i) cases in which proteins play
an auxiliary role; and (ii) cases in which transcrip-
tional repressor proteins play a real regulatory role.
This latter type of regulation is relatively new and
seems to be widespread among plasmids from Gram-
positive bacteria, at least for the rolling circle-
replicating plasmids of the pMV158 family and the
theta-replicating plasmids of the Inc18 streptococcal
family.
Introduction
Bacterial plasmids are extrachromosomal genomes that
replicate autonomously and in a controlled manner. Many
plasmids are self-transmissible or mobilizable by other
replicons, thus having the ability to colonize new bacterial
species. Plasmids may provide the host with valuable
functions, such as drug resistance(s) or metabolic path-
ways useful under certain environmental conditions,
Accepted 9 May, 2000. *For correspondence. E-mail gdelsolar@
cib.csic.es; Tel. (134) 91 561 1800; Fax (134) 91 562 7518.
Q 2000 Blackwell Science Ltd
although they are likely to constitute a slight metabolic
burden to the host. To co-exist stably with their hosts and
to minimize the metabolic load, plasmids must control
their replication, so that the copy number (N) of a given
plasmid is usually fixed within a given host and under
defined cell growth conditions.
It is convenient to distinguish two different stages of the
plasmid life cycle. The first, termed establishment, occurs
when a plasmid copy enters a new permissive host. A
successful establishment may depend upon the plasmid's
ability to replicate rapidly before the host divides (High-
lander and Novick, 1987). This may result in an overshoot
in N before it reaches the characteristic value (Nav). In the
second stage, the replicon enters into the steady state in
which Nav is maintained because, on average, there is
one replicative event per plasmid copy and cell cycle
(NordstroÈm and Wagner, 1994). To maintain this average
rate, plasmids use self-encoded negative control systems
that are able to `sense' and correct up and down
fluctuations from the Nav in individual cells (Summers,
1996). These control systems adjust the rate of replication
per plasmid copy, so that it becomes higher or lower than
1, depending on whether there has been a decrease or an
increase, respectively, in copy number with respect to the
Nav in a given cell. Thus, the N-values from individual cells
of a population should follow a narrow Gaussian distribu-
tion when the regulatory circuits are functioning under
optimal conditions. This would not be the case for a
replicon lacking specific control functions, such as the
oriC minichromosomes (the cloned origin of replication of
Escherichia coli). Narrow distribution of N in several
plasmids and broad deviations in oriC minichromosomes
have been measured by a combination of flow cytometry
and plasmid-driven expression of the green fluorescent
protein (Lùbner-Olesen, 1999).
There are three general types of plasmid copy number
control systems, depending on the type of negative
control element used: (i) directly repeated sequences
(iterons) that complex with cognate replication (Rep)
initiator proteins; (ii) antisense RNAs that hybridize to a
complementary region of an essential RNA, therefore
termed countertranscribed (ct) RNAs; and (iii) ctRNA and
a protein. Within this last group, there are two categories.
In one of them, the ctRNA plays the main regulatory role,
whereas the protein has been proposed as only an
 Control of plasmid replication 493

I Fig. 1. Control of replication by pseudoknot

formation in ColIb-P9. Genes repY (for leader
peptide) and repZ (for initiator of replication)
are translationally coupled. On the mRNA, the
Shine±Dalgarno sequence (SD) of repY is
III exposed, whereas the SD and the initiation
start repZ end repZ codon of repZ can be occluded within
SD repZ
structure III. If there is no interaction with Inc
RNA (left), repY translation takes place.
rep mRNA Stalled ribosomes at the end of repY unfold
SD structure III, allowing the formation of the
repY repY
repZ pseudoknot between the region located on the
loop of structure I (double lines) and its
start repY end repY Replication complementary region (stippled rectangle).
This facilitates binding of the ribosomes
(ellipses) to the repZ SD sequences, followed
by translation of repZ. The antisense Inc RNA
(right) would hinder pseudoknot formation and
Inc RNA translation of repY, leading to inhibition of
repZ translation. Events and elements acting
repY translation positively in replication are indicated in blue,
whereas negative ones are depicted in red.

unfolding of structure III Inc RNA

INHIBITION OF repZ TRANSLATION
tion
la
trans

I
T
O
repZ

KN
O
D

III
EU
PS

I
O
T
O

N
KN
O
D
EU
PS

Inc RNA-mRNA duplex formation

Inhibition of repY translation

auxiliary element. Within the second category, both they will be found to be more widespread. Finally, a single
elements, acting on different targets, could correct instance of a new control mechanism, involving only the
fluctuations in the N-value at the steady state (del Solar Rep protein in plasmids lacking iterons, will be discussed
et al., 1995; 1998). Replication control by iterons will be briefly (Burian et al., 1999).
covered in the accompanying review by Chattoraj, this
issue pp. 467±476. Here, we will consider the other two
Control by ctRNAs
types of mechanisms of replication control, focusing on
cases in which significant advances have recently been Control systems using ctRNAs (Wagner and Brantl, 1998)
achieved. The first case, including plasmids that use only are widespread within plasmids replicating by different
ctRNA as a control element, involves the inhibition of the mechanisms, but sharing a similar genetic structure in the
formation of a long-range RNA structure (the pseudo- control region: two oppositely oriented promoters direct,
knot), which actively enhances translation of an essential respectively, the synthesis of an RNA essential for
replication initiator gene. The second focus will be on replication and of the inhibitor ctRNA. The ctRNAs are
those systems that include proteins playing an auxiliary complementary to a region (the target) near the 5 0 end of
regulatory role. Next, we will pay attention to those the essential RNA. An important feature of this kind of
replicons with two plasmid-encoded control elements. control system is that the rate of synthesis of the inhibitor
There are few reports on plasmids using this last control ctRNA is much higher than that of the essential RNA. In
mechanism but, considering their efficiency, it is likely that addition, the ctRNAs are synthesized from a constitutive
Q 2000 Blackwell Science Ltd, Molecular Microbiology, 37, 492±500
494 G. del Solar and M. Espinosa
promoter and have a short half-life, so that their
intracellular concentration stays nearly proportional to N.
Replication is inhibited by RNA±ctRNA pairing and
abolition of the essential RNA activity. Differences
between various replicons regulated by ctRNAs are
found based on how the inhibition occurs and can be
exemplified as follows: (i) inhibition of maturation of the
primer essential for replication, as in plasmid ColE1
(Tomizawa and Itoh, 1981); (ii) premature termination of
the synthesis of the essential rep-mRNA (pT181; Novick
et al., 1989); and (iii) inhibition of rep translation. This last
method of ctRNA-dependent copy number control can be
achieved by three mechanisms, namely inhibition of
translation of a leader peptide (plasmid R1; Blomberg
et al., 1992), inhibition of both translation of a leader
peptide and formation of an activator pseudoknot (ColIb-
P9; Asano et al., 1991; Asano and Mizobuchi, 1998a) and
direct inhibition of translation of the essential rep gene. A
direct inhibition mechanism has been proposed for
plasmid ColE2, in which the formation of a stable complex
between the ctRNA and the complementary region in the
rep mRNA would be able to inhibit rep expression, even
though the ctRNA does not overlap the putative transla-
tion initiation region of the essential gene (T. Itoh,
personal communication).
For a number of plasmids, studies on the kinetics of
interaction between ctRNAs and their RNA targets have
shown that stable duplex formation is not required for
inhibition of rep expression, as this inhibition takes place
faster than stable binding (reviewed by Wagner and
Brantl, 1998). In contrast, a recent report on the ctRNA±
rep mRNA interactions of the staphylococcal plasmid
pT181 shows that the rate constant of stable complex
formation is similar to the inhibition rate constant (Brantl
and Wagner, 2000).
New findings on the molecular and biochemical bases
for the control mechanism involving an intramolecular
RNA±RNA interaction within the leader region of the rep
mRNA have been provided for the IncIa ColIb-P9 and the
IncL/M pMU604 plasmids (Asano and Mizobuchi, 1998a;
2000; Athanasopoulos et al., 1999). Replication depends
on the expression of the essential rep gene, which
requires coupled translation of a gene encoding a leader
peptide and the formation of a pseudoknot by intra-
molecular pairing of two complementary sequences of the
rep mRNA (Asano et al., 1991; Wilson et al., 1993).
The model for the control of replication of ColIb-P9
(reviewed by Wagner and Simons, 1994) involves two
stem±loop structures in the repZ mRNA that have been
mapped in vitro (Asano and Mizobuchi, 1998b) and are
located upstream (structure I) and in the middle (structure
III) of the leader repY gene (Fig. 1). Structure III occludes
both the translation initiation sequences of the essential
repZ gene and a short sequence complementary to a
region in the loop of structure I. Appropriate termination of
translation of the leader repY gene unfolds structure III,
thus inducing the formation of the pseudoknot by
intramolecular pairing of loop I with its complementary
sequence. Pseudoknot formation facilitates the binding of
the ribosomes to the Shine±Dalgarno sequence of the
essential repZ gene. Mutations causing disruption of
structure III have allowed the mapping of the pseudoknot
in vitro (Asano and Mizobuchi, 1998b). Translation of repY
and pseudoknot formation are inhibited by the interaction
of the ctRNA (Inc RNA) with the complementary region of
the rep mRNA. Generation of the duplex Inc RNA±target
mRNA occludes the repY ribosome binding site, leading
to an indirect inhibition of repZ translation. In addition,
pairing between the single-stranded loops on the com-
plementary structures of rep mRNA (structure I) and Inc
RNA is sufficient to block generation of the pseudoknot,
so that Inc RNA can repress repZ translation at the level
of a transient interaction with its target before a stable
duplex is detected. The same region in the loop of
structure I of the rep mRNA is involved in the initial
interactions that take place in pseudoknot formation and
in Inc RNA binding. These initial interactions are
stimulated by the presence of a hexanucleotide (con-
served in various antisense systems; Asano and
Mizobuchi, 1998a), which presumably supports the
element termed U-turn and includes the structure I
sequence involved in the initial interactions. The U-turn
loop structures are general binding rate enhancers that
facilitate rapid RNA±RNA interactions (reviewed by
Franch and Gerdes, 2000). As the early stages in
pseudoknot formation and in the binding of the inhibitory
RNA are similar, an explanation as to how these two
processes can compete with each other is easily deduced
(Asano and Mizobuchi, 1998a). Although Inc RNA
represses the translation of repY and repZ genes,
repression of the former is much less efficient than that
of the latter. Repression of repZ and repY expression are
accomplished at different stages during the pairing
between Inc RNA and rep mRNA (Asano and Mizobuchi,
2000). This differential repression allows the Inc RNA to
keep the total level of repZ expression constant. A
constant total rate of synthesis of the initiator was
demonstrated early for IncFII plasmids (Nielsen and
Molin, 1984). As the level of repZ expression is rate-
limiting for replication, the Inc RNA-based regulation
mechanism maintains a constant N-value.
A similar genetic structure and control mechanism were
shown to exist in the IncB plasmid pMU720, closely
related to the IncIa-ColIb-P9 (Siemering et al., 1994;
Wilson et al., 1994). However, some differences are found
in plasmid pMU604, belonging to the IncL/M group
(Athanasopoulos et al., 1999). In this case, replication
control also involves translational coupling of a leader
Q 2000 Blackwell Science Ltd, Molecular Microbiology, 37, 492±500

CopB Tap RepA
ori
P1 P2
CopA
Fig. 2. Auxiliary elements controlling plasmid copy number: the R1
paradigm. The initiator RepA protein acts on the origin of replication
(ori) located downstream of the repA gene. Expression of repA is
translationally coupled to that of tap (encoding a small leader
peptide). CopB protein represses transcription from promoter P2,
so that this promoter is normally silent. Transcription of the copB±
tap±repA mRNA is mainly directed by the constitutive weak
promoter P1. The antisense RNA, CopA, acts by blocking the
translation of tap. Negatively acting elements are depicted in red;
those acting positively are shown in blue.
gene and the formation of a pseudoknot, although, unlike
ColIb-P9, the positioning of sequences involved in the
expression of the essential gene repA is different.
The most important variation is the spacer between the
pseudoknot and the translation initiation region of the
essential repA gene, which seems to be suboptimal in
pMU604. Mutational analyses showed that the require-
ment for pseudoknot formation in pMU604 could be
partially replaced by improving the initiation of translation
signals of the essential gene repA. Interestingly,
RNA II pre-primer
RNAP Origin rom
RNA I
Rom
RNA I
Origin
RNase H
RNA I
Rom
Primer maturation No DNA-RNA hybrid
No primer maturation
Replication
Replication inhibited
Fig. 3. Copy number control in ColE1. Synthesis of the preprimer
RNA II by RNAP (stippled circle) is essential for replication. In the
absence of interaction with the RNA I (left), the RNA II forms a
stable hybrid with the template DNA at the origin of replication. This
hybrid is cleaved by RNase H to generate the 3 0 -OH end of the
RNA primer, from which replication starts. Interaction between the
inhibitor RNA I and the complementary region in the RNA II
preprimer (right) is aided by Rom protein (ellipse). RNA I±RNA II
interaction inhibits the formation of the DNA±RNA II hybrid at the
origin region, preventing maturation of the RNA II into the
replication primer. Colours as in the legends to Figs 1 and 2.
Q 2000 Blackwell Science Ltd, Molecular Microbiology, 37, 492±500
Control of plasmid replication 495
pseudoknot formation inhibited translation of the leader
gene in pMU604 (Athanasopoulos et al., 1999), a feature
that has not been reported to exist in ColIb-P9. The
existence of such complex regulatory circuits in other
plasmids awaits further investigation.
Accessory proteins as inhibitory elements
Plasmids R1 and ColE1, which control their N-values by
ctRNAs, also encode a protein as a second inhibitory
element that, however, plays only an auxiliary role. In the
case of R1, expression of the essential gene repA
requires the translation of the leader gene tap, which is
translationally coupled to repA (Blomberg et al., 1992).
The main replication control element is the ctRNA, CopA,
which inhibits translation of tap and, indirectly, that of
repA. The second inhibitory element of R1 is the product
of the copB gene, which is co-transcribed with tap and
repA from promoter P1 (Fig. 2). CopB protein is a
transcriptional repressor of a second promoter, P2,
located downstream of copB, which directs the synthesis
of a tap±repA mRNA. At steady state, CopB is present at
saturating concentrations, blocking transcription from P2,
so that repA is expressed almost exclusively from P1.
Deletion of the entire copB gene (including promoter P1)
results in plasmids with an eightfold increase in N (Riise
et al., 1982). The CopB regulatory loop has been
suggested to serve as a rescue mechanism that prevents
plasmid loss in newborn cells harbouring very few plasmids.
However, computer simulation of mini-R1 plasmid replica-
tion indicated that the CopB regulatory circuit contributes
little to the stability of these replicons (Rosenfeld and
Grover, 1993).
The second instance of plasmids with auxiliary proteins
is ColE1 (Fig. 3). In this case, replication is mediated by
the synthesis of a preprimer RNA (RNA II) by the host
RNA polymerase (RNAP) and the formation of a DNA±
RNA hybrid between the RNA II and the template DNA
strand at the origin region. This hybrid is cleaved by
RNase H, generating a 3 0 -OH end, which is used by DNA
polymerase I to initiate leading strand synthesis. The
availability of the primer 3 0 -OH end is rate-limiting for
initiation, and it is modulated by the ctRNA I control
element. Interaction between ctRNA I and its comple-
mentary region in the preprimer alters the secondary
structure of the latter, leading to the inhibition of stable
DNA±RNA hybrid formation. This, in turn, leads to
inhibition of replication. The second element of this circuit
is protein Rom (Rop), which enhances the rate of
formation of a stable complex between the ctRNA I and
the preprimer RNA. Rom does not seem to be an
essential component of the ColE1 control system.
Deletion of the rom gene leads to a two- to threefold
increase in N in slowly growing cells, but it has no
496 G. del Solar and M. Espinosa

pMV158 Fig. 4. Dual regulation of plasmid copy

number, as exemplified by pMV158 (top) and
by two members of the Inc18 plasmid family
(middle and bottom). In pMV158, the initiator
protein, RepB, acts on the leading strand
CopG RepB origin, dso. The CopG repressor protein
blocks transcription from promoter Pcr, which
directs synthesis of the copG±repB mRNA, so
that CopG autoregulates its own synthesis.
dso Pcr Pctll From promoter PctII, the ctRNA II is
RNA II synthesized, which, in turn, blocks translation
of repB. In the case of pIP501, the initiator
RepR protein acts on the origin (oriR) located
downstream of repR. The transcriptional
repressor CopR regulates transcription from
pII, which is the only promoter that directs the
expression of repR. CopR is synthesized from
promoter pI, which would be constitutive.
From promoter pIII, the long-lived inhibitor
pIP501
ctRNA III is synthesized. Interaction of ctRNA
CopR RepR III with its target in the leader region of the
+ oriR repR mRNA induces (%) generation of an
mRNA secondary structure, which acts as a
plII transcriptional attenuator, avoiding repR
pl plI
expression. CopR-dependent inhibition of pII
RNA III + increases transcription from pIII (%) by
alleviating the effect of convergent
transcription. The replication and control
regions of pSM19035 share similarities in the
genes, promoters and regulatory circuits to
those of pIP501. The segB region is shown as
pSM19035 composed of three genes, their expression
being controlled by a single promoter (Pv).
The product of gene v regulates this operon
and also the promoter of gene copS. The
CopS RepS main regulatory circuits would be the same as
+ oriS ω ε ζ in pIP501 (Brantl et al., 1990). Colours as in
the legends to previous figures.
Pcop1 Prep P
III Pω
RNA +

phenotypic consequences on the N-value in fast-growing
bacteria (Atlung et al., 1999). When cloned on a
compatible multicopy plasmid, the rom gene is able to
complement a rom2 derivative, although it has no further
effect upon the replication of a co-resident ColE1. The
absence of incompatibility caused by extra copies of rom
shows that Rom is not a primary inhibitor of ColE1
replication, as it exerts its maximum effect at the wild-type
concentration (Summers, 1996). Mathematical models of
the dynamic features of copy number control in ColE1 and
the experimental observations about the small effect
caused by variations in the dosage of the rom gene have
left open the question of why there is a Rom protein
(Paulson et al., 1998). At least three theoretical proposals
have been made to account for an important role of Rom
in the dynamics of ColE1 copy number control (Paulson
et al., 1998). First, Rom concentration would be propor-
tional to the N-value, so that the response in replication
frequency to variations in the N-value would be sharper
than if RNA I acted alone. This hypothesis requires that
Rom is rapidly degraded, which has not been tested
experimentally. Secondly, Rom would act by making the
probability of plasmid replication very close to zero at high
ctRNA I concentration because, in the absence of Rom,
the intrinsic rate of ctRNA I±RNA II duplex formation
would be too slow to ensure total inhibition of replication.
Thus, Rom would ensure an efficient copy number control
system. Thirdly, Rom could act as a back-up system when
N (and, subsequently, Rom concentration) is greatly
reduced: under normal conditions, the replication fre-
quency would not depend on small deviations in Rom
concentration but, if this concentration decreases greatly
(as a result of a large reduction in N), inhibition of primer
formation would decrease, thus leading to an increase in
the replication frequency. However, and as far as we are
aware, no experiments have been performed to clarify
these hypotheses. On the other hand, experimental
evidence has shown that the presence of ColE1 deriva-
tives lacking rom reduced bacterial growth in medium
impoverished in carbon sources, whereas rom1 deriva-
tives did not show an adverse effect on cell growth (Atlung
et al., 1999). This is thought to be related to the

Q 2000 Blackwell Science Ltd, Molecular Microbiology, 37, 492±500


amplification of ColE1-type plasmids in slowly growing
cells, which is higher in plasmids lacking rom. From these
observations, a key role for Rom protein has been
envisaged, that is to prevent ColE1-type plasmids from
representing a metabolic burden to their hosts in natural
habitats, in which cells grow slowly because of nutrient
limitation.
In conclusion, CopB and Rom are inhibitory elements
that limit plasmid replication, as their inactivation leads to
an increase in the N-value of R1 or ColE1 respectively.
However, the genes encoding these proteins do not
constitute an incompatibility determinant against the wild-
type plasmid, so that both proteins would be auxiliary
negative elements unable to correct up-fluctuations in N.
However, the influence of these proteins on the dynamics
of plasmid copy number control remains to be solved.
Dual regulation by ctRNA and inhibitory protein
A novel mode of regulation of replication, involving a
ctRNA and a transcriptional repressor protein, has been
shown for two plasmid families represented by pMV158
(del Solar and Espinosa, 1992) and pIP501 (Brantl, 1994).
Unlike plasmids controlling their replication only by
ctRNAs, the promoter directing the expression of the
essential rep gene in these plasmids is not constitutive,
but regulated by a Cop protein. Although the dual circuits
of control of N in these two plasmids share similarities,
significant differences exist with respect to the features
and inhibitory mechanisms of both regulatory elements.
In the case of pMV158, characterization of mutations
that result in increased N-values and definition of DNA
regions (inc determinants) that affect plasmid replication
in trans showed a third general mechanism of replication
control involving the transcriptional repressor protein
CopG and the ctRNA (RNA II), both involved in regulation
of synthesis of the initiator RepB protein (del Solar et al.,
1995). Gene copG is co-transcribed with repB from a
single promoter (Pcr), and the copG product binds to a
DNA region that includes promoter Pcr, thus inhibiting
transcription of the cop±rep operon (Fig. 4). Competition
between purified CopG and RNAP proteins indicates that
CopG impairs the binding of RNAP to promoter Pcr (our
unpublished observations). CopG protein is very stable, at
least when overproduced in Escherichia coli. Promoter
PctII directs the synthesis of RNA II, which has a short
half-life (<2 min, G. del Solar, unpublished) and seems to
inhibit repB expression by direct pairing with the transla-
tion initiation signals of repB. Both CopG and RNA II are
able to sense and correct up-fluctuations in the N-value at
the steady state, and together constitute an inc determi-
nant much more powerful than each element alone. This
synergistic effect argues in favour of a primary role for
both elements in pMV158 control of N and allowed us to
Q 2000 Blackwell Science Ltd, Molecular Microbiology, 37, 492±500
Control of plasmid replication 497
suggest the importance of derepression of the CopG-
regulated Pcr promoter to overcome severe decreases in
N (del Solar et al., 1995). Owing to its autorepressor
capacity, CopG would keep the level of synthesis of the
cop±rep mRNA within narrow limits. Variable concen-
trations of RNA II (the concentration depending upon the
N-value because of the constitutive synthesis of RNA II
and of its short half-life) would act on these practically
constant levels of mRNA. A similar genetic structure
exists in all plasmids from the pMV158 family.
Three closely related streptococcal plasmids belonging
to the Inc18 family (pIP501, pAMb1 and pSM19035) have
been studied. In the case of pIP501, control of N involves
the transcriptional repressor CopR and the antisense
RNA III (Fig. 4). Promoter pI directs the expression of
gene copR, whose product inhibits transcription from
promoter pII located downstream of pI (Brantl, 1994).
CopR-mediated repression does not completely inhibit
transcription from pII, which is the only promoter that
directs the expression of gene repR, essential for plasmid
replication. A third promoter, pIII, directs the synthesis of
RNA III, a ctRNA that inhibits the expression of repR by a
mechanism of transcriptional attenuation (Brantl et al.,
1993), similar to that postulated for plasmids pAMb1 (Le
Chatelier et al., 1996) and pT181 (Novick et al., 1989), so
that interaction between repR mRNA and RNA III leads to
premature termination of the mRNA synthesis. Interest-
ingly, RNA III has an unusually long half-life of about
30 min, which would make it unable to correct down-
fluctuations in N rapidly. To explain how the control
system of pIP501 copes with a reduction in N, a dual
function has been assigned to CopR (Brantl and Wagner,
1997). A decrease in N would result in a decrease in
CopR concentration, which would lead to derepression of
promoter pII and to the concomitant increase in the
expression of repR. In addition, transcription from
promoter pIII is inhibited by the increased convergent
transcription from pII, so that the slow reduction in the
amounts of intracellular stable RNA III inhibitor could be
slightly accelerated.
In the case of plasmid pAMb1, the situation may be
more complex. The repressor CopF (equivalent to CopR)
not only inhibits RepF (equivalent to RepR) synthesis, but
it may also decrease primer formation, as the activating
transcription fork that passes through the origin of
replication originates from the repF promoter (L. JannieÁre
in Espinosa et al., 2000). A further degree of complexity
arises from new findings on plasmid pSM19035. In this
plasmid, part of the segB region (involved in better-than-
random segregation) contains genes v , 1 and z (Fig. 4).
These genes are organized in an operon controlled by the
product of gene v . Purified Omega protein not only binds
to its promoter region, but also to a DNA fragment
containing the promoter region of gene copS (the
498 G. del Solar and M. Espinosa
homologue of pIP501-copR and pAMb1-copF). Conse-
quently, it would appear that plasmid partition and plasmid
copy number control are under a common pathway of
regulation (de la Hoz et al., 2000).
There is a fundamental difference between the control
mechanisms governed only by ctRNA and those
governed by dual regulators. In the former, the rate of
synthesis per plasmid copy of the essential RNA needed
for replication is constant but rather low compared with
that of the ctRNA. In the latter, expression of the rep gene
is directed by a strong and Cop-regulated promoter so
that, when the regulatory protein does not operate, there
is a high rate of rep transcription. This seems to be at
least the case for the pMV158-Pcr promoter, which is the
strongest promoter of the plasmid (our unpublished
results). A high potential to transcribe the essential rep
gene would represent an advantage for these plasmids
during the establishment stage, as they would replicate
rapidly to reach Nav, thus decreasing the frequency of
appearance of plasmid-free cells from newly colonized
bacteria. We have preliminary evidence suggesting that
the efficiency of establishment of pMV158 derivatives is
influenced by the activity of the promoter directing the
expression of the repB gene. It is worth pointing out that
plasmids from the Inc18 family and several replicons of
the pMV158 family are mobilizable, and that these
replicons have a very broad host range. Plasmids with
auxiliary proteins involved in their control may also share
the same advantage during the establishment stage
(Summers, 1996).
Non-iteron non-ctRNA plasmid control?
It seems that a new mechanism for copy number control
independent of antisense RNA and iterons operates in
small cryptic plasmids of E. coli, with the Rep protein
being the only plasmid-encoded factor involved in initia-
tion and repression of replication (Burian et al., 1999). The
copy number control (cop) region, which also constitutes
an inc determinant, includes the rep promoter and two
Rep binding sites (BD1 and BD2). BD1 is close to the
replication initiation region and seems to bind preferen-
tially monomers/dimers of the Rep protein, whereas BD2
is adjacent to the rep promoter region and would appear
to bind Rep oligomers preferentially. Initiation of replica-
tion would require binding of the Rep protein to both
binding sites. Increases in N, and subsequent elevation of
the Rep concentration, would favour the formation of Rep
oligomers, displacing Rep monomers/dimers from BD2,
thereby autorepressing rep transcription. Thus, control of
replication is proposed to be exerted by the monomer±
multimer Rep equilibrium and, unlike the iteron-based
control systems, incompatibility in these replicons would
easily be over-ridden by an excess of Rep protein.
Whether this mechanism is widespread among small
plasmids of enterobacteria remains to be determined.
Perspectives
The recently reported new mechanisms controlling
plasmid copy number show that these processes are
more complex than previously envisaged. Pseudoknot
formation, known for a number of years, has only recently
been characterized in some depth. Mechanisms control-
ling the replication of newly described replicons, espe-
cially those isolated from extremophiles, are mostly
unknown, and a novel control mechanism (still to be
explored in some depth), independent of iterons and
ctRNAs, seems to exist in small cryptic plasmids of E. coli.
New and exciting findings will be derived from a general
mechanism for interactions between antisense and target
RNAs in prokaryotes. This mechanism involves rapid
interaction between both RNAs mediated by a motif (Pyr-
U-N-Pur, the so-called U-turn) that is conserved in RNA
recognition loops in either the antisense or the target
RNAs in many antisense RNA-regulated gene systems
(Franch et al., 1999; Franch and Gerdes, 2000). However,
and in spite of the knowledge accumulated during recent
years, there is a lack of structural studies on the elements
involved in the regulation of replication, with a few
exceptions. The structure of pMV158-CopG repressor
has been solved both alone and in complex with its target
DNA, and it has been shown that the CopG dimer has a
ribbon±helix±helix structure, resembling that of the Arc
repressor of phage P22 (Gomis-RuÈth et al., 1998).
Modelling of the first 62 residues of the pIP501-CopR
repressor indicates that it belongs to the helix±turn±helix
family of DNA-binding proteins (Steinmetzer et al., 2000),
and preliminary X-ray diffraction data are available from
the pSM19035-Omega repressor (Murayama et al.,
1999). In addition, a three-dimensional model of the
antisense RNA of plasmid R1, together with in vitro
structural determinations, have revealed that the anti-
sense-mRNA binding product is not a full duplex, but a
complicated extended complex that is stabilized by distal
basepairings (Kolb et al., 2000). Determination of the
crystal structure of this antisense RNA may throw new
light on the nature of the contacts between the two RNA
species. Finally, and in spite of the early characterization
of auxiliary proteins such as CopB or Rom, there seems
not to be enough information on the precise role of these
proteins in replication control at the steady state and
during the plasmid establishment stage.
Acknowledgements
Thanks are due to D. Chattoraj for his critical reading of the
manuscript, and T. Itoh for communicating his unpublished
Q 2000 Blackwell Science Ltd, Molecular Microbiology, 37, 492±500

results. The authors' laboratories were financed by CICYT
(grants BIO97-0347 and 2FD97-0518, to M.E.) and by
Comunidad AutoÂnoma de Madrid (grant 07B/0049/1999 to
G.d.S.).
References
Asano, K., and Mizobuchi, K. (1998a) Copy number control of
IncIa plasmid ColIb-P9 by competition between pseudo-
knot formation and antisense RNA binding at a specific
RNA site. EMBO J 17: 5201±5213.
Asano, K., and Mizobuchi, K. (1998b) An RNA pseudoknot as
the molecular switch for translation of the repZ gene
encoding the replication initiator of IncIa plasmid ColIb-P9.
J Biol Chem 273: 11815±11825.
Asano, K., and Mizobuchi, K. (2000) Structural analysis of
late intermediate complex formed between plasmid ColIb-
P9 Inc RNA and its target RNA. How does a single
antisense RNA repress translation of two genes at different
rates? J Biol Chem 275: 1269±1274.
Asano, K., Moriwaki, H., and Mizobuchi, K. (1991) Positive
and negative regulations of plasmid ColIb-P9 repZ gene
expression at the translational level. J Biol Chem 266:
3774±3781.
Athanasopoulos, V., Praszkier, J., and Pittard, A.J. (1999)
Analysis of elements involved in pseudoknot-dependent
expression and regulation of the repA gene of an IncL/M
plasmid. J Bacteriol 181: 1811±1819.
Atlung, T., Christensen, B.B., and Hansen, F.G. (1999) Role
of the Rom protein in copy number control of plasmid
pBR322 at different growth rates in Escherichia coli K-12.
Plasmid 41: 110±119.
Blomberg, P., Nordstrom, K., and Wagner, E.G.H. (1992)
Replication control of plasmid R1: RepA synthesis is
regulated by CopA RNA through inhibition of leader peptide
translation. EMBO J 11: 2675±2683.
Brantl, S. (1994) The copR gene product of plasmid pIP501
acts as a transcriptional repressor at the essential repR
promoter. Mol Microbiol 14: 473±483.
Brantl, S., and Wagner, E.G.H. (1997) Dual function of the
copR gene product of plasmid pIP501. J Bacteriol 179:
7016±7024.
Brantl, S., and Wagner, E.G.H. (2000) Antisense RNA-
mediated transcriptional attenuation: an in vitro study of
plasmid pT181. Mol Microbiol 35: 1469±1482.
Brantl, S., Behnke, D., and Alonso, J.C. (1990) Molecular
analysis of the replication region of the conjugative
Streptococcus agalactiae plasmid pIP501 in Bacillus
subtilis. Comparison with plasmids pAMb1 and
pSM19035. Nucleic Acids Res 18: 4783±4790.
Brantl, S., Birch-Hirschfeld, E., and Behnke, D. (1993) RepR
protein expression on plasmid pIP501 is controlled by an
antisense RNA-mediated transcription attenuation
mechanism. J Bacteriol 175: 4052±4061.
Burian, J., StuchlõÂk, S., and Kay, W.W. (1999) Replication
control of a small cryptic plasmid of Escherichia coli. J Mol
Biol 294: 49±65.
Espinosa, M., Cohen, S.N., Couturier, M., del Solar, G., DõÂaz-
Orejas, R., Giraldo, R., et al. (2000) Plasmid replication and
copy number control. In The Horizontal Gene Pool:
Bacterial Plasmids and Gene Spread. Thomas, C.M.
Q 2000 Blackwell Science Ltd, Molecular Microbiology, 37, 492±500
Control of plasmid replication 499
(ed.). Amsterdam: Harwood Academic Publishers, pp. 1±
47.
Franch, T., Petersen, M., Wagner, E.G.H., Jacobsen, J.P.,
and Gerdes, K. (1999) Antisense RNA regulation in
prokaryotes: rapid RNA/RNA interaction facilitated by a
general U-turn loop structure. J Mol Biol 294: 1115±1125.
Franch, T., and Gerdes, K. (2000) U-turns and regulatory
RNAs. Curr Opin Microbiol 3: 159±164.
Gomis-RuÈth, F.X., SolaÂ, M., Acebo, P., PaÂrraga, A., Guasch,
A., Eritja, R., et al. (1998) The structure of plasmid-
encoded transcriptional repressor CopG unliganded and
bound to its operator. EMBO J 17: 7404±7415.
Highlander, S.K., and Novick, R.P. (1987) Plasmid repopulation
kinetics in Staphylococcus aureus. Plasmid 17: 210±221.
de la Hoz, A.B., Ayora, S., Sitkiewicz, I., Pankiewicz, R.,
Alonso, J.C., and Ceglowski, P. (2000) Plasmid copy-
number control and better-than-random segregation genes
of pSM19035 share a common regulator. Proc Natl Acad
Sci USA 97: 728±733.
Kolb, F.A., Malmgren, C., Westhof, E., Ehresmann, C.,
Ehresmann, B., Wagner, E.G.H., et al. (2000) An unusual
structure formed by antisense-target RNA binding involves
an extended kissing complex with a four-way junction and a
side-by-side helical alignment. RNA 6: 311±324.
Le Chatelier, E., Ehrlich, S.D., and JannieÁre, L. (1996)
Countertranscript-driven attenuation system of the pAMb1
repE gene. Mol Microbiol 20: 1099±1112.
Lùbner-Olesen, A. (1999) Distribution of minichromosomes in
individual Escherichia coli cells: implications for replication
control. EMBO J 18: 1712±1721.
Murayama, K., de la Hoz, A.B., Alings, C., LoÂpez, G., Orth,
P., Alonso, J.C., et al. (1999) Crystallization and pre-
liminary X-ray diffraction studies of Streptococcus pyo-
genes plasmid pSM19035-encoded v transcriptional
repressor. Acta Crystallog D-55: 2041±2042.
Nielsen, P.F., and Molin, S. (1984) How the R1 replication
control system responds to copy number deviations.
Plasmid 11: 264±267.
NordstroÈm, K., and Wagner, E.G.H. (1994) Kinetic aspect of
control of plasmid replication by antisense RNA. Trends
Biol Sci 19: 294±300.
Novick, R.P., Iordanescu, S., Projan, S.J., Kornblum, J., and
Edelman, I. (1989) pT181 plasmid replication is regulated
by a countertranscript-driven transcriptional attenuator.
Cell 59: 395±404.
Paulson, J., NordstroÈm, K., and Ehrenberg, M. (1998)
Requirements for rapid plasmid ColE1 copy number
adjustments: a mathematical model of inhibition modes
and RNA turnover rates. Plasmid 39: 215±234.
Riise, E., Stougaard, P., Bindslev, B., NordstroÈm, K., and
Molin, S. (1982) Molecular cloning and functional char-
acterization of a copy number control gene (copB) of
plasmid R1. J Bacteriol 151: 1136±1145.
Rosenfeld, R., and Grover, N.B. (1993) Control of mini-R1
plasmid replication: a computer simulation. Plasmid 29:
94±116.
Siemering, K.R., Praszkier, J., and Pittard, A.J. (1994)
Mechanisms of binding of the antisense and target RNAs
involved in the regulation of IncB plasmid replication. J
Bacteriol 176: 2677±2688.
del Solar, G., and Espinosa, M. (1992) The copy number of
500 G. del Solar and M. Espinosa
plasmid pLS1 is regulated by two trans-acting plasmid
products: the antisense RNA II and the repressor protein,
RepA. Mol Microbiol 6: 83±94.
del Solar, G., Acebo, P., and Espinosa, M. (1995) Replication
control of plasmid pLS1: efficient regulation of plasmid copy
number is exerted by the combined action of two plasmid
components, CopG and RNA II. Mol Microbiol 18: 913±924.
del Solar, G., Giraldo, R., Ruiz-Echevarria, M.J., Espinosa, M.,
and Diaz-Orejas, R. (1998) Replication and control of circular
bacterial plasmids. Microbiol Mol Biol Rev 62: 434±464.
Steinmetzer, K., Hillisch, A., Behlke, J., and Brantl, S. (2000)
Transcriptional repressor CopR: structure model based
localization of the DNA binding motif. Proteins 38: 393±406.
Summers, D.K. (1996). The Biology of Plasmids. Oxford:
Blackwell Science.
Tomizawa, J., and Itoh, T. (1981) Plasmid ColE1 incom-
patibility determined by interaction of RNA1 with primer
transcript. Proc Natl Acad Sci USA 78: 6096±6100.
Wagner, E.G.H., and Brantl, S. (1998) Kissing and RNA
stability in antisense control of plasmid replication. Trends
Biochem Sci 23: 451±454.
Wagner, E.G.H., and Simons, R.W. (1994) Antisense RNA
control in bacteria, phages, and plasmids. Annu Rev
Microbiol 48: 713±742.
Wilson, I.W., Praszkier, J., and Pittard, A.J. (1993) Mutations
affecting pseudoknot control of the replication of B group
plasmids. J Bacteriol 175: 6476±6483.
Wilson, I.W., Praszkier, J., and Pittard, A.J. (1994) Molecular
analysis of RNAI control repB translation in IncB plasmids.
J Bacteriol 176: 6497±6508.
Q 2000 Blackwell Science Ltd, Molecular Microbiology, 37, 492±500
CONJUGACION EN BACTERIAS GRAM NEGATIVAS
Fig. 4. Model for regulation of T-strand production and delivery to the translocation machinery of the MOBF T4SS (indicated by ‘uptake of DNA’ in abox). Steps in the
initiation reaction cascade (I–VII) are described in DNA processing of MOBF relaxosomes. Protein components are illustrated as follows:the bifunctional TraI
phosphodiesterase (blue) with active site tyrosine (Tyr) and TraI helicase domains (light blue), IHF (orange), TraY (purple), TraM (green) and the T4CP TraD
(yellow). See DNA processing of MOBF relaxosomes for details. r Journal of Bacteriology 19: 6888-6899, doi:10.1128/JB.00920-09, reproduced with permission from
the American Society for Microbiology.
fig07: Stages of T4 nucleoprotein transfer initiation mediated by plasmid R1-16.A and B. The T4 transfer apparatus is
constitutively expressed and assembled. Stage 1: The relaxosome, containing oriT bound by TraM (green), TraY
(violet), IHF (orange) and TraI (blue) is docked to the T4CP (yellow) via TraI translocation signals (TSA, TSB) and TraM
binding. Stage 2: Productive pilus contacts with adherent R17 phage (A) or another cell (B) produces signals (lightning
bolt) conveyed over the pilus to the T4CP-relaxosome receptor. Processing of those distinct signals into nucleoprotein
translocation requires at a minimum, a conserved relaxase activation domain. TraI N1-992 comprises this domain for R1
(inset). Stage 3: During phage infection (A) downstream activation of the T4CP ATPase activity is essential for uptake of
the R17 RNA–protein A complex. Alternatively, when cell contact permits plasmid self-transfer (B) initiation still requires
productive interactions between the T4CP and R1 TraM bound to oriT. Signal processing is performed by the relaxase
activation domain occupying the T4CP sites (inset). If TraI is bound to TraD, then the TraI N1-992 domain converts the
cell contact signal into localized oriT melting (inset) and helicase activation in a separate downstream step.

Mentions: In this model (Fig. 7) TraD is anchored to the base of the transfer channel while its cytosolic domain binds TraM, TraI and oriT
DNA (Stage 1). The relaxosome is catalytically active at nic in the absence of TraD, but cleavage is stimulated by its presence (Mihajlovic
et al., 2009). NTP hydrolysis by TraD appears to be silent. Progression from this stage requires signals communicated over the pilus from
the cell exterior (Stage 2). In the case of R17 phage adsorption the productive receptor for the incoming signal is TraD docked by the R1
relaxosome (Fig. 7A). The accessory factors bound at oriT are important, but the key component is the TraI N1-992 docking and
activation domain (inset). We propose that processing of the external signal through this mechanism depends on the physical link
between catalytic activity at the plasmid nick site and the T4CP to modulate TraD conformation and thereby activate the essential
ATPase. The activation initiates translocation of the RNA–protein A complex into the host cell (Stage 3). Activities related to TraI helicase
and its C-terminal domain are dispensable. The fact that this form of translocation activation is independent of the helicase domain and
that the helicase itself is not activated on the docked oriT under these conditions is logical, since phage penetration would otherwise
result in plasmid DNA being extruded pointlessly into the medium. This has never been observed.

BACTERIAS hfr (HIGH FREQUENCY RECOMBINATION)

Map eo de
genes mediante Hfrs
 crossmark
MINIREVIEW

Enterococcal Sex Pheromones: Evolutionary Pathways to Complex,

Two-Signal Systems
Gary M. Dunny,a Ronnie Per-Arne Berntssonb
Department of Microbiology and Immunology, University of Minnesota, Minneapolis, Minnesota, USAa; Department of Medical Biochemistry and Biophysics, Umeå
University, Umeå, Swedenb

Gram-positive bacteria carry out intercellular communication using secreted peptides. Important examples of this type of com-
munication are the enterococcal sex pheromone systems, in which the transfer of conjugative plasmids is controlled by intercel-
lular signaling among populations of donors and recipients. This review focuses on the pheromone response system of the con-
jugative plasmid pCF10. The peptide pheromones regulating pCF10 transfer act by modulating the ability of the PrgX
transcription factor to repress the transcription of an operon encoding conjugation functions. Many Gram-positive bacteria reg-
ulate important processes, including the production of virulence factors, biofilm formation, sporulation, and genetic exchange

Downloaded from http://jb.asm.org/ on March 7, 2017 by guest

using peptide-mediated signaling systems. The key master regulators of these systems comprise the RRNPP (RggRap/NprR/
PlcR/PrgX) family of intracellular peptide receptors; these regulators show conserved structures. While many RRNPP systems
include a core module of two linked genes encoding the regulatory protein and its cognate signaling peptide, the enterococcal sex
pheromone plasmids have evolved to a complex system that also recognizes a second host-encoded signaling peptide. Additional
regulatory genes not found in most RRNPP systems also modulate signal production and signal import in the enterococcal pher-
omone plasmids. This review summarizes several structural studies that cumulatively demonstrate that the ability of three
pCF10 regulatory proteins to recognize the same 7-amino-acid pheromone peptide arose by convergent evolution of unrelated
proteins from different families. We also focus on the selective pressures and structure/function constraints that have driven the
evolution of pCF10 from a simple, single-peptide system resembling current RRNPPs in other bacteria to the current complex
inducible plasmid transfer system.

BACKGROUND AND SIGNIFICANCE
I n 1965, Tomasz (1) described “a new type of regulatory mecha-
nism in bacteria,” in which the control of competent cell genetic
transformation in pneumococci was expressed in a density-de-
pendent fashion (1). He reported that the culture medium of cells
grown to the optimal density for maximum competence con-
tained a soluble factor capable of inducing competence expression
when added to low-density noncompetent cultures. Conceptu-
ally, the phenomenon of density-dependent pneumococcal com-
petence expression mediated by intercellular signaling molecules
is very similar to the autoinduction of light production in marine
Vibrio species described a few years later by Nealson (2). These
seminal studies initiated a paradigm shift in microbial research,
changing the concept of normal bacterial behavior from single
cells acting independently to coordinated behaviors of microbial
populations via communication between individuals. Quorum
sensing, in which a single cell type monitors its population density
to coordinate activity (3), is perhaps the best studied mechanism
for the modulation of multicellular behaviors by intercellular sig-
naling, which is more broadly termed sociomicrobiology (4).
Enterococcus faecalis is a major cause of opportunistic infec-
tions of hospital patients, and E. faecalis clinical isolates are noto-
rious for their carriage of antibiotic resistance genes (5, 6). These
are frequently disseminated by conjugation. In 1978, Dunny et al.
(7) reported that donor/recipient clumping and conjugative
transfer of plasmids in Enterococcus (formerly Streptococcus)
faecalis could be induced by low-molecular-weight signaling mol-
ecules excreted by recipient cells and sensed by plasmid-contain-
ing donor cells; it was suggested that these signals served as bacte-
rial sex pheromones. A few years later, the Clewell et al. (8) and
Suzuki et al. (9) research groups reported the identification of
several different molecules that mediated signaling for various
plasmids; these signals were unmodified hydrophobic peptides 7
to 8 amino acid residues in length. These studies were the first
demonstrations that the prevalent extracellular signaling mole-
cules of Gram-positive bacteria were oligopeptides, in contrast
to the acyl-homoserine-lactone signals that frequently mediate
quorum sensing in Gram-negative microbes (10). Both the pep-
tide-mediated signaling mechanisms and the peptide signals
themselves fall into two categories. Some signals are secreted as
unmodified peptides processed from longer precursors, while
others are both processed and posttranslationally modified (11–
13). Likewise, sensing of peptide signals can involve either signal
transduction across the membrane or signal import, followed by
binding to a cytoplasmic receptor protein, which is often a tran-
scription factor (14).
The enterococcal sex pheromone systems function by import
of a signaling pheromone peptide encoded by the chromosome.
For simplicity, we use “C” as an abbreviation for all conjugation/
clumping-inducing peptide pheromones, where cCF10 is the pep-
tide that specifically induces cells carrying pCF10, cAD1 induces
Accepted manuscript posted online 28 March 2016
Citation Dunny GM, Berntsson RP-A. 2016. Enterococcal sex pheromones:
evolutionary pathways to complex, two-signal systems. J Bacteriol
198:1556 –1562. doi:10.1128/JB.00128-16.
Editor: W. Margolin, University of Texas Medical School at Houston
Address correspondence to Gary M. Dunny, dunny001@umn.edu.
Copyright © 2016, American Society for Microbiology. All Rights Reserved.

1556 jb.asm.org Journal of Bacteriology June 2016 Volume 198 Number 11

 Minireview

FIG 1 Diagram of the signaling circuits in the E. faecalis pCF10 conjugation system (adapted from Annual Review of Genetics [19]). Recipient and donor have
similar chromosomes, but the donor also carries pCF10. The plasmid confers a response to the chromosomally encoded peptide C, which induces conjugation.

Downloaded from http://jb.asm.org/ on March 7, 2017 by guest

The plasmid encodes the antagonistic peptide I, which inhibits C competitively. Two constitutively expressed pCF10 gene products, PrgZ and PrgY, function in
pheromone import and in reduction of the amount of active C excreted by plasmid-carrying cells, respectively, as detailed in the text. Imported C interacts with
PrgX (not shown) in the cytoplasm to induce a conjugation response. Pheromone induction of donor cells results in the synthesis of conjugation-related gene
products, including surface adhesin proteins, type 4 secretion proteins (T4SS), and DNA transfer proteins (DTR).

those carrying pAD1, etc. Mature C is processed by host-encoded
proteins, and all known members of the sex pheromone family are
processed from the cleaved signal peptides of secreted lipoproteins
(15, 16). Binding of imported C by its cytoplasmic receptor initi-
ates the pheromone response in the donor; the presence of C in the
growth medium of donor cells thus serves as a cue for the presence
of recipients (Fig. 1). Peptide binding modulates the ability of the
C receptor (PrgX in the case of pCF10) to regulate the transcrip-
tion of an operon containing conjugation genes (17). However,
the enterococcal sex pheromone systems have several additional
layers of complexity, including a second plasmid-encoded peptide
(inhibitor [I]) that competes directly with C for binding to the
same receptor (17, 18). In addition, several layers of posttranscrip-
tional regulation greatly amplify the direct effects of the peptides
on the expression of conjugation genes (17). The remainder of this
essay will focus on the tetracycline-resistant pheromone-respon-
sive plasmid pCF10 to illustrate the salient features of many sex
pheromone plasmids (19) and to explore how the current com-
plex systems may have evolved from simpler progenitor systems
similar to the peptide-regulated RRNPP signaling systems that
have now been implicated in the control of virulence, develop-
mental processes, and horizontal gene transfer in numerous
Gram-positive pathogens (20, 21).
OVERVIEW OF THE PEPTIDE-MEDIATED REGULATION OF
pCF10 CONJUGATION
Figure 2 depicts a simplified map of the pheromone-inducible
conjugation genes of pCF10 (22). The prgQ operon confers pro-
duction of ⬎30 polypeptides and regulatory RNAs required for
regulated expression of conjugation. The pheromone receptor

FIG 2 Genetic organization of pheromone-inducible conjugation genes found on enterococcal plasmids (approximate size of the entire region indicated at the
top). This map depicts the prg genes of pCF10 with single-letter designations, but similar gene content and organization are found on other well-studied plasmids,
such as pAD1 and pPD1 (17). The left portion of the map shows conserved genes involved in pheromone sensing, and the relative locations of the genes of the
pheromone-inducible prgQ operon encoding the I peptide, surface adhesin gene module (ABUC), downstream type IV secretion system (T4SS) genes, and
conjugative DNA transfer genes (Dtr) are shown. The prgQ gene encodes the production of I, whereas an ⬃1-kb segment between prgQ and prgA encodes two
small open reading frames (ORFs) and small RNAs (sRNAs) that regulate the expression of downstream genes posttranscriptionally (65). The sizes of the
individual genes are not drawn to scale. I, the putative origin of the system as a surface protein module negatively regulated by quorum sensing through the X/Q
cassette; this gene pair resembles RRNPP systems recently identified in numerous Gram-positive pathogens (21, 31). II shows how the system became more
complex as it acquired the ability to enable its host cell to recognize C as an indicator of close proximity of plasmid-free recipients (mate sensing). At the
mechanistic level, the C peptide competes with I, which functions as a classic quorum-sensing signal of donor density (self-sensing) (64). Evolution of the ability
to differentially respond to these two antagonistic peptides was accompanied by the acquisition of genes encoding an oligopeptide binding protein, PrgZ, which
binds both C and I with high affinity and increases their import via the Opp ABC transporter (37, 38), and PrgY, a predicted membrane peptidase that reduces
the production of endogenous C by the host cell (36). III depicts the acquisition of T4SS and Dtr genes conferring conjugative transfer ability. There is high
conservation of the regions indicated by I and II among many pheromone plasmids, suggesting that they all arose from a common ancestor, but step III likely
occurred multiple times to link different conjugation gene cassettes to the pheromone-inducible aggregation module.

June 2016 Volume 198 Number 11 Journal of Bacteriology jb.asm.org 1557

Minireview
PrgX controls the initiation of transcription of this long operon
from the prgQ promoter; the interaction of I with PrgX reduces
transcription, whereas the interaction of C with PrgX allows for
increased transcription. It is important to note that the direct ef-
fects of the peptides on control of the prgQ promoter by PrgX are
actually quite modest, but they are greatly amplified by several
posttranscriptional mechanisms, which are described elsewhere
(23–27). Determination of the structures of Apo-PrgX and of
PrgX bound to I or C, along with extensive genetic and biochem-
ical analyses, indicates that Apo-PrgX and PrgX-I complexes re-
press transcription from the prgQ promoter, while PrgX-C com-
plexes are impaired in repression (28, 29). It was originally
suggested that the replacement of I by C in PrgX-DNA complexes
disrupts PrgX tetramers within repressing complexes, allowing
RNA polymerase to access the prgQ promoter (28–30). Very re-
cent data (Y. Chen, A. Bandyopadhyay, B. K. Kozlowicz, H. A. H.
Haemig, A. Tai, W.-S. Hu, and G. Dunny, unpublished data) sug-
gest that PrgX forms tetramers when complexed with either pep-
tide, but conformational differences of the DNA-bound tetramers
account for differential repression. In both models, the ultimate
induction state of a donor cell is dependent on the relative intra-
cellular levels of I and C in donor cells. Interestingly, all of the
peptide-controlled transcription factors of the RRNPP family ap-
pear to have a structure that is very similar to that of PrgX (20, 21,
31), and in most cases, the gene organization of the determinants
for regulatory protein and the cognate regulatory peptide is simi-
lar to that of prgX and prgQ. Below, we focus on the evolutionary
processes that likely shaped the emergence of the dual-peptide-
controlled pCF10 system and how it may have evolved from a
simple RRNPP-like system to its present complex state.
HOW AND WHY DID THE pCF10 SYSTEM BECOME SO
COMPLEX?
The key functional components of the pCF10 system, which are
also found in other pheromone plasmids (17, 32), are illustrated in
Fig. 2. It is likely that current pheromone-inducible conjugation
systems originated from a system with a single I-regulated Q-X-
like module. This module likely controlled the expression of ad-
jacent genes for surface adhesins, as similar surface adhesion gene
content and organization are conserved with other pheromone-
controlled systems (17). Contemporary pheromone plasmids may
have a common ancestor that includes contiguous genes corre-
sponding to prgZ through prgQ and extending through the down-
stream cassette of LPXTG-anchored cell surface protein genes
(prgABC) and the small regulatory prgU gene (22, 33); this gene
cluster is indicated by the Roman numeral II in Fig. 2. Prior1 to the
acquisition of the ability to recognize the peptide signal C, the
I-autoregulated X-Q surface protein cluster may have functioned
to increase the ability of the host bacterium to attach to other
bacterial or metazoan host cells at low density while reducing
these interactions at high bacterial density to enable escape from
stagnant communities and recolonization of new niches.
The next major event in the evolution of the system was prob-
ably the ability to recognize the host-encoded C peptide as an
indicator of the presence of plasmid-free enterococci in close
proximity. In strict evolutionary biology parlance (34, 35), C
would be classified as a cue rather than a signal, since the sensing
system seems to have hijacked this molecule produced from a gene
not linked to the sensing genes in physical proximity or in func-
tion. In the case of pCF10, the C peptide is produced by processing
1558 jb.asm.org Journal of Bacteriology
the cleaved signal peptide of a predicted secreted lipoprotein
CcfA, whose function has not been demonstrated (15); likewise,
all known pheromone-responsive plasmids analyzed to date en-
code a response to a specific peptide encoded by one of the ⬎50
potential lipoprotein genes in the organism (16). As indicated in
step II, the system acquired additional components that recognize
C; PrgY prevents self-induction of donors by decreasing the
amount of mature C released (36), and PrgZ binds both C and I
and facilitates their import into the cell via a chromosomally en-
coded peptide transporter (37, 38). PrgX also needed to evolve to
recognize C and I. These 3 proteins are all from different families
and share only 9 to 13% sequence identity and no significant ho-
mology at the structural level. We have structural data on the
interactions of PrgZ with C (37) and of PrgX with both C and I (28,
29), but to date, there are no structural data available on PrgY.
PrgZ belongs to the family of substrate-binding proteins found
Downloaded from http://jb.asm.org/ on March 7, 2017 by guest
in ABC transporters, G protein-coupled receptors (GPCRs), and
DNA binding proteins (39, 40). It is likely that PrgZ evolved from
a chromosomal oligopeptide-binding protein. Previous experi-
ments have shown that the oligopeptide-binding protein OppA of
E. faecalis can facilitate the import of C, even though a higher
concentration of C is required than is produced by recipients un-
der normal physiological conditions (38). PrgZ can bind both C
and I and has a typical Venus flytrap fold of a cluster C substrate-
binding protein (37), with C bound within an internal cavity (Fig.
3a). C is firmly bound to PrgZ via 10 direct hydrogen bonds. These
bonds are mostly formed with the peptide backbone, with one
exception, an H-bond that is formed with the side chain of Thr3,
giving an explanation for results from genetic screens that Thr3 of
C was important for PrgZ binding (41). Further H-bonds between
C and PrgZ are formed via bridging water molecules, and there is
also a salt bridge that anchors the N terminus of C. Although no
structure of PrgZ complexed with I is available, it is highly likely
that I binds in the same way as C (37). This is expected due to the
similarities of PrgZ to other oligopeptide-binding proteins (37,
40, 42, 43).
As noted previously, PrgX serves as the primary cytoplasmic
receptor for both peptides and acts as the master transcription
regulator of the prgQ operon. PrgX has a conserved helix-turn-
helix (HTH) domain responsible for its interaction with DNA (28,
29). The binding site for C and I is located in the larger dimeriza-
tion domain (Fig. 3b and c), where both peptides form ␤-sheet-
like interactions with PrgX and have a similar amount of H-bonds
to PrgX. There is no conservation in the binding site between PrgX
and PrgZ at either the sequence level or structurally. The two
peptides favor different conformational states of PrgX, with the
consequence that one alpha-helix is preferentially stabilized (I
bound) or unwound (C bound) (28, 29). These different confor-
mations either favor transcriptional repression when inhibitor is
bound, or they reduce the level of repression of PrgX when C is
bound. While the basis for the differential effects of the two pep-
tides on the conformation of PrgX is not completely understood,
it probably relates to the N-terminal regions of the two peptides.
The bulky leucine residue at the N terminus of C likely crowds the
surrounding PrgX residues in the binding pocket to a greater ex-
tent than the alanine at the N terminus of I (for a partial illustra-
tion, compare the PrgX residues surrounding the bound peptide
N termini in Fig. 3b and c). These differences may indirectly affect
the conformation of the C terminus of PrgX.
PrgY is required to prevent cells carrying pCF10 from being
June 2016 Volume 198 Number 11
 Minireview

Downloaded from http://jb.asm.org/ on March 7, 2017 by guest

FIG 3 Comparison of the peptide binding of PrgZ (gray) complexed with C (a) and PrgX (dark red) complexed with C (b) or I (c). Each upper subfigure shows
the full protein structure in a cartoon representation, with the bound ligand in spheres, C in teal, and I in green. The lower enlarged representations at the bottom
show the ligand (as sticks) with interacting protein residues. As can be seen by comparing panels b and c, the C peptide has a very different structure when bound
to PrgZ compared with its structure when bound to PrgX.

self-induced by their own endogenous pheromone (36). Its amino stream genes predicted to encode the type 4 secretion systems
acid sequence, in conjunction with genetic and biochemical stud- (T4SSs) and DNA transfer (Dtr) machinery required for conjuga-
ies, suggest that a C-terminal subdomain anchors the protein in tion are located immediately downstream from the surface pro-
the membrane with the N-terminal region outside the cell; the tein cassettes in other pheromone plasmids, but these T4SS loci
external N-terminal subdomain confers the ability to specifically
bind the mature C peptide (36, 44) and may contribute to its
degradation. Initial studies of PrgY suggested that similar pro-
teins, none with known functions, were present in organisms from
all kingdoms, and that the protein phylogenies correlated with
those of the host organisms (36). Recently, an important new
study provided new insights into the structure/function relation-
ships of these proteins. Zhang et.al. (45) identified Tiki as a pro-
tease family playing a critical role in cell growth and development
via specific cleavage of the Wnt protein (45). PrgY is homologous
to the human Tiki metalloprotease, both having a pair of GX2H
motifs and a conserved glutamate residue, and it is predicted to
have structural similarity to the so-called EraA/ChaN-like family
of proteins (46). The structure of PrgY has not been determined,
but structural modeling using Phyre2 gives a model with a 96%
confidence level over most of the extracellular domain (Fig. 4).
This model does not contain any structural motif that resembles
the pheromone-binding site of either PrgZ or PrgX. From the
homology to the Tiki metalloproteases, we can deduce which res-
idues likely form the active site in PrgY, with some of those specific
residues, like His21, having previously been verified to be impor-
tant for function (36, 44). To date, only PrgY and Tiki are known
to have specific interactions with polypeptide substrates.
The cumulative analysis suggests that the pCF10 system did
not independently evolve these 3 different components from a FIG 4 Predicted structure of PrgY. The extracellular part of PrgY, here shown
single protein with a peptide-binding motif. More likely, an an- as a cartoon representation, was modeled using Phyre2 and colored from the N
cestral system, i.e., the inhibitor-regulated Q-X module, at some terminus (blue) toward the C-terminal end of the model (yellow). The C-ter-
minal domain, which could not be modeled, is predicted to contain 4 trans-
point acquired genes that coded for the early versions of PrgY and membrane helices, shown here as rectangles in a membrane. The predicted
PrgZ, and those proteins then evolved specific binding affinity to active site, based on the homology of PrgY to the Tiki metalloproteases (46), is
the cognate C and I peptides, as illustrated in Fig. 2. The down- shown within the dashed line.

June 2016 Volume 198 Number 11 Journal of Bacteriology jb.asm.org 1559

Minireview
show considerable divergence (22). This suggests that phero-
mone-inducible aggregation cassettes became linked to the addi-
tional components required for conjugation on multiple occa-
sions (Fig. 2, step III). Interestingly, the available data suggest that
the downstream conjugation functions for all known plasmids are
transcriptionally regulated by the peptide signals even though they
became linked to the upstream regions in multiple events (22).
REMAINING QUESTIONS AND FUTURE DIRECTIONS
While significant questions about the molecular mechanisms of
pheromone-mediated control of conjugation remain, the most
compelling areas for future study may be the analysis of structure/
function relationships of the key regulatory components and in-
vestigations of how these systems function in the natural environ-
ment, including their impacts on maintenance and dissemination
of the plasmid itself, and on the fitness of the bacterial hosts. The
significance of such studies is heightened by the fact that consid-
erable experimental and theoretical investigations of the evolu-
tionary aspects of the well-studied acyl-homoserine-lactone auto-
inducer systems in Gram-negative alphaproteobacteria have been
carried out (reviewed in reference 47). For example, a recent study
by Cornforth et al. (48) provided evidence that combining two
separate quorum-sensing systems allowed for greater resolution
of the local environment of a bacterium, both in terms of sensing
social (microbial cell population density) and physical (diffusion,
etc.) parameters. In contrast, much less attention has been di-
rected toward the naturally occurring and more complex intercel-
lular communication system represented by pCF10. While pCF10
was discovered because of its role in the transmission of antibiotic
resistance (49), enterococcal pheromone-responsive plasmids fre-
quently do not carry resistance genes (50, 51), suggesting that they
may encode other traits that increase the fitness of their host cells,
relative to the costs of plasmid maintenance. It is interesting to
consider how the expression of the pheromone-inducible conju-
gation genes may impact host fitness and how this relates to the
regulatory properties of the system. Published data from our
group (52–57) and others (58–61) suggest that the expression of
aggregation substance proteins, such as PrgB, can increase colo-
nization and virulence by promoting biofilm formation and at-
tachment to host tissues, and by increasing resistance to phago-
cytic killing. Notably, there are still no direct data on how PrgB or
other inducible proteins might impact fitness in the gut. On the
other hand, the overexpression of these genes likely has very high
costs for the induced cell, including the energy required for syn-
thesizing conjugation proteins, the likely inhibition of growth in
cells trapped in large aggregates, and cell death and lysis due to
toxic effects of overexpressed gene products on highly induced
cells (62). Interestingly, clusters of genes related to the plasmid-
encoded pheromone-inducible adhesins/transfer determinants
have been identified within genomic islands in the chromosomes
of some strains, but these chromosomal determinants are not ca-
pable of transfer unless a coresident pheromone plasmid inte-
grates and mobilizes them via an Hfr-like mechanism (63).
Numerous studies have documented the extremely tight regu-
lation of the pheromone system (19). The system not only avoids
spurious induction but also limits the duration of induction due
to the fact that the induction process itself dramatically increases
inhibitor production, leading to rapid shut off of the response
after a short period of induction (64). Furthermore, the inhibitor
can function as a classic quorum sensor of donor density; at high
1560 jb.asm.org Journal of Bacteriology
donor densities, donors are poorly induced even by high concen-
trations of C (64). These cumulative effects of the inhibitor appar-
ently limit the extent of induction in mixed populations of donors
and recipients. This raises the question of whether the system may
have evolved to maintain mixed populations of donors and recip-
ients in shared niches in the natural environment of the bacteria,
e.g., the intestinal tract. The maintenance of recipient populations
by limiting their conversion to donors should result in a steady
supply of C within the niche, whose inducing capacity is limited by
the inhibitor. In this scenario, basal levels of expression of the
inducible genes could be maintained within the mixed popula-
tion, providing the previously described benefits (note that induc-
tion of a few donors can coaggregate recipients and uninduced
donors in close proximity) while minimizing costs of overexpres-
sion. The pheromone system has thus evolved under strong con-
flicting selective pressures for an extremely sensitive detection sys-
Downloaded from http://jb.asm.org/ on March 7, 2017 by guest
tem to induce expression while simultaneously limiting the extent
and duration of induction. This may have driven convergent evo-
lution of the three unrelated proteins with vital, but distinct, reg-
ulatory functions to recognize the same peptide. Direct experi-
mental testing of these ideas in the mammalian gastrointestinal
(GI) tract, along with further mechanistic and structural studies of
regulatory components, is in progress and might yield insights
into more effective approaches to reduce the spread of antibiotic
resistance and to impair the ability of resistant strains to overgrow
and disrupt the gut microbiota of hospital patients.
ACKNOWLEDGMENTS
The research for this review was supported by U.S. PHS grant GM49530 to
G.M.D. and by the Kempe Foundations grant JCK-1524 to R.P.-A.B.
FUNDING INFORMATION
This work, including the efforts of Gary M. Dunny, was funded by HHS |
National Institutes of Health (NIH) (GM49530). This work, including the
efforts of Ronnie Per-Arne Berntsson, was funded by Kempestiftelserna
(Kempe Foundations) (JCK-1524).
REFERENCES
1. Tomasz A. 1965. Control of the competent state in Pneumococcus by a
hormone-like cell product: an example for a new type of regulatory
mechanism in bacteria. Nature 208:155–159. http://dx.doi.org/10.1038
/208155a0.
2. Nealson KH. 1999. Early observations defining quorum-dependent gene
expression, p 277–290. In Dunny GM, Winans SC (ed), Cell-cell signaling
in bacteria. ASM Press, Washington, DC.
3. Fuqua WC, Winans SC, Greenberg EP. 1994. Quorum sensing in bac-
teria: the LuxR-LuxI family of cell density-responsive transcriptional reg-
ulators. J Bacteriol 176:269 –275.
4. Parsek MR, Greenberg EP. 2005. Sociomicrobiology: the connections
between quorum sensing and biofilms. Trends Microbiol 13:27–33. http:
//dx.doi.org/10.1016/j.tim.2004.11.007.
5. Kristich CJ, Rice LB, Arias CA. 6 February 2014. Enterococcal infec-
tion—treatment and antibiotic resistance. In Gilmore MS, Clewell DB, Ike
Y, Shankar N (ed), Enterococci: from commensals to leading causes of
drug resistant infection. Massachusetts Eye & Ear Infirmary, Boston, MA.
http://www.ncbi.nlm.nih.gov/books/NBK190420/.
6. Miller WR, Munita JM, Arias CA. 2014. Mechanisms of antibiotic resis-
tance in enterococci. Expert Rev Anti Infect Ther 12:1221–1236. http://dx
.doi.org/10.1586/14787210.2014.956092.
7. Dunny GM, Brown BL, Clewell DB. 1978. Induced cell aggregation and
mating in Streptococcus faecalis: evidence for a bacterial sex pheromone.
Proc Natl Acad Sci USA 75:3479 –3483. http://dx.doi.org/10.1073/pnas.75
.7.3479.
8. Mori M, Sakagami Y, Narita M, Isogai A, Fujino M, Kitada C, Craig RA,
Clewell DB, Suzuki A. 1984. Isolation and structure of the bacterial sex
June 2016 Volume 198 Number 11

pheromone, cAD1, that induces plasmid transfer in Streptococcus faecalis.
FEBS Lett 178:97–100. http://dx.doi.org/10.1016/0014-5793(84)81248-X.
9. Suzuki A, Mori M, Sakagami Y, Isogai A, Fujino M, Kitaga C, Craig RA,
Clewell DB. 1984. Isolation and structure of bacterial sex pheromone,
cPD1. Science 226:849 – 850. http://dx.doi.org/10.1126/science.6436978.
10. Fuqua C, Parsek MR, Greenberg EP. 2001. Regulation of gene expression
by cell-to-cell communication: acyl-homoserine lactone quorum sensing.
Annu Rev Genet 35:439 – 468. http://dx.doi.org/10.1146/annurev.genet
.35.102401.090913.
11. Chandler JR, Dunny GM. 2004. Enterococcal peptide sex pheromones:
synthesis and control of biological activity. Peptides 25:1377–1388. http:
//dx.doi.org/10.1016/j.peptides.2003.10.020.
12. Dunny GM, Leonard BAB. 1997. Cell-cell communication in Gram-
positive bacteria. Annu Rev Microbiol 51:527–564.
13. Dunny GM, Winans SC. 1999. Cell-cell signaling in bacteria. ASM Press,
Washington, DC.
14. Waters CM, Bassler BL. 2005. Quorum sensing: cell-to-cell communica-
tion in bacteria. Annu Rev Cell Dev Biol 21:319 –346. http://dx.doi.org/10
.1146/annurev.cellbio.21.012704.131001.
15. Antiporta MH, Dunny GM. 2002. ccfA, the genetic determinant for the
cCF10 peptide pheromone in Enterococcus faecalis OG1RF. J Bacteriol
184:1155–1162. http://dx.doi.org/10.1128/jb.184.4.1155-1162.2002.
16. Clewell DB, An FY, Flannagan SE, Antiporta M, Dunny GM. 2000.
Enterococcal sex pheromone precursors are part of signal sequences for
surface lipoproteins. Mol Microbiol 35:246 –248. http://dx.doi.org/10
.1046/j.1365-2958.2000.01687.x.
17. Clewell DB, Weaver KE, Dunny GM, Coque TM, Francia MV, Hayes F.
9 February 2014. Extrachromosomal and mobile elements in enterococci:
transmission, maintenance, and epidemiology. In Gilmore MS, Clewell
DB, Ike Y, Shankar N (ed), Enterococci: from commensals to leading
causes of drug resistant infection. Massachusetts Eye & Ear Infirmary,
Boston, MA. http://www.ncbi.nlm.nih.gov/books/NBK190430/.
18. Nakayama J, Ruhfel RE, Dunny GM, Isogai A, Suzuki A. 1994. The prgQ
gene of the Enterococcus faecalis tetracycline resistance plasmid pCF10
encodes a peptide inhibitor, iCF10. J Bacteriol 176:7405–7408.
19. Dunny GM. 2013. Enterococcal sex pheromones: signaling, social behav-
ior, and evolution. Annu Rev Genet 47:457– 482. http://dx.doi.org/10
.1146/annurev-genet-111212-133449.
20. Cook LC, Federle MJ. 2014. Peptide pheromone signaling in Streptococ-
cus and Enterococcus. FEMS Microbiol Rev 38:473– 492. http://dx.doi.org
/10.1111/1574-6976.12046.
21. Parashar V, Aggarwal C, Federle MJ, Neiditch MB. 2015. Rgg protein
structure-function and inhibition by cyclic peptide compounds. Proc
Natl Acad Sci U S A 112:5177–5182. http://dx.doi.org/10.1073/pnas
.1500357112.
22. Hirt H, Manias DA, Bryan EM, Klein JR, Marklund JK, Staddon JH,
Paustian ML, Kapur V, Dunny GM. 2005. Characterization of the pher-
omone response of the Enterococcus faecalis conjugative plasmid pCF10:
complete sequence and comparative analysis of the transcriptional and
phenotypic responses of pCF10-containing cells to pheromone induction.
J Bacteriol 187:1044 –1054. http://dx.doi.org/10.1128/JB.187.3.1044-1054
.2005.
23. Bae T, Kozlowicz BK, Dunny GM. 2004. Characterization of cis-acting
prgQ mutants: evidence for two distinct repression mechanisms by Qa
RNA and PrgX protein in pheromone-inducible enterococcal plasmid
pCF10. Mol Microbiol 51:271–281.
24. Chatterjee A, Johnson CM, Shu CC, Kaznessis YN, Ramkrishna D,
Dunny GM, Hu WS. 2011. Convergent transcription confers a bistable
switch in Enterococcus faecalis conjugation. Proc Natl Acad Sci U S A
108:9721–9726. http://dx.doi.org/10.1073/pnas.1101569108.
25. Dunny GM, Johnson CM. 2011. Regulatory circuits controlling entero-
coccal conjugation: lessons for functional genomics. Curr Opin Microbiol
14:174 –180. http://dx.doi.org/10.1016/j.mib.2011.01.008.
26. Johnson CM, Haemig HH, Chatterjee A, Wei-Shou H, Weaver KE,
Dunny GM. 2011. RNA-mediated reciprocal regulation between two bac-
terial operons is RNase III dependent. mBio 2(5):e00189-11. http://dx.doi
.org/10.1128/mBio.00189-11.
27. Johnson CM, Manias DA, Haemig HA, Shokeen S, Weaver KE, Henkin
TM, Dunny GM. 2010. Direct evidence for control of the pheromone-
inducible prgQ operon of Enterococcus faecalis plasmid pCF10 by a coun-
tertranscript-driven attenuation mechanism. J Bacteriol 192:1634 –1642.
http://dx.doi.org/10.1128/JB.01525-09.
28. Kozlowicz BK, Shi K, Gu ZY, Ohlendorf DH, Earhart CA, Dunny GM.
June 2016 Volume 198 Number 11 Journal of Bacteriology
Minireview
2006. Molecular basis for control of conjugation by bacterial pheromone
and inhibitor peptides. Mol Microbiol 62:958 –969. http://dx.doi.org/10
.1111/j.1365-2958.2006.05434.x.
29. Shi K, Brown CK, Gu ZY, Kozlowicz BK, Dunny GM, Ohlendorf DH,
Earhart CA. 2005. Structure of peptide sex pheromone receptor PrgX and
PrgX/pheromone complexes and regulation of conjugation in Enterococ-
cus faecalis. Proc Natl Acad Sci U S A 102:18596 –18601. http://dx.doi.org
/10.1073/pnas.0506163102.
30. Kozlowicz BK. 2005. The molecular mechanism and peptide signaling
response of PrgX used to control pheromone-induced conjugative
transfer of pCF10. Ph.D. dissertation. University of Minnesota, Min-
neapolis, MN.
31. Declerck N, Bouillaut L, Chaix D, Rugani N, Slamti L, Hoh F, Lereclus D,
Arold ST. 2007. Structure of PlcR: insights into virulence regulation and
evolution of quorum sensing in Gram-positive bacteria. Proc Natl Acad Sci
U S A 104:18490 –18495. http://dx.doi.org/10.1073/pnas.0704501104.
32. Clewell DB, Dunny GM. 2002. Conjugation and genetic exchange in
enterococci, p 265–300. In Gilmore MS, Clewell DB, Courvalin P, Dunny
GM, Murray BE, Rice LB (ed), The enterococci: pathogenesis, molecular
biology and antibiotic resistance. ASM Press, Washington, DC.
Downloaded from http://jb.asm.org/ on March 7, 2017 by guest
33. Kozlowicz BK, Dworkin M, Dunny GM. 2006. Pheromone-inducible
conjugation in Enterococcus faecalis: a model for the evolution of biological
complexity? Int J Med Microbiol 296:141–147. http://dx.doi.org/10.1016
/j.ijmm.2006.01.040.
34. Platt TG, Fuqua C. 2010. What’s in a name? The semantics of quorum
sensing. Trends Microbiol 18:383–387.
35. Keller L, Surette MG. 2006. Communication in bacteria: an ecological
and evolutionary perspective. Nat Rev Microbiol 4:249 –258. http://dx.doi
.org/10.1038/nrmicro1383.
36. Chandler JR, Flynn AR, Bryan EM, Dunny GM. 2005. Specific control of
endogenous cCF10 pheromone by a conserved domain of the pCF10-
encoded regulatory protein PrgY in Enterococcus faecalis. J Bacteriol 187:
4830 – 4843. http://dx.doi.org/10.1128/JB.187.14.4830-4843.2005.
37. Berntsson RP, Schuurman-Wolters GK, Dunny G, Slotboom DJ, Pool-
man B. 2012. Structure and mode of peptide binding of pheromone re-
ceptor PrgZ. J Biol Chem 287:37165–37170. http://dx.doi.org/10.1074/jbc
.M112.386334.
38. Leonard BA, Podbielski A, Hedberg PJ, Dunny GM. 1996. Enterococcus
faecalis pheromone binding protein, PrgZ, recruits a chromosomal oligo-
peptide permease system to import sex pheromone cCF10 for induction of
conjugation. Proc Natl Acad Sci U S A 93:260 –264. http://dx.doi.org/10
.1073/pnas.93.1.260.
39. Berntsson RP, Doeven MK, Fusetti F, Duurkens RH, Sengupta D,
Marrink SJ, Thunnissen AM, Poolman B, Slotbloom DJ. 2009. The
structural basis for peptide selection by the transport receptor OppA.
EMBO J 28:1332–1340. http://dx.doi.org/10.1038/emboj.2009.65.
40. Berntsson RP, Smits SH, Schmitt L, Slotbloom DJ, Poolman B. 2010. A
structural classification of substrate-binding proteins. FEBS Lett 584:
2606 –2617. http://dx.doi.org/10.1016/j.febslet.2010.04.043.
41. Fixen KR, Chandler JR, Le T, Kozlowicz BK, Manias DA, Dunny GM.
2007. Analysis of the amino acid sequence specificity determinants of the
enterococcal cCF10 sex pheromone in interactions with the pheromone-
sensing machinery. J Bacteriol 189:1399 –1406. http://dx.doi.org/10.1128
/JB.01226-06.
42. Levdikov VM, Blagova EV, Brannigan JA, Wright L, Vagin AA, Wilkin-
son AJ. 2005. The structure of the oligopeptide-binding protein, AppA,
from Bacillus subtilis in complex with a nonapeptide. J Mol Biol 345:879 –
892. http://dx.doi.org/10.1016/j.jmb.2004.10.089.
43. Tame JR, Murshudov GN, Dodson EJ, Neil TK, Dodson GG, Higgins
CF, Wilkinson AJ. 1994. The structural basis of sequence-independent
peptide binding by OppA protein. Science 264:1578 –1581. http://dx.doi
.org/10.1126/science.8202710.
44. Chandler JR, Dunny GM. 2008. Characterization of the sequence speci-
ficity determinants required for processing and control of sex pheromone
by the intramembrane protease Eep and the plasmid-encoded protein
PrgY. J Bacteriol 190:1172–1183. http://dx.doi.org/10.1128/JB.01327-07.
45. Zhang X, Abreu JG, Yokota C, MacDonald BT, Singh S, Coburn KL,
Cheong SM, Zhang MM, Ye QZ, Hang HC, Steen H, He X. 2012. Tiki1
is required for head formation via Wnt cleavage-oxidation and inactiva-
tion. Cell 149:1565–1577. http://dx.doi.org/10.1016/j.cell.2012.04.039.
46. Bazan JF, Macdonald BT, He X. 2013. The TIKI/TraB/PrgY family: a
common protease fold for cell signaling from bacteria to metazoa? Dev
Cell 25:225–227. http://dx.doi.org/10.1016/j.devcel.2013.04.019.
jb.asm.org 1561
Minireview
47. Stevens AM, Schuster M, Rumbaugh KP. 2012. Working together for the
common good: cell-cell communication in bacteria. J Bacteriol 194:2131–
2141. http://dx.doi.org/10.1128/JB.00143-12.
48. Cornforth DM, Popat R, McNally L, Gurney J, Scott-Phillips TC, Ivens
A, Diggle SP, Brown SP. 2014. Combinatorial quorum sensing allows
bacteria to resolve their social and physical environment. Proc Natl Acad
Sci U S A 111:4280 – 4284. http://dx.doi.org/10.1073/pnas.1319175111.
49. Dunny G, Funk C, Adsit J. 1981. Direct stimulation of the transfer of
antibiotic resistance by sex pheromones in Streptococcus faecalis. Plasmid
6:270 –278. http://dx.doi.org/10.1016/0147-619X(81)90035-4.
50. Clewell DB. 1993. Bacterial sex pheromone-induced plasmid transfer.
Cell 73:9 –12. http://dx.doi.org/10.1016/0092-8674(93)90153-H.
51. Clewell DB. 1999. Sex pheromone systems in enterococci, p 47– 66. In
Dunny GM, Winans SC (ed), Cell-cell signaling in bacteria. ASM Press,
Washington, DC.
52. Chandler JR, Hirt H, Dunny GM. 2005. A paracrine peptide sex phero-
mone also acts as an autocrine signal to induce plasmid transfer and vir-
ulence factor expression in vivo. Proc Natl Acad Sci U S A 102:15617–
15622. http://dx.doi.org/10.1073/pnas.0505545102.
53. Chuang ON, Schlievert PM, Wells CL, Manias DA, Tripp TJ, Dunny
GM. 2009. Multiple functional domains of Enterococcus faecalis aggrega-
tion substance Asc10 contribute to endocarditis virulence. Infect Immun
77:539 –548. http://dx.doi.org/10.1128/IAI.01034-08.
54. Chuang-Smith ON, Wells CL, Henry-Stanley MJ, Dunny GM. 2010.
Acceleration of Enterococcus faecalis biofilm formation by aggregation
substance expression in an ex vivo model of cardiac valve colonization.
PLoS One 5:e15798. http://dx.doi.org/10.1371/journal.pone.0015798.
55. Hirt H, Erlandsen SL, Dunny GM. 2000. Heterologous inducible expres-
sion of Enterococcus faecalis pCF10 aggregation substance Asc10 in Lacto-
coccus lactis and Streptococcus gordonii demonstrates contribution to cell
hydrophobicity and adhesion to fibrin. J Bacteriol 182:2299 –2306. http:
//dx.doi.org/10.1128/JB.182.8.2299-2306.2000.
56. Olmsted SB, Dunny GM, Erlandsen SL, Wells CL. 1994. A plasmid-
encoded surface protein on Enterococcus faecalis augments its internaliza-
tion by cultured intestinal epithelial cells. J Infect Dis 170:1549 –1556.
57. Schlievert PM, Gahr PJ, Assimacopoulos AP, Dinges MM, Stoehr JA,
Harmala JW, Hirt H, Dunny GM. 1998. Aggregation and binding sub-
Gary M. Dunny received his B.S. and Ph.D.
from the University of Michigan and spent 11
years at Cornell University as a postdoctoral fel-
low and as a faculty member before moving to
the University of Minnesota in 1989, where he is
currently professor of microbiology. He has
studied conjugation, cell signaling, and adapta-
tion in enterococci using genetics, biochemis-
try, and microscopic imaging for his entire
career.
1562 jb.asm.org Journal of Bacteriology
stances enhance pathogenicity in rabbit models of Enterococcus faecalis
endocarditis. Infect Immun 66:218 –223.
58. Kreft B, Marre R, Schramm U, Wirth R. 1992. Aggregation substance of
Enterococcus faecalis mediates adhesion to cultured renal tubular cells.
Infect Immun 60:25–30.
59. Süssmuth SD, Muscholl-Silberhorn A, Wirth R, Susa M, Marre R,
Rozdzinski E. 2000. Aggregation substance promotes adherence, phago-
cytosis, and intracellular survival of Enterococcus faecalis within human
macrophages and suppresses respiratory burst. Infect Immun 68:4900 –
4906. http://dx.doi.org/10.1128/IAI.68.9.4900-4906.2000.
60. Chow JW, Thal LA, Perri MB, Vazquez JA, Donabedian SM, Clewell
DB, Zervos MJ. 1993. Plasmid-associated hemolysin and aggregation
substance production contribute to virulence in experimental enterococ-
cal endocarditis. Antimicrob Agents Chemother 37:2474 –2477. http://dx
.doi.org/10.1128/AAC.37.11.2474.
61. Rakita RM, Vanek NN, Jacques-Palaz K, Mee M, Mariscalco MM,
Dunny GM, Snuggs M, van Winkle WB, Simon SI. 1999. Enterococcus
faecalis bearing aggregation substance is resistant to killing by human neu-
trophils despite phagocytosis and neutrophil activation. Infect Immun
67:6067– 6075.
Downloaded from http://jb.asm.org/ on March 7, 2017 by guest
62. Bhatty M, Cruz MR, Frank KL, Gomez JA, Andrade F, Garsin DA,
Dunny GM, Kaplan HB, Christie PJ. 2015. Enterococcus faecalis pCF10-
encoded surface proteins PrgA, PrgB (aggregation substance) and PrgC
contribute to plasmid transfer, biofilm formation and virulence. Mol Mi-
crobiol 95:660 – 677. http://dx.doi.org/10.1111/mmi.12893.
63. Manson JM, Hancock LE, Gilmore MS. 2010. Mechanism of chromo-
somal transfer of Enterococcus faecalis pathogenicity island, capsule, anti-
microbial resistance, and other traits. Proc Natl Acad Sci U S A 107:
12269 –12274. http://dx.doi.org/10.1073/pnas.1000139107.
64. Chatterjee A, Cook LCC, Shu C-C, Chen Y, Manias DA, Ramkrishna D,
Dunny GM, Hu W-S. 2013. Antagonistic self-sensing and mate-sensing
signaling controls antibiotic-resistance transfer. Proc Natl Acad Sci U S A
110:7086 –7090. http://dx.doi.org/10.1073/pnas.1212256110.
65. Bensing BA, Manias DA, Dunny GM. 1997. Pheromone cCF10 and
plasmid pCF10-encoded regulatory molecules act post-transcriptionally
to activate expression of downstream conjugation functions. Mol Micro-
biol 24:285–294. http://dx.doi.org/10.1046/j.1365-2958.1997.3301710.x.
Ronnie Per-Arne Berntsson studied biotech-
nology at Chalmers University in Gothenburg,
Sweden. In 2010, he received his Ph.D. in bio-
chemistry from the University of Groningen,
the Netherlands, after working in the groups of
Bert Poolman and Dirk-Jan Slotboom on stud-
ies of ABC transporters and their domains. Af-
ter his Ph.D., he moved to Stockholm Univer-
sity, Sweden, where he received an EMBO
fellowship to do postdoctoral research in the
group of Pål Stenmark on botulinum neurotox-
ins and their receptors. In 2015, he became an assistant professor at the
Department of Medical Biochemistry and Biophysics at Umeå University,
Sweden. His laboratory studies the function, structure, and regulation of
type 4 secretion systems in Gram-positive bacteria.
June 2016 Volume 198 Number 11
 ANRV260-GE39-18 ARI 15 October 2005 13:5

Switches in Bacteriophage
Lambda Development∗
Amos B. Oppenheim,1,3,4 Oren Kobiler,1
Joel Stavans,2 Donald L. Court,3
Annu. Rev. Genet. 2005.39:409-429. Downloaded from www.annualreviews.org
by University of California - San Diego on 03/29/11. For personal use only.

and Sankar Adhya4

1
Department of Molecular Genetics and Biotechnology, The Hebrew
University-Hadassah Medical School, Jerusalem 91120
2
Department of Physics of Complex Systems, Weizmann Institute of Science,
Rehovot, Israel 76100
3
Gene Regulation and Chromosome Biology Laboratory, National Cancer Institute at
Frederick, Frederick, Maryland 21702
4
Laboratory of Molecular Biology, National Cancer Institute, Bethesda, Maryland
20892-4264; email: oppenhea@mail.nih.gov

Annu. Rev. Genet. Key Words

2005. 39:409–29
genetic circuits, gene modules, developmental pathways,
First published online as a
Review in Advance on developmental decisions, threshold effects, host factors
July 26, 2005
Abstract
The Annual Review of
Genetics is online at The lysis-lysogeny decision of bacteriophage lambda (λ) is a
http://genet.annualreviews.org paradigm for developmental genetic networks. There are three key
doi: 10.1146/ features, which characterize the network. First, after infection of
annurev.genet.39.073003.113656 the host bacterium, a decision between lytic or lysogenic develop-
Copyright

c 2005 by ment is made that is dependent upon environmental signals and the
Annual Reviews. All rights number of infecting phages per cell. Second, the lysogenic prophage
reserved
state is very stable. Third, the prophage enters lytic development in
∗
The U.S. Government response to DNA-damaging agents. The CI and Cro regulators de-
has the right to retain a
nonexclusive, royalty-free fine the lysogenic and lytic states, respectively, as a bistable genetic
license in and to any switch. Whereas CI maintains a stable lysogenic state, recent studies
copyright covering this indicate that Cro sets the lytic course not by directly blocking CI
paper.
expression but indirectly by lowering levels of CII which activates cI
0066-4197/05/1215- transcription. We discuss how a relatively simple phage like λ employs
0409$20.00
a complex genetic network in decision-making processes, providing
a challenge for theoretical modeling.

409
 ANRV260-GE39-18
Annu. Rev. Genet. 2005.39:409-429. Downloaded from www.annualreviews.org
by University of California - San Diego on 03/29/11. For personal use only.
ARI 15 October 2005 13:5
Contents
INTRODUCTION . . . . . . . . . . . . . . . .
The λ System . . . . . . . . . . . . . . . . . . . .
GENE ORGANIZATION
AND REGULATION . . . . . . . . . . .
The Lytic Transcription Cascade .
The Lysogenic Process . . . . . . . . . . .
THE PROPHAGE STATE
AND ITS MAINTENANCE . . . . .
THE DEFAULT LYTIC COURSE:
Cro AND N FUNCTIONS . . . . . .
The N Antiterminator . . . . . . . . . . .
SWITCHING THE DEFAULT
LYTIC MODE TO LYSOGENY:
ROLE OF CII . . . . . . . . . . . . . . . . . . .
Regulation of CII Activity . . . . . . . .
Regulation of CIII . . . . . . . . . . . . . . . .
THE DECISION PROCESS:
GENETIC FUNCTIONS AND
EXTERNAL INFLUENCES . . . .
Multiplicity Effect . . . . . . . . . . . . . . . .
Physiological Effects . . . . . . . . . . . . . .
COEXPRESSION OF BOTH
LYTIC AND LYSOGENIC
GENES FROM THE SAME
OPERON: A PARADOX . . . . . . . .
THEORETICAL STUDIES . . . . . . . .
OPEN QUESTIONS, SUMMARY,
AND CONCLUSION . . . . . . . . . . .
Counting Infecting Phage. . . . . . . . .
The Distinction Between Phage
Infection and Prophage
Induction . . . . . . . . . . . . . . . . . . . . .
Role of Cro in Lytic Growth . . . . .
Evolution of Temperate Phages . . .
EPILOG . . . . . . . . . . . . . . . . . . . . . . . . . . .
For the concert of life no-one gets a program.
[From a Dutch tile (81)]
INTRODUCTION
A central challenge in the post genomic era
is to understand processes governing the
dynamics of highly complex genetic regula-
410 Oppenheim et al.
tory networks. A growing number of the-
oretical modeling investigators are attempt-
ing to explain the underlying principles of
410
complex regulatory networks involved in nor-
410
mal mammalian development, including al-
terations that can result in a disease state
411
such as cancer. Analysis of such complex sys-
413
tems would be greatly facilitated by simi-
413
lar studies using an ideal paradigm in which
most if not all of the elements composing
413
the system were known. Phage λ, the most
comprehensively studied bacteriophage, is the
415
prototype of a class of lambdoid phages with
415
whom it shares similar genome organization
and functions (38, 82). Studies of λ that began
in the 1950s continue to reveal key molecu-
417
lar processes in gene regulatory mechanisms
417
and development. However, despite years of
418
study, many genetic interactions still remain
to be uncovered and those that we already
know require reexamination. For an accurate,
419
complete, and quantitative analysis of the ge-
419
netic network, in particular its temporal pro-
419
gression, these remaining questions need to
be addressed. In this review we summarize
a systems biology approach to the study of
genetic regulatory circuits of phage λ. We
421
define the individual components of the cir-
421
cuits and switches, describe the kinetics of
their interactions, and explain how the in-
422
teractions achieve robustness in the perfor-
422
mance of the circuits. We also stress some
puzzles that still exist in lambda’s regulatory
system.
423
Our citation of literature is not exhaus-
423
tive but provides examples to illustrate spe-
423
cific points. We direct the reader to several
423
comprehensive reviews on phage biology (13,
15, 24, 31, 34, 37, 41, 82, 107).
The λ System
Bacteria and their temperate phages, like Es-
cherichia coli and λ, exist in symbiotic rela-
tionships. These phages can be present in a
dormant, lysogenic (prophage) state replicat-
ing passively with the host or they can de-
velop lytically, producing progeny phages and
 ANRV260-GE39-18 ARI 15 October 2005 13:5
Annu. Rev. Genet. 2005.39:409-429. Downloaded from www.annualreviews.org
by University of California - San Diego on 03/29/11. For personal use only.

Figure 1
Decision-making steps by the temperate λ phage. A cell infected with phage λ can follow (denoted as
Decision I) the lytic response (left) or the lysogenic response (right). The resulting lysogenic cell carries a
repressed prophage shown in orange. The prophage is irreversibly induced only when a threshold
amount of DNA damage causes an SOS response, leading to lytic development (Decision II). The
prophage is normally extremely stable and rarely undergoes this type of induction by random DNA
damage to produce progeny phage in a very small fraction of the lysogenic cell (basal or spontaneous
induction). Some of the spontaneously induced cells enter the lytic cycle abortively, lose the prophage
(curing), and become nonlysogens (Decision III) (65).

killing their hosts. λ phage infecting an E.
coli cell makes a decision to follow either a
lytic or a lysogenic pathway (Figure 1). If the
lytic pathway is followed, the phage replicates
its DNA autonomously, expresses the mor-
phogenetic genes, assembles virions, and ly-
ses the host. If the lysogenic course ensues,
a stable lysogen is established in which the
prophage is integrated into the host chro-
mosome with lytic gene expression turned
off. The prophage DNA replicates as part
of the bacterial genome during subsequent
cell divisions, and confers immunity to the
cell against infection by another λ. Treatment
with DNA-damaging agents, which leads to
an SOS response, causes the lysogenic state
to irreversibly switch into lytic development,
mimicking the lytic infection. Otherwise, the
prophage state is extremely stable, rarely un-
dergoing induction by DNA damage. Some
of these rarely induced cells enter an abortive
lytic cycle by losing the prophage and be-
coming nonlysogens (curing). Similarly, af-
ter infection abortive lytic or lysogenic events
may also occur (62). The importance of
these abortive events has not been thoroughly
studied.
GENE ORGANIZATION
AND REGULATION
The λ genetic map and transcription profile
involved in early developmental processes are
shown in Figure 2. The gene organization
of lambdoid phages is based on a number of
recurring principles. Phage λ and its many
relatives have genomes that evolved as highly
mosaic, modular structures. This property has
long been recognized and led to the formula-
tion of the “modular genome hypothesis” (10,
14, 38, 101). A short summary of the λ phage
modules and submodules is given in Table 1.
Thus, for example, it is possible to replace
the “immunity” module of λ by that of an-
other lambdoid phage. The organization of
the gene modules allows a typical cascade of

www.annualreviews.org • Switches in Bacteriophage Lambda Development 411

 ANRV260-GE39-18 ARI 15 October 2005 13:5

Figure 2
Genetic map and transcriptional units of the phage regulatory region. Key genes and signals discussed in
the text are shown in their map order between the parallel lines. The early transcripts, the extended
delayed early transcripts, and the late transcripts are shown in black arrows. The transcripts initiated
from the pI, pRE, and paQ that are required for lysogeny are shown in blue. The pRM promoter, which is
activated by the CI regulator, is required for maintenance of the lysogenic state, which is shown in green.
Critical transcription terminators are marked in orange, including the sib region containing the tI
Annu. Rev. Genet. 2005.39:409-429. Downloaded from www.annualreviews.org

terminator. Leftward promoters are indicated above and the rightward ones below the map. pL and pR
by University of California - San Diego on 03/29/11. For personal use only.

are the early promoters and pR is the late lytic promoter. The role of the pOOP promoter is not fully
understood. The operators oL and oR, cognate to pL and pR respectively, are also shown. The immunity
module of the λ chromosome encompasses pLoL, rex, cI, oRpR, and cro. ori is the origin of O- and
P-mediated phage DNA replication (Table 1). Int carries the site-specific recombination reaction, and
Int and Xis support the excision reaction.

phage gene expression in lytic growth delin-
eating the early, delayed early, and late stages
of transcription. This modular and temporal
expression facilitates the alternative λ devel-
opmental pathways. Because of the transcrip-
tional cascade, the repression of the early
phage promoters pR and pL prevents expres-
sion of all lytic genes (Figure 2, Figure 3).
In a prophage, this repression is carried out
by the phage CI protein, which acts by bind-
ing to the oL and oR operators that overlap
the pL and pR promoters, allowing the main-
tenance of the lysogenic state to be governed
by CI alone; when CI is inactivated, e.g., by
the SOS response, the lytic development fol-
lows. By blocking the pL and pR promoters
of an incoming phage genome, the CI reg-
ulator confers immunity against further pro-
ductive infection by another λ (superinfection
immunity).

Table 1 A list of phage λ modules

Module Genes and sites Function
CI regulator pRM-cI CI activates pRM
Immunity pLoL-rex-cI-pRM-oRpR-cro CI represses pL and pR
CII regulator cII-pRE CII activates pRE
Site-specific recombination xis-int-att-tI Int and Xis catalyze recombination
General recombination gam-exo-bet Control and catalyze recombination
DNA replication ori-O-P Control of initiation of DNA replication
control
Early antitermination NUTL-N-tL1 N-antitermination and translational
control
Late antitermination Q-pR  -tR  Q-antitermination for late gene
expression
DNA packaging cos-Nu1-A Cleavage of head-full genomes
Head genes Nu1 to FII Morphogenesis of head particles
Tail genes Z to J Morphogenesis of tails

Description of the functions of genes not discussed in this review can be found in (38).

412 Oppenheim et al.

 ANRV260-GE39-18 ARI 15 October 2005 13:5
The Lytic Transcription Cascade
The gene expression cascade that leads to the
lytic mode of growth is the default mode of λ
development. It is carried out in three stages.
(i) Transcription is initiated with the synthe-
sis of the early transcripts from the pL and
pR promoters (Figure 2). Early transcripts,
which encode two regulators, N and Cro, are
attenuated at the tL1 and tR1 terminators, re-
spectively. These transcriptional terminators
play an important role in controlling the cas-
cade of gene expression. By acting as a weak
repressor for both pL and pR promoters, Cro
Annu. Rev. Genet. 2005.39:409-429. Downloaded from www.annualreviews.org
facilitates the lytic mode (described below).
by University of California - San Diego on 03/29/11. For personal use only.
The N protein is an antitermination factor
that promotes the assembly of a transcription
complex (9, 35). This assembly occurs on the
RNA at the nutL and nutR sites and is made up
of RNA polymerase and a number of host pro-
teins called Nus. The N- and Nus-modified
RNA polymerase can overcome the tL1 and
tR1 transcription terminators, resulting in ex-
pression of the distal delayed early functions
(30, 84). (ii) The delayed early functions in-
clude the lysogenic regulators CII and CIII,
as well as the lytic DNA replication functions
O and P, and the late gene regulator Q. (iii)
After sufficient accumulation, the Q protein
modifies RNA polymerase that has just initi-
ated transcription from the pR late promoter
(66). This modification causes the RNA poly-
merase to become resistant to transcription
terminators present downstream to pR , al-
lowing the expression of the late genes, which
encode proteins for phage morphogenesis and
host cell lysis. There is a kinetic separation
between the expression of delayed early and
late genes. This is caused by the location of
the Q gene at the very end of the delayed
early cascade and the high threshold level of
Q protein needed for its activity (52, 63, 109).
During the late stage of the cascade, the late
gene products assemble phage virions and lyse
the host. A similar temporal lytic cascade of
gene expression follows prophage induction
(38).
www.annualreviews.org • Switches in Bacteriophage Lambda Development
The Lysogenic Process
The lysogenic pathway is also initiated during
the transcription cascade of the delayed early
genes. Under conditions leading to lysogeny,
the expression of the lytic regulators fails be-
cause Q is actively switched off by the accumu-
lation of the critical lysogenic regulator CII,
thereby blocking the default lytic pathway
and switching to the lysogenic course. As dis-
cussed in greater detail below, an accumula-
tion of the CII protein above a threshold level
is critical for initiating the lysogenic switch.
During this active switching from the default
lytic to the lysogenic mode, CII stimulates the
synthesis of Int, which catalyzes the insertion
of the phage DNA into the host chromosome,
and of the CI regulator, which binds to oL and
oR to repress the early promoters. CII also in-
hibits Q function (see 52). These three activ-
ities are mediated by transcription activation
of three promoters, pI, pRE, and paQ, respec-
tively (Figure 2; see Figure 6 below). Activa-
tion of all three promoters is critical for the
establishment of a stable prophage state. Once
the prophage genome integrates into the bac-
terial chromosome and CI protein represses
pL and pR, the lysogenic state is established.
The prophage state is extremely stable and is
maintained through many generations of divi-
sion (8). How CI repressor synthesis is main-
tained in the absence of CII is discussed later.
THE PROPHAGE STATE
AND ITS MAINTENANCE
In the prophage state, the CI regulator con-
trols the expression of three promoters. It
represses transcription from the pL and pR
promoters and positively and negatively reg-
ulates its own synthesis from the pRM pro-
moter (Figure 3; see Figure 6 below). In the
prophage state, pRM is responsible for CI syn-
thesis; CI expression from pRE is prevented
owing to repression of CII in a lysogen.
Figure 3 illustrates a set of cooperative in-
teractions of CI binding to DNA, which lead
413
 ANRV260-GE39-18 ARI 15 October 2005 13:5
Annu. Rev. Genet. 2005.39:409-429. Downloaded from www.annualreviews.org
by University of California - San Diego on 03/29/11. For personal use only.
Figure 3
A scheme showing the repression of pL and pR by CI induced DNA
looping [adapted from (24, 82)]. RNA polymerase is symbolized as a green
ellipse; CI regulator monomers (depicted as orange dumbbells) have two
domains: the N-terminal DNA binding domain and the C-terminal
oligomerization domain. The top line shows a linear arrangement of the
DNA. In the absence of CI, transcription is initiated at both pL and pR (top
line). The second line shows how repressor molecules bind cooperatively
to oL1/oL2 and to oR1/oR2 prior to DNA looping repressing pL and pR
and activating the pRM promoter. The third line shows the first stage of
DNA looping. RNA polymerase bound to pRM either before (second line)
or after (third line) looping occurs transcribes pRM. Finally, further
stabilization of DNA looping occurs by binding of CI to oL3 and oR3
together with repression of pRM when CI levels become very high ( fourth
line). The additional CI molecules bound to oL3 and oR3 interact to form a
tetramer. The SOS response reverses the repression process (see text).
414 Oppenheim et al.
to extremely efficient repression (25, 64, 82).
The CI repressor is made of two domains teth-
ered by a short linker, an N-terminal DNA
binding domain with a helix-turn-helix mo-
tif, and a C-terminal oligomerization domain.
The intrinsic relative affinities of CI to its op-
erators at oR and oL are as follows: oR1 >
oR2 > oR3 and oL1 > oL2 > oL3. Dimers of
CI cooperatively bind to oL1 and oL2 on the
left and oR1 and oR2 on the right by forming
tetramers repressing pL and pR, respectively.
Furthermore, another cooperative interaction
between these two sets of tetramers bound to
oL and oR, 2.3 Kb apart, leads to the formation
of a DNA loop held by a CI octamer, i.e., two
interacting tetramers, that enhances repres-
sion. In this DNA-multiprotein complex, the
CI dimer bound at oR2 also stimulates pRM
transcription, thus activating CI synthesis in
a repressed prophage by a positive autoregu-
latory loop (44, 75). As the CI concentration
increases because of pRM activation, two ad-
ditional CI dimers are recruited to bind at oL3
and oR3 to further stabilize the oL-oR loop. In
this context, the binding of CI to oR3 represses
pRM and prevents CI overexpression. Based
on direct measurements of CI (86), recent cal-
culation shows that a lysogenic cell contains
about 250 CI monomers when the cells are
growing in rich media (23). This translates
into about 30 dimers per prophage copy, as-
suming an average of four chromosomes per
cell. This number of CI molecules not only
achieves a strong repression of the phage pro-
moters but also sets the threshold level for
SOS induction. Note that cell division does
not randomly reduce the CI concentration
to a point low enough to cause prophage
induction (8).
Repression of transcription from the pR
promoter inhibits not only expression of genes
in that operon but also phage DNA repli-
cation by preventing “transcriptional activa-
tion” of λ ori, the site where phage DNA
replication is initiated (26, 33). This inhibition
occurs even if O and P functions are present.
(71, 106). This regulation appears to be crit-
ical for establishing a lysogen. If the λ origin
 ANRV260-GE39-18 ARI 15 October 2005 13:5
replicates after the phage is integrated because
O and P are still available, the replication
event is lethal to the cell, resulting in abortive
lytic infection (11). Thus, it is important to
immediately block phage replication as the
choice for lysogeny is made.
THE DEFAULT LYTIC COURSE:
Cro AND N FUNCTIONS
The gene coding for Cro, which is the essen-
tial lytic regulator, is the first one to be tran-
scribed from the pR promoter following phage
infection or prophage induction. It is a weak
Annu. Rev. Genet. 2005.39:409-429. Downloaded from www.annualreviews.org
repressor of the pR promoter, allowing contin-
by University of California - San Diego on 03/29/11. For personal use only.
uous expression from pR during lytic infection
(29, 94, 95). Cro is a single-domain protein
that binds as a dimer to the oL and oR opera-
tors (60, 82). The intrinsic affinity of Cro for
the individual operators is oR3 > oR1 ≥ oR2,
and, unlike CI, higher-ordered structures, i.e.,
tetramers or octamers of Cro have not been
detected (19). However, Ackers and cowork-
ers found a small amount of cooperative bind-
ing of Cro dimers to adjacent sites in the
oR complex (19). Whether such cooperative
binding is due to dimer-dimer interaction or
to changes in DNA needs to be investigated. It
was shown, however, that changes in DNA ac-
company the cooperative binding of CI (21).
The higher affinity of Cro binding to oR3
supports the hypothesis that during lytic de-
velopment Cro binds first to oR3, repressing
transcription of the cI gene from the pRM
promoter (45). It was therefore proposed that
Cro binding to oR3 represses pRM and com-
pletes the switch to the lytic pathway as well
as sustaining it after an initial SOS-mediated
inactivation of CI in a lysogen (28, 46, 74).
However, recent experiments suggest that the
presence of Cro might be unimportant for the
lysogenic to lytic switch during induction of
the prophage (94). Furthermore, Little and
coworkers showed that replacing oR3 by the
weaker oR1 Cro-binding site has a marginal
effect on prophage induction (64). Thus, it
appears that the critical role of Cro in lytic
development may be only to turn down the
www.annualreviews.org • Switches in Bacteriophage Lambda Development
pL and pR promoters and that its role in turn-
ing down pRM is not critical but supplemen-
tary (52, 94). An argument that Cro binding
to oR3 sets the lytic course has also been
made to explain the role of Cro following
phage infection (82). This interpretation of
the role of Cro is also unlikely because a phage
carrying the same oR3 to oR1 variant (see
above) that reduces Cro but allows CI bind-
ing still shows lytic growth (64). However, un-
der certain conditions Cro binding to oR3 may
contribute to lytic growth after prophage in-
duction. The function of Cro in lytic devel-
opment is addressed below.
The N Antiterminator
The N protein, like Cro, is a critical lytic
regulator but acts by a completely different
mechanism. It antiterminates transcription at
termination signals, allowing expression of
distal genes in the pL and pR operons. N pro-
tein, once made, binds to the specific RNA
sites NUTL and NUTR. The NUTL site
is located between the N gene and the pL
promoter, and NUTR is downstream of cro,
the first gene in the pR operon (Figure 1).
The Nus factors, a complex set of host pro-
teins, take part in cellular transcriptional and
translational processes and interact with the
N system (18, 93). Purified in vitro studies
have defined NusA, NusB, NusE, and NusG
as the host components of N-antitermination
(22, 68). Each NUT site can be divided into
two parts, BOXA and BOXB (32). BOXB is
a stem-loop structure in the RNA and is spe-
cific for N binding (16). As it binds BOXB,
N associates with NusA, NusG, and RNAP.
BOXA RNA is specific for NusB and NusE
binding (77, 80). It is proposed that N, Nus
factors, and Nut interact and complex with
RNA polymerase while tethered on the same
RNA (76, 102, 104, 105). Although N is the
essential factor for antitermination, NUT and
the Nus factors confer stability and full activ-
ity to the antitermination complex (22, 85).
N is a critical regulatory protein of phage
λ, which is reflected in the number of
415
 ANRV260-GE39-18 ARI 15 October 2005 13:5
Annu. Rev. Genet. 2005.39:409-429. Downloaded from www.annualreviews.org
by University of California - San Diego on 03/29/11. For personal use only.
Figure 4
Models for N-mediated translation repression. The complex of RNAP is
modified by NUT, N, and Nus factors to form an antitermination
complex. Translation repression of N takes place on the uncleaved RNA
in the absence of RNase III action (a). In the presence of RNase III
cleavage occurs and the Shine–Dalgarno (SD) sequence for N is no longer
held in close contact to the antitermination complex, translational
repression is prevented, and efficient expression of N can take place (b).
ways its expression is controlled. At the
transcription level, N is regulated from oLpL
by CI- and Cro-mediated repression. At
the translation level, N represses and au-
toregulates its own translation, and the en-
doribonuclease RNase III modulates this
repression (Figure 4). By acting at their
respective sites located upstream of the N
gene, N and RNase III control N translation
(103, 104). High concentrations of RNase III
as found in cells grown in rich media pre-
vent N-mediated translation repression alto-
gether, whereas low concentrations of RNase
III found in cells grown in poor media all-
ow translation repression. At the posttrans-
lational level, the Lon protease degrades N
causing a relatively short half-life.
416 Oppenheim et al.
N expression from the pL-antiterminated
transcripts is dependent upon the level of
RNase III in cells. If RNaseIII concentration
is high, processing of the RNase III site (RIII
site) between NUTL and the N message is
rapid, causing high rates of N translation. If
RNase III concentration is low or zero, pro-
cessing of the RIII site occurs slowly if at
all, allowing repression of N translation. Cel-
lular RNase III concentrations rise and fall
with growth rate (12), and correspondingly,
N translation rises and falls with growth rate
(104).
Evidence suggests that the N translation
repression complex is dependent on E. coli
RNAP and the N-transcription antitermina-
tion complex (105). Under conditions of lim-
iting RNase III, the antiterminating complex
prevents translation of N from its antitermi-
nated message. This means that N is expressed
from RNA transcripts that terminate at tL1,
i.e., do not form an antitermination com-
plex. Thus, under minimal growth conditions
where RNase III activity is limiting, N con-
centrations would be stringently controlled by
negative autoregulation. On the other hand,
in rich media, RNase III processing would
ensure high N concentrations. Viral lytic de-
velopment predominates in rich media, and
lysogenic development is enhanced under
limiting growth or starvation. High N con-
centration enhances lytic growth of the phage,
as evidenced by λ forming clear plaques on
cells expressing N from a plasmid (104). Clear
plaques indicate that only few cells follow the
lysogenic pathway. Low levels of N are made
following λ infection of cells grown in poor
carbon conditions; these same conditions have
been shown to enhance lysogen formation fol-
lowing infection at low multiplicity of infec-
tion (56). These latter conditions are likely to
reduce Q concentrations and provide a bet-
ter opportunity for lysogeny. Regulation of N
also appears to be a way in which λ senses envi-
ronmental conditions through RNase III and
responds by increasing or decreasing N con-
centrations and altering the lytic/lysogenic
response. We note, however, that at high
 ANRV260-GE39-18 ARI 15 October 2005 13:5
multiplicity of infection, there is no effect
of growth media on lysogenization efficiency
(27).
SWITCHING THE DEFAULT
LYTIC MODE TO LYSOGENY:
ROLE OF CII
The inhibition, or absence, of lytic functions
is not sufficient for the switching to the lyso-
genic mode. Rather, as noted above, infec-
tion resulting in a lysogenic response proceeds
through a number of required events: inte-
gration of the DNA, efficient repression of
Annu. Rev. Genet. 2005.39:409-429. Downloaded from www.annualreviews.org
the early promoters, as well as timely inhibi-
by University of California - San Diego on 03/29/11. For personal use only.
tion of the lytic genes expression. These re-
quirements are met by CII turning on pI, pRE,
and paQ promoters to express Int and CI, and
to inhibit Q function, respectively. All three
promoters contain a direct repeat TTGC-
N6-TTGC sequence that binds to CII. The
N6 region corresponds to the –35 elements
of these promoters. Expression from all three
CII-activated promoters is coordinated by the
CII protein during infection. However, the
mechanism of activation of the promoters by
CII is not well understood. Specific muta-
tions in the α or σ subunits of RNAP prevent
CII-mediated activation from these promot-
ers in vitro (48, 67). Consistently, these RNA
polymerase mutants prevent the establish-
ment of lysogeny in vivo (78).
The quaternary structure of the CII pro-
tein has been solved recently (20a, 81a). The
structure shows that a CII tetramer is made of
two nearly equivalent dimers. Each of the four
monomers contains a helix-turn-helix DNA-
binding motif but only two of the monomers
appear, by modeling, to be involved in actual
DNA binding at the direct repeat sequence.
The function of the other two helix-turn-helix
motifs in the tetramer is not known. The di-
rect repeat sequence of pRE is located within
the N-terminal coding sequence of CII. Inci-
dentally, in an elegant genetic study, Friedman
and coworkers isolated and analyzed a CII
mutant defective in transcription activation
of the pE promoter in lambdoid phage P22
www.annualreviews.org • Switches in Bacteriophage Lambda Development
(88). This mutation affects the N-terminal se-
quence and is not in the helix-turn-helix do-
main of CII. This mutation also modified the
pRE promoter so that CII can activate the
mutant pRE but not the wild-type pRE. This
suggests that this N-terminal sequence of the
protein may play a role in determining the
specificity of CII binding to DNA (88).
The pI, pRE, and paQ promoters are lo-
cated within the Xis, CII, and Q protein cod-
ing sequences, respectively. The Xis protein
is needed for excision of the prophage DNA
from the chromosome after induction. For
rapid synthesis of Int, which facilitates inte-
gration after infection, CII activates the pI
promoter. The integration reaction also re-
quires the integration host factor, IHF. IHF
is critical for generating the multicomponent
Intasome structure, which catalyzes the in-
tegration reaction (38). Since pI is located
within the xis gene, the pI transcript synthe-
sizes Int but not Xis, helping to ensure that
integration will not be accompanied by the
presence of Xis function. By the same crite-
rion, CII activates the pRE promoter to direct
rapid synthesis of the CI regulator following
infection. The amount of CI made from pRE
during lysogenic response was found to be
as much as 10- to 20-fold higher than the
amount made from pRM in an established
lysogen (86). The initial high concentrations
of CI may guarantee that all infecting and
replicating phage genomes become repressed.
But the CII-dependent paQ promoter, which
lies within the Q gene, was found to reduce
Q function (52), providing a mechanism by
which CII reduces late gene expression to en-
hance lysogeny (17, 70). Mutations affecting
the ability of the pRE or paQ promoter to re-
spond to CII prevent the lysogenic response
resulting in clear plaque formation (42, 52).
Regulation of CII Activity
CII, which plays a key role in the lysis-
lysogeny decision, is regulated at numerous
levels (Figure 5) (39, 43, 50, 83, 91): (i) The
transcription of the cII gene is inhibited both
417
 ANRV260-GE39-18 ARI 15 October 2005 13:5
Annu. Rev. Genet. 2005.39:409-429. Downloaded from www.annualreviews.org

Figure 5
by University of California - San Diego on 03/29/11. For personal use only.

Multilevel regulation of CII activity. CI and Cro negatively regulate the CII gene at transcription
initiation. tR1 aided by Rho factor reduce transcription elongation into the CII gene. The N
antitermination factor allows the extension of transcription beyond tR1. IHF stimulates CII translation
initiation. The antisense OOP RNA together with RNase III reduces CII mRNA stability, and FtsH
protease acting at the C terminus of CII is responsible for the rapid proteolysis of CII. The center bar
represents the DNA; the CII gene and the positive controls are shown in blue and negative controls in
red. The direction of CII and OOP transcriptions are shown in blue and red, respectively.

by Cro and CI binding to oR, and is stimu-
lated by the N antitermination factor acting
at NUTR. High rates of CII synthesis take
place only for a limited period before being
repressed, by Cro or CI. (ii) Translation initi-
ation of CII is stimulated by IHF (43). An IHF
binding site is located immediately upstream
of CII, which has been proposed to stimu-
late CII translation in the presence of IHF
by an unknown mechanism (72a, 80a). (iii)
The stability of CII mRNA is affected by the
OOP RNA, a short antisense transcript com-
plementary to the 3 end of the cII mRNA (57,
58). RNase III recognizes and cleaves the CII-
OOP double-stranded RNA, thereby initiat-
ing rapid CII mRNA degradation (57). The
DNA coding for the protease target is also the
target of CII mRNA degradation mediated by
the antisense OOP RNA. It was reported that
the stability of the OOP RNA is reduced by
polyadenylation but whether this process reg-
ulates CII concentration has not been clarified
(108). (iv) The ATP-dependent protease FtsH
is responsible for the rapid degradation of CII
(39, 50, 91). Host mutations in ftsH lead to
stabilization of CII and thereby an increased
lysogenization frequency. A C-terminal flex-
ible tail of CII, which is not required for
CII activity, acts as a target for initiating
rapid CII proteolysis by FtsH (20, 51). (v)
The long leader CI RNA initiated from pRE
is antisense to cro, which could prevent the
translation of Cro (92). Indeed, an ftsH host
mutant that results in higher concentrations
of CII was found to be defective in Cro (79,
83, 96). Unfortunately, the concentration of
Cro as a function of pRE activity has not
been directly measured. (vi) CIII, which is
a 54-residue long peptide and required for
lysogeny, controls the rate of CII degradation
by acting as an inhibitor of the FtsH protease
(40, 51, 53).
Regulation of CIII
The cIII gene expression is also subject to mul-
tiple controls by λ CI, Cro, and N and by the
host RNase III (Figure 5). CI and Cro in-
hibit CIII synthesis by binding to oL, and the
N-antiterminator stimulates CIII expression
by acting at NUTL. Unlike its negative effect
on CII, RNase III stimulates CIII translation.

418 Oppenheim et al.

 ANRV260-GE39-18 ARI 15 October 2005 13:5
It was shown that the mRNA that codes for
the amino terminal residues of CIII is present
in two conformations (3, 54). In one con-
formation, the translation initiation region is
occluded, preventing cIII translation. In the
other, the mRNA is open allowing efficient
translation. Point mutations that favor one or
the other structures have been described. It
appears that RNase III regulates cIII trans-
lation by acting as an RNA chaperone to af-
fect CIII mRNA structure without processing
(2).
Annu. Rev. Genet. 2005.39:409-429. Downloaded from www.annualreviews.org
THE DECISION PROCESS:
by University of California - San Diego on 03/29/11. For personal use only.
GENETIC FUNCTIONS AND
EXTERNAL INFLUENCES
We now summarize the critical events that
take place in the decision making process fol-
lowing infection by λ (Decision I in Figure 1,
and shown schematically in Figure 6). Fol-
lowing infection, λ begins a cascade of tran-
scription destined for the lytic mode. Under
appropriate conditions the lytic cascade can be
switched off, allowing entry into the lysogenic
pathway. The switch requires high thresh-
old concentrations of CII, which facilitates
CI and Int synthesis and reduces the regu-
lator function Q. CI, once made, maintains
the repressed state of the prophage. In es-
tablishing repression, CII and CI act tempo-
rally to inhibit lytic functions. CII acts first
by direct reduction of Q function, and as CII
function ceases, CI takes over by repressing
the pR promoter. CI executes repression after
the CII-mediated switch and appears not to
participate in initiating the switch. The activ-
ity of CII is programmed to allow an initial
overshoot of CI expression from the pRE pro-
moter and a preemptive inhibition of Q activ-
ity (52, 86). The overshoot may be required
to ensure the repression of all lytic genes of
the infecting and replicating phage genomes
in the cell during the establishment of repres-
sion. Repression of pR by CI at this stage
also ensures immediate cessation of phage
DNA replication by preventing transcription
through the ori site (97).
www.annualreviews.org • Switches in Bacteriophage Lambda Development
Following infection, transcription from
early pL and pR promoters would start the lytic
pathway by default with N antitermination of
transcription and subsequent expression of Q.
Q in turn antiterminates transcription, lead-
ing to late lytic gene expression, cell lysis, and
phage release. To set the course for lysogeny,
CII reduces Q function in two ways. First,
CII activates antisense paQ RNA inhibiting
Q. Second, CII activates pRE for synthesis of
CI, which represses pR and thus Q transcrip-
tion. CII continues to repress Q expression
via paQ until cII transcription is repressed by
CI at pR. This kinetic coordination of Q shut-
off by CII and CI ensures the switch from the
default lytic pathway to lysogeny. Q protein
must build up to a high threshold concen-
tration to become functional, thus providing
a window for CII to exert its effect on Q
through paQ. In this way, CII prevents Q from
reaching its threshold concentration. If CI ac-
tion is prevented, CII inhibition of Q via paQ
is not sustained because of repression of CII
by Cro acting at pR and by rapid CII degra-
dation. Concordantly, λ phage mutants defec-
tive in CI, oR, or CII follow exclusively lytic
growth, whereas mutants defective in Cro are
unable to follow the lytic pathway.
Multiplicity Effect
The lysis-lysogeny decision is influenced by
the number of phage particles infecting the
cell as well as by the cell physiology (55, 56).
High multiplicity of infection favors lysogeny.
Direct measurements of CII activity showed,
in wild-type infection, functional CII activ-
ity in cells infected with two or more phages
(52). This threshold level of CII is reached
when two or more phage infect a cell. In sin-
gle infection this critical level of CII is not
attained, as mentioned above, allowing lytic
development.
Physiological Effects
In cells grown in rich media the lytic course
predominates, whereas growing cells in poor
419
 ANRV260-GE39-18 ARI 15 October 2005 13:5
Annu. Rev. Genet. 2005.39:409-429. Downloaded from www.annualreviews.org
by University of California - San Diego on 03/29/11. For personal use only.

Figure 6
Description of the λ lytic and lysogenic genetic network. Arrows mark the positive effects between
elements that make up the genetic network, whereas bars denote inhibitory effects. Promoters and
operators are shown in green, the phage genes in light orange, and cis antisense RNA is shown in light
blue (Cis acting has been used to note that it acts on the RNA from the opposite strand and not on an
RNA originating from different sequences in the genome). Arrows emanating from the promoters
denote transcription of specific functions. The OOP antisense RNA and the yet unknown threshold
effect of Q activity are not shown. For simplicity, the developmental network is divided into early gene
expression (a), delayed gene expression (b), lytic gene expression (c), lysogenic establishment (d ), and
lysogenic maintenance (e). The dotted line in (b) leading to Int expression signifies reduced level of Int
expression due to retroregulation of int mRNA.

media increases the chances of the lyso-
genic pathway after single infection (56). Cells
growing in rich media have higher concen-
tration of the host global regulator RNase
III, which leads to high rates of expression of
the protein N favoring lytic growth [see (104)
and references therein]. In carbon-starved
cells, on the other hand, RNase III and con-
sequently N concentrations are low. Under
these conditions N translation is repressed.

420 Oppenheim et al.

 ANRV260-GE39-18 ARI 15 October 2005 13:5
This reduction of N concentration would re-
duce Q expression to a level that provides
more opportunity for lysogenic response.
Nevertheless, when cells are infected by two
or more phages, CII function is epistatic to
the effects of growth conditions preferring
lysogeny.
Unlike the decision that λ makes after the
infection process described above whereby in-
fected cells follow either lytic or lysogenic
development, prophage induction leads ex-
clusively to the lytic course (Decision II in
Figure 1). It was proposed that Cro is respon-
sible for repressing CI expression following
Annu. Rev. Genet. 2005.39:409-429. Downloaded from www.annualreviews.org
induction to keep the developmental switch
by University of California - San Diego on 03/29/11. For personal use only.
in the default lytic mode (82), although the
action of Cro in keeping the lytic course on
track is simply to reduce the concentration of
the unstable master lysogenic regulator CII.
COEXPRESSION OF BOTH
LYTIC AND LYSOGENIC GENES
FROM THE SAME OPERON: A
PARADOX
The pL operon encodes a large number of
ORFs of which only N, CIII, Xis, and Int are
essential for either the lytic or lysogenic re-
sponse (see Figure 2). N is a key regulator
for lytic growth of the phage. On the other
hand, the lysogenic function CIII acts as an
inhibitor of a host protease (FtsH, also called
HflB) that destabilizes the critical lysogenic
regulator CII (40, 47, 51). Proteins Int and Xis
carry out site-specific recombination (59, 73).
During lysogenization, Int catalyzes the inte-
gration of the phage genome into the bacterial
chromosome site, att, whereas Int together
with Xis catalyze the reverse reaction to excise
the λ chromosome during prophage induc-
tion. The pR promoter also transcribes both
lytic and lysogenic genes, cro, cII, O, P, and Q.
Whereas Cro is another critical regulator for
lytic growth of the phage, the very next gene
in the operon encodes the CII protein, which
coordinates the lysogenic pathway. O and P
are needed for phage DNA replication prior
to morphogenesis, whereas the second lytic
www.annualreviews.org • Switches in Bacteriophage Lambda Development
regulator Q activates the synthesis of proteins
for phage morphogenesis and cell lysis. Thus
functions necessary for lysogenic and lytic de-
velopment are both expressed from the pL and
pR promoters. The location of both lysogenic
and lytic genes in the same operons creates an
apparent paradox in our mind about the lysis-
lysogeny decision. This paradox is resolved
by regulating the synthesis and degradation
of critical RNA and proteins. This regula-
tion provides the catalytic or stoichiometric
amounts of functions as required to pursue
either the lytic or lysogenic course. As exam-
ples, during lytic response, the CII regulator is
rapidly eliminated by proteolysis, whereas in
the lysogenic response, CII accumulates and
limits Q activity. Presumably, the high thresh-
old requirement for Q function allows the in-
fected complex a time window for controlling
the concentration of CII.
In the lysogenic response, high concentra-
tions of Int are made from pI under the control
of CII, while the expression of Int from the pL
is greatly reduced by retro-regulation because
of Int mRNA degradation by RNase III from
a site called sib, located on the other side of att
(see Figure 2) (36). Prophage induction leads
to the expression of both Int and Xis from
the pL promoter because the sib regulator is
detached from the int gene in the prophage
DNA.
THEORETICAL STUDIES
The lambda genetic network has been a fertile
ground for theoretical modeling of decision-
making processes during the regulation of
development, and for testing new modeling
methodologies of genetic networks in general.
Models have been constructed addressing the
lysis-lysogeny decision in terms of the lambda
genetic switch describing the competition
between CI and Cro, using statistical mechan-
ics with the explicit goal of obtaining bista-
bility. The earlier models focused on the (i)
probabilities of the occupancy of the pR and
pRM promoters by different binding configu-
rations of CI and Cro, and of pR and pRM by
421
 ANRV260-GE39-18
Annu. Rev. Genet. 2005.39:409-429. Downloaded from www.annualreviews.org
by University of California - San Diego on 03/29/11. For personal use only.
ARI 15 October 2005 13:5
RNAP, and (ii) Gibbs free energies of binding
characterizing the different configurations as
free parameters (1, 87, 90). These models
assumed that the behavior of the reactants
studied under in vitro conditions reflects
in vivo situations, and did not incorporate the
possible existence of other levels of regulation,
e.g., regulation of translation of CI and Cro,
degradation of CI and Cro, and anticooper-
ative binding of the two proteins to adjacent
operators (87).
More recent models, based on the same ap-
proach, also incorporated stochastic effects in
the concentrations of the regulatory proteins
and, as a result, in the selection between lysis
and lysogeny (5, 7, 49, 69, 89, 98, 100, 110).
Although apparent agreement between the
theoretical calculations and the experimen-
tally observed values was noted, these efforts
did not lead to a theoretical description with
improved predictive values. However, the the-
oretical analysis of the exceptional stability the
prophage state of λ led to the conclusion, now
confirmed by experimental evidence, that a
view of the switch focusing only on oR to ex-
plain the stability is incomplete, and that oL
participates as well (7, 23).
An advantage of computer simulations is
their ability to take into account multiple ele-
ments and variables within the decision mod-
ule of the λ network (5, 49, 69). Furthermore,
they can predict, for example, the values of
reactant concentrations during the temporal
execution of the lytic and lysogenic pathways,
which can be readily compared with biological
experiments. Nevertheless, at present there
are only limited experimental data on the ki-
netic changes in the concentration and activ-
ity of the regulatory elements in the network
and the strength of their interactions, which
are crucial to achieve real agreement between
theoretical results and experimental observa-
tions. Furthermore, the in vivo values of the
parameters may differ quantitatively by orders
of magnitude from in vitro ones owing to such
factors as macromolecular crowding, varia-
tion in local concentrations, and yet unknown
functions, as alluded to above. By exposing
422 Oppenheim et al.
inconsistencies with the observed behavior,
future theoretical studies with predictive val-
ues are expected to play a more important role.
OPEN QUESTIONS, SUMMARY,
AND CONCLUSION
Counting Infecting Phage
The decision made by a phage-infected cell
is dependent on the multiplicity of infec-
tion. When one phage infects, most infec-
tion shows lytic development. However, when
two phage infect, the lysogenic pathway pre-
vails. What is the molecular basis for such a
dramatic response to a small change, which
to some may be counterintuitive? Further-
more, the network response also suggests
tight communication between two coinfect-
ing phage genomes. This suggests that repli-
cating phage genomes after single phage
infection have little effect on the decision pro-
cess. This issue requires further investigation.
However, there are conditions when a single
infecting phage enters the lysogenic pathway.
When infecting an hflA or hflB mutant host,
efficient lysogenization takes place. As dis-
cussed, these mutant hosts increase the level
of CII function.
A possible parsimonious model to explain
the multiplicity response runs as follows: First,
multiple infection results in the titration of a
critical regulatory host factor that is present
at a very low concentration. One such can-
didate is FtsH, the product of the hflB gene
of which there are less than 100 molecules
per cell (99). Second, the phage CIII protein
acting as an inhibitor of FtsH plays a critical
role in the decision. Indeed, mutants that re-
sult in a elevated CIII translation (by about
threefold) no longer respond to the multi-
plicity of infection and can efficiently lyso-
genize upon single infection (4; A. Rattray,
unpublished). Thus, we expect that a small in-
crease in multiplicity from one to two would
raise CIII to concentrations that inhibit FtsH
increasing CII expression to allow lysogenic
development.
 ANRV260-GE39-18 ARI 15 October 2005 13:5

The Distinction Between Phage Evolution of Temperate Phages

Infection and Prophage Induction
The idea that two infecting phage are critical
for switching on the lysogenic course raises
interesting questions. A lysogen contains a
number of host chromosomes at various
stages of replication, and accordingly more
than one prophage. Would induction result
in high CII activity as is found in multiple
infections? If a high concentration of CII ac-
cumulates, we would expect inhibition of Q
function by CII, even after CI inactivation by
induction. It remains to be seen whether such
Annu. Rev. Genet. 2005.39:409-429. Downloaded from www.annualreviews.org
Phage λ lifestyles have been a textbook case
for complex genetic control circuits. Gene or-
der, modular construct, recruitment of spe-
cific host factors, and the presence of a
complex genetic network allow the infect-
ing phage to make decisions and to proceed
through alternative developmental pathways.
The lysis-lysogeny bistable decision is clearly
much more complex than originally por-
trayed. This complexity was most probably
reached through evolutionary forces tinker-
ing with specific elements present in the indi-

high concentrations of CII if they occur would vidual functional modules that evolved to act
by University of California - San Diego on 03/29/11. For personal use only.

affect Q function and thus reduce lytic phage
yield.
Role of Cro in Lytic Growth
An open question is the role of the high-
affinity binding of Cro to the oR3 operator
site in lytic growth after phage infection or
prophage induction. It is clear that Cro is
essential to allow lytic development. Is this
high-affinity binding critical to regulate pRM
following prophage induction or does it have
a critical role in lytic decision? This issue
needs to be addressed by the use of phage
mutants defective in oR3 that uniquely affect
Cro binding and do not affect CI binding or
pRM activity. Would such mutants affect lytic
growth after phage infection or prophage in-
duction? The intercalation of CI, Cro, and
RNAP binding site at this locus may make the
isolation of such phage mutants problematic.
in concert. Of interest is a recent approach
of tinkering with λ modules that revealed the
robustness of its genetic network (6, 72). Al-
though the λ lysogenic promoter pRM can tol-
erate a number of mutational changes, we note
that such tolerance may be limited to specific
environmental conditions.
EPILOG
In summary, a small set of regulatory proteins
in an organism as simple as a bacteriophage
function through a diverse set of macro-
molecular interactions in a temporal fashion.
Future investigations into the detailed molec-
ular aspects of the functions of specific mod-
ules coupled with kinetic and quantitative
analysis of the phage genetic network will
yield a realistic picture of this important
paradigm for more complex developmental
processes (61).

ACKNOWLEDGMENTS
We apologize to our colleagues whose work we have omitted. We thank John Little, Ian Dodd,
Ted Cox, Pradeep Parrack, and Grzes Wegrzyn for communication of manuscripts before
publication and David Friedman for critical reading of the manuscript. The research carried
out by O. K. and A. B. O. was supported in part by The Israel Science Foundation (grants #
489/01–1 and 340/04).

LITERATURE CITED
1. Ackers GK, Johnson AD, Shea MA. 1982. Quantitative model for gene regulation by
lambda phage repressor. Proc. Natl. Acad. Sci. USA 79:1129–33

www.annualreviews.org • Switches in Bacteriophage Lambda Development 423

 ANRV260-GE39-18 ARI 15 October 2005 13:5

2. Altuvia S, Kornitzer D, Kobi S, Oppenheim AB. 1991. Functional and structural elements
of the mRNA of the cIII gene of bacteriophage lambda. J. Mol. Biol. 218:723–33
3. Altuvia S, Kornitzer D, Teff D, Oppenheim AB. 1989. Alternative mRNA structures of
the cIII gene of bacteriophage lambda determine the rate of its translation initiation. J.
Mol. Biol. 210:265–80
4. Altuvia S, Oppenheim AB. 1986. Translational regulatory signals within the coding region
of the bacteriophage lambda cIII gene. J. Bacteriol. 167:415–19
5. Arkin A, Ross J, McAdams HH. 1998. Stochastic kinetic analysis of developmental path-
way bifurcation in phage lambda-infected Escherichia coli cells. Genetics 149:1633–48
6. Atsumi S, Little JW. 2004. Regulatory circuit design and evolution using phage lambda.
Genes Dev. 18:2086–94
7. Aurell E, Brown S, Johanson J, Sneppen K. 2002. Stability puzzles in phage lambda. Phys.
Rev. E 65:051914-1-9
8. Baek K, Svenningsen S, Eisen H, Sneppen K, Brown S. 2003. Single-cell analysis of
Annu. Rev. Genet. 2005.39:409-429. Downloaded from www.annualreviews.org

lambda immunity regulation. J. Mol. Biol. 334:363–72

by University of California - San Diego on 03/29/11. For personal use only.

9. Barik S, Ghosh B, Whalen W, Lazinski D, Das A. 1987. An antitermination protein

engages the elongating transcription apparatus at a promoter-proximal recognition site.
Cell 50:885–99
10. Botstein D. 1980. A theory of modular evolution for bacteriophages. Ann. NY Acad. Sci.
354:484–90
11. Brachet P, Eisen H, Rambach A. 1970. Mutations of coliphage lambda affecting the
expression of replicative functions O and P. Mol. Gen. Genet. 108:266–76
12. Britton RA, Powell BS, Dasgupta S, Sun Q, Margolin W, et al. 1998. Cell cycle arrest
in Era GTPase mutants: a potential growth rate-regulated checkpoint in Escherichia coli.
Mol. Microbiol. 27:739–50
13. Calender RL, ed. 2005. The Bacteriophages. New York: Oxford Univ. Press. 2nd ed.
14. Campbell A. 1994. Comparative molecular biology of lambdoid phages. Annu. Rev. Mi-
crobiol. 48:193–222
15. Campbell A. 2003. The future of bacteriophage biology. Nat. Rev. Genet. 4:471–77
16. Chattopadhyay S, Garcia-Mena J, DeVito J, Wolska K, Das A. 1995. Bipartite function
of a small RNA hairpin in transcription antitermination in bacteriophage lambda. Proc.
Natl. Acad. Sci. USA 92:4061–65
17. Court D, Green L, Echols H. 1975. Positive and negative regulation by the cII and cIII
gene products of bacteriophage lambda. Virology 63:484–91
18. Court DL, Patterson TA, Baker T, Costantino N, Mao X, Friedman DI. 1995. Struc-
tural and functional analyses of the transcription-translation proteins NusB and NusE.
J. Bacteriol. 177:2589–91
19. Darling PJ, Holt JM, Ackers GK. 2000. Coupled energetics of lambda cro repressor self-
assembly and site-specific DNA operator binding I: analysis of cro dimerization from
nanomolar to micromolar concentrations. Biochemistry 39:11500–7
20. Datta AB, Roy S, Parrack P. 2005. Role of C-terminal residues in oligomerization and
stability of lambda CII: implications for lysis-lysogeny decision of the phage. J. Mol. Biol.
345:315–24
20a. Datta AB, Panjikar S, Weiss MS, Chakrabarti P, Parrack P. 2005. Structure of λCII:
Implications for recognition of direct repeat DNA by an unusual tetrameric organization.
Proc. Natl. Acad. Sci. USA. In press
21. Deb S, Bandyopadhyay S, Roy S. 2000. DNA sequence dependent and independent
conformational changes in multipartite operator recognition by lambda-repressor. Bio-
chemistry 39:3377–83

424 Oppenheim et al.

 ANRV260-GE39-18 ARI 15 October 2005 13:5

22. DeVito J, Das A. 1994. Control of transcription processivity in phage lambda: Nus factors
strengthen the termination-resistant state of RNA polymerase induced by N antitermi-
nator. Proc. Natl. Acad. Sci USA 91:8660–64
23. Dodd IB, Perkins AJ, Tsemitsidis D, Egan JB. 2001. Octamerization of lambda CI re-
pressor is needed for effective repression of P(RM) and efficient switching from lysogeny.
Genes Dev. 15:3013–22
24. Dodd IB, Shearwin KE, Egan JB. 2005. Revisited gene regulation in bacteriophage λ.
Curr. Opin. Genet. Dev. 15
25. Dodd IB, Shearwin KE, Perkins AJ, Burr T, Hochschild A, Egan JB. 2004. Cooperativity
in long-range gene regulation by the lambda CI repressor. Genes Dev. 18:344–54
26. Dove WF, Inokuchi H, Stevens WF. 1971. In The Bacteriophage Lambda, ed. AD Hershey,
pp. 747–71. Cold Spring Harbor, NY: Cold Spring Harbor Lab. Press
27. Echols H, Green L, Kudrna R, Edlin G. 1975. Regulation of phage lambda development
with the growth rate of host cells: a homeostatic mechanism. Virology 66:344–46
Annu. Rev. Genet. 2005.39:409-429. Downloaded from www.annualreviews.org
by University of California - San Diego on 03/29/11. For personal use only.

28. Eisen H, Brachet P, Pereira da Silva L, Jacob F. 1970. Regulation of repressor expression
in lambda. Proc. Natl. Acad. Sci. USA 66:855–62
29. Folkmanis A, Maltzman W, Mellon P, Skalka A, Echols H. 1977. The essential role of
the cro gene in lytic development by bacteriophage lambda. Virology 81:352–62
30. Friedman DI, Court DL. 1995. Transcription antitermination: the lambda paradigm
updated. Mol. Microbiol. 18:191–200
31. Friedman DI, Court DL. 2001. Bacteriophage lambda: alive and well and still doing its
thing. Curr. Opin. Microbiol. 4:201–7
32. Friedman DI, Gottesman M. 1983. Lytic mode of lambda development. See Ref. 38,
pp. 21–51
33. Furth ME, Dove WF, Meyer BJ. 1982. Specificity determinants for bacteriophage lambda
DNA replication. III. Activation of replication in lambda ric mutants by transcription
outside of ori. J. Mol. Biol. 154:65–83
34. Gottesman ME, Weisberg RA. 2004. Little lambda, who made thee? Microbiol. Mol. Biol.
Rev. 68:796–813
35. Greenblatt J, Mah TF, Legault P, Mogridge J, Li J, Kay LE. 1998. Structure and mech-
anism in transcriptional antitermination by the bacteriophage lambda N protein. Cold
Spring Harb. Symp. Quant. Biol. 63:327–36
36. Guarneros G. 1988. Retroregulation of bacteriophage lambda int gene expression. Curr.
Top. Microbiol. Immunol. 136:1–19
37. Hendrix RW, Lawrence JG, Hatfull GF, Casjens S. 2000. The origins and ongoing
evolution of viruses. Trends Microbiol. 8:504–8
38. Hendrix RW, Roberts JW, Stahl FW, Weisberg RA. 1983. Lambda II. Cold Spring Har-
bor, NY: Cold Spring Harbor Lab. Press
39. Herman C, Ogura T, Tomoyasu T, Hiraga S, Akiyama Y, et al. 1993. Cell growth
and lambda phage development controlled by the same essential Escherichia coli gene,
ftsH/hflB. Proc. Natl. Acad. Sci. USA 90:10861–65
40. Herman C, Thevenet D, D’Ari R, Bouloc P. 1997. The HflB protease of Escherichia coli
degrades its inhibitor lambda cIII. J. Bacteriol. 179:358–63
41. Herskowitz I, Hagen D. 1980. The lysis-lysogeny decision of phage lambda: explicit
programming and responsiveness. Annu. Rev. Genet. 14:399–445
42. Hoopes BC, McClure WR. 1985. A cII-dependent promoter is located within the Q gene
of bacteriophage lambda. Proc. Natl. Acad. Sci. USA 82:3134–38

www.annualreviews.org • Switches in Bacteriophage Lambda Development 425

 ANRV260-GE39-18 ARI 15 October 2005 13:5

43. Hoyt MA, Knight DM, Das A, Miller HI, Echols H. 1982. Control of phage lambda
development by stability and synthesis of cII protein: role of the viral cIII and host hflA,
himA and himD genes. Cell 31:565–73
44. Jain D, Nickels BE, Sun L, Hochschild A, Darst SA. 2004. Structure of a ternary tran-
scription activation complex. Mol. Cell 13:45–53
45. Johnson A, Meyer BJ, Ptashne M. 1978. Mechanism of action of the cro protein of
bacteriophage lambda. Proc. Natl. Acad. Sci. USA 75:1783–87
46. Johnson AD, Poteete AR, Lauer G, Sauer RT, Ackers GK, Ptashne M. 1981. λ Repressor
and cro—components of an efficient molecular switch. Nature 294:217–23
47. Kaiser AD. 1957. Mutations in a temperate bacteriophage affecting its ability to lysogenize
Escherichia coli. Virology 3:42–61
48. Kedzierska B, Lee DJ, Wegrzyn G, Busby SJ, Thomas MS. 2004. Role of the RNA
polymerase alpha subunits in CII-dependent activation of the bacteriophage lambda pE
promoter: identification of important residues and positioning of the alpha C-terminal
Annu. Rev. Genet. 2005.39:409-429. Downloaded from www.annualreviews.org

domains. Nucleic Acids Res. 32:834–41

by University of California - San Diego on 03/29/11. For personal use only.

49. Kiehl TR, Mattheyses RM, Simmons MK. 2004. Hybrid simulation of cellular behavior.
Bioinformatics 20:316–22
50. Kihara A, Akiyama Y, Ito K. 1997. Host regulation of lysogenic decision in bacteriophage
lambda: transmembrane modulation of FtsH (HflB), the cII degrading protease, by HflKC
(HflA). Proc. Natl. Acad. Sci. USA 94:5544–49
51. Kobiler O, Koby S, Teff D, Court D, Oppenheim AB. 2002. The phage lambda CII
transcriptional activator carries a C-terminal domain signaling for rapid proteolysis. Proc.
Natl. Acad. Sci. USA 99:14964–69
52. Kobiler O, Rokney A, Friedman N, Court DL, Stavans J, Oppenheim AB. 2005. Quanti-
tative kinetic analysis of the bacteriophage lambda genetic network. Proc. Natl. Acad. Sci.
USA 102:4470–75
53. Kornitzer D, Altuvia S, Oppenheim AB. 1991. Genetic analysis of the cIII gene of bac-
teriophage HK022. J. Bacteriol. 173:810–15
54. Kornitzer D, Teff D, Altuvia S, Oppenheim AB. 1989. Genetic analysis of bacteriophage
lambda cIII gene: mRNA structural requirements for translation initiation. J. Bacteriol.
171:2563–72
55. Kourilsky P. 1973. Lysogenization by bacteriophage lambda. I. Multiple infection and
the lysogenic response. Mol. Gen. Genet. 122:183–95
56. Kourilsky P, Knapp A. 1974. Lysogenization by bacteriophage lambda. III. Multiplicity
dependent phenomena occurring upon infection by lambda. Biochimie 56:1517–23
57. Krinke L, Mahoney M, Wulff DL. 1991. The role of the OOP antisense RNA in coliphage
lambda development. Mol. Microbiol. 5:1265–72
58. Krinke L, Wulff DL. 1990. RNase III-dependent hydrolysis of lambda cII-O gene mRNA
mediated by lambda OOP antisense RNA. Genes Dev. 4:2223–33
59. Landy A. 1993. Mechanistic and structural complexity in the site-specific recombination
pathways of Int and FLP. Curr. Opin. Genet. Dev. 3:699–707
60. LeFevre KR, Cordes MH. 2003. Retroevolution of lambda Cro toward a stable monomer.
Proc. Natl. Acad. Sci. USA 100:2345–50
61. Levine M, Davidson EH. 2005. Gene regulatory networks for development. Proc. Natl.
Acad. Sci. USA 102:4936–42
62. Lieb M. 1953. The establishment of lysogenicity in Escherichia coli. J. Bacteriol. 65:642–51
63. Little JW. 2005. Threshold effects in gene regulation: When some is not enough. Proc.
Natl. Acad. Sci. USA 102:5310–11

426 Oppenheim et al.

 ANRV260-GE39-18 ARI 15 October 2005 13:5

64. Little JW, Shepley DP, Wert DW. 1999. Robustness of a gene regulatory circuit. EMBO
J. 18:4299–307
65. Livny J, Friedman DI. 2004. Characterizing spontaneous induction of Stx encoding
phages using a selectable reporter system. Mol. Microbiol. 51:1691–704
66. Marr MT, Datwyler SA, Meares CF, Roberts JW. 2001. Restructuring of an RNA poly-
merase holoenzyme elongation complex by lambdoid phage Q proteins. Proc. Natl. Acad.
Sci. USA 98:8972–78
67. Marr MT, Roberts JW, Brown SE, Klee M, Gussin GN. 2004. Interactions among
CII protein, RNA polymerase and the lambda PRE promoter: contacts between RNA
polymerase and the -35 region of PRE are identical in the presence and absence of CII
protein. Nucleic Acids Res. 32:1083–90
68. Mason SW, Greenblatt J. 1991. Assembly of transcription elongation complexes con-
taining the N protein of phage lambda and the Escherichia coli elongation factors NusA,
NusB, NusG, and S10. Genes Dev. 5:1504–12
Annu. Rev. Genet. 2005.39:409-429. Downloaded from www.annualreviews.org

69. McAdams HH, Shapiro L. 1995. Circuit simulation of genetic networks. Science 269:650–
by University of California - San Diego on 03/29/11. For personal use only.

66
70. McMacken R, Mantei N, Butler B, Joyner A, Echols H. 1970. Effect of mutations in the
c2 and c3 genes of bacteriophage lambda on macromolecular synthesis in infected cells.
J. Mol. Biol. 49:639–55
71. Mensa-Wilmot K, Carroll K, McMacken R. 1989. Transcriptional activation of bacte-
riophage lambda DNA replication in vitro: regulatory role of histone-like protein HU of
Escherichia coli. EMBO J. 8:2393–402
72. Michalowski CB, Short MD, Little JW. 2004. Sequence tolerance of the phage lambda
PRM promoter: implications for evolution of gene regulatory circuitry. J. Bacteriol.
186:7988–99
72a. Mahajna J, Oppenheim AB, Rattray A, Gottesman M. 1986. Translation initiation of
bacteriophage lambda gene cII requires integration host factor. J. Bacteriol. 165:167–74
73. Nash HA. 1981. Integration and excision of bacteriophage lambda: the mechanism of
conservation site specific recombination. Annu. Rev. Genet. 15:143–67
74. Neubauer Z, Calef E. 1970. Immunity phase-shift in defective lysogens: non-mutational
hereditary change of early regulation of lambda prophage. J. Mol. Biol. 51:1–13
75. Nickels BE, Dove SL, Murakami KS, Darst SA, Hochschild A. 2002. Protein-protein
and protein-DNA interactions of sigma70 region 4 involved in transcription activation
by lambda cI. J. Mol. Biol. 324:17–34
76. Nodwell JR, Greenblatt J. 1991. The nut site of bacteriophage lambda is made of RNA
and is bound by transcription antitermination factors on the surface of RNA polymerase.
Genes Dev. 5:2141–51
77. Nodwell JR, Greenblatt J. 1993. Recognition of boxA antiterminator RNA by the E. coli
antitermination factors NusB and ribosomal protein S10. Cell 72:261–68
78. Obuchowski M, Giladi H, Koby S, Szalewska-Palasz A, Wegrzyn A, et al. 1997. Impaired
lysogenisation of the Escherichia coli rpoA341 mutant by bacteriophage lambda is due to
the inability of CII to act as a transcriptional activator. Mol. Gen. Genet. 254:304–11
79. Oppenheim A, Honigman A, Oppenheim AB. 1974. Interference with phage lambda cro
gene function by a colicin-tolerant Escherichia coli mutant. Virology 61:1–10
80. Patterson TA, Zhang Z, Baker T, Johnson LL, Friedman DI, Court DL. 1994. Bacte-
riophage lambda N-dependent transcription antitermination. Competition for an RNA
site may regulate antitermination. J. Mol. Biol. 236:217–28
80a. Peacock S, Weissbach H. 1985. IHF stimulation of lambda cII gene expression is inhibited
by the E coli NusA protein. Biochem. Biophys. Res. Commun. 127:1026–31

www.annualreviews.org • Switches in Bacteriophage Lambda Development 427

 ANRV260-GE39-18 ARI 15 October 2005 13:5

81. Perutz M. 2003. I Wish I’d Made You Angry Earlier. Cold Spring Harbor, NY: Cold Spring
Harbor Lab. Press
81a. Jain D, Youngchang K, Maxwell KL, Beasley S, Rongguang Z, et al. 2005. Crystal struc-
ture of bacteriophage λ cII and its DNA complex. Mol. Cell 19:1–11
82. Ptashne M. 2004. Genetic Switch: Phage Lambda Revisited. Cold Spring Harbor, NY: Cold
Spring Harbor Lab. Press
83. Rattray A, Altuvia S, Mahajna G, Oppenheim AB, Gottesman M. 1984. Control of bacte-
riophage lambda CII activity by bacteriophage and host functions. J. Bacteriol. 159:238–
42
84. Reed MR, Shearwin KE, Pell LM, Egan JB. 1997. The dual role of Apl in prophage
induction of coliphage 186. Mol. Microbiol. 23:669–81
85. Rees WA, Weitzel SE, Yager TD, Das A, von Hippel PH. 1996. Bacteriophage lambda
N protein alone can induce transcription antitermination in vitro. Proc. Natl. Acad. Sci.
USA 93:342–46
Annu. Rev. Genet. 2005.39:409-429. Downloaded from www.annualreviews.org

86. Reichardt L, Kaiser AD. 1971. Control of lambda repressor synthesis. Proc. Natl. Acad.
by University of California - San Diego on 03/29/11. For personal use only.

Sci. USA 68:2185–89

87. Reinitz J, Vaisnys JR. 1990. Theoretical and experimental analysis of the phage lambda
genetic switch implies missing levels of co-operativity. J. Theor. Biol. 145:295–318
88. Retallack DM, Johnson LL, Ziegler SF, Strauch MA, Friedman DI. 1993. A single-base-
pair mutation changes the specificities of both a transcription activation protein and its
binding site. Proc. Natl. Acad. Sci. USA 90:9562–65
89. Santillan M, Mackey MC. 2004. Why the lysogenic state of phage lambda is so stable: a
mathematical modeling approach. Biophys. J. 86:75–84
90. Shea MA, Ackers GK. 1985. The OR control system of bacteriophage lambda. A physical-
chemical model for gene regulation. J. Mol. Biol. 181:211–30
91. Shotland Y, Koby S, Teff D, Mansur N, Oren DA, et al. 1997. Proteolysis of the phage
lambda CII regulatory protein by FtsH (HflB) of Escherichia coli. Mol. Microbiol. 24:1303–
10
92. Spiegelman WG, Reichardt LF, Yaniv M, Heinemann SF, Kaiser AD, Eisen H. 1972.
Bidirectional transcription and the regulation of phage lambda repressor synthesis. Proc.
Natl. Acad. Sci. USA 69:3156–60
93. Squires CL, Zaporojets D. 2000. Proteins shared by the transcription and translation
machines. Annu. Rev. Microbiol. 54:775–98
94. Svenningsen SL, Costantino N, Court DL, Adhya S. 2005. On the role of Cro in λ
prophage induction. Proc. Natl. Acad. Sci. USA. In press
95. Takeda Y, Folkmanis A, Echols H. 1977. Cro regulatory protein specified by bacterio-
phage lambda. Structure, DNA-binding, and repression of RNA synthesis. J. Biol. Chem.
252:6177–83
96. Teff D, Koby S, Shotland Y, Ogura T, Oppenheim AB. 2000. A colicin-tolerant Escherichia
coli mutant that confers hfl phenotype carries two mutations in the region coding for the
C-terminal domain of FtsH (HflB). FEMS Microbiol. Lett. 183:115–17
97. Thomas R. 1971. Control circuits. In The Bacteriophage Lambda, ed. AD Hershey,
pp. 211–20. Cold Spring Harbor, NY: Cold Spring Harbor Lab. Press
98. Tian T, Burrage K. 2004. Bistability and switching in the lysis/lysogeny genetic regulatory
network of bacteriophage lambda. J. Theor. Biol. 227:229–37
99. Tomoyasu T, Yamanaka K, Murata K, Suzaki T, Bouloc P, et al. 1993. Topology
and subcellular localization of FtsH protein in Escherichia coli. J. Bacteriol. 175:1352–
57

428 Oppenheim et al.

 ANRV260-GE39-18 ARI 15 October 2005 13:5

100. Vohradsky J. 2001. Neural model of the genetic network. J. Biol. Chem. 276:36168–73
101. Weisberg RA, Gottesmann ME, Hendrix RW, Little JW. 1999. Family values in the age
of genomics: comparative analyses of temperate bacteriophage HK022. Annu. Rev. Genet.
33:565–602
102. Whalen WA, Das A. 1990. Action of an RNA site at a distance: role of the nut genetic signal
in transcription antitermination by phage-lambda N gene product. New Biol. 2:975–91
103. Wilson HR, Kameyama L, Zhou JG, Guarneros G, Court DL. 1997. Translational re-
pression by a transcriptional elongation factor. Genes Dev. 11:2204–13
104. Wilson HR, Yu D, Peters HK 3rd, Zhou JG, Court DL. 2002. The global regulator
RNase III modulates translation repression by the transcription elongation factor N.
EMBO J. 21:4154–61
105. Wilson HR, Zhou JG, Yu D, Court DL. 2004. Translation repression by an RNA poly-
merase elongation complex. Mol. Microbiol. 53:821–28
106. Wold MS, Mallory JB, Roberts JD, LeBowitz JH, McMacken R. 1982. Initiation of
Annu. Rev. Genet. 2005.39:409-429. Downloaded from www.annualreviews.org

bacteriophage lambda DNA replication in vitro with purified lambda replication proteins.
by University of California - San Diego on 03/29/11. For personal use only.

Proc. Natl. Acad. Sci. USA 79:6176–80

107. Wommack KE, Colwell RR. 2000. Virioplankton: Viruses in aquatic ecosystems. Micro-
biol. Mol. Biol. Rev. 64:69–114
108. Wrobel B, Herman-Antosiewicz A, Szalewska-Palasz S, Wegrzyn G. 1998. Polyadenyla-
tion of oop RNA in the regulation of bacteriophage lambda development. Gene 212:57–65
109. Yang XJ, Hart CM, Grayhack EJ, Roberts JW. 1987. Transcription antitermination by
phage-lambda gene-Q protein requires a DNA segment spanning the RNA start site.
Genes Dev. 1:217–26
110. Zhu XM, Yin L, Hood L, Ao P. 2004. Calculating biological behaviors of epigenetic
states in the phage lambda life cycle. Funct. Integr. Genomics 4:188–95

www.annualreviews.org • Switches in Bacteriophage Lambda Development 429

 Contents ARI 20 October 2005 19:18

Annual Review of

Contents Genetics

Volume 39, 2005

John Maynard Smith

Richard E. Michod p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 1
The Genetics of Hearing and Balance in Zebrafish
Teresa Nicolson p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 9
Annu. Rev. Genet. 2005.39:409-429. Downloaded from www.annualreviews.org
by University of California - San Diego on 03/29/11. For personal use only.

Immunoglobulin Gene Diversification

Nancy Maizels p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p23
Complexity in Regulation of Tryptophan Biosynthesis in
Bacillus subtilis
Paul Gollnick, Paul Babitzke, Alfred Antson, and Charles Yanofsky p p p p p p p p p p p p p p p p p p p p p p p47
Cell-Cycle Control of Gene Expression in Budding and Fission Yeast
Jürg Bähler p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p69
Comparative Developmental Genetics and the Evolution of Arthropod
Body Plans
David R. Angelini and Thomas C. Kaufman p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p95
Concerted and Birth-and-Death Evolution of Multigene Families
Masatoshi Nei and Alejandro P. Rooney p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 121
Drosophila as a Model for Human Neurodegenerative Disease
Julide Bilen and Nancy M. Bonini p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 153
Molecular Mechanisms of Germline Stem Cell Regulation
Marco D. Wong, Zhigang Jin, and Ting Xie p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 173
Molecular Signatures of Natural Selection
Rasmus Nielsen p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 197
T-Box Genes in Vertebrate Development
L.A. Naiche, Zachary Harrelson, Robert G. Kelly, and Virginia E. Papaioannou p p p p p p 219
Connecting Mammalian Genome with Phenome by ENU Mouse
Mutagenesis: Gene Combinations Specifying the Immune System
Peter Papathanasiou and Christopher C. Goodnow p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 241
Evolutionary Genetics of Reproductive Behavior in Drosophila:
Connecting the Dots
Patrick M. O’Grady and Therese Anne Markow p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 263
v
Contents ARI 20 October 2005 19:18

Sex Determination in the Teleost Medaka, Oryzias latipes

Masura Matsuda p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 293
Orthologs, Paralogs, and Evolutionary Genomics
Eugene V. Koonin p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 309
The Moss Physcomitrella patens
David Cove p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 339
A Mitochondrial Paradigm of Metabolic and Degenerative Diseases,
Aging, and Cancer: A Dawn for Evolutionary Medicine
Douglas C. Wallace p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 359
Switches in Bacteriophage Lambda Development
Amos B. Oppenheim, Oren Kobiler, Joel Stavans, Donald L. Court,
Annu. Rev. Genet. 2005.39:409-429. Downloaded from www.annualreviews.org

and Sankar Adhya p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 409

by University of California - San Diego on 03/29/11. For personal use only.

Nonhomologous End Joining in Yeast

James M. Daley, Phillip L. Palmbos, Dongliang Wu, and Thomas E. Wilson p p p p p p p p p p 431
Plasmid Segregation Mechanisms
Gitte Ebersbach and Kenn Gerdes p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 453
Use of the Zebrafish System to Study Primitive and Definitive
Hematopoiesis
Jill L.O. de Jong and Leonard I. Zon p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 481
Mitochondrial Morphology and Dynamics in Yeast and Multicellular
Eukaryotes
Koji Okamoto and Janet M. Shaw p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 503
RNA-Guided DNA Deletion in Tetrahymena: An RNAi-Based
Mechanism for Programmed Genome Rearrangements
Meng-Chao Yao and Ju-Lan Chao p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 537
Molecular Genetics of Axis Formation in Zebrafish
Alexander F. Schier and William S. Talbot p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 561
Chromatin Remodeling in Dosage Compensation
John C. Lucchesi, William G. Kelly, and Barbara Panning p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 615

INDEXES

Subject Index p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p p 653

ERRATA

An online log of corrections to Annual Review of Genetics

chapters may be found at http://genet.annualreviews.org/errata.shtml

vi Contents
 Modern Microbial Genetics, Second Edition. Edited by Uldis N. Streips, Ronald E. Yasbin
Copyright # 2002 Wiley-Liss, Inc.
ISBNs: 0-471-38665-0 (Hardback); 0-471-22197-X (Electronic)

21
Transduction in Gram-Negative Bacteria
GEORGE M. WEINSTOCK
Department of Biochemistry and Molecular Biology, University of Texas Medical School,
Houston, Texas 77225

I. Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 561
II. Generalized Transduction . . . . . . . . . . . . . . . . . . . . . . . . 562
A. Events in the Donor: Bacteriophage P22 . . . . . . . . 563
1. P22 DNA Metabolism . . . . . . . . . . . . . . . . . . . . . 564
2. Formation of P22 Transducing Particles . . . . . . 565
B. Formation of Generalized Transducing Particles
by Other Phages. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 566
1. Phage P1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 566
2. Phage T4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 567
3. Phage l. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 567
4. Phage Mu. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 567
C. Events in the Recipient . . . . . . . . . . . . . . . . . . . . . . . 568
1. Homologous Recombination With the
Recipient's Chromosome . . . . . . . . . . . . . . . . . . . 568
D. Measuring Transduction. . . . . . . . . . . . . . . . . . . . . . 568
1. Cotransduction of Markers . . . . . . . . . . . . . . . . . 570
2. Marker Effects. . . . . . . . . . . . . . . . . . . . . . . . . . . . 571
3. Abortive Transduction . . . . . . . . . . . . . . . . . . . . . 572
4. Transduction of Plasmids . . . . . . . . . . . . . . . . . . 572
E. Uses for Generalized Transduction . . . . . . . . . . . . . 573
III. Specialized Transduction. . . . . . . . . . . . . . . . . . . . . . . . . 573
A. Bacteriophage l DNA Metabolism. . . . . . . . . . . . . 573
B. Events in the Recipient . . . . . . . . . . . . . . . . . . . . . . . 575
C. Uses for Specialized Transduction. . . . . . . . . . . . . . 576
APPENDIX: Bacteriophage l . . . . . . . . . . . . . . . . . . . . . . . . 577

I. INTRODUCTION carry parts of bacterial or plasmid chromo-

somes from one bacterial host to another. For
Transduction is the mode of genetic exchange
some phages this DNA is joined to the viral
between cells that is mediated by viruses. The
chromosome in the particle (as in specialized
term transduction refers to the transfer of cel-
transductionbyphagel),whileforothertrans-
lular DNA or RNA from one cell to another by
ducing phages only DNA of host origin is pre-
infection with a viral vector. Virus particles
sent in the particle (as in generalized
that carry and introduce cellular nucleic acids
on infection are called transducing particles.
Bacteriophages are transducing viruses if they
562 WEINSTOCK
transduction by phages P1 or P22). The forma-
tion of transducing particles is a part of the
normal DNA metabolism of viral infection.
In addition, when foreign DNA is introduced
into a virus chromosome by recombinant
DNA techniques, the resulting phages that
carry the cloned foreign DNA are transducing
particles. Finally, RNA viruses, such as retro-
viruses, that carry host RNA and introduce it
into a recipient cell are transducing viruses.
The process of transduction is divided into
two types. In generalized transduction, the
virus can introduce any region from the donor
chromosome into a recipient. In specialized
transduction, the virus always carries the
same segment into the recipient. A phage that
randomly packages host DNA into particles is
a generalized transducing phage. A phage that
has had a particular gene stably introduced
into its chromosome is a specialized transdu-
cing phage. For detailed reviews of these pro-
cesses, see Margolin (1987) and Weisberg
(1987); for a summary of the early literature,
see Low and Porter (1978) (also see Hendrix,
this volume).
II. GENERALIZED
TRANSDUCTION
In 1952 Zinder and Lederberg first reported
generalized transduction in Salmonella tym-
phimurium. They were able to show that
genetic exchange occurred in Salmonella
by mixing different auxotrophic mutants to-
gether and isolating prototrophic recombin-
ants. To differentiate this mechanism of
genetic exchange from conjugation, they
showed that transfer of the genetic material
did not require physical contact between
the donor and recipient bacteria. The agent
responsible for transfer was able to pass
through the pores of a filter, pores that were
too small to allow bacteria through. Moreover
the genetic exchange was not due to transform-
ation (see Streips, this volume), since the pro-
cess was resistant to DNase treatment. It was
eventuallyshownthattheagentthatcarriedthe
genesfromthedonortotherecipientbacterium
wasthetemperatebacteriophageP22.Thisdis-
covery was made possible because fortuitously
one of the Salmonella strains that Zinder and
Lederberg had used in the mixing experiment
was a lysogen for P22. During their experi-
ments the P22 prophage had spontaneously
induced, forming particles that infected the
second Salmonella strain. The infection of the
second strain then produced particles that
carried genes of the second host. When these
particlesinfectedtheoriginalauxotrophicbac-
terium, they introduced the wild-type gene,
which could recombine and replace the defect-
ivesegmentofthechromosome,formingapro-
totrophic transductant bacterium.
Subsequently this process of bacterio-
phage-mediated genetic transduction was
shown to occur in many other Gram-negative
bacteria, including Escherichia coli, Myxo-
bacteria, Rhizobium, Caulobacter, and Pseu-
domonas. In E. coli, for example, transduction
was shown to occur with bacteriophage P1. In
addition it was possible to transfer any host
gene from one bacterium to another with
these phages, and hence this process came to
be known as generalized transduction. That
this was due to physical transfer of DNA
from the donor bacterium to the recipient
was demonstrated for both P22 (Ebel-Tsipis
et al., 1972) and P1 (Ikeda and Tomizawa,
1965a) transduction. To show this (Fig. 1),
the donor bacteria were grown in a medium
containing heavy isotopes of nitrogen and
hydrogen (15 N and 2 H) to make the donor
DNA more dense. Then the bacteria were
shifted to a normal medium (14 N and 1 H)
and infected with the phage. During growth,
the newly synthesized phage DNA was light,
while the preexisting bacterial DNA was
heavy. It was then shown that the phage pre-
pared in this way contained heavy DNA; that
is, they had packaged the donor bacterial
DNA into particles, and when they were
used to infect a recipient bacterium that had
light DNA, some of the heavy DNA became
incorporated into the recipient chromosome.
These experiments illustrate the overall pro-
cess of generalized transduction. They show
that there are two parts to the mechanism:
the packaging of donor DNA into a phage
 GENETIC TRANSDUCTION 563

Fig. 1. Physical transfer of DNA in generalized transduction. An example of generalized transduction by

phage P22 is shown. The donor bacteria are grown in a medium containing heavy isotopes of nitrogen and
hydrogen so that their DNA has a heavier density than normal (the heavy DNA is represented by the thick
circular line in the cell at the upper left). Following P22 infection of this cell, some heavy chromosomal DNA
is packaged into phage particles. These can then be used to infect a cell grown in medium with light isotopes
of nitrogen and hydrogen, and thus containing DNA with a lighter density than the donor's. When heavy
donor DNA is introduced, it can recombine and replace some light DNA in the recipient's chromosome.
When the chromosome is broken into small pieces, the heavy DNA can be detected by equilibrium gradient
centrifugation because it bands at a different density than the bulk of the light DNA.

particle and the stable introduction of this
packaged DNA into the recipient cell, usually
through genetic recombination with the re-
cipient chromosome (Fig. 2). The ability of a
phage to perform generalized transduction
thus depends on the mechanism of packaging
DNA into phage particles. If this mechanism
allows host as well as phage DNA to be en-
capsidated, generalized transduction invari-
ably occurs. As we will see, there are
numerous types of DNA metabolism that
lead to generalized transduction. In each
case the capability for generalized transduc-
tion is a result of the mode of packaging
phage DNA.
A. Events in the Donor: Bacteriophage
P22
A well-studied example of generalized trans-
duction is that of the Salmonella phage P22.
This is a useful paradigm for DNA
metabolism leading to generalized transduc-
tion, since it is not unique and is representa-
tive of a number of other phages (see
Guttman and Kutter, this volume, for phage
T4). The DNA metabolism of phage P22 has
been reviewed by Susskind and Botstein
(1978).
The phage DNA in a P22 particle is ter-
minally redundant and circularly permuted
(Fig. 3). Each phage chromosome has the
564 WEINSTOCK
Fig. 2. Overview of generalized transduction. The
scheme shown applies to phages like P22 and P1.
During normal productive infection of bacteria two
kinds of particles are produced: those with a viral
chromosome and those containing a fragment of the
host chromosome. When the phage with a viral
chromosome infects a cell, another round of pro-
ductive infection occurs. However, when the particle
with the host DNA injects this DNA into a cell,
the fragment can recombine with the recipient's
chromosome, resulting in a transductant.
leftmost sequence of its chromosome
repeated at the right end (terminally redun-
dant). However, this repetitious segment
(about 2% of the chromosome) is not unique
and differs for each chromosome (circularly
permuted). This chromosome structure is a
result of the mechanism of packaging DNA
employed by P22, the headful packaging
mechanism (Fig. 3).
1. P22 DNA metabolism
When P22 infects a cell, its DNA must circu-
larize for the infection to be productive and
the chromosome to replicate. This obligatory
circularization occurs by homologous re-
combination between the repetitious ends of
the chromosome. This is the significance
of the terminal redundancy of the chromo-
someÐit provides the substrate for the cir-
cularization reaction. The circular molecule
replicates and ultimately produces a conca-
temer, a long DNA molecule containing
multiple copies of the genome. This probably
arises from rolling circle replication, an effi-
cient process where a single replication fork
continually circles a circular chromosome
synthesizing a long concatemeric tail (see
Fig. 3). Packaging of this DNA initiates at
Fig. 3. Scheme of phage P22 DNA metabolism. At
the top is shown the linear DNA molecule carried by
the phage particle. This chromosome is terminally
redundant, containing a copy of the ``a'' region of the
genome at each end. The two copies of the ``a'' region
can recombine to circularize the chromosome. Then
a concatemer is formed by rolling circle replication.
This concatemeric DNA is packaged into phage par-
ticles by sequential cuts starting at a pac site. Each
packaging event measures out a ``headful'' of DNA,
which is slightly longer than the size of the genome.
As a result a series of chromosomes is produced.
Each chromosome is terminally redundant, but the
redundant sequence is different for each one. Thus
the chromosomes are said to be circularly permuted.
 GENETIC TRANSDUCTION
a specific region in the genome (pac) and
proceeds processively along the concatemer
from this site. The first packaging cut, by
a phage-encoded enzyme, is site specific, at
pac, and a defined length of DNA is encap-
sidated. When this length (a ``head-ful'') is
packaged, the second cut is made and a
phage head containing DNA is formed.
This second cut also defines the start of
packaging for the next phage head to be
formed. The same headful length of DNA
is encapsulated starting at the second cut,
and when this length is reached a third cut
is made. The third cut plays the same role as
the second. It terminates packaging of the
second headful and starts the packaging
into the next head. This process continues.
Thus, although DNA packaging starts at a
defined sequence, the subsequent cuts are
not sequence specific. Rather, it is the same
length of DNA that is packaged each time.
For phage P22, the size of the genome is
slightly less than the length of the headful.
Thus, each time a head is formed, the en-
capsidated DNA contains a short terminal
repeat. Since each packaging event starts
at a different sequence, the repeats are dif-
ferent, producing permuted chromosomes.
Headful packaging thus provides the mech-
anism for generating the essential terminal
repeats.
2. Formation of P22 transducing particles
During P22 infection the host DNA is not
degraded (in contrast to infections with viru-
lent phages like T4; see Guttman and Kutter,
this volume), and thus is potentially also a
substrate for DNA packaging. As just de-
scribed, the characteristics of DNA pack-
aging are that a concatemeric molecule is
encapsidated in processive headfuls starting
from a unique site. Exactly the same process
operating on the host DNA generates trans-
ducing particles (Fig. 4). When a sequence
similar to the pac sequence occurs in the host
chromosome, the phage packaging appar-
atus uses this analogous pac site and the
host chromosome can be packaged. When a
pac site occurs in the host DNA, packaging
565
Fig. 4. Packaging of host DNA by the phage P22
headful packaging mechanism. A: A wild-type bacter-
ial chromosome contains a site, pac, that is recog-
nized by the phage DNA packaging apparatus. The
bacterial chromosome is packaged into phage par-
ticles starting at pac and proceeding sequentially in
one direction. Each packaging event cuts a headful of
DNA and each of the resulting transducing frag-
ments carries a characteristic set of genes. Genes
on the same fragment are cotransducible because
they can be introduced into a cell together. B: The
packaging of a bacterial chromosome that is deleted
for gene E is shown. The pac site is still recognized
and the overall process of forming transducing par-
ticles is the same as in A. However, because of the
deletion, the cotransduction of genes is altered. This
is because each packaging event measures the same
amount of DNA and the deletion has changed the
distance of distal genes from the pac site.
initiates and proceeds processively, generat-
ing a series of phage particles that carry
different parts of the host chromosome.
These transducing particles carry DNA frag-
ments that are all the same size but obviously
do not have the terminal repetition that the
phage chromosomes have, since the genome
size of the host is much greater than the
headful size that is packaged. For P22, al-
though there are roughly equal amounts of
phage and host DNA in the cell during
infection, only a few percent of the host
chromosome is packaged and only a few
percent of the particles are transducing par-
ticles. What limits formation of transducing
566 WEINSTOCK
particles is the number of pac sites in the host
DNA, as well as the fact that these chromo-
somal sites are probably not identical in se-
quence to the phage pac site and are thus
inefficiently used. It has been estimated
(Chelala and Margolin, 1974) that there are
10 to 15 of these pac-like sites in the Salmon-
ella chromosome. Other studies (reviewed in
Margolin, 1987) indicate that at least 5±10%
of the bacterial chromosome is packaged in
the sequential headfuls from each pac site.
However, the efficiency of packaging de-
clines the more distant a gene is from a pac
site, and this contributes to the wide range in
transduction frequencies seen for different
genes transduced with P22. This variation
in frequency can be two or three orders of
magnitude between different genes.
One experimental approach (Chelala and
Margolin, 1974) that provided support for
this view of transducing particle formation
comes from the study of the effects of dele-
tions in the bacterial chromosome on trans-
duction (Fig. 4). One effect of deletions is on
the cotransduction of genes, the frequency at
which two nearby genes are transduced to-
gether on the same fragment. Some deletions
were found to alter this frequency, presum-
ably by altering the register of subsequent
packaging events. Effects on transduction
have also been observed for insertions; these
are due to an analogous mechanism. The
studies imply that host DNA packaging ini-
tiates from a limited number of sites and
occurs processively with a characteristic
register for each region.
Another important approach to under-
standing this mechanism of generalized
transduction was the isolation and analysis
of HT mutants of P22: mutants that perform
generalized transduction at a higher fre-
quency than a wild-type phage (Schmieger,
1972). HT mutants, which can increase the
frequency of transduction of some genes by
as much as 10,000-fold, have an alteration in
the phage protein involved in recognizing the
pac site (Raj et al., 1974). As a consequence
of this mutation, the HT mutants show an
altered specificity in making the first cut:
packaging initiates at a different, more
common site (Casjens et al., 1987). With
respect to packaging of the phage chromo-
some, it is observed that in HT mutants there
is a different distribution of cut sites. With
respect to generalized transduction, in the
HT mutants other, more numerous sites in
the bacterial chromosome can be used to
initiate packaging and hence the sites are no
longer as limiting a factor in the initiation of
packaging. This means that the host DNA
competes more effectively with phage DNA
for the packaging apparatus, and there is a
higher proportion (up to 50%) of phage par-
ticles that carry host DNA, resulting in a
higher frequency of transduction. In add-
ition there is less variation in the transduc-
tion frequency of different genes. Because of
these properties the HT mutants are ex-
tremely useful tools when transduction is
needed for genetic manipulations.
B. Formation of Generalized
Transducing Particles by Other Phages
Phage P22 is a useful model, both to illus-
trate a common mode of DNA metabolism
as well as its consequences for transduction.
However, among the phages that perform
generalized transduction, there are both
minor variations on this theme as well as
completely different modes of transduction.
Some examples are given below.
1. Phage P1
The E. coli phage P1 is in many ways as well
studied as P22 (see Sternberg and Hoess,
1983, for a review). Phage P1 is noteworthy
because it is the major generalized transdu-
cing phage used in E. coli and can infect a
broad range of hosts, making it an important
tool for genetic manipulation. In general, P1
is similar to P22 in using a processive headful
packaging process initiating at a specific site
to produce terminally redundant, circularly
permuted phage DNA. The P1 pac region
has been analyzed and is complex, compris-
ing a number of short sequence elements in a
161 bp region (Sternberg and Coulby, 1987).
The size of the P1 headful is about twice that
 GENETIC TRANSDUCTION
of P22, thus larger segments of the bacterial
chromosome can be transduced. There is less
variation in transduction frequency between
different genes with P1 than seen with P22.
This could result from the larger headful as
well as there being more packaging sites
in the bacterial chromosome. These results
have also been interpreted to mean that ini-
tiation of packaging of the bacterial chromo-
some by phage P1 does not use pac sites
like those found in the phage chromosome
(Sternberg, 1986).
2. Phage T4
As detailed by Guttman and Kutter (this
volume), phage T4 is not usually thought of
as a generalized transducing phage. T4 is a
virulent phage that degrades the host DNA
during infection and normally packages only
phage DNA by a headful packaging mech-
anism. This degradation is a result of an-
other feature of T4 DNA metabolism. T4
DNA is modified and contains glycosylated
hydroxymethylcytosine, a modification that
results from phage-encoded enzymes. The
degradation is due to a phage-encoded
nuclease that cuts DNA at unmodified cyto-
sines, generating substrates for other DNases
in the cell. A multiple mutant phage was
constructed (Wilson et al., 1979) with muta-
tions to prevent degradation of host DNA.
This phage can perform generalized trans-
duction, presumably packaging host DNA
as well as phage DNA because the mechan-
ism to distinguish these two has been re-
moved.
3. Phage l
Although l is not usually thought of as a
generalized transducing phage, it can pack-
age host DNA (Sternberg, 1986). This phage
does not use a headful packaging mechan-
ism. During infection, replication of l DNA
generates concatemers from a rolling circle,
like the phages described above. However
the packaging mechanism is site specific. A
particular sequence, cos, is recognized and
the DNA is cut (discussed below; Fig. 9).
Two cos sites must be present in the chromo-
567
some for packaging to occur. Cutting at cos
produces a double-stranded break that is not
blunt ended; a 12 nucleotide single-strand
end is produced that is necessary for the
chromosome to circularize in the next infec-
tion. This DNA processing system, although
suited for processing of the phage's DNA, is
apparently only loosely related to the mech-
anism of formation of generalized transdu-
cing particles. A study of the formation
of transducing particles suggests that l
packages host DNA without recognition
of specific cos sites (Sternberg, 1986). More-
over formation of transducing particles
is not seen until 60 to 90 minutes after in-
fection and is inhibited by the phage
exonuclease synthesized during infection.
Because of these properties, l transducing
particles are difficult to detect in a wild-
type infection, where cell lysis occurs after
about 60 minutes. Lysis defective mutants
(also mutant for the exonuclease) must be
used for maximal transducing particle for-
mation.
4. Phage Mu
Mu has quite a different DNA metabolism
from the phages discussed so far (see Whittle
and Salyers, this volume). Mu is a transpos-
able element that replicates by copying its
genome and integrating this copy at random
sites in bacterial DNA (replicative transpos-
ition; Fig. 5). During infection the cell thus
accumulates a number of different insertions
of the Mu chromosome. Packaging of these
Mu chromosomes is by a headful mechan-
ism. In this case the pac site is at the c end
and one cut is made in bacterial DNA about
100 nucleotides outside of this end while the
other cut is a fixed length in the direction of
the S end, from 500 to 3200 nucleotides out-
side of this end for the wild-type phage.
Hence the cuts are made in the bacterial
DNA that flanks the integrated Mu genome.
This genome structure is important for the
initial transposition event in subsequent
infections. Genetic studies (Howe, 1973)
showed that Mu is capable of generalized
transduction. Presumably pac sites in the
568 WEINSTOCK
Fig. 5. Phage Mu DNA metabolism. Shown at the
top is the linear phage Mu chromosome as it is carried
in the phage particle. The central (wavy) portion
represents phage genes while the terminal (straight)
portions flanking the phage c and s regions are bac-
terial DNA segments. Upon infection, the Mu seg-
ment of the chromosome transposes into the circular
bacterial chromosome. This integrated genome then
replicates and transposes to additional sites in the
chromosome. Finally, the integrated genomes are
packaged by a process that involves cutting in the
bacterial DNA that flanks the integrated Mu
genomes. This produces chromosomes that contain
different regions of the bacterial genome.
host chromosome are also recognized, lead-
ing to the formation of transducing particles.
C. Events in the Recipient
As stated earlier, generalized transduction
can be thought of as a two-stage process.
The varied ways in which the first stage,
packaging of host DNA into phage particles,
can occur has been described. In this section
the second stage of transduction is de-
scribedÐthe fate of the host DNA after it
has been introduced into a recipient cell.
When the packaged host DNA is injected
into a recipient cell, it can become stably
integrated into the recipient cell's chromo-
some by homologous recombination, or it
can remain unintegrated in the cytoplasm.
In this latter unintegrated case the DNA
will eventually be lost if it cannot replicate
in the cytoplasm (abortive transduction),
whereas it will be maintained as an extra-
chromosomal element if it can replicate as a
plasmid.
1. Homologous recombination with the
recipient's chromosome
Most of what is known about integration of
transducing fragments by recombination
comes from studies of phages P22 and P1.
In the case of P1 for example, linear double-
stranded DNA molecules are injected into
the recipient where they may assume a circu-
lar conformation due to the presence of a
bound protein. Studies with isotopically
labeled DNA showed that there is a physical
replacement of the recipient's DNA by the
transduced fragment, as described earlier. In
the case of transduction by both P1 and P22,
it was determined that this is a double-strand
replacement. That is, a double-stranded seg-
ment of the recipient's chromosome is re-
placed by the homologous double-stranded
segment from the transducing fragment.
Such an event requires two crossovers, one
at each end of the replaced segment (Fig. 6).
It has been estimated that only about 5% of
the DNA introduced as transducing frag-
ments ends up being incorporated into the
bacterial chromosome. This is in part be-
cause some fragments do not recombine at
all and remain in the cytoplasm. In addition,
for those fragments that do recombine, the
entire transducing fragment that is intro-
duced into the recipient need not integrate
into the recipient chromosome. The portion
that is integrated is determined by where the
two crossovers occur.
D. Measuring Transduction
In discussing the various fates of the DNA
in the recipient, it is also useful to consider
the practical problems in assaying for
 GENETIC TRANSDUCTION
Fig. 6. Recombination in the recipient: possible
outcomes of a three-factor transductional cross.
Shown at the top is a recipient chromosome (thin
lines) carrying mutations in the ilv, cya, and metE
genes and a donor transducing DNA fragment (thick
lines) that is wild-type at these three loci. In order to
integrate donor DNA into the recipient chromo-
some, two crossovers must occur, and these can
be in any of the four regions defined by the three
genes. At the bottom is shown the genotypes
resulting from crossovers in different regions. The
donor DNA (thick line) substituted for that of the
recipient (thin line) is shown for each case.
transduction. Imagine an experiment to
measure the transduction of a specific gene,
metE, that encodes the enzyme homocysteine
methylase, required for methionine biosyn-
thesis. The donor is metE ‡ and the recipient
is metE . After growth of a phage like P1 or
P22 on the donor strain, the resulting phage
lysate contains two kinds of particles (Fig.
2). The majority of particles carry phage
DNA and will cause phage production and
cell death upon subsequent infection. The
remaining particles carry host DNA from
all regions of the bacterial chromosome, so
only a small fraction of these carry the
metE ‡ gene. For P22, only a few percent of
the particles carry host DNA, and of these,
at most only a few percent carry metE. In the
transduction experiment, the phage lysate is
569
mixed with the recipient cells and, after the
phages have adsorbed to the bacteria, the
mixture is plated on a medium that selects
for metE ‡ cells. Most of the cells fail to grow
on this medium because they remain metE ,
but a few cells can grow and these form
colonies. These are the cells that have re-
ceived a transducing fragment carrying the
metE ‡ gene and this wild-type gene has
become integrated into the recipient chromo-
some by recombination, replacing the metE
gene and making the cell metE ‡ (Fig. 6). The
problem is to quantitate the frequency of
transducing the metE ‡ gene.
One factor that complicates this quantita-
tion is the multiplicity of infection (MOI), the
ratio of phage particles to bacterial cells
(Fig. 7). At low multiplicities (MOI  1)
most bacteria receive no phage particles and
those that are infected receive only a single
particle. Very few cells receive more than one
particle and thus the number of metE ‡ trans-
ductants observed is simply related to the
number of particles carrying metE ‡ DNA.
The number of transductants will be less
than the number of particles carrying metE ‡
gene because, as mentioned above, not all of
Fig. 7. Effect of multiplicity of infection (MOI) on
transduction frequency. This curve illustrates the fact
that at high MOI, where all cells are infected by a
particle containing phage DNA, there will be killing
of cells receiving transducing DNA, leading to a
decrease in frequency. Note that this effect will not
be as great if the phage can form lysogens. In that
case there will be some lysogen formation at high
MOI and correspondingly less killing. However, mu-
tations are often introduced into transducing phages
to prevent lysogenization.
570 WEINSTOCK
these fragments recombine with the bacterial
chromosome and those that do will some-
times not integrate the metE region of the
fragment. At high multiplicities (MOI  1)
all cells become infected and most cells re-
ceive more than one particle. In this case,
when a cell is infected by a particle that
carries a metE ‡ gene, it is simultaneously
infected by other particles, and these will
carry phage DNA since this is the major
class of particles present. The particles intro-
ducing phage DNA will usually cause the cell
to produce more phage and lyse, thus
making it impossible for the cell to become
a transductant. At high MOI the only cells
that can become transductants are those
that receive a transducing fragment and
either do not become infected by particles
carrying phage DNA, or if they do, the
phage infection leads to lysogeny rather
than lytic growth. As a result of this, fewer
transductants may be observed when large
amounts of lysate are used for transduction
(high MOI) than when lower amounts of
lysate are used (low MOI). The optimum
amount of lysate is somewhere near an
MOI of 1, where the largest amount of
lysate (and transducing particles) has been
added that does not cause significant loss of
transductants by killing due to multiple in-
fection with particles containing phage
DNA.
In measuring transduction frequencies,
two approaches can be used. In one method,
after the growth of the phage on the donor
strain, the resulting lysate is titered for par-
ticles containing phage DNA by measuring
the number of plaque-forming units (PFUs)
on a suitable indicator bacterium. Next the
phage lysate is titered for transducting par-
ticles by measuring the number of trans-
ductants in the volume of lysate added to
recipient cells. The ratio of transductants to
PFUs is then calculated, using the titer of
transductants determined in the low MOI
portion of the titration. This method allows
the frequency of transduction of two differ-
ent markers to be compared, by normalizing
them both to the total number of phage
DNA-containing particles. The alternative
method is to measure the frequency of trans-
duction of several different markers by
always using the same volume of lysate (i.e.,
the same MOI) in the transductions and ig-
noring the titer of phage DNA-containing
particles. In this case the ratio of transduc-
tants for each marker can be normalized to
that for a common marker, allowing them to
be compared.
1. Cotransduction of markers
One of the most important uses for general-
ized transduction is in measuring how far
apart markers are or in determining the
order of three or more markers in the
chromosome. An important concept for this
analysis is that of contransduction, the abil-
ity of two markers to be simultaneously inte-
grated into the recipient's chromosome on
the same transducting fragment (Figs. 4, 6).
Since phages like P1 and P22 package a
discrete length of DNA in their particles,
for two markers to be contransduced at all
requires that they not be farther apart in
the bacterial chromosome than the headful
length of DNA that the transducing phage
will package. Furthermore the closer to-
gether two markers are located, the higher
is the probability that they can be integrated
into the recipient chromosome together. This
probability is manifested operationally as the
contransduction frequency. In the example
of Figure 6, the donor strain is ilv‡ cya‡
metE ‡ , while the recipient is ilv cya
metE . When metE ‡ transductants are
selected, more of these will also be cya‡
(cotransduction of cya and metE) than will
be ilv‡ (cotransduction of ilv and metE) be-
cause cya is closer to metE than ilv is. More-
over the metE ‡ transductants that are also
ilv‡ will almost always be cya‡ because this
marker is in the middle (only rare double
recombinants will be ilv‡ cya metE ‡ ). In
contrast, of the metE ‡ transductants that
are also cya‡ , some will be ilv‡ and some
will be ilv . These results would suggest the
gene order ilv-cya-metE. This mapping tech-
nique is very similar to the three-factor cross
 GENETIC TRANSDUCTION
procedure using bacterial transformation (see
Streips, this volume).
There are actually two factors that contrib-
ute to the frequency of cotransduction of two
markers. One is the probability that they will
reside on the same transducting fragment
formed in the donor. The closer two markers
are to each other, the greater the chance that
they will be packaged together on a single
fragment. In addition, since the two cross-
overs resulting in integration in the recipient
can occur anywhere along the transducing
fragment, the closer together the markers
are, the greater the chance that they will be
integrated together. A formula relating dis-
tance to cotransduction frequency has been
derived, and the relation is nonlinear, reflect-
ing these complexities. A theoretical treat-
ment of this mapping is presented in
Mandecki et al. (1986).
2. Marker effects
Although generalized transduction is ``gen-
eral'' because any gene in the donor chromo-
some can be introduced into the recipient
chromosome, there are quantitative differ-
ences in the frequency with which different
genes can be transduced. This variation is
as much as 1000-fold for P22 transduction
and 10-fold for P1. Based on the previous
discussion of transduction, there are two
principal sources for variation. The first is
variation due to differences in the packaging
efficiency of different regions of the chromo-
some. This results from the requirement
for specific pac sequences, which would not
be expected to be placed periodically around
the chromosome. This appears to be the
major source of variation in P22 transduc-
tion. It is estimated that there are only 10 to
15 pac sites in the Salmonella chromosome,
and thus the transduction frequency of a
gene is strongly dependent on how close it
is to a pac site. In P22 HT mutants, which
can package from other sites, the variation
in transduction frequencies is only about
10-fold, and markers that are poorly trans-
duced by wild-type P22 often show the
greatest increases with HT. The second
571
source of variability is due to differences
in the efficiency of recombination of trans-
ducing fragments from different regions of
the chromosome. This appears to be the pri-
mary cause of variation in P1 transduction.
Treatments that stimulate recombination,
such as UV irradiation (see chapters by Yas-
bin and Levene and Huffman, this volume),
reduce the variability of P1 transduction fre-
quencies.
As a result of these factors, quantitating
transduction frequencies can be complex. An
example of this is the asymmetry of cotrans-
duction frequencies. In an experiment where
an ilv‡ metE ‡ donor and an ilv metE re-
cipient are used, when ilv‡ is selected, about
30% of these transductants are also found
to be metE ‡ . In contrast, when metE ‡ is
selected, about 70% of these are found to
be ilv‡ . Thus, although the distance is the
same between these markers in the two
crosses, different cotransduction frequencies
are obtained depending on the selection.
This could reflect either asymmetries in
packaging or recombination. In the case
of a packaging asymmetry (Fig. 8), it is pos-
sible that there are two types of transducing
fragments carrying the ilv‡ gene: a major
Fig. 8. Hypothetical classes of transducing frag-
ments in the ilv-metE region and their relative
abundance. When an ilv =metE recipient is trans-
duced to ilv ‡ , about 30% of these transductants
are also metE‡ . In contrast, when metE‡ is
selected, about 70% of the transductants are ilv ‡ .
To account for the asymmetry by differential pack-
aging in the donor, it is proposed that there are
three types of transducing fragments packaged
from this region: a major one containing only ilv,
a minor one carrying both ilv and metE, and a very
minor one with only metE.
572 WEINSTOCK
one that does not contain the metE gene and
a minor one that carries metE. Hence the
majority of the recombinants formed when
ilv‡ is selected will retain the recipient's
metE marker. On the other hand, there
may be more metE ‡ transducing fragments
that carry the ilv‡ marker than those that
do not carry ilv, resulting in the higher
frequency of ilv‡ metE ‡ transductants. It
is also possible to account for this difference
in cotransduction based on hot spots in
recombination. From these considerations
it is clear that care must be taken in inter-
preting data based solely on transduction
frequencies. At the minimum, reciprocal
crosses must be performed to determine if
there are marker effects. Ideally multiple-
factor crosses should be used (like the ilv
cya metE example above) to determine gene
order.
3. Abortive transduction
Not all of the DNA fragments introduced
into a cell by a generalized transducing
phage recombine and integrate into the
bacterial chromosome; the fraction that do
recombine are in fact the minority. The ma-
jority of transducing fragments remain in the
cytoplasm of the recipient cell, although the
efficiency of recombination depends on the
region (Schmieger, 1982). These fragments
do not replicate, since they do not contain
an origin of replication and thus are in-
herited by only one cell at each division.
Eventually these fragments are lost from
the culture because very few cells carry an
unintegrated, nonreplicating fragment and
because they can be degraded. This phenom-
enon of creating unstable transductants with
unintegrated fragments is called abortive
transduction.
The genes in an unintegrated fragment can
be expressed and thus the cell is functionally
diploid for this region of its chromosome. If
the cell is mutant for a gene in the diploid
region, while the fragment is wild-type, com-
plementation will be observed. At each cell
division the functional product generated in
the diploid can be distributed into the two
daughter cells, and thus the daughter cell
that does not receive the fragment may
remain phenotypically wild-type until the
functional product is degraded or diluted
through subsequent cell divisions. Typically,
in a transduction of an auxotrophic recipient
to prototrophy, a large number of microco-
lonies will be observed on minimal medium.
These are abortive transductants that grow
by virtue of the small amount of nutrient
made by the diploid. In contrast, the stable
transductants (where the fragment has re-
combined into the bacterial chromosome)
form normal-sized colonies. Another classic
and beautiful example of this phenomenon is
the behavor of nonmotile mutants (Leder-
berg, 1956; Stocker, 1956). In this case the
diploid formed by abortive transduction is
motile. On a plate with a low agar concen-
tration the diploid can move, but at each cell
division a nonmotile daughter is produced
that will grow into a colony. Eventually a
trail of these nonmotile daughters is made,
called a flare, which marks the path of the
original abortive transductant.
The unintegrated DNA introduced by
transduction is quite stable in the cell. In
the case of P1 transduction, when this
DNA is isolated from a recipient it is found
to be circular, even though the DNA was
linear in the phage particle (Sandri and Ber-
ger, 1980). The circles can be disrupted by
treatments with detergent or protease, imply-
ing that the circle is held together by a pro-
tein linker. Evidence for protein attached to
the DNA in P1 transducing particles has in
fact been reported (Ikeda and Tomizawa,
1965b). Normally, a linear double-stranded
DNA molecule is rapidly degraded in vivo
by the action of the RecBCD nuclease. Thus
it appears that the protein at the DNA end
may be a special mechanism to protect trans-
ducing DNA from this degradation.
4. Transduction of plasmids
Just as the bacterial chromosome is a sub-
strate for packaging by generalized transdu-
cing phages, so is any other DNA that is
inside an infected cell, including plasmid
 GENETIC TRANSDUCTION
chromosomes. When the plasmid is larger
than a phage headful, only a part of the plas-
mid can be packaged. This is the case with R
factors, which are much larger than a headful
of a phage like P22. When this occurs, the
fragment of the plasmid that is introduced
into the recipient cannot circularize efficiently
because there is no homology between the
fragment ends or with cellular sequences to
allow recombination. Thus the majority of
transduced plasmid fragments will be lost
from the recipient. Nevertheless, at a low fre-
quency it is possible to recover deleted ver-
sions of the plasmid following transduction.
These represent fragments that retain the plas-
mid origin of replication. It is possible that
these are plasmids that acquired a deletion in
the donor that made them small enough to fit
completely into a phage head. Alternatively,
these may be fragments that were introduced
into the recipient where they circularized by
some inefficient ``illegitimate'' recombination
event. This process is called transductional
shortening (Shipley and Olsen, 1975; also see
the review by Low and Porter, 1978).
A different situation is found when the
plasmid is much smaller than the phage
headful. In this case, the plasmid is packaged
inefficiently because the chromosome is too
small to form a stable head. However, multi-
meric forms of the plasmid can be packaged
efficiently, as shown for transduction of
plasmids by P22HT mutants (for a review
see Margolin, 1987) and T4 transducing
phages (Takahashi and Saito, 1982). The
multimeric plasmid chromosomes may be
preexisting in the donor or may be caused
by the phage infection, either by phage-
promoted recombination between plasmids
or rolling circle replication of the plasmid.
In these cases the multimers that are intro-
duced into the recipient circularize by recom-
bination between the repeated plasmid units.
A final case of plasmid transduction in-
volves plasmids that are smaller than a
phage headful but contain sequences that
are homologous to the bacterial chromo-
some. Examples of this are small plasmid
vectors carrying cloned genes. These can re-
573
combine with the donor's chromosome and
become integrated. This integrated plasmid
may then be packaged and transduced by the
usual mechanism (Trun and Silhavy, 1987).
Once in the recipient, the plasmid can recom-
bine out of the chromosome by the reverse
of the integration process, leading to a trans-
ductant containing the free plasmid. This
is a useful mechanism for selecting for re-
combination between a plasmid and the host
chromosome.
E. Uses for Generalized Transduction
In the preceding sections several uses for
generalized transduction have been described.
These include genetic mapping, complemen-
tation analysis, transduction of plasmids,
selection for deletions in plasmids, and selec-
tion for recombination between cloned genes
and the bacterial chromosome. Other
common uses include strain constructions,
delivery of transposons, and isolation of
chromosome rearrangements such as dupli-
cations (see Margolin, 1987, for a review).
III. SPECIALIZED
TRANSDUCTION
Specialized transduction is the second major
class of virus-mediated genetic exchange. It
differs from generalized transduction in two
respects. First, in this mode of transduction
the transduced genes are covalently joined to
the viral chromosome, allowing them to be
replicated, packaged, and introduced into a
recipient with the rest of the viral chromo-
some. Second, a specialized transducing par-
ticle carries a specific chromosome segment,
and consistently only introduces this set of
genes into the recipient. Hence the name,
specialized transduction (see Hendrix, this
volume, for a more complete discussion).
A. Bacteriophage l DNA Metabolism
The temperate phage l (Fig. 9 and Appen-
dix) is the classic example of a specialized
transducing phage. The l phage particle
contains a linear double-stranded DNA
molecule with complementary 12 nucleo-
tide single-stranded ends (see Furth and
574 WEINSTOCK
Fig. 9. Phage l DNA metabolism. When l infects a
cell (1) the phage DNA circularizes by virtue of the
single-stranded cohesive ends (cos) and the cell can
have two fates. If the phage grows lytically (2), the
chromosome replicates to form concatemers, which
are then packaged to form new phage (3). Alterna-
tively (4), the phage chromosome can integrate into
the host chromosome during the lysogenic response.
This is accomplished by site-specific recombination
between the attP site of the phage chromosome and
the attB site of the host (5). When a lysogen induces,
the phage genome excises by a reverse of the integra-
tion process (6), leading to a wild-type chromosome
that can replicate to form concatemers (8) and
normal phage. However, on rare occasions the exci-
sion process is aberrant (7) and the recombination
does not occur between the att sites. As a result a
chromosome is formed that has lost some phage
material and acquired some host sequences. As long
as none of the missing phage genes were essential, this
hybrid chromosome is capable of replicating normally
(8) and producing new phage. These phages are all
specialized transducing phages since they carry host
genes in their chromosomes.
Wickner, 1983, for a review of l DNA me-
tabolism). When l infects a cell, its chromo-
some circularizes by hybridization of these
single-strand extensions (cos sites), and the
phage then chooses between two alternative
life cycles. In the lytic response, the chromo-
some replicates to form concatemers that are
packaged into particles by cutting at the cos
sites, reforming the single-stranded ends.
This packaging differs from the headful
mechanism in that the size of the DNA that
is encapsidated is determined by the distance
between the cos sites and can vary consider-
ably. This is the basis for cosmid cloning
vectors (see Geoghegan, this volume). How-
ever, there are limits on the size of the pack-
aged chromosome. DNA molecules that are
larger than about 50,000 bp, or smaller than
about 35,000 bp, produce unstable phage
particles. The alternative life cycle is the ly-
sogenic response. In this case the phage
chromosome integrates into the bacteria
chromosome and a repressor is produced
that prevents expression of the lytic gene
products. The phage chromosome is repli-
cated and passively carried by the bacterium
as a latent prophage. The integration reac-
tion is performed by a phage-encoded site-
specific recombination system that breaks
the phage chromosome at a specific se-
quence, attP, and joins it to the host at an-
other specific sequence, attB (see Weisberg
and Landy, 1983, for a review). When the
prophage becomes induced, the phage re-
pressor is inactivated and the reverse reac-
tion occurs. The site-specific recombination
system acts at the attP/attB joints to reform
a circular chromosome that undergoes the
lytic cycle.
Specialized transducing particles are
formed in vivo when an integrated l proph-
age excises from the bacterial chromosome
imprecisely (Fig. 9). In this case the breaking
and joining events do not occur at the attP/
attB joints, but are the result of ``illegitim-
ate'' recombination events occurring at other
sites in the vicinity of the prophage. This
process occurs at low frequency and has little
sequence specificity. Markers either to the
left or right of the prophage can be trans-
duced. The result is a circular chromosome
that has lost some of the phage DNA se-
quences from one end of the prophage and
in their place carries bacterial DNA that was
 GENETIC TRANSDUCTION
adjacent to the other prophage end. This
hybrid chromosome can replicate and be
packaged, provided that the required lytic
functions and cos site are present, and results
in the formation of transducing particles that
carry and can introduce the incorporated
bacterial genes. Some specialized transdu-
cing phages carry the complete array of
genes and sites necessary for lytic growth.
These are ``plaque-forming'' phages, and
this property is usually designated by includ-
ing a ``p'' in their name, such as l pgal, a
plaque-forming gal transducing phage. There
can also be defective transducing phages.
These are phages that have lost essential
genes from their chromosome in the process
of excision. The minimum requirement for a
phage to be packaged and maintained is that
it contain the cos site and the origin of repli-
cation. All other functions can be provided
by helper phages. To designate that a trans-
ducing phage is defective and requires a
helper for growth, a ``d'' is included in its
name, such as lpgal, a defective gal transdu-
cing phage.
From this mechanism it is clear that only a
restricted set of bacterial genes can be trans-
duced. This is a consequence of the fact that
only genes near the attB site can be incorpor-
ated, as well as the limits on the size of the
chromosome that can give a stable phage
particle. However, a number of methods
have been used to allow a broader range of
genes to be transduced by l. Rearrange-
ments of the bacterial chromosome, such as
deletions, can bring different genes close
enough to the attB site to be transduced. A
second method relies on the observation that
when the attB site itself is deleted, l now
integrates at a large number of secondary
attachment sites. These sites are not used as
efficiently as attB, presumably because their
sequences deviate from attB. However, lyso-
gens with prophages located at these second-
ary sites can be used to isolate specialized
transducing phages carrying the neighboring
genes. In a third method, the site of integra-
tion of the phage chromosome does not
depend on the site-specific recombination
575
system at all, but rather utilizes transposons
(see Whittle and Salyers, this volume). If a
transposon insertion can be isolated in the
bacterial chromosome near a gene of inter-
est, then a phage chromosome that also
carries an insertion of the transposon can
integrate at this site by homologous recom-
bination. This is similar to generating F plas-
mid integration sites using transposon Tn10
insertions (see Porter, this volume). Finally,
derivatives of l have been developed that
contain the transposition system from phage
Mu (see Whittle and Salyers, this volume)
instead of the l site-specific recombination
system (Bremer et al., 1984). These phages
behave like l except that integration during
lysogenization uses the Mu system and there
is no sequence specificity. As a result pro-
phages can be isolated anywhere around the
bacterial chromosome. This is the most gen-
eral in vivo method, and it is as general as the
in vitro method for forming specialized
transducing phages: molecular cloning using
recombinant DNA methodology.
B. Events in the Recipient
When a specialized transducing phage like l
stably transduces a recipient cell, it forms a
lysogen that is, it integrates into the chromo-
some and expression of lytic functions is
repressed. There are several different ways
this can occur. If the phage has an intact
attP site and int gene (the gene needed for
site-specific integration at attB), and the re-
pression system of the phage is intact, lyso-
geny can follow the same steps as for a wild-
type l phage. Often, however, the transdu-
cing phage does not have attP or int, these
having been lost during formation of the
transducing phage genome. In this case inte-
gration can occur through homologous
recombination between the bacterial chro-
mosomal sequences carried by the phage
and its counterpart in the host chromosome.
In some cases, if the recipient is lysogenized
by another wild-type helper l phage, this
prophage can also provide a homologous
region for recombination with the transdu-
cing phage. Finally, under some conditions,
576 WEINSTOCK
the transducing phage can be maintained as
an un-integrated plasmid in the cell (see
Weisberg, 1987).
C. Uses for Specialized Transduction
Specialized transducing phages provide one
of the most convenient ways of manipulating
genes. They represent the original method
for gene cloning, using in vivo genetic tech-
niques. Because these phages carry a small
region of the chromosome, genes carried on
phages can be mapped with great precision,
as can mutations within these genes. More-
over mutagenesis can be targeted to specific
genes. These phages also provide a con-
venient means for constructing strains that
are diploid for a very limited region of the
chromosome, useful for complementation
analysis.
REFERENCES
Bremer E, Silhavy TJ, Weisemann JM, Weinstock GM
(1984): l placMu: A transposable derivative of bac-
teriophage lambda for creating lacZ protein fusions in
a single step. J Bacteriol 158:1084±1093.
Casjens S, Huang WM, Hayden M, Parr R (1987):
Initiation of bacteriophage P22 packaging series. An-
alysis of a mutant that alters the DNA target specifi-
city of the packaging apparatus. J Mol Biol 194:411±
422.
Chelala CA, Margolin P (1974): Effects of deletions on
cotransduction linkage in Salmonella typhimurium:
Evidence that bacterial chromosome deletions affect
the formation of transducing DNA fragments. Mol
Gen Genet 131:97±112.
Ebel-Tsipis J, Botstein D, Fox MS (1972): Generalized
transduction by phage P22 in Salmonella typhimur-
ium. I. Molecular origin of transducing DNA. J Mol
Biol 71:433±448.
Furth M, Wickner S (1983): Lambda DNA replication.
In Hendrix R, Roberts J, Stahl F, Weisberg R (eds):
``Lambda II.'' Cold Spring Harbor, NY: Cold Spring
Harbor Laboratory, pp 145±174.
Hendrix R, Roberts J, Stahl F, Weisberg R (eds) (1983):
``Lambda II.'' Cold Spring Harbor, NY: Cold Spring
Harbor Laboratory.
Howe M (1973): Transduction by bacteriophage Mu-1.
Virology 55:103±117.
Ikeda H, Tomizawa J (1965a): Transducing fragments in
generalized transduction by phage P1. I. Molecular
origin of the fragments. J Mol Biol 14:85±109.
Ikeda H, Tomizawa J (1965b): Transducing fragments
in generalized transduction by phage P1. II. Associ-
ation of protein and DNA in the fragments. J Mol
Biol 14:110±119.
Lederberg J (1956): Linear inheritance in transductional
clones. Genetics 41:845±871.
Low KB, Porter DD (1978):Modes of gene transfer and
recombination in bacteria. Annu Rev Genet 12:249±
287.
Mandecki W, Krajewska-Grynkiewicz K, Klopotowski
T (1986): A quantitative model for nonrandom gener-
alized transduction, applied to the phage P22-Salmon-
ella typhimurium system. Genetics 114:633±657.
Margolin P (1987): Generalized transduction. In Ingra-
ham JL, Low KB, Magasanik B, Neidhardt FC,
Schaechter M, Umbarger HE (eds): ``Escherichia coli
and Salmonella typhimurlum: Cellular and Molecular
Biology.'' Washington, DC: American Society for
Microbiology, pp 1154±1168.
Ptashne M (1985): ``A Genetic Switch.'' Boston: Black-
well Scientific Publications.
Raj AS, Raj AY, Schmieger H (1974): Phage genes
involved in the formation of generalized transducing
particles in Salmonella-phage P22. Mol Gen Genet
135:175±184.
Sandri RM, Berger H (1980): Bacteriophage P1-
mediated generalized transduction in Escherichia
coli: Structure of abortively transduced DNA. Vir-
ology 106:30±40.
Schmieger H (1972): Phage P22 mutants with increased
or decreased transduction abilities. Mol Gen Genet
119:75±88.
Schmieger H (1982): Packaging signals for phage P22 on
the chromosome of Salmonella typhimurium. Mol Gen
Genet 187:516±518.
Shipley PL, Olsen RH (1975): Isolation of a nontrans-
missible antibiotic resistance plasmid by transduc-
tional shortening of R factor RP1. J Bacteriol
123:20±27.
Sternberg N (1986): The production of generalized
transducing phage by bacteriophage lambda. Gene
50:69±85.
Sternberg N, Coulby J (1987): Recognition and cleavage
of the bacteriophage P1 packaging site (pac). II. Func-
tional limits of pac and location of pac cleavage ter-
mini. J Mol Biol 194:469±479.
Sternberg N, Hoess R (1983): The molecular genetics of
bacteriophage P1. Annu Rev Genet 17:123±154.
Stocker BAD (1956): Abortive transduction of motility
in Salmonella: A nonreplicated gene transmitted
through many generations to a single descendant. J
Gen Microbiol 15:575±598.
Susskind MM, Botstein D (1978): Molecular genetics of
bacteriophage P22. Microbiol Rev 42:385±413.
Takahashi H, Saito H (1982): High-frequency transduc-
tion of pBR322 by cytosine-substituted T4 bacterio-
phage: Evidence for encapsulation and transfer of
head-to-tail plasmid concatemers. Plasmid 8:29±35.
 GENETIC TRANSDUCTION
Trun NJ, Silhavy TJ (1987): Characterization and in vivo
cloning of prlC, a suppressor of signal sequence mu-
tations in Escherichia coli K12. Genetics 116:513±521.
Weisberg RA (1987): Specialized transduction. In Ingra-
ham JL, Low KB, Magasanik B, Neidhardt FC,
Schaechter M, Umbarger HE (eds): ``Escherichia coli
and Salmonella tymphimurium: Cellular and Molecu-
lar Biology.'' Washington, DC: American Society for
Microbiology, pp 1169±1176.
Weisberg R, Landy A (1983): Site-specific recombin-
ation. In Hendrix R, Roberts J, Stahl F, Weisberg R
(eds): ``Lambda II.'' Cold Spring Harbor, NY: Cold
Spring Harbor Laboratory, pp 211±250.
Wilson GG, Young KKY, Edlin GJ, Konigsberg W
(1979): High frequency generalized transduction by
bacteriophage T4. Nature 280:80±82.
Zinder ND, Lederberg J (1952): Genetic exchange in
Salmonella. J Bacteriol 64:679±699.
APPENDIX:
BACTERIOPHAGE l
The bacteriophage l is one of the most thor-
oughly studied systems in biology. The DNA
sequence of the entire phage l chromosome
has been determined (over 48,500 bp), and
this organism has served as a paradigm for
many phenomena in addition to genetic
transduction. Much of the information
about phage l is summarized in the mono-
graph Lambda II (Hendrix et al., 1983). An-
other good reference is the book A Genetic
Switch (Ptashne, 1985). The metabolism of l
DNA during infection has been described
above. Here we present an overview of the
genetic organization of the phage and some
of the important aspects of the control of
phage l gene expression to complement the
chapter by Hendrix, this volume.
Genetic Organization
The l chromosome encodes about 50 genes
that are organized in functionally related
clusters. At the left end of the chromosome
(Fig. A1) is a block of 10 genes that is re-
quired for production of the phage head
followed by 11 genes needed to make the
phage tail. The central region of the chromo-
some contains 19 nonessential genes that are
also functionally clustered. Included in this
region are the genetic elements of the site-
specific (int, xis, and the att site) and hom-
577
ologous (exo and bet) recombination systems
of phage l as well as a number of genes
whose products are involved in interactions
with the host. Next is a short stretch contain-
ing the important control genes N, cl, cro,
and cll, followed by the DNA replication
region (genes O and P and the origin of
replication), the controller of late gene ex-
pression, Q, and the genes for cell lysis, S,
R, and Rz. This overall functional arrange-
ment along the chromosome is followed by a
number of other phages of E. coli (e.g.,
phages 434 and 21) as well as phage P22 of
Salmonella. Because of this relatedness, these
phages are collectively referred to as the lam-
boid phages, and it has been possible to
create hybrids in which functions in one
lambdoid phage are replaced with the analo-
gous genes from another phage in this race.
Transcriptional Units
There are four major transcriptional units in
the phage (Fig. A1). The promoter PL is used
for the early leftward operon and the pro-
moter PR is used for the early rightward
operon. The genes in these two operons are
made early in infection. The PR0 promoter
(sometimes called the late promoter) drives a
large operon consisting of genes that are
expressed late in infection, such as those for
head and tail production and cell lysis. Last
is the transcription unit that contains the cl
gene, encoding the l repressor. There are
two promoters for this purpose: PRE , which
is used early in infection, and PRM , which is
used in a repressed lysogenic cell. In addition
to these is the P1 promoter, which produces a
small RNA to augment the expression of the
int gene, and the Po promoter, which makes
another small RNA of unknown function.
Lytic Growth
In the lytic response, l replicates its DNA,
packages the new chromosomes into phage
heads, adds phage tails to these, and finally
lyses the cell. The genes for head and tail
production and cell lysis are contained in
the later operon, expressed from the PR0 pro-
moter. However, essential control genes and
Fig. A1. Phage l genetic map. The details of the map are described in the appendix. Arrows indicate direction of transcriHption of the different operons.
 GENETIC TRANSDUCTION
the DNA replication functions, located in
the early leftward and rightward operons,
must be expressed first.
The leftward operon contains the essential
N gene, which is an antiterminator of tran-
scription. Transcription termination signals
are located throughout the leftward and right-
ward operons. There is a terminator immedi-
ately after the N gene as well as at other
downstream sites in this operon. In the right-
ward operon there are terminators after the
cro and P genes. The N gene product prevents
transcription from terminating at these sites
and thus allows expression of downstream
genes. This provides a timing mechanism.
Genes preceding the terminators (N and cro)
are early genes, while the others are sometimes
called delayed-early genes, since their expres-
sion is delayed until N protein is available.
In the rightward operon there are several
essential functions for lytic development.
First are the DNA replication genes O and
P and the origin of replication. These are
located after the first transcription termin-
ation signal in this operon. In addition is
the Q gene, located after the second termin-
ator, which thus delays its expression fur-
ther. The Q gene product is required for the
expression of all late proteins (i.e., produc-
tion of heads, tails, and cell lysis). This is
because there is a transcription terminator
just after the late promoter P0R and immedi-
ately before all of the late genes. The Q
protein is a transcription antiterminator
that is specific for this site and thus controls
late gene expression. The delay in expression
of Q contributes to the delay in expression of
the late genes and ensures that plenty of
DNA has accumulated before the packaging
and lysis processes begin.
The Lysogenic Response
In the lysogenic response l turns off the lytic
genes by the action of a single repressor and
integrates its chromosome into the bacterial
579
chromosome. It is passively carried in this
latent state as a prophage until it becomes
induced. The repressor that turns off all of
the lytic genes is the product of the cl gene. It
binds to the phage chromosome at two
regions: OL and OR (Fig. A1). Each of these
is a cluster of three operator sites and bind-
ing of the l repressor causes repression of
transcription from the PL and PR promoters.
This in turn prevents all lytic gene expression
by preventing expression of the control genes
N and Q.
Expression of the cl gene involves two
promoters. The PRE promoter (for the estab-
lishment of repression) is used early in infec-
tion when the decision between the lytic and
lysogenic choices has not been resolved. This
is a strong promoter that is stimulated by the
product of the cll gene, located in the early
rightward operon. This promoter allows a
lot of repressor to be made quickly. Once a
lysogen has been established, PRE is not used
and expression switches to the PRM pro-
moter (for maintenance of repression). This
is a weaker promoter that produces enough
repressor to maintain the repressed state. It
can be either stimulated or inhibited by l
repressor bound at OR (depending on the
repressor concentration) and thus allows
for buffering of the amount of repressor pro-
duced in the lysogen.
The other important control protein in the
l system is the cro gene product, which
serves to counteract the effect of cl. The cro
protein is also a repressor and it also binds at
the sites at OR and OL . However, its binding
characteristics are different from cl and as a
result it can antagonize cl binding and lead
to a lytic response. The decision between the
lytic and lysogenic response involves the
competition for control of the OR and OL
sites by cl and cro. This competition is ultim-
ately decided by the physiology of the host,
which influences expression of these genes in
ways that are not at present clear.
isotope effect, reported by Baker (6, 7),
for B-NADD of 0.50 + 0.10 based on
kinetic measurements, it is in complete
agreement with our most recent kinetic
results (9). Our observations also indi-
cate that the B-NADD preparation is
more sensitive to deterioration on stor-
age than the usual NADH sample. This
is probably the cause of the discrepancy
between Baker's results and those re-
ported here.
The isotope effect for equilibrium
dissociation constants with A-NADD
was determined by both the Sephadex
and the fluorimetric methods, and the
two results are identical within experi-
mental error:
Kd -KdE-NADIT/KdE-(A-NADD)
= 0.65 ± 0.14 (Sephadex method)
= 0.52 ± 0.07 (fluorimetric method)
This kind of ("inverse") isotope effect
indicates that A-NADD is bound less
tightly to the enzyme than is NADH:
ki
LADH + NADH = LADH - NADH
k2
and that the zero point energies for the
isotopically sensitive vibrations are low-
er in the complex than in the free
NADD.
Since Kd (observed) = KH/KD =
k2/k1; k, - klHIk1D, k2 -- k2E/k2D,
then if Kd < 1 and I2 > 1 it follows
that k1 > k2 > 1 (conversely for k2 < 1,
k, < k2). Therefore if the isotope effect
on the dissociation reaction (rate con-
stant k2) is normal (k2 > 1), a possi-
bility suggested from kinetic data (6,
7, 9), then the isotope effect for the
binding step (rate constant kl) must
also be normal and even larger, thus
indicating an ability of the enzyme to
discriminate kinetically against the co-
enzyme deuterated at the transferable
hydrogen.
Previously Mahler et al. (7) reported
Kd for the LADH-(A-NADD) dissocia-
tion constant to be 1.31 + 0.64; how-
ever, this value was based on an extrap-
olation from kinetic constants derived
from initial rate measurement and as
indicated, exhibited a large experimental
uncertainty. Isotope effects for the cor-
responding equilibrium constants for
yeast alcohol dehydrogenase (1) and
malate dehydrogenase (3) were reported
to be of the same magnitude as that
for LADH. However, "inverse" deu-
terium isotope effects similar to the one
reported here, but based on inverse
rate effects for kl, were reported from
Thomson's laboratory (4) for the bind-
480
ing of A-NADD to heart lactate de-
hydrogenase (Kd = 0.69) and muscle
lactate dehydrogenase (Kd = 0.43).
These various results in conjunction
with the absence of any real difference
in the binding constant between NADH
and B-NADD, which eliminates the
possibility of secondary interaction be-
tween the B proton and the enzyme,
indicate that dehydrogenases are capa-
ble of recognizing and discriminating
against coenzymes isotopically substi-
tuted in the appropriate position ("A"
for A enzymes) even in the initial bind-
ing reaction. The properly deuterated
coenzyme is bound less tightly than is
its protonated counterpart, so that dur-
ing the binding reaction the equilibrium
(and perhaps also the rate) constant
already reflects the specificity of the
subsequent hydrogen transfer step.
KAREN BUSH, H. R. MAHLER
V. J. SHINER, JR.
Department of Chemistry,
Indiana University, Bloomington 47401
References and Notes
1. V. J. Shiner, Jr., H. R. Mahler, R. H. Baker,
Jr., R. R. Hiatt, Ann. N.Y. Acad. Sci. 84,
583 (1960).
2. H. R. Mabler and J. Douglas, J. Amer.
Chem. Soc. 79, 1159 (1957); D. Palm, Z.
Naturforsch. 21b, 540, 547 (1966).
3. J. F. Thomson, D. A. Bray, J. J. Bummert,
Biochenm. Pharmacol. 11, 943 (1962).
4. J. F. Thomson, J. J. Darling, L. F. Bordner,
Biochim. Biophys. Acta 85, 177 (1964); J. F.
Thomson and S. L. Nance, ibid. 99, 369
(1965).
5. D. Palm, Euir. J. Biochem. 5, 270 (1968).
Bacteriophage and the Toxigenicity of
Clostridium botulinum Type C
Abstract. Nontoxigenic and bacteriophage-sensitive bacterial cultures have been
isolated from toxigenic Clostridium botulinum type C, strain 468C, after treat-
ment with either ultraviolet light or acridine orange. Two bacteriophages, desig-
nated CE/3 and CE-y, were isolated from toxigenic strain 468C. Both of these
bacteriophages were capable of infecting the nontoxigenic type C cultures, but
only bacteriophage CE/ was involved in the change from nontoxigenicity to
toxigenicity.
Nontoxigenic clostridia resembling
toxigenic Clostridium botulinum have
frequently been isolated from marine
and terrestrial environments (1); oc-
casionally, pure cultures of toxigenic C.
botulinum have also become nontoxi-
genic (2). It has been suggested that the
production of toxin by C. botulinum
might be governed by the presence of
a specific converting bacteriophage in a
system analogous to the production of
diphtheria toxin by Corynebacterium
diphtheriae (3, 4). In a recent report,
6. R. H. Baker, Jr., Biochemistry 1, 41 (1962).
7. H. R. Mahler, R. H. Baker, V. J. Shiner, Jr.,
ibid., p. 47.
8. D. Palm, T. Fielder, D. Ruhrseitz, Z.
Naturforsch. 23b, 623 (1968).
9. K. Bush, H. R. Mahler, V. J. Shiner, Jr.,
in preparation.
10. Abbreviations: LADH, horse liver alcohol
dehydrogenase; A-NADD and B-NADD,
NADH with deuterium in the 4-A and 4-B
nicotinamide positions, respectively; Kd, dis-
sociation constant.
11. H. Theorell, A. Ehrenberg, C. de Zalenski.
Biochem. Biophys. Res. Commun. 27, 309
(1967).
12. R. Pietruszko, H. J. Ringold, T. K. Li, B.
L. Vallee, A. Akeson, H. Theorell, Nature
221, 440 (1969).
13. S. Taniguchi, H. Theorell, A. Akeson, Acta
Chem. Scand. 21, 1903 (1967); H. Theorell,
A. Akeson, B. Liszka-Kopec, C. de Zalenski,
Arch. Biochem. Biophys. 139, 241 (1970).
14. K. Dalziel, Acta Chem. Scand. 12, 459 (1958).
15. S. Taniguchi, ibid. 21, 1511 (1967).
16. S. P. Colowick and N. 0. Kaplan, Methods
Enzymol. 4, 840 (1957).
17. A. D. Winer, J. Biol. Chem. 239, PC 3598
(1964).
18. J. P. Hummel and W. J. Dreyer, Biochim.
Biophys. Acta 63, 530 (1962).
19. R. W. Green and R. H. McKay, J. Biol.
Chem. 244, 5034 (1969); H. Jornwall, Eur.
J. Biochem. 16, 25 (1970).
20. G. F. Fairclough, Jr., and J. S. Fruton, Bio-
chemistry 5, 673 (1966).
20. H. Theorell and A. D. Winer, Arch. Bio-
chem. Biophys. 83, 291 (1959).
22. S. R. Anderson and G. Weber, Biochemistry
4, 1948 (1965).
23. K. Dalziel, Acta Chem. Scand. 17, S27 (1963).
24. H. Theorell and J. S. McKinley-McKee, ibid.
15, 1811 (1961).
25. We thank Mrs. Genevieve Adams for techni-
cal assistance in the preparation of the deu-
terated coenzymes and the laboratory of Dr.
Drew Schwartz for perforniing the starch-
gel electrophoresis. We thank Dr. Henry
Weiner of Purdue University who donated his
laboratory facilities and computer program
for the fluorimetric determinations. Supported
in part by grant RG-8502 from NIH and by
grant AT(11-1)-1008 from AEC (Document
No. COO-1008-17).
9 November 1970; revised 14 December 1970 U
Inoue and lida (5) presented data which
strongly suggested that bacteriophage
may be involved in the toxigenicity of C.
botulinum type C. They were able to
recover toxigenic isolates from nontoxi-
genic cultures incubated in broth with
filtrates of a toxigenic type C (strain
Stockholm). Marked lysis occurred in
the cultures; however, plaques were not
demonstrated on solid medium.
In our studies, two bacteriophages,
designated CE,8 and CE-y, were isolated
from toxigenic C. botulinum type C,
SCIENCE, VOL. 172
 strain 468C (received from W. P. Seg- Table 1. Reinfecting cured nontoxigenic bac- of its prophage. After a 60-second treat-
ner, Continental Can Co., Chicago). terial cultures with bacteriophage CEj8 and the ment with ultraviolet light, 15 of the 106
effect on toxigenicity. UV, ultraviolet.
This report presents conclusive evidence surviving colonies were cured of their
for the involvement of CE/3 in the toxi- Nontoxi- Method of No. of cultures
genic obtaining prophages and concomitantly became
genicity of C. botulinum type C. Bac- cured toxigenic Toxi- Tested nontoxigenic. In comparison, only 2 of
teriophage-producing cultures were ob- culture cultures genic the 68 colonies surviving the AO treat-
tained from nontoxigenic type C cultures A028 Plaques* 40 40 ment were cured of their prophages and
that had been reinfected with either bac- A028 30-minute 4 40 toxigenic character. The cured cultures
teriophage. However, only bacteriophage contactt were subcultured ten times over a period
CE,8 caused the change from nontoxi- A028 240-minute 40 40 of 2 months, and all of them remained
contactt
genicity to toxigenicity, and these toxi- UV171 Plaques 20 20 nontoxigenic and sensitive to the bac-
genic isolates invariably carried bac- UV171 30-minute 2 40 teriophages of the toxigenic parent strain
teriophage CE,8. Cultivation of the contact 468C.
CE,8-infected isolates in medium con- UV171 240-minute 40 40 Each nontoxigenic culture cured of its
contact
taining antiserum against bacteriophage prophage was tested for its sensitivity to
* Material from plaques produced by bacteriophage
CE,B resulted in the loss of bacteriophage CEfl on culture A028 or culture UV171 was lysates of the other cured nontoxigenic
CE,8 and loss of toxigenicity. These iso- checked for toxigenicity and bacteriophage produc- cultures by the agar-layer procedure.
lates were then reinfected with bac- tion. t Cultures A028 and UV171 were ex-
posed to bacteriophage and isolated colonies were Only one culture, A028, was sensitive to
teriophage CE,8 and they again became checked for toxigenicity and bacteriophage pro- the lysates of the other nontoxigenic
duction.
toxigenic. cultures. Culture A028, therefore, had
Nontoxigenic cultures were isolated been cured of two of its prophages,
after treatment of toxigenic strain 468C whereas the other cultures continued to
with either ultraviolet irradiation or (pH 7.4) containing 20 ,tug of AO per carry a second prophage.
acridine orange (AO). Irradiation with milliliter and incubated at 37°C for 18 Bacteriophages were isolated from an
ultraviolet light is a standard method for hours. The AO culture was streaked 18-hour broth culture of the toxigenic
inducing lysis of bacteria that harbor onto TYG blood agar plates, and iso- parent strain 468C. The broth culture
prophages (the form in which bac- lated colonies were tested for toxigenicity was centrifuged at 6000g for 10 minutes
teriophages are perpetuated in lysogenic and sensitivity to bacteriophages by the and sterilized by filtration. Colony-
bacteria), whereas AO has been used same procedures that were described for centered plaques and turbid plaques pro-
to remove extra chromosomal elements the colonies treated with ultraviolet light. duced by bacteriophages designated as
from bacteria. Both procedures can re- Ultraviolet treatment was more effi- CE,B and CEy, respectively, were ob-
sult in bacterial isolates that are "cured" cient than AO was in curing strain 468C served when the filtrate was plated with
of either prophages in the former or
extra chromosomal elements in the
latter.
For irradiation, about 106 cells of
strain 468C in the logarithmic phase of
growth were spread onto the surfaces of
trypticase, yeast-extract, glucose (TYG)
blood agar plates and irradiated for 60
seconds at a distance of 20 cm (General
Electric germicidal lamp, 15 watts).
The agar plates were incubated anaero-
bically in Brewer jars for 48 hours at
37°C. Isolated colonies surviving the
treatment (approximately 102 cells)
were tested for toxigenicity by the mouse
assay (6) and for sensitivity to the
bacteriophages of the toxigenic parent
culture by the agar-layer procedure (7).
The base agar used in the agar-layer pro-
cedure was TYG agar, and the overlay
was soft agar (0.7 percent) prepared
from the filtered broth of Segner's forti-
fied egg meat (SFEM) medium (8).
The addition of blood (0.2 ml) to the
overlays was essential for confluent
growth. Cysteine hydrochloride (0.1
percent) was used as the reducing agent
in all media.
For curing with AO, about 105 cells
of strain 468C in the logarithmic phase Fig. 1. Bacteriophage CE,8, from lysate of Clostridium botulinum type C, strain 468C,
of growth were added to TYG broth which induces toxigenicity (X 247,500). Scale, 20 nm.
30 APRIL 1971 481
culture A028. These two different bac-
teriophages were purified by five succes-
sive single-plaque isolations on culture
A028.
Electron micrographs were prepared
of the two bacteriophages according to
the procedure of Eklund et al. (9). Bac-
teriophage CE/? exhibited a hexagonal
head 100 nm in diameter and a tail 400
nm long and 11 nm in diameter sur-
rounded by a contracted sheath 30 nm
in diameter (Fig. 1). Bacteriophages
comparable to CE/3 in size and mor-
phology have been observed in lysates
of other strains of type C (5, 10). Bac-
teriophage CE-y exhibited a hexagonal
head 60 nm in diameter and a tail 185
nm long and 6 nm in diameter sur-
rounded by a contracted sheath 20 nm
in diameter (Fig. 2).
Both bacteriophages were studied for
their relation to the toxigenicity of type
C strain 468C. Bacteriophage stocks
used in these studies were produced by
propagating the purified bacteriophages
with culture A028 in TYG broth at
30(C. Bacteriophage stocks were
treated with 25 ug of crystalline deoxy-
ribonuclease II (Sigma) per milliliter
for 3 hours at 37°C and then sterilized
by filtration. This step was used to rule
out the possibility of a transformation
principle of the deoxyribonucleic acid
type. Sterility checks were made by
inoculating several milliliters of filtrate
into SFEM broth and then incubating
the medium for several weeks at 37°C.
Two techniques were used to test the
relation of the two bacteriophages to
the toxigenicity of cultures A028 and
UV171 (which carries bacteriophage
CEy). In the first, dilutions of bacterio-
phage CE/I were plated with the recipi-
ent nontoxigenic culture according to
the agar-layer procedure. Material from
the plaques was transferred into TYG
broth and incubated at 30°C for 5 days.
Cultures arising from the plaques pro-
duced by CE/I on cultures A028 and
UV171 produced bacteriophage and
were toxigenic (Table 1). Cultures aris-
ing from the plaques produced by CEy
on culture A028 also produced bac-
teriophage but were nontoxigenic (data
not shown).
In the second procedure, bacterio-
phage CE/I was added to actively grow-
ing nontoxigenic cultures of A028 and
UVI 71 at a multiplicity of infection
(ratio of adsorbed bacteriophage to
bacteria) of 1.8 and 2.6, respectively.
After 30 and 240 minutes of exposure
to the bacteriophage, the cultures were
482
Fig. 2. Bacteriophage CEy, from lysate of
Clostridium botulinium type C, strain 468C,
which does not induce toxigenicity (X
247,500). Scale, 20 nm.
streaked onto TYG blood agar plates,
and isolated colonies were transferred
into TYG broth, incubated, assayed for
toxin, and tested for bacteriophage pro-
duction, with culture A028 as the indi-
cator strain.
After a 30-minute contact with bac-
teriophage CEj/, 10 percent of the colo-
nies of culture A028 and 5 percent of
the colonies of culture UV171 were
toxigenic (Table 1). The toxigenic cul-
tures were resistant to bacteriophage
CE,I. The nontoxigenic cultures con-
tinued to be sensitive to bacteriophage
CE/I? but when they were reinfected
with this bacteriophage they also be-
came toxigenic and produced bacterio-
phage. After a 240-minute contact be-
tween bacteriophage and bacteria, all
of the isolates were toxigenic and pro-
duced bacteriophage (Table 1).
Several of the CE/I-infected cultures
were subcultured in TYG broth con-
taining 50 percent antiserum against
bacteriophage CE,B (produced by im-
munizing rabbits with CE,/ bacterio-
phage). Four successive transfers in
antiserum resulted in the simultaneous
loss of bacteriophage CE,8 and toxi-
genicity. When these nontoxigenic cul-
tures were reinfected with bacterio-
phage CE,/, they again became toxi-
genic.
The toxins of the different isolates
were neutralized with type C antitoxin.
These cultures were subcultured in
SFEM six times over a period of 2
months, and all of them maintained
their toxigenic and bacteriophage-pro-
ducing characteristics.
Experiments reported in Table 1 were
repeated with bacteriophage CEy and
with bacterial culture A028 (data not
shown). Colonies from a 240-minute
contact between bacteriophage CEy and
culture A028 were all bacteriophage-
producers but nontoxigenic.
There is little doubt that in our ex-
periments toxigenicity of C. botulinumn
type C (strain 468C) is in some manner
dependent on the presence of bacterio-
phage CE/B. Groman (4) also reported
that conversion of toxigenicity of
Corynebacterium diphtheriae is bac-
teriophage specific. The curing of CE,B-
infected cultures by antiserum against
the bacteriophage indicates that this
relation between bacteriophage CE/l
and host is probably pseudolysogeny,
similar to that which occurs with Bacil-
lus subtilis (11), instead of true
lysogeny.
Some of the other types of C. botu-
linum that we have studied have been
cured of their prophage but still remain
toxigenic. It is possible either that they
carry other bacteriophages or that not
all C. botulinum toxins are induced by
bacteriophage.
M. W. EKLUND
F. T. POYSKY
S. M. REED
C. A. SMITH
National Marine Fisheries Service
Technology Laboratory,
2725 Montlake Boulevard East,
Seattle, Washington 98102
References and Notes
1. C. E. Dolman, M. Tomsich, C. C. R. Camp-
bell, W. B. Laing, J. Infect. Dis. 106, 5
(1960).
2. C. E. Dolman and L. Murakami, ibid. 109,
107 (1961).
3. V. J. Freeman, J. Bacteriol. 61, 675 (1951).
4. N. B. Groman, ibid. 69, 9 (1955).
5. K. Inoue and H. lida, Jap. J. Microbiol. 14,
87 (1970).
6. M. W. Eklund, F. T. Poysky, D. I. Wieler,
Appl. Microbiol. 15, 1316 (1967).
7. M. H. Adams, Bacteriophage (Interscience,
New York, 1959).
8. W. P. Segner, C. F. Schmidt, J. K. Boltz,
Report No. COO-1183-32 (U.S. Atomic Energy
Commission, Washington, D.C., 1970).
9. M. W. Eklund, F. T. Poysky, E. S. Boatman,
J. Virol. 3, 270 (1969).
10. G. Vinet and V. Fredette, Rev. Ca,t. Biol. 27,
73 (1968).
11. K. Bott and B. Strauss, Virology 25, 212
(1965).
12. We thank Dr. E. S. Boatman, University of
Washington, Seattle, for preparation of the
electron micrographs, and Miss Doris Huff
for technical assistance. Supported by AEC
contract No. AT(949-7)-2442.
18 January 1971
SCIENCE, VOL. 172
INFECTION AND IMMUNITY, Aug. 2002, p. 3985–3993
0019-9567/02/$04.00⫹0 DOI: 10.1128/IAI.70.8.3985–3993.2002
Copyright © 2002, American Society for Microbiology. All Rights Reserved.
MINIREVIEW
Bacteriophage Control of Bacterial Virulence
Patrick L. Wagner and Matthew K. Waldor*
Howard Hughes Medical Institute and New England Medical Center, Boston, Massachusetts 02110
In 1930, Felix d’Herelle wrote “. . .the actions and reactions
are not solely between these two beings, man and bacterium,
for the bacteriophage also intervenes; a third living being and,
hence, a third variable is introduced” (19). The contribution of
bacteriophages to the pathogenicity of their bacterial hosts
began to be uncovered as early as 1927, when Frobisher and
Brown discovered that nontoxigenic streptococci exposed to
filtered supernatants of toxigenic streptococcal cultures ac-
quired the ability to produce scarlatinal toxin (28). We now
know that these supernatants contained a bacteriophage en-
coding the scarlatinal toxin and that these investigators were
describing transduction, i.e., the transfer of genetic material to
a bacterial cell via phage infection. Although these early in-
vestigators lacked a mechanistic explanation for their observa-
tions, they postulated that the bacterium was “within certain
limits, perhaps a matter of secondary importance and that
toxicogenicity might be a property that could be acquired by
different types of organisms.” (28). Their hypothesis that bac-
teria acquire virulence properties has since gained widespread
acceptance, as many virulence genes have been shown to un-
dergo transfer among bacteria by phages (via transduction)
and other mobile genetic elements such as plasmids (via con-
jugation).
Over time, a number of toxin genes were found to be phage
encoded, and consideration of the role of phages in bacterial
pathogenesis emphasized the dissemination of toxin genes
among bacterial strains (10). However, it has become increas-
ingly clear that toxin genes are only a subset of the diverse
virulence factors encoded by bacteriophages. For example,
some phages encode regulatory factors that increase expres-
sion of virulence genes not encoded by the phage (84), while
others encode enzymes that alter bacterial components related
to virulence (31, 58). Furthermore, phages have unique prop-
erties that enable them to contribute more directly to bacterial
virulence than via transduction. Structural components of
virion particles, for example, may be directly pathogenic (5, 6,
99). Additionally, phage-encoded genes frequently undergo
replication and transcriptional activation following prophage
induction, a process that was speculated to have a role in the
production of diphtheria toxin by Corynebacterium diphtheriae
as early as 1960 (3).
Since d’Herelle’s time, his notion of the phage as a third
* Corresponding author: Division of Geographic Medicine and In-
fectious Diseases, Tufts University School of Medicine and New En-
gland Medical Center, NEMC #041, 750 Washington St., Boston, MA
02110. Phone: (617) 636-7618. Fax: (617) 636-5292. E-mail: mwaldor
@lifespan.org.
3985
Vol. 70, No. 8
variable in bacterial pathogenesis has proven correct. How-
ever, while d’Herelle emphasized the diminution of bacterial
virulence by bacteriophages (19), we have instead come to
learn that phages serve as a driving force in bacterial patho-
Downloaded from http://iai.asm.org/ on August 12, 2014 by guest
genesis, acting not only in the evolution of bacterial pathogens
through gene transfer, but also contributing directly to bacte-
rial pathogenesis at the time of infection. This review provides
a discussion of (i) the discovery of phage-encoded virulence
factors, (ii) bacterial virulence properties altered by phages,
(iii) regulation of phage-encoded virulence factors, and (iv) the
role of in situ prophage induction in the control of bacterial
virulence.
DISCOVERY OF PHAGE-ENCODED
VIRULENCE FACTORS
Historically, the discovery that a virulence factor was phage
encoded relied upon observation of transmissibility of the fac-
tor among bacterial strains, followed by demonstration, some-
times much later, that the relevant genes were located on a
phage chromosome. The scarlatinal and diphtheria toxins, for
example, were proven to be phage encoded many years after
their production was initially reported to be transmissible (27,
28, 38, 44, 90, 102, 108). Later, genetic tools such as trans-
posons TnphoA and Tn10d-bla (4, 73) were developed to aid in
the discovery of phage-encoded virulence factors. Since PhoA
and Bla must be secreted in order to be active, screening fusion
libraries for transducible PhoA or Bla activity allowed for the
rapid identification of secreted, phage-encoded putative viru-
lence factors, including the ␭-encoded lom and bor (4) genes of
Escherichia coli and the phage K139-encoded glo gene of Vibrio
cholerae (73).
Phage-encoded genes are not always transmissible, however,
either due to technical limitations in detecting transduction or
due to the fact that integrated prophages frequently become
defective. In the present era, analysis of the genome sequences
of bacterial pathogens can expeditiously reveal whether viru-
lence factors are associated with phage-like DNA sequences
regardless of whether they are transmissible. For example, the
nontransmissible (86) stx genes in Shigella dysenteriae are ad-
jacent to lambdoid phage-like sequences interrupted by nu-
merous insertion sequences, suggesting that the toxin genes lie
in a prophage that has been rendered defective by the insertion
sequences (59). Moreover, transduction of a virulence gene is,
in and of itself, insufficient evidence that the gene resides in a
phage genome. Generalized transducing phages can package
and transmit any chromosomal locus, and specific interactions
between generalized transducing phages and chromosomal vir-
3986 MINIREVIEW
ulence genes have been described. For example, the Staphylo-
coccus aureus pathogenicity island SapI1, which contains the
gene for toxic shock syndrome toxin (tst), is mobilized at high
frequency by the generalized staphylococcal transducing phage
80␣ (76); a similar mechanism has not been ruled out for the
transmission of the V. cholerae pathogenicity island (VPI),
which was reported to correspond to the genome of a phage
(49).
Currently, analysis of sequences surrounding virulence fac-
tor genes offers the most direct and sensitive method of deter-
mining whether virulence genes are associated with phage-like
sequences. While this approach can reveal whether a virulence
gene is associated with phage sequences, it cannot reveal
whether the gene is part of a prophage capable of transducing
it or of influencing its expression. Mutational analyses can
complement sequence analysis in this respect, revealing, for
example, the necessity of phage morphogenesis genes for
transduction (100) or the necessity of phage regulatory se-
quences for virulence factor production (98).
BACTERIAL VIRULENCE PROPERTIES ALTERED
BY PHAGES
Historically, exotoxin production has been the most widely
recognized bacterial characteristic linked to bacteriophage in-
fection (10). However, phage infection of bacteria is increas-
ingly associated with additional effects on bacterial virulence.
Phages can alter host bacterial properties relevant to all stages
of the infectious process (Table 1), including bacterial adhe-
sion, colonization, invasion, and spread through human tissues;
resistance to immune defenses; exotoxin production; sensitivity
to antibiotics; and transmissibility among humans.
Phages influence bacterial adhesion, colonization, and in-
vasion. Early events during bacterial infection include adhe-
sion, colonization, and sometimes, invasion. Phages have been
shown to be involved in each of these steps in different bacte-
rial pathogens. Phage-mediated adhesion is illustrated by
Streptococcus mitis, a causative agent of infective endocarditis.
Blood-borne streptococci adhere to platelets, fibrin, and ex-
posed valvular matrix components, giving rise to the vegeta-
tions that damage heart valves and provide a source of bacte-
remia and septic emboli. Recently, Bensing et al. (7) used
transposon mutagenesis to identify bacterial loci involved in
platelet adherence. They identified a locus encoding two sur-
face proteins, PblA and PblB, the gene sequences of which
resembled phage capsid and tail fiber genes and were sur-
rounded by other phage-like gene sequences (7). These genes
reside on an inducible prophage, SM1 (8), and encode proteins
present in the SM1 phage particle (8). Disruption of either
pblA or pblB resulted in decreased platelet binding by S. mitis,
although polar effects on other loci were not strictly ruled out
(7). Whether and by what mechanisms the phage-particle-as-
sociated and/or the bacterial-membrane-bound fractions of
PblA and PblB contribute directly to vegetation formation by
S. mitis remain unknown.
Bacterial invasion of human tissues involves the elaboration
of bacterial enzymes, including collagenases, hyaluronidases,
and hemolysins. The group A streptococcal (GAS) hyaluroni-
dase is phage encoded (43), and hyaluronidase activity is as-
sociated with phage particles themselves (5, 6). Since the cap-
INFECT. IMMUN.
sule of GAS is composed solely of hyaluronic acid, the
hyaluronidase presumably benefits the phage by aiding in cap-
sule penetration during its infection of or release from strep-
tococci. This phage-associated hyaluronidase activity may also
aid the streptococci in their spread through human connective
tissues. Although this hypothesis is untested, antibody to
phage-encoded hyaluronidase is detectable in sera of patients
infected by GAS (32).
The invasive properties of Salmonella enterica serovar Ty-
phimurium require the Salmonella pathogenicity island 1
(SPI1)-encoded type III secretion system, which injects effec-
tor proteins directly into the cytoplasm of host cells (40).
Among the numerous effector proteins exported via this secre-
tion system is SopE, which activates human Rho GTPases and
contributes to efficient entry of Salmonella into tissue culture
Downloaded from http://iai.asm.org/ on August 12, 2014 by guest
cells (33, 34, 105). The sopE gene is located on a temperate
Salmonella phage, designated SopE␾ (62). Since effector pro-
teins such as SopE interact intimately with mammalian intra-
cellular components, lysogeny thus aids in the adaptation of
Salmonella to its mammalian host. In this instance, two acces-
sory genetic elements (SPI1 and SopE␾) interact to enhance
the virulence of a bacterial pathogen. Following epithelial
translocation, Salmonella preferentially localizes to Peyer’s
patches within the bowel wall. Stanley et al. identified a Sal-
monella gene, gipA, encoded by phage Gifsy-1, that is ex-
pressed specifically in the Peyer’s patch and is necessary for
optimal survival in this environment (85). Therefore, two dif-
ferent phage-encoded genes facilitate two separate early steps
during Salmonella infection, namely, uptake by intestinal epi-
thelial cells (sopE) and survival in the Peyer’s patch (gipA).
Phages enhance bacterial resistance to serum and phago-
cytes. During infection, bacterial pathogens encounter the se-
rum- and phagocyte-mediated elements of the innate immune
system. Staphylococci produce a number of proteins involved
in phagocyte evasion, including a recently discovered chemo-
taxis inhibitory protein (CHIPS) that binds to and attenuates
the activity of the neutrophil receptors for complement and
formylated peptides (93; W. Van Wamel, A. Peshel, K. Van
Kessel, and J. Van Strijp, Abstr. 101st Gen. Meet. Am. Soc.
Microbiol. 2001, abstr. B-98, p. 62, 2001). This function is
proposed to protect S. aureus from neutrophil-mediated killing
(93; Van Wamel et al., Abstr. 101st Gen. Meet. Am. Soc.
Microbiol.), an important host defense against staphylococci.
The gene encoding CHIPS (chp) has been shown to reside on
a functional phage that also transduces the staphylokinase
(sak) and enterotoxin A (sea) genes and eliminates ␤-hemoly-
sin production, presumably by insertional inactivation (Van
Wamel et al., Abstr. 101st Gen. Meet. Am. Soc. Microbiol.
2001). Staphylococci also produce the phage-encoded Panton-
Valentine leukocidin (PVL), a cytotoxin with direct activity
against human phagocytes (48, 92). Thus, by inhibiting phago-
cytosis (CHIPS) and by directly attacking phagocytes (PVL),
two different phage gene products counteract phagocyte-me-
diated destruction of their staphylococcal hosts.
Inside phagocytes, bacteria are subjected to oxidative stress
within organelles that produce superoxide radical and hydro-
gen peroxide. Bacterial survival in the presence of these mol-
ecules is enhanced by enzymes such as superoxide dismutase,
which catalyzes the conversion of superoxide ion into hydrogen
peroxide and molecular oxygen (21, 23). Figueroa-Bossi and
VOL. 70, 2002 MINIREVIEW 3987

TABLE 1. Bacterial virulence properties altered by bacteriophages

Bacterial property altered Mechanism Reference(s) or Source

Colonization/adhesion
E. coli The ␭-encoded lom gene promotes adhesion to buccal epithelial cells. 4, 69, 72
P. aeruginosa Phage FIZ15 promotes adhesion to buccal epithelial cells. 91
S. mitis The SM1-encoded PblA and PblB surface proteins promote adhesion to 7, 8
platelets.
V. cholerae The toxin-coregulated pilus may be phage encoded. 49

Invasion
S. enterica Phage SopE␾ transduces a type III secretion system effector that promotes 62
entry into epithelial cells.
Phage Gifsy-1 encodes gipA, a gene that enhances survival in the Peyer’s patch. 85
S. pyogenes Hyaluronidase is phage encoded. 43
S. aureus Fibrinolysin is phage encoded. 77

Resistance to serum/phagocytes

Downloaded from http://iai.asm.org/ on August 12, 2014 by guest

E. coli The ␭-encoded bor gene confers a survival advantage in animal serum. 4
P. aeruginosa Phages encode enzymes that alter the O antigen. 37
S. enterica Phage Gifsy-2 encodes SodC, a superoxide dismutase. 24
Phages encode enzymes that alter the O antigen. 74, 75, 106
S. dysenteriae Phages encode enzymes that alter the O antigen. 31, 58
S. aureus Phages encode CHIPS, a phagocytotoxin. Van Wamel et al.a
Phage ␾PVL encodes the Panton-Valentine leukocidin. 46
S. pyogenes Lysogeny up-regulates the antiphagocytic M protein. 84

Exotoxin production
B. avium Pertussis toxin is phage encoded in B. avium. van Horne et al.b
C. botulinum Botulinum toxin is phage encoded. 29
C. diphtheriae Diphtheria toxin is phage encoded. 38, 90
E. coli The Shiga toxins are phage encoded. 39, 66, 103
P. aeruginosa Pseudomonas cytotoxins are phage encoded. 35
S. dysenteriae The Shiga toxin genes are associated with phage sequences, probably a defective 59
prophage.
S. aureus Staphylococcal enterotoxins are phage encoded. Staphylococcal exfoliative toxins 9, 17, 18
are phage encoded. Toxic shock syndrome toxin is encoded by SapI, a mobile
pathogenicity island transduced at high frequency by phage 80␣.
107
55
S. pyogenes Streptococcal pyrogenic (erythrogenic, scarlatinal) exotoxins are phage encoded. 30, 44, 102
V. cholerae Cholera toxin is phage encoded. 100

Susceptibility to antibiotic
S. aureus Generalized transduction contributes to horizontal transmission of gram-positive 15, 26
antibiotic-resistance genes.
S. pyogenes

Transmission
V. cholerae Phage-encoded cholera toxin likely promotes transmission by stimulating 100
copious amounts of watery diarrhea.
a
Van Wamel et al., Abstr. 101st Gen. Meet. Am. Soc. Microbiol. 2001.
b
S. J. van Horne, D. Bjornsen, P. Carpentier, and L. M. Temple, Abstr. 101st Gen. Meet. Am. Soc. Microbiol. 2001, abstr. B-109, p. 64–65, 2001.

Bossi showed that a superoxide dismutase (SodC) of S. enterica
serovar Typhimurium is encoded by a functional bacterio-
phage, Gifsy-2; isogenic Salmonella derivatives lacking Gifsy-2
were attenuated in a mouse infection model (24). Intriguingly,
hydrogen peroxide is a highly effective chemical inducer of
Gifsy-2, suggesting a possible relationship between SodC ac-
tivity, which results in hydrogen peroxide production, and in-
duction of the Gifsy-2 prophage (24).
Phages encode bacterial exotoxins. The most widely recog-
nized examples of phage-encoded virulence factors are exotox-
ins. Exotoxin production is the major pathogenic mechanism of
several bacterial pathogens, including V. cholerae, C. diphthe-
riae, and Clostridium botulinum. Phage-encoded exotoxins
work by a variety of mechanisms that have been the subject of
a previous review of the involvement of phages in bacterial
pathogenesis (10).
Phages alter bacterial susceptibility to antibiotics. Most mo-
bile antibiotic resistance genes are encoded on plasmids or
transposons, and no examples of phage-encoded resistance
genes are known. However, phages may play an important role,
via transduction, in the mobility of these resistance plasmids
among staphylococci (26) and streptococci (15). Streptococcal
phages have transferred resistance to tetracycline, chloram-
phenicol, macrolides, lincomycin, and clindamycin (89) and
streptomycin (42), probably via generalized transduction of
non-phage-encoded resistance genes.
3988 MINIREVIEW
Phage-encoded products enhance transmission of bacterial
pathogens. Cholera toxin up-regulates enterocyte adenylate
cyclase activity, leading to profuse watery diarrhea (25) that is
widely assumed to contribute significantly to the fecal-oral
transmission of V. cholerae (61). Since the cholera toxin genes
are phage encoded (100), this system is an example of a bac-
teriophage contributing to the transmission of its bacterial host
among humans.
REGULATION OF PHAGE-ENCODED
VIRULENCE FACTORS
While phages encode virulence genes that influence virtually
all aspects of bacterial pathogenesis, phages have been thought
to have little role in regulating the expression of these genes.
Three main observations account for the predominance of this
idea. First, virulence factors are often produced during lysog-
eny, when most phage-encoded genes are not efficiently tran-
scribed. Some virulence genes, including the ␭-encoded bor of
E. coli and the phage-encoded vir of Mycoplasma arthritidis
(94), lie on the noncoding strand relative to lytic phage genes,
suggesting that phage transcriptional regulation has little, if
any, influence over virulence factor production.
Second, many phage-encoded virulence genes are located
near the attachment (att) sites of their respective phages, the
portion of the integrated phage adjacent to the host chromo-
some. This finding suggests that they were acquired from the
ancestral bacterial chromosome via aberrant phage excision
events, perhaps as independently regulated operons. This
property is true of all of the gram-positive exotoxin-encoding
bacteriophages in which the att site is known, including diph-
theria toxin (52), staphylococcal enterotoxin A (18), staphy-
lokinase (18), PVL (47), staphylococcal exfoliative toxin (107),
CHIPS (Van Wamel et al., Abstr. 101st Gen. Meet. Am. Soc.
Microbiol. 2001), streptococcal erythrogenic (scarlatinal) toxin
(45), and streptococcal pyrogenic exotoxin C (30). Interest-
ingly, many phage-encoded virulence genes in gram-negative
organisms are not located near the att site, including the Shiga
toxin genes of E. coli (64, 71), the cytotoxin gene of Pseudo-
monas aeruginosa (35), and the SodC (24) and SopE (62) genes
of Salmonella.
Third, many phage-encoded virulence genes are regulated
by chromosome-encoded transcription factors. For example,
the iron-dependent transcriptional repressor DtxR, encoded
on the C. diphtheriae chromosome, represses diphtheria toxin
production by binding to an operator overlapping the toxin
gene promoter (12, 81, 83). Since DtxR is known to have
numerous chromosomal target genes involved in iron acquis-
tion (54, 78–80, 82), diphtheria toxin production is apparently
regulated as part of a global cellular response to environmental
stimuli. Similarly, the cholera toxin genes (ctx) are up-regu-
lated by the V. cholerae virulence transcriptional activators
ToxR and ToxT (22, 60). While toxR appears to be a compo-
nent of the ancestral V. cholerae genome, toxT is encoded on
VPI (36). Thus, the CTX␾-encoded ctx genes are regulated
both by a chromosomal locus (toxR) and by a locus encoded on
another presumedly mobile genetic element (toxT). Likewise,
phage-encoded staphylococcal toxins are regulated by the
chromosomal accessory gene regulatory locus (agr), which co-
ordinates gene expression during the post-exponential growth
INFECT. IMMUN.
phase in culture (67). Since these examples of phage-encoded
virulence factors are coordinately regulated with other genes
not encoded on phages and not solely involved in pathogenesis
per se, they are consequently not thought to be regulated in
concert with other phage-encoded genes.
In contrast to these examples in which phages encode
virulence factors but are thought to play little role in their
regulation, our recent work on the Shiga toxin-encoding
phages of E. coli has established the principle that the bac-
teriophage life cycle can exert control over virulence factor
production by bacterial pathogens. The Shiga toxins (Stx1
and Stx2) are the principal virulence factors of enterohem-
orrhagic E. coli (EHEC), accounting for the most-severe
consequences of EHEC infection, including hemorrhagic
colitis and hemolytic uremic syndrome (11, 68, 95). The stx
Downloaded from http://iai.asm.org/ on August 12, 2014 by guest
genes in EHEC strains are located within prophages related
to ␭. Like ␭, these prophages contain transcriptional units
for various functions, such as replication, morphogenesis,
lysis, etc., that are coordinately expressed during specific
intervals following prophage induction due to the regulatory
influence of phage promoters, phage repressor, transcrip-
tional terminators, and transcriptional antiterminators (Fig.
1). The stx genes were found within the late operons of the
Stx-encoding phages, suggesting that they could be tran-
scribed from the late phage promoter (pR⬘) along with lysis
and morphogenesis genes following prophage induction
(Fig. 1) (64, 71). Indeed, phage-inducing agents such as
mitomycin C have been shown to increase Stx production by
E. coli (2, 41; A. E. Hull, D. W. K. Acheson, A. Donohue-
Rolfe, G. T. Keusch, and P. Echeverria, Abstr. 91st Gen.
Meet. Am. Soc. Microbiol. 1991, abstr., p. 390, 1991), and
subsequently, it was demonstrated that overexpression of Q,
a gene that encodes an antiterminator necessary for efficient
late gene expression in ␭-like phages, resulted in increased
Stx production by E. coli strains harboring Stx-encoding
phages (65).
We characterized the mechanisms by which prophage in-
duction augments Stx production by constructing a mutant
derivative of the Stx2-encoding phage ␾361 in which the pR⬘
promoter and a portion of the Q antiterminator gene nec-
essary for efficient transcription from pR⬘ were deleted. A
lysogen of the mutant phage was drastically impaired in Stx2
production both in vitro and in a mouse model of intestinal
infection (98), implying that the primary regulatory element
in Stx2 production is the late phage promoter pR⬘ and that
Stx2 production relies ultimately upon prophage induction.
The previously characterized pStx2 promoter adjacent to stx2
(87) is apparently not a dominant regulatory element in Stx2
production.
We also studied the Stx1-encoding phage H-19B. Like the
stx2 genes in ␾361, the stx1 genes in H-19B are located between
the late promoter pR⬘ and the lysis genes. However, the pR⬘
promoter is not necessary for Stx1 production by lysogens of
H-19B (97). Instead, multiple promoters contribute to stx1
transcription, including the adjacent iron-regulated promoter
pStx1 (14), as well as upstream phage promoters (pR and pR⬘)
that are regulated by prophage induction. Other phage-related
mechanisms besides control of stx1 transcription contribute to
Stx1 production following H-19B induction. Replication of the
H-19B genome, with the associated increase in stx1 copy num-
VOL. 70, 2002 MINIREVIEW 3989

Downloaded from http://iai.asm.org/ on August 12, 2014 by guest

FIG. 1. Phage control of Stx production and release by E. coli. (Top) A partial map (not to scale) of an integrated Stx-encoding prophage is
shown. Transcription from phage promoters is limited both by direct phage repressor activity at pL and pR and by transcriptional terminators
(downstream of pL, pR, and pR⬘). Following prophage induction, repressor cleavage allows for increased transcription from pL and pR and
production of antiterminators N and Q. These proteins facilitate transcription initiated at pR and pR⬘, respectively, by allowing readthrough at
several terminators, including tL, tR1, and tR⬘. (Bottom) Both the pR and pR⬘ promoters, as well as the toxin-associated promoter (pStx) can
contribute directly to stx transcription following prophage induction. Additionally, the phage replication gene products (O and P) contribute to Stx
production by amplifying stx copy numbers. The phage lysis gene products set a limit on the duration of Stx production and provide a mechanism
of toxin release from the cell (97).

ber, was the most quantitatively important mechanism of in-
creased Stx1 production. Phage-mediated lysis regulates the
quantity of Stx1 produced by limiting the duration of Stx1
accumulation following prophage induction and provides a
mechanism for toxin release (97). In summary, by amplifying
stx copy number, by contributing to stx transcription, and by
allowing for Stx release, the Stx-encoding phages are intimately
involved in the regulation of Stx production and thereby the
pathogenicity of EHEC.
The Stx-encoding prophages are not the only prophages
implicated in the regulation of virulence factor production.
Some 40 years ago, Barksdale et al. discovered that UV light,
a potent prophage-inducing agent, greatly enhanced diphthe-
ria toxin production, suggesting that prophage induction and
replication could amplify toxin production (3). Similarly, UV
light enhances phage and streptococcal pyrogenic exotoxin A
(scarlatinal toxin) production by GAS (108), as well as produc-
tion of the phage-encoded platelet-binding proteins of S. mitis
(7, 8). Mitomycin C increases production of CTX␾ virions and
cholera toxin by V. cholerae (B. Davis and M. Waldor, unpub-
lished observations), and ToxR-independent cholera toxin pro-
duction has been detected from the replicative form of CTX␾,
indicating the possible activity of alternative induction-related
phage promoters for ctx transcription (53). Although muta-
tional analyses have not been carried out in these cases, the
results are highly suggestive that, as for Stx1 and Stx2, the
production of other virulence factors may be controlled, in
part, by prophage induction.
ROLE OF IN SITU PROPHAGE INDUCTION IN
BACTERIAL VIRULENCE
Although prophage induction has been shown to contribute
to the production of several virulence factors in vitro, we have
little knowledge regarding the contribution of the phage life
cycle to the regulation of virulence factor production during
infection. It is conceivable that certain bacterial pathogens
exist in a less virulent state until they encounter a stimulus for
prophage induction within the human body, at which time
production of a virulence factor, regulated as part of the phage
life cycle, could contribute to the pathogenesis of the bacte-
rium. We propose that this in situ prophage induction could
help to explain when and where certain virulence factors are
produced during the course of bacterial infection. Moreover,
virion production and release within the human body, with
subsequent phage infection of other resident bacteria, could
contribute to pathogenesis by amplifying the number of viru-
lence gene-encoding organisms in the body during infection
(1).
What are the stimuli for in situ prophage induction? Many
prophages are induced by environmental conditions that lead
to bacterial DNA damage and, in the case of E. coli, activation
of RecA, which in turn catalyzes cleavage of phage repressors
(56). Such conditions include exposure to agents produced by
human cells, including the reactive oxygen species generated
and released by leukocytes, and to exogenous agents, such as
antibiotics. Many antibiotics commonly used to treat diarrhea,
3990 MINIREVIEW
for example, are known to induce Stx-encoding phages and
therefore to promote toxin production by EHEC (51, 57, 101,
109). Perhaps not coincidentally, numerous epidemiological
studies have detected an association between increased sever-
ity of EHEC infection and treatment with antibiotics (13, 16,
50, 70, 104). This observation has obvious clinical ramifica-
tions, in that antibiotics may actually worsen the clinical course
of infection by bacteria such as EHEC.
Phage-inducing agents derived from human cells have been
implicated in the pathogenesis of at least three organisms. We
showed in EHEC that H2O2 (a known inducer of ␭-like
phages) and neutrophils, an endogenous source of H2O2, can
induce Stx-encoding prophages, which enhances toxin produc-
tion by EHEC cells (96). The superoxide dismutase (SodC)-
encoding Salmonella phage Gifsy-2 (described above) is also
induced by H2O2 (24). H2O2 encountered by Salmonella in the
phagosome could result in prophage induction and subsequent
up-regulation of SodC production. In this way, prophage in-
duction could serve as an adaptive response by coupling pro-
duction of the protective SodC enzyme with exposure to reac-
tive oxygen species. In a third example, Broudy et al.
discovered a soluble phage-inducing factor (SPIF), elaborated
by human pharyngeal epithelial cells, which induced the strep-
tococcal pyrogenic exotoxin C (SpeC)-encoding phage of GAS
and subsequent toxin production by this pathogen (12a). Al-
though the SPIF molecule and its mechanism of prophage
induction remain to be characterized, these observations may
reflect an important role for an epithelium-derived prophage-
inducing factor in the pathogenesis of GAS.
We believe that the examples described here demonstrate
the need for a systematic evaluation of the extent to which the
induction of prophages influences expression of the virulence
factors that they encode. For example, consider the pathogen-
esis of endocarditis caused by S. mitis. It is important to learn
whether the organisms encounter prophage-inducing agents in
the bloodstream or in the endocardium that lead to increased
expression of the phage-encoded platelet-binding factors. Such
information could have important clinical implications in the
treatment of bacterial diseases. For instance, patients suffering
from or at risk for diseases in which prophage induction plays
a pathogenic role might benefit from anti-inflammatory agents
that reduce the generation of prophage-inducing antibacterial
factors by human cells and from avoiding the use of prophage-
inducing antibiotics.
EVOLUTIONARY CONSIDERATIONS
AND CONCLUSIONS
Bacteriophages are intimately associated with bacterial vir-
ulence genes. What factors account for this association? This
relationship likely reflects the mobility that virulence genes
derive from being phage encoded, allowing for their wide-
spread dissemination among bacterial populations. Further-
more, virulence genes encoded by phages may withstand envi-
ronmental exposure better than those encoded by bacteria. For
example, Muniesa et al. found that Stx2-encoding phages per-
sisted in river water longer and were more resistant to chlori-
nation and pasteurization than Stx2-encoding bacteria, sug-
gesting that phage particles serve as a more durable
environmental reservoir of the stx2 genes than bacteria (63).
INFECT. IMMUN.
Thus, enhanced mobility and persistence in the environment
may provide a selective advantage to virulence genes that are
phage encoded.
This relationship may be of benefit to phages as well. For
example, the hyaluronidase encoded by and incorporated into the
particles of S. aureus phages may facilitate phage penetration of
the S. aureus capsule during phage infection or release. Similarly,
O-antigen alteration by phages of certain gram-negative patho-
gens, which may aid in bacterial evasion of immunity, presum-
ably benefits the phage by conferring superinfection resistance
on the bacterial host. Moreover, in some cases, virulence fac-
tors may actually be integral components of the phage particle.
The CTX␾-encoded Ace protein, reported to account in part
for V. cholerae enterotoxicity (88), is thought to be a minor coat
protein of the CTX␾ virion (100). Thus, these putative viru-
Downloaded from http://iai.asm.org/ on August 12, 2014 by guest
lence factors may ultimately be phage attributes that inciden-
tally contribute to the pathogenicity of the bacterial host.
Other virulence factors may bestow upon phages unknown
advantages, a possibility that merits future investigation.
One paradoxical implication of in situ prophage induction in
bacterial pathogenesis is that virulence factor production could
be linked to the death of the bacterial cell via phage-mediated
lysis. This consideration would not apply to pathogens such as
V. cholerae, since production of the filamentous CTX␾ does
not destroy its host bacterial cell. In fact, CTX␾ and cholera
toxin share a mechanism of nonlethal secretion from the cell
(20). In contrast, we have learned that in E. coli lysogens of
phage ␾361, Stx2 is apparently produced only when the late
phage promoter, responsible for lysis gene transcription, is
active (98). This finding implies that toxin production involves
simultaneous and obligatory phage-mediated cell death. One
rationale for the tight linkage between Stx2 production and
phage induction is that production of Stx2 by a small number
of doomed cells contributes in some way to the survival of the
remaining lysogenic population whose prophages were not in-
duced (98). Alternatively, if the unit of selection is ␾361 or the
stx2 genes themselves, death of the host bacterium could be
irrelevant.
In summary, bacteriophages are intimately involved in bac-
terial pathogenesis, including adhesion and invasion, evasion
of immunity, and exotoxin production, largely by transducing a
diverse set of virulence genes. In some cases, the expression of
these genes may be driven in part by induction of the proph-
ages that encode them. Since prophage induction contributes
to virulence gene expression and since the human body is
replete with prophage-inducing chemicals, prophage induction
in the human body may be an important, underrecognized
mechanism of bacterial pathogenesis.
ACKNOWLEDGMENTS
We thank our collaborators David Friedman, Melody Neely, and
Jonathan Livny for educating us about lambdoid prophage induction
and David Acheson and Xiaoping Zhang for their insights into the
epidemiology and pathogenesis of EHEC. We also thank Brigid Davis
and Anne Kane for critical reading of the manuscript.
P.L.W. was supported by the Harvard Medical School and the
Howard Hughes Medical Institute. M K.W.’s laboratory is supported
by the National Institute of Allergy and Infectious Diseases, the
Howard Hughes Medical Center, the Pew Foundation, and the New
England Medical Center GRASP Center.
VOL. 70, 2002
REFERENCES
1. Acheson, D. W., J. Reidl, X. Zhang, G. T. Keusch, J. J. Mekalanos, and
M. K. Waldor. 1998. In vivo transduction with Shiga toxin 1-encoding
phage. Infect. Immun. 66:4496–4498.
2. Al-Jumaili, I., D. A. Burke, S. M. Scotland, H. M. Al-Mardini, and C. O.
Record. 1992. A method of enhancing verocytotoxin production by Esche-
richia coli. FEMS. Microbiol. Lett. 93:121–126.
3. Barksdale, L., L. Garmise, and K. Horibata. 1960. Virulence, toxinogeny
and lysogeny in Corynebacterium diphtheriae. Ann. N. Y. Acad. Sci. 88:
1093–1098.
4. Barondess, J. J., and J. Beckwith. 1990. A bacterial virulence determinant
encoded by lysogenic coliphage ␭. Nature 346:871–874.
5. Benchetrit, L. C., E. D. Gray, R. D. Edstrom, and L. W. Wannamaker. 1978.
Purification and characterization of a hyaluronidase associated with a tem-
perate bacteriophage of group A, type 49 streptococci. J. Bacteriol. 134:
221–228.
6. Benchetrit, L. C., E. D. Gray, and L. W. Wannamaker. 1977. Hyaluronidase
activity of bacteriophages of group A streptococci. Infect. Immun. 15:527–
532.
7. Bensing, B. A., C. E. Rubens, and P. M. Sullam. 2001. Genetic loci of
Streptococcus mitis that mediate binding to human platelets. Infect. Immun.
69:1373–1380.
8. Bensing, B. A., I. R. Siboo, and P. M. Sullam. 2001. Proteins PblA and PblB
of Streptococcus mitis, which promote binding to human platelets, are en-
coded within a lysogenic bacteriophage. Infect. Immun. 69:6186–6192.
9. Betley, M. J., and J. J. Mekalanos. 1985. Staphylococcal enterotoxin A is
encoded by phage. Science 229:185–187.
10. Bishai, W. R., and J. R. Murphy. 1988. Bacteriophage gene products that
cause human disease, p. 683–724. In R. Calendar (ed.), The bacteriophages.
Plenum Press, New York, N.Y.
11. Boerlin, P., S. A. McEwen, F. Boerlin-Petzold, J. B. Wilson, R. P. Johnson,
and C. L. Gyles. 1999. Associations between virulence factors of Shiga
toxin-producing Escherichia coli and disease in humans. J. Clin. Microbiol.
37:497–503.
12. Boyd, J., M. N. Oza, and J. R. Murphy. 1990. Molecular cloning and DNA
sequence analysis of a diphtheria tox iron-dependent regulatory element
(dtxR) from Corynebacterium diphtheriae. Proc. Natl. Acad. Sci. USA 87:
5968–5972.
12a.Broudy, T. B., V. Pancholi, and V. A. Fischetti. 2001. Induction of lysogenic
bacteriophage and phage-associated toxin from group A streptococci dur-
ing coculture with human pharyngeal cells. Infect. Immun. 69:1440–1443.
13. Butler, T., M. R. Islam, M. A. K. Azad, and P. K. Jones. 1987. Risk factors
for development of hemolytic uremic syndrome during shigellosis. J. Pedi-
atr. 110:894–897.
14. Calderwood, S. B., and J. J. Mekalanos. 1987. Iron regulation of Shiga-like
toxin expression in Escherichia coli is mediated by the fur locus. J. Bacteriol.
169:4759–4764.
15. Caparon, M. 2000. Genetics of group A streptococci, p. 53–64. In V. A.
Fischetti, R. P. Novick, J. J. Ferretti, D. A. Portnoy, and J. I. Rood (ed.),
Gram-positive pathogens. ASM Press, Washington, D.C.
16. Carter, A. O., A. A. Borczyk, J. A. Carlson, B. Harvey, J. C. Hockin, M. A.
Karmali, C. Krishnan, D. A. Korn, and H. Lior. 1987. A severe outbreak of
Escherichia coli O157:H7-associated hemorrhagic colitis in a nursing home.
N. Engl. J. Med. 317:1496–1500.
17. Casman, E. P. 1965. Staphylococcal enterotoxin. Ann. N. Y. Acad. Sci.
128:124–131.
18. Coleman, D. C., D. J. Sullivan, R. J. Russell, J. P. Arbuthnott, B. F. Carey,
and H. M. Pomeroy. 1989. Staphylococcus aureus bacteriophages mediating
the simultaneous lysogenic conversion of ␤-lysin, staphylokinase and en-
terotoxin A: molecular mechanism of triple conversion. J. Gen. Microbiol.
135:1679–1697.
19. d’Herelle, F. 1930. The bacteriophage and its clinical applications. Charles
C. Thoams, Baltimore, Md.
20. Davis, B. M., E. H. Lawson, M. Sandkvist, A. Ali, S. Sozhamannan, and
M. K. Waldor. 2000. Convergence of the secretory pathways for cholera
toxin and the filamentous phage, CTX␾. Science 288:333–335.
21. De Groote, M. A., U. A. Ochsner, M. U. Shiloh, C. Nathan, J. M. McCord,
M. C. Dinauer, S. J. Libby, A. Vazquez-Torres, Y. Xu, and F. C. Fang. 1997.
Periplasmic superoxide dismutase protects Salmonella from products of
phagocyte NADPH-oxidase and nitric oxide synthase Proc. Natl. Acad. Sci.
USA 94:13997–14001.
22. DiRita, V. J., C. Parsot, G. Jander, and J. J. Mekalanos. 1991. Regulatory
cascade controls virulence in Vibrio cholerae. Proc. Natl. Acad. Sci. USA
88:5403–5407.
23. Farrant, J. L., A. Sansone, J. R. Canvin, M. J. Pallen, P. R. Langford, T. S.
Wallis, G. Dougan, and J. S. Kroll. 1997. Bacterial copper- and zinc-
cofactored superoxide dismutase contributes to the pathogenesis of sys-
temic salmonellosis. Mol. Microbiol. 25:785–796.
24. Figueroa-Bossi, N., and L. Bossi. 1999. Inducible prophages contribute to
Salmonella virulence in mice. Mol. Microbiol. 33:167–176.
25. Finkelstein, R. A. 1992. Cholera enterotoxin (choleragen): a historical per-
MINIREVIEW 3991
spective, p. 155–187. In D. Barua and W. B. Greenough III (ed.), Cholera.
Plenum Press, New York, N.Y.
26. Firth, N., and R. A. Skurray. 2000. Genetics: accessory elements and ge-
netic exchange, p. 326–338. In V. A. Fischetti, R. P. Novick, J. J. Ferretti,
D. A. Portnoy, and J. I. Rood (ed.), Gram-positive pathogens. ASM Press,
Washington, D.C.
27. Freeman, V. J. 1951. Studies on the virulence of bacteriophage-infected
strains of Corynebacterium diphtheriae. J. Bacteriol. 61:675–688.
28. Frobisher, M., and J. Brown. 1927. Transmissible toxicogenicity of strep-
tococci. Bull. Johns Hopkins Hosp. 41:167–173.
29. Fujii, N., K. Oguma, N. Yokosawa, K. Kimura, and K. Tsuzuki. 1988.
Characterization of bacteriophage nucleic acids obtained from Clostridium
botulinum types C and D. Appl. Environ. Microbiol. 54:69–73.
30. Goshorn, S. C., and P. M. Schlievert. 1989. Bacteriophage association of
streptococcal pyrogenic exotoxin type C. J. Bacteriol. 171:3068–3073.
31. Guan, S., D. A. Bastin, and N. K. Verma. 1999. Functional analysis of the
O-antigen glucosylation gene cluster of Shigella flexneri bacteriophage SfX.
Microbiology 145:1263–1273.
32. Halperin, S. A., P. Ferrieri, E. D. Gray, E. L. Kaplan, and L. W. Wanna-
maker. 1987. Antibody response to bacteriophage hyaluronidase in acute
glomerulonephritis after group A streptococcal infection. J. Infect. Dis.
Downloaded from http://iai.asm.org/ on August 12, 2014 by guest
155:253–261.
33. Hardt, W. D., L. M. Chen, K. E. Schuebel, X. R. Bustelo, and J. E. Galan.
1998. S. typhimurium encodes an activator of Rho GTPases that induces
membrane ruffling and nuclear responses in host cells. Cell 93:815–826.
34. Hardt, W. D., H. Urlaub, and J. E. Galan. 1998. A substrate of the centi-
some 63 type III protein secretion system of Salmonella typhimurium is
encoded by a cryptic bacteriophage. Proc. Natl. Acad. Sci. USA 95:2574–
2579.
35. Hayashi, T., T. Baba, H. Matsumoto, and Y. Terawaki. 1990. Phage-con-
version of cytotoxin production in Pseudomonas aeruginosa. Mol. Microbiol.
4:1703–1709.
36. Heidelberg, J. F., J. A. Eisen, W. C. Nelson, R. A. Clayton, M. L. Gwinn,
R. J. Dodson, D. H. Haft, E. K. Hickey, J. D. Peterson, L. Umayam, S. R.
Gill, K. E. Nelson, T. D. Read, H. Tettelin, D. Richardson, M. D. Ermo-
laeva, J. Vamathevan, S. Bass, H. Qin, I. Dragoi, P. Sellers, L. McDonald,
T. Utterback, R. D. Fleishmann, W. C. Nierman, and O. White. 2000. DNA
sequence of both chromosomes of the cholera pathogen Vibrio cholerae.
Nature 406:477–483.
37. Holloway, B. W., and G. N. Cooper. 1962. Lysogenic conversion in Pseudo-
monas aeruginosa J. Bacteriol. 84:1321–1324.
38. Holmes, R. K., and L. Barksdale. 1969. Genetic analysis of tox⫹ and tox-
bacteriophages of Corynebacterium diphtheriae. J. Virol. 3:586–598.
39. Huang, A., S. De Grandis, J. Friesen, M. Karmali, M. Petric, R. Congi, and
J. L. Brunton. 1986. Cloning and expression of the genes specifying Shiga-
like toxin production in Escherichia coli H19. J. Bacteriol. 166:375–379.
40. Hueck, C. J. 1998. Type III protein secretion systems in bacterial pathogens
of animals and plants. Microbiol. Mol. Biol. Rev. 62:379–433.
41. Hull, A. E., D. W. K. Acheson, P. Echeverria, A. Donohue-Rolfe, and G. T.
Keusch. 1993. Mitomycin immunoblot colony assay for detection of Shiga
like toxin-producing Escherichia coli in fecal samples: comparison with
DNA probes. J. Clin. Microbiol. 31:1167–1172.
42. Hyder, S. L., and M. M. Streitfeld. 1978. Transfer of erythromycin resis-
tance from clinically isolated lysogenic strains of Streptococcus pyogenes via
their endogenous phage. J. Infect. Dis. 138: 281–286.
43. Hynes, W. L., and J. J. Ferretti. 1989. Sequence analysis and expression in
Escherichia coli of the hyaluronidase gene of Streptococcus pyogenes bacte-
riophage H4489A. Infect. Immun. 57:533–539.
44. Johnson, L. P., and P. M. Schlievert. 1984. Group A streptococcal phage
T12 carries the structural gene for pyrogenic exotoxin type A. Mol. Gen.
Genet. 194:52–56.
45. Johnson, L. P., M. A. Tomai, and P. M. Schlievert. 1986. Bacteriophage
involvement in group A streptococcal pyrogenic exotoxin A production. J.
Bacteriol. 166:623–627.
46. Kaneko, J., K. Muramoto, and Y. Kamio. 1997. Gene of LukF-PV-like
component of Panton-Valentine leukocidin in Staphylococcus aureus P83 is
linked with lukM. Biosci. Biotechnol. Biochem. 61:541–544.
47. Kaneko, J., T. Kimura, S. Narita, T. Tomita, and Y. Kamio. 1998. Complete
nucleotide sequence and molecular characterization of the temperate
staphylococcal bacteriophage ␾PVL carrying Panton-Valentine leukocidin
genes. Gene 215:57–67.
48. Kaneko, J., T. Kimura, Y. Kawakami, T. Tomita, and Y. Kamio. 1997.
Panton-valentine leukocidin genes in a phage-like particle isolated from
mitomycin C-treated Staphylococcus aureus V8 (ATCC 49775). Biosci. Bio-
technol. Biochem. 61:1960–1962.
49. Karaolis, D. K. R., S. Somara, D. R. Maneval, J. A. Johnson, and J. B.
Kaper. 1999. A bacteriophage encoding a pathogenicity island, a type-IV
pilus and a phage receptor in cholera bacteria. Nature 399:375–379.
50. Kimmit, P. T., C. R. Harwood, and M. R. Barer. 2000. Toxin gene expres-
sion by Shiga toxin-producing Escherichia coli: the role of antibiotics and
the bacterial SOS response. Emerg. Infect. Dis. 6:458–465.
51. Kohler, B., H. Karch, and H. Schmidt. 2000. Antibacterials that are used as
3992 MINIREVIEW
growth promoters in animal husbandry can affect the release of Shiga-toxin-
2-converting bacteriophages and Shiga toxin 2 from Escherichia coli strains.
Microbiology 146:1085–1090.
52. Laird, W., and N. B. Groman. 1976. Prophage map of converting coryne-
bacteriophage beta. J. Virol. 19:208–219.
53. Lazar, S., and M. K. Waldor. 1998. ToxR-independent expression of chol-
era toxin from the replicative form of CTX␾. Infect. Immun. 66:394–397.
54. Lee, J. H., T. Wang, K. Ault, J. Liu, M. P. Schmitt, and R. K. Holmes. 1997.
Identification and characterization of three new promoter/operators from
Corynebacterium diphtheriae that are regulated by the diphtheria toxin re-
pressor (DtxR) and iron. Infect. Immun. 65:4273–4280.
55. Lindsay, J. A., A. Ruzin, H. F. Ross, N. Kurepina, and R. P. Novick. 1998.
The gene for toxic shock toxin is carried by a family of mobile pathogenicity
islands in Staphylococcus aureus. Mol. Microbiol. 29:527–543.
56. Little, J. W. 1996. The SOS regulatory system, p. 453–479. In E. C. C. Lynn
and A. S. Lynch (ed.), Regulation of gene expression in Escherichia coli.
R. G. Landis Co., Georgetown, Tex.
57. Matsushiro, A., K. Sato, H. Miyamoto, T. Yamamura, and T. Honda. 1999.
Induction of prophages of enterohemorrhagic Escherichia coli O157:H7
with norfloxacin. J. Bacteriol. 181:2257–2260.
58. Mavris, M., P. A. Manning, and R. Morona. 1997. Mechanism of bacterio-
phage SfII-mediated serotype conversion in Shigella flexneri. Mol. Micro-
biol. 26:939–950.
59. McDonough, M. A., and J. R. Butterton. 1999. Spontaneous tandem am-
plification and deletion of the Shiga toxin operon in Shigella dysenteriae 1.
Mol. Microbiol. 34:1058–1069.
60. Miller, V. L., R. K. Taylor, and J. J. Mekalanos. 1987. Cholera toxin
transcriptional activator toxR is a transmembrane DNA binding protein.
Cell 48:271–279.
61. Mintz, E. D., T. Popovic, and P. A. Blake. 1994. Transmission of Vibrio
cholerae O1, p. 345–356. In I. K. Wachsmuth, P. A. Blake, and O. Olsvik
(ed.), Vibrio cholerae and cholera: molecular to global perspectives. Amer-
ican Society for Microbiology, Washington, D.C.
62. Mirold, S., W. Rabsch, M. Rohde, S. Stender, H. Tschape, H. Russmann, E.
Igwe, and W. D. Hardt. 1999. Isolation of a temperate bacteriophage en-
coding the type III effector protein SopE from an epidemic Salmonella
typhimurium strain. Proc. Natl. Acad. Sci. USA 96:9845–9850.
63. Muniesa, M., F. Lucena, and J. Jofre. 1999. Comparative survival of free
Shiga toxin 2-encoding phages and Escherichia coli strains outside the gut.
Appl. Environ. Microbiol. 65:5615–5618.
64. Neely, M. N., and D. I. Friedman. 1998. Arrangement and functional iden-
tification of genes in the regulatory region of lambdoid phage H-19B, a
carrier of a Shiga-like toxin. Gene 223:105–113.
65. Neely, M. N., and D. I. Friedman. 1998. Functional and genetic analysis of
regulatory regions of coliphage H-19B: location of Shiga-like toxin and lysis
genes suggest a role for phage functions in toxin release. Mol. Microbiol.
28:1255–1267.
66. Newland, J. W., N. A. Strockbine, S. F. Miller, A. D. O’Brien, and R. K.
Holmes. 1985. Cloning of Shiga-like toxin structural genes from a toxin
converting phage of Escherichia coli. Science 230:179–181.
67. Novick, R. P. 2000. Pathogenicity factors and their regulation, p. 392–407.
In V. A. Fischetti, R. P. Novick, J. J. Ferretti, D. A. Portnoy, and J. I. Rood
(ed.), Gram-positive pathogens. ASM Press, Washington, D.C.
68. Ostroff, S. M., P. I. Tarr, M. A. Neill, J. H. Lewis, N. Hargrett-Bean, and
J. M. Kobayashi. 1989. Toxin genotypes and plasmid profiles as determi-
nants of systemic sequelae in Escherichia coli O157:H7 infections. J. Infect.
Dis. 160: 994–998.
69. Pacheco, S. V., O. G. Gonzalez, and G. L. P. Contreras. 1997. The lom gene
of bacteriophage ␭ is involved in Escherichia coli K12 adhesion to human
buccal epithelial cells. FEMS Microbiol. Lett. 156:129–132.
70. Pavia, A. T., C. R. Nichols, D. P. Green, R. V. Tauxe, S. Mottice, K. D.
Greene, J. G. Wells, R. L. Siegler, E. D. Brewer, D. Hannon, et al. 1990.
Hemolytic-uremic syndrome during an outbreak of Escherichia coli
O157:H7 infections in institutions for mentally retarded persons: clinical
and epidemiologic observations. J. Pediatr. 116:544–551.
71. Plunkett, G., 3rd, D. J. Rose, T. J. Durfee, and F. R. Blattner. 1999.
Sequence of Shiga toxin 2 phage 933W from Escherichia coli O157:H7:
Shiga toxin as a phage late-gene product. J. Bacteriol. 181:1767–1778.
72. Reeve, J. N., and J. E. Shaw. 1979. Lambda encodes an outer membrane
protein: the lom gene. Mol. Gen. Genet. 172:243–248.
73. Reidl, J., and J. J. Mekalanos. 1995. Characterization of Vibrio cholerae
bacteriophage K139 and use of a novel mini-transposon to identify a phage-
encoded virulence factor. Mol. Microbiol. 18:685–701.
74. Robbins, P. W., and T. Uchida. 1965. Chemical and macromolecular struc-
ture of O antigens from Salmonella anatum strains carrying mutants of
bacteriophage ε34. J. Biol. Chem. 240:375–383.
75. Robbins, P. W., and T. Uchida. 1962. Studies on the chemical basis of the
phage conversion of O-antigens in the E-group Salmonellae. Biochemistry
1:323–335.
76. Ruzin, A., J. A. Lindsay, and R. P. Novick. 2001. Molecular genetics of
SaPI1–a mobile pathogenicity island in Staphylococcus aureus. Mol. Micro-
biol. 41:365–377.
INFECT. IMMUN.
77. Sako, T., S. Sawaki, T. Sakurai, S. Ito, Y. Yoshizawa, and I. Kondo. 1983.
Cloning and expression of the staphylokinase gene of Staphylococcus aureus
in Escherichia coli. Mol. Gen. Genet. 190:271–277.
78. Schmitt, M. P. 1997. Transcription of the Corynebacterium diphtheriae
hmuO gene is regulated by iron and heme. Infect. Immun. 65:4634–4641.
79. Schmitt, M. P. 1997. Utilization of host iron sources by Corynebacterium
diphtheriae: identification of a gene whose product is homologous to eu-
karyotic heme oxygenases and is required for acquisition of iron from heme
and hemoglobin. J. Bacteriol. 179:838–845.
80. Schmitt, M. P., and R. K. Holmes. 1994. Cloning, sequence, and footprint
analysis of two promoter/operators from Corynebacterium diphtheriae that
are regulated by the diphtheria toxin repressor (DtxR) and iron. J. Bacte-
riol. 176:1141–1149.
81. Schmitt, M. P., and R. K. Holmes. 1991. Iron-dependent regulation of
diphtheria toxin and siderophore expression by the cloned Corynebacterium
diphtheriae repressor gene dtxR in C. diphtheriae C7 strains. Infect. Immun.
59:1899–1904.
82. Schmitt, M. P., B. G. Talley, and R. K. Holmes. 1997. Characterization of
lipoprotein IRP1 from Corynebacterium diphtheriae, which is regulated by
the diphtheria toxin repressor (DtxR) and iron. Infect. Immun. 65:5364–
5367.
Downloaded from http://iai.asm.org/ on August 12, 2014 by guest
83. Schmitt, M. P., E. M. Twiddy, and R. K. Holmes. 1992. Purification and
characterization of the diphtheria toxin repressor. Proc. Natl. Acad. Sci.
USA 89:7576–7580.
84. Spanier, J. G., and P. P. Cleary. 1980. Bacteriophage control of anti-
phagocytic determinants in group A streptococci. J. Exp Med. 152:1393–
1406.
85. Stanley, R. L., C. D. Ellermeier, and J. M. Slauch. 2000. Tissue-specific
gene expression identifies a gene in the lysogenic phage Gifsy-1 that affects
Salmonella enterica serovar typhimurium survival in Peyer’s patches. J.
Bacteriol. 182:4406–4413.
86. Strockbine, N. A., M. P. Jackson, L. M. Sung, R. K. Holmes, and A. D.
O’Brien. 1988. Cloning and sequencing of the genes for Shiga toxin from
Shigella dysenteriae type 1. J. Bacteriol. 170:1116–1122.
87. Sung, L. M., M. P. Jackson, A. D. O’Brien, and R. K. Holmes. 1990.
Transcription of the Shiga-like toxin type II and Shiga-like toxin type II
variant operons of Escherichia coli J. Bacteriol. 172:6386–6395.
88. Trucksis, M., J. E. Galen, J. Michalski, A. Fasano, and J. B. Kaper. 1993.
Accessory cholera enterotoxin (Ace), the third toxin of a Vibrio cholerae
virulence cassette. Proc. Natl. Acad. Sci. USA 90:5267–5271.
89. Ubukata, K., M. Konno, and R. Fujii. 1975. Transduction of drug resistance
to tetracycline, chloramphenicol, macrolides, lincomycin and clindamycin
with phages induced from Streptococcus pyogenes. J. Antibiot. (Tokyo)
28:681–688.
90. Uchida, T., D. M. Gill, and A. M. Pappenheimer. 1971. Mutation in the
structural gene for diphtheria toxin carried by temperate phage ␤. Nature
New Biol. 233:8–11.
91. Vaca, S., J. Arce, G. Oliver, D. Arenas, and F. Arguello. 1989. FIZ15
bacteriophage increases the adhesion of Pseudomonas aeruginosa to human
buccal epithelial cells. Rev. Lat. Am. Microbiol. 31:1–5.
92. van der Vijver, J. C., M. van Es-Boon, and M. F. Michel. 1972. Lysogenic
conversion in Staphylococcus aureus to leucocidin production. J. Virol.
10:318–319.
93. Veldkamp, K. E., H. C. Heezius, J. Verhoef, J. A. van Strijp, and K. P. van
Kessel. 2000. Modulation of neutrophil chemokine receptors by Staphylo-
coccus aureus supernate. Infect. Immun. 68:5908–5913.
94. Voelker, L. L., and K. Dybvig. 1999. Sequence analysis of the Mycoplasma
arthritidis bacteriophage MAV1 genome identifies the putative virulence
factor. Gene 233:101–107.
95. Wadolkowski, E. A., L. M. Sung, J. A. Burris, J. E. Samuel, and A. D.
O’Brien. 1990. Acute renal tubular necrosis and death of mice orally in-
fected with Escherichia coli strains that produce Shiga-like toxin type II.
Infect. Immun. 58:3959–3965.
96. Wagner, P. L., D. W. K. Acheson, and M. K. Waldor. 2001. Human neu-
trophils and their products induce Shiga toxin production by enterohem-
orrhagic Escherichia coli. Infect. Immun. 69:1934–1937.
97. Wagner, P. L., J. Livny, M. N. Neely, D. W. K. Acheson, D. I. Friedman, and
M. K. Waldor. 2002. Bacteriophage control of Shiga toxin 1 production and
release by Escherichia coli. Mol. Microbiol. 44:957–970.
98. Wagner, P. L., M. N. Neely, X. Zhang, D. W. Acheson, M. K. Waldor, and
D. I. Friedman. 2001. Role for a phage promoter in Shiga toxin 2 expression
from a pathogenic Escherichia coli strain. J. Bacteriol. 183:2081–2085.
99. Waldor, M. K. 1998. Bacteriophage biology and bacterial virulence. Trends
Microbiol. 6:295–297.
100. Waldor, M. K., and J. J. Mekalanos. 1996. Lysogenic conversion by a
filamentous phage encoding cholera toxin. Science 272:1910–1914.
101. Walterspiel, J. N., S. Ashkenazi, A. L. Morrow, and T. G. Cleary. 1992.
Effect of subinhibitory concentrations of antibiotics on extracellular Shiga-
like toxin I. Infection 20:25–29.
102. Weeks, C. R., and J. J. Ferretti. 1984. The gene for type A streptococcal
exotoxin (erythrogenic toxin) is located in bacteriophage T12. Infect. Im-
mun. 46:531–536.
VOL. 70, 2002
103. Willshaw, G. A., H. R. Smith, S. M. Scotland, and B. Rowe. 1985. Cloning
of genes determining the production of vero cytotoxin by Escherichia coli.
J. Gen. Microbiol. 131:3047–3053.
104. Wong, C. S., S. Jelacic, R. L. Habeeb, S. L. Watkins, and P. I. Tarr. 2000.
The risk of the hemolytic-uremic syndrome after antibiotic treatment of
Escherichia coli O157:H7 infections. N. Engl. J. Med. 342:1930–1936.
105. Wood, M. W., R. Rosqvist, P. B. Mullan, M. H. Edwards, and E. E. Galyov.
1996. SopE, a secreted protein of Salmonella dublin, is translocated into the
target eukaryotic cell via a sip-dependent mechanism and promotes bacte-
rial entry. Mol. Microbiol. 22:327–338.
106. Wright, A. 1971. Mechanism of conversion of the salmonella O antigen by
Editor: D. A. Portnoy
MINIREVIEW 3993
bacteriophage ε34. J. Bacteriol. 105:927–936.
107. Yamaguchi, T., T. Hayashi, H. Takami, K. Nakasone, M. Ohnishi, K.
Nakayama, S. Yamada, H. Komatsuzawa, and M. Sugai. 2000. Phage con-
version of exfoliative toxin A production in Staphylococcus aureus. Mol.
Microbiol. 38:694–705.
108. Zabriskie, J. B. 1964. The role of temperate bacteriophage in the production
of erythrogenic toxin by group A streptococci. J. Exp. Med. 119:761–779.
109. Zhang, X., A. D. McDaniel, L. E. Wolf, G. T. Keusch, M. K. Waldor, and
D. W. Acheson. 2000. Quinolone antibiotics induce Shiga toxin-encoding
bacteriophages, toxin production, and death in mice. J. Infect. Dis. 181:
664–670.
Downloaded from http://iai.asm.org/ on August 12, 2014 by guest

Edited by:
Yongqun “Oliver” He,
University of Michigan, USA
Reviewed by:
Jianguo Jian Zhu,
Shanghai Jiao Tong University, China
Menachem Banai,
Kimron Veterinary Institute, Israel
Jiabo Ding,
China Institute of Veterinary Drugs
Control, China
*Correspondence:
Bu-Yun Cui
cuibuyun@icdc.cn;
Yung-Fu Chang
yc42@cornell.edu;
Xiao-Kui Guo
microbiology@sjtu.edu.cn;
Yong-Zhang Zhu
yzhzhu@hotmail.com
Specialty section:
This article was submitted to
Infectious Diseases,
a section of the journal
Frontiers in Microbiology
Received: 30 June 2015
Accepted: 19 November 2015
Published: 08 December 2015
Citation:
Kang Y-X, Li X-M, Piao D-R, Tian G-Z,
Jiang H, Jia E-H, Lin L, Cui B-Y,
Chang Y-F, Guo X-K and Zhu Y-Z
(2015) Typing Discrepancy Between
Phenotypic and Molecular
Characterization Revealing an
Emerging Biovar 9 Variant of Smooth
Phage-Resistant B. abortus Strain
8416 in China.
Front. Microbiol. 6:1375.
doi: 10.3389/fmicb.2015.01375
Frontiers in Microbiology | www.frontiersin.org
CASE REPORT
published: 08 December 2015
doi: 10.3389/fmicb.2015.01375
Typing Discrepancy Between
Phenotypic and Molecular
Characterization Revealing an
Emerging Biovar 9 Variant of Smooth
Phage-Resistant B. abortus Strain
8416 in China
Yao-Xia Kang 1,2 , Xu-Ming Li 1 , Dong-Ri Piao 3 , Guo-Zhong Tian3 , Hai Jiang 3 , En-Hou Jia 3 ,
Liang Lin 2 , Bu-Yun Cui 3*, Yung-Fu Chang 4* , Xiao-Kui Guo 1* and Yong-Zhang Zhu1,4*
1
Department of Immunology and Microbiology, Institutes of Medical Sciences, Shanghai Jiao Tong University School of
Medicine, Shanghai, China, 2 Baotou Municipal Center for Disease Control and Prevention, Baotou, China, 3 State Key
Laboratory for Infectious Disease Prevention and Control, Collaborative Innovation Center for Diagnosis and Treatment of
Infectious Diseases, National Institute for Communicable Disease Control and Prevention, Beijing, China, 4 Department of
Population Medicine and Diagnostic Sciences, College of Veterinary Medicine, Cornell University, Ithaca, NY, USA
A newly isolated smooth colony morphology phage-resistant strain 8416 isolated from a
45-year-old cattle farm cleaner with clinical features of brucellosis in China was reported.
The most unusual phenotype was its resistance to two Brucella phages Tbilisi and
Weybridge, but sensitive to Berkeley 2, a pattern similar to that of Brucella melitensis
biovar 1. VITEK 2 biochemical identification system found that both strain 8416 and
B. melitensis strains shared positive ILATk, but negative in other B. abortus strains.
However, routine biochemical and phenotypic characteristics of strain 8416 were most
similar to that of B. abortus biovar 9 except CO2 requirement. In addition, multiple PCR
molecular typing assays including AMOS-PCR, B. abortus special PCR (B-ab PCR) and
a novel sub-biovar typing PCR, indicated that strain 8416 may belong to either biovar
3b or 9 of B. abortus. Surprisingly, further MLVA typing results showed that strain 8416
was most closely related to B. abortus biovar 3 in the Brucella MLVA database, primarily
differing in 4 out of 16 screened loci. Therefore, due to the unusual discrepancy between
phenotypic (biochemical reactions and particular phage lysis profile) and molecular
typing characteristics, strain 8416 could not be exactly classified to any of the existing
B. abortus biovars and might be a new variant of B. abortus biovar 9. The present
study also indicates that the present phage typing scheme for Brucella sp. is subject to
variation and the routine Brucella biovar typing needs further studies.
Keywords: B. abortus, smooth phage-resistant (SPR), MLVA typing, unusual biochemical reactions
Brucellosis is one of the most common zoonotic infectious diseases, causing enormous economic
loss in domestic animals and public health problems worldwide (Adone and Pasquali, 2013; Van der
Henst et al., 2013). Transmission from animals to human occurs primarily through direct contact
with infected animals and ingestion of raw milk or unpasteurized cheese. On the basis of obviously
1 December 2015 | Volume 6 | Article 1375
Kang et al.
different phenotypic characteristics, host preference, growth
and biochemical characteristics including CO2 requirement,
substrate utilization and growth on dyes and agglutination with
monospecific sera as well as Brucella phage lysis profiles, four
main Brucella pathogenic species including Brucella melitensis
(sheep and goat), B. suis (pigs), B. abortus (cattle), and B. canis
(dogs), a taxonomic scheme can be defined and further divided
into multiple biovars. For example, B. abortus is subdivided into
eight biovars (biovar 1–7 and 9) (Van der Henst et al., 2013).
Because of unstable phenotypic characteristics among Brucella
strains, it is somewhat difficult to define atypical strains into
standard biovars. For instance, the susceptibility of smooth
B. abortus strains to lysis by most of brucella phages, such
as Tbilisi (Tb), Firenze (Fi), Weybridge (Wb), and Berkeley 2
(BK2 ), is commonly regarded as one of the routine criteria to
differentiate this organism from other Brucella species. However,
the majority of B. abortus strains resistant to Brucella phage
have been currently reported primarily due to variation from
smooth to rough form during normal in vitro culture. Since the
first smooth phage-resistant strain (SPR) of B. abortus isolated
from bovine tissue was reported in 1973 (Corbel and Morris,
1974, 1975), a similar study describing SPR strains has not been
reported yet. In this study, we report a newly isolated SPR
strain, strain 8416 from a patient with brucellosis in the Inner
Mongolia Autonomous Region of China on 2012. Actually, it
was the only B. abortus strain among a total of 197 Brucella
strains isolated and authenticated by Chinese CDC during this
year. The Inner Mongolia Autonomous Region has the highest
incidence, responsible for about more than 40% of reported cases
in China (Zhang et al., 2010; Chen et al., 2013). Interestingly, the
unique phenotypical characteristics of the B. abortus SPR strain
8416, determined by routine biotyping for the identification of
Brucella species and biovars, did not completely fit into any
of the recognized classification biovars, indicating the potential
presence of a new variant of B. abortus biovar 3.
MATERIALS AND METHODS
Bacterial Isolation and Used Strains
The protocol for this study was approved by ethics committee of
local disease control and Prevention Research Center of the Inner
Mongolia Autonomous Region and Baotou Municipal Center
for Disease Control and Prevention. In June 2012, two workers
from a cattle farm in Sichuan province, presenting fever, night
sweat and soreness of waist, arthralgia and muscle weakness,
were admitted to one local hospital in the Inner Mongolia
Autonomous Region. The serum samples from these two patients
were strongly positive to Brucella by both Rose-Bengal-plate-
agglutination-test (RBPT) and Serum Agglutination Test (SAT)
with titers of 1/320 according to standard procedures. Moreover,
the two serum samples were also confirmed by positive ELISA
results with Brucella IgG (>150 U/ml) and IgM (>60 U/ml)
(Brucella IgG and IgM ELISA kits, IBL Germany). At the same
time, the blood culture of the two patients were inoculated in
a dual-phase coloration blood culture bottle (BioMerieux Inc.,
Durham, USA) at 37◦ C for 2–3 weeks at the diagnostic laboratory
Frontiers in Microbiology | www.frontiersin.org
Identification of an Unusual B. abortus SPR Strain
of Baotou Municipal Center for Disease Control and Prevention,
the Inner Mongolia Autonomous Region of China. However,
only one blood sample from a 45-year-old male janitor yielded a
positive culture result. The isolated strain 8416 displayed smooth,
tiny, white, shiny and translucent colonies on solid agar after
3 days of incubation. The strain 8416 was sub-cultured on blood
plate with 5% CO2 and displayed typical colonies with small
Gram-negative coccobacilli. The strain was sent to department
of brucellosis, Chinese Communicable Disease Control and
Prevention (Chinese CDC) for further analysis and identification.
The reference strains including B. abortus biovar 1 to 7 and
9, strains: 544A (ATCC 23448), 86/8/59 (ATCC 23449), Tulya
(ATCC 23450), 292 (ATCC 23451), B3196 (ATCC 23452), 870
(ATCC 23453), 63/75, and C68 (ATCC 23455), B. melitensis
biovar 1 to 3, strains: 16M (ATCC 23456), 63/9 (ATCC 23457)
and Ether (ATCC 23458)), B. suis biovar 1 to 5, strains: 1330S
(ATCC 23444), Thomsen (ATCC 23445), 686 (ATCC 23446), 40
(ATCC 23447), and 513, B. neotomae RM6/66 (ATCC 23365),
B. ovis 63/290 (ATCC 25840), and B. canis 5K33 (ATCC 23459)
were used as controls for phenotype typing, biochemical and/or
molecular analysis.
Analysis of Phenotypic Characteristics
At first, to exclude mixed cultures of different biovars and phage
carrier state, the strain used in this study was subjected to a single
cloned isolation for successive three times to confirm no variable
colonial morphology as described by Jones et al. (1962). The
strain was further characterized by using the classical Brucella
phenotypic identification procedures, such as CO2 requirement,
H2 S production, dye sensitivity by basic fuchsin and thionin,
agglutination with monospecific antisera, and phage typing as
described by Alton GG (Alton et al., 1975). Brucella monospecific
antisera to A, M, and R (rough) and Brucella phages Tb, Wb,
and Bk2 were used according to standard protocol of the Chinese
CDC (Jiang et al., 2013) to characterize this strain. All of
phenotypic characterizations in this study were repeated at least
three times to make sure the results are repeatable.
Molecular Typing Identification
Brucella strains were inactivated by suspending one loop from
a solid bacterial culture in 200 µl DNA storage buffer. Total
genomic DNA was extracted using the DNeasy Blood &
Tissue Kit (Qiagen China Ltd., Beijing, China) following the
manufacture’s instruction. The PCR assay targeting bcsp31, was
performed to confirm the Brucella genus as previously described
(Bounaadja et al., 2009), and species-level using the routine
Abortus-Melitensis-Ovis-Suis PCR (AMOS-PCR) (Bricker and
Halling, 1994). Furthermore, B. abortus B-ab PCR and a novel
PCR to differentiate B. abortus biovar 3a, 3b, 5, 6, and 9 were
performed as previously described (Ocampo-Sosa et al., 2005;
Huber et al., 2009).
Multiple Locus Variable Number Tandem
Repeat Analysis (MLVA) Genotyping
Multiple locus variable number tandem repeat analysis (MLVA)
was performed as previously described by Le Fleche et al. (2006)
2 December 2015 | Volume 6 | Article 1375
Kang et al. Identification of an Unusual B. abortus SPR Strain

and by Jiang et al. (2013), respectively. The 16 primer pairs

B. abortus biovar 9 variant

comprised three main groups: panel 1 including bruce06, 08, 11,

B. melitensis biovar 1
B. melitensis biovar 3
B. abortus biovar 1
B. abortus biovar 3
B. abortus biovar 6
B. abortus biovar 9
12, 42, 43, 45, and 55 for species identification, panel 2A (bruce18,

B. suis biovar 1
Interpretation 19, and 21), and panel 2B (bruce04, 07, 09, 16, and 30) for further
subspecies differentiation were used.

Biochemical Identification by VITEK 2

bruce30 System
3
5
5
3
3
6
3
3
A total of 47 biochemical reactions of the Brucella strains

11
bruce16
were analyzed using the standard Gram-negative bacteria
3
4

3
2
3
3
5
bruce09 identification card on automatic VITEK 2 system according to
7
3
3
2
2
8
1
5
the manufacturer’s instructions.
TABLE 1 | Comparison of phenotypic characteristics and Brucella phage lysis profiles of Brucella abortus strain 8416 and other Brucella reference strains.

bruce07
6
5
5
2
2
5
1
6
bruce04
6
3
6
3
3
2
3
6
bruce21 RESULTS
8
8
8
8
8
6
8
9
Routine Phenotypic Typing
42
21
20
42
42
18
42
brucel9 19
Brucella MLVA16

bruce18 Characteristics
6
5
8
7
7
5
7
4

bruce55 According to routine phenotypic analysis, strain 8416 was anti-

2
3
3
3
3
3
9
2

R negative and H2 S positive, agglutination with anti-M serum

but not anti-A serum and grew in the presence of thionine and
12

bruce45
12

13
3
3
3

bruce43 fuchsin dyes (Table 1). Moreover, it was not lysed by Tb and
2
2
2
3
3
2
3
1

Wb phages both in 1× RTD (Routine Test Dilution) and 104 ×

bruce42
RTD, but lysed by BK2 phage both in 1× RTD and 102 × RTD
12
2
2
2
3
3
4

bruce12 (Figure 1A). Thus, the particular phenotypic profiles of the strain
12
12
11

13

10
5
5

8416 were more similar to that of the classic characteristics of

bruce11
B. abortus biovars 9.
4
4
4
6
6
2
5
6

bruce08
5
5
5
3
3
4
3
3

bruce06
Biochemical Identification of Automatic
4
4
3
3
6
3
7
2

VITEK 2 System
Four biochemical indicators ProA (L-pyrrolydonyl-arylamidase),
BK2
Mono specific phage at

+
+
+
+
+
+
+
+

TyrA (tyrose arylamidase), URE (urease), and GlyA could be

RTD

used to distinguish Brucella species. All of eight B. abortus

Wb

–
–
+
+
+
+

reference strains and 21 field strains were positive in ILATk (L-

Tb

lactate alkalization), but it was negative in strain 8416, three

–

–
–
–
+
+
+
+

B. melitensis reference strains and 92 field strains (Cui BuYun’s

M

–
–
–

–
+

+
+
+
Sera

unpublished data). This result indicated that strain 8416 showed

special biochemical characteristics distinct from that of B. abortus
–

–
A

+
+
+

+
+
+

strains.
Fuschin

–
+
+
+
+
+
+
+

Molecular Typing Identification

Strain 8416 was identified as B. abortus by the combination of
Thionin

bcsp31 PCR (223-bp, data not shown) and B-ab PCR (370-bp)
Growth characteristics

–
+

+
+
+
+
+
+

(Figure 1B) but not as biovar 1, 2, and 4 of B. abortus according

to AMOS-PCR (Figure 1C). The novel PCR assay was used to
production

compare strain 8416 to B. abortus biovar 3b, 5, 6, and 9, and

H2 S

found that the PCR product of 1.7 kb from strain 8416 was similar
++
–

–
–
+
+
+

to B. abortus biovar 3b, 5, 6, and 9, but not to other B. abortus

biovars (Figure 1D).
requirement

MLVA Genotyping
±

±
–

–
–
–
–
–
CO2

According to Brucella MLVA typing database (Grissa et al., 2008),

16 loci of MLVA matching results displayed that strain 8416
was closely related to B. abortus biovar 3 (Jiang et al., 2013),
1330S
Strain

but primarily different in four variable loci, bruce04, bruce07,

544A

Ether
8416

Tulya

16M
C68
870

bruce11, and bruce55 (Table 2).

Frontiers in Microbiology | www.frontiersin.org 3 December 2015 | Volume 6 | Article 1375

Kang et al. Identification of an Unusual B. abortus SPR Strain

FIGURE 1 | (A) The lysis patterns of phage Tb, Wb and Bk2 to Brucella abortus strain 8416, B. abortus biovar 1 strain 544A (A is indicated as B. abortus),
B. melitensis biovar 1 strain 16M (M is indicated as B. melitensis), and B. suis biovar 1 strain 1330S (S is indicated as B. suis) as well as B. abortus biovar 6 strain
870 and biovar 9 strain C68; (B) Amplification of DNA fragments from different Brucella strains. Genomic DNA was amplified by the B-ab PCR assay. 1: strain 8416;
2–5: four B. melitensis field strains; 104 M: B. melitensis biovar 1 strain 104M; 544A: B. abortus biovar 1 strain 544A; (C) Amplification of DNA fragments from
different Brucella strains. Genomic DNA was amplified by AMOS-PCR assay. 1: strain 8416; 2–5: four B. melitensis field strains; 104 M: B. melitensis biovar 1 strain
104M; 544A: B. abortus biovar 1 strain 544A; (D) Amplification of DNA fragments from different Brucella strains. Genomic DNA was amplified by new PCR assay
identifying B. abortus biovar 3b, 5, 6, and 9. 1: B. melitensis biovar 1 strain 16M; 2: B. abortus biovar 1 strain 544A; 3: B. suis biovar 1 strain 1330S; 4: B. abortus
biovar 9 strain C68; 5: B. abortus biovar 3a strain Tulya; 6: strain 8416.

Finally, based on these typing results, strain 8416 might be a
new variant of B. abortus biovar 9.
DISCUSSION
Until now, the phage resistance mechanism from Brucella SPR
strains was poorly understood. In this study, a natural SPR strain
of B. abortus isolated from a patient in China was identified.
Although SPR strains of B. abortus were rarely isolated from
patients, a SPR strain was isolated from a B. abortus phage
sensitive parent strain 544 in 1974 and a SPR variant of B. abortus
strain 19 was identified in 1976 through the manipulation
of laboratory cultures (Corbel and Morris, 1974; Corbel and
Thomas, 1976). Compared to the parent strain 544, the SPR strain
FS showed no differences in virulence, morphological, cultural,
biochemical or metabolic, and serological reactions, but with an
altered phage resistance profile (Corbel and Morris, 1974). The
potential mechanism of the phage resistance may be due to its
failure to penetrate the FS cell wall since the strain FS is more
resistant to lysis by phage lysozymes than that of the phage-
sensitive parent strain 544 (Corbel and Morris, 1975). Strain
544-FS showed a complete resistance to lysis by many Brucella
phages except Bk2 at 1× RTD and 104 × RTD. Subsequently,
another B. abortus SPR strain with resistance to phage Tb, was
isolated from a supramammary lymph node of a cow and it is
virulent to guinea-pigs (Harrington et al., 1977). Interestingly,
these B. abortus SPR strains mentioned above belonging to
B. abortus biovar 1 were identified. However, strain 8416 was
significantly different from all of B. abortus biovars by using

Frontiers in Microbiology | www.frontiersin.org 4 December 2015 | Volume 6 | Article 1375

 TABLE 2 | Comparison of Brucella MLVA typing results of B. abortus strain 8416 and the most closely related B. aborus biovar 3 field strains in the Brucella MLVA database.
Kang et al.

Distance
MLVA8
MLVA11
MLVA16
Bruce06- 1322
Bruce08- 1134
Bruce11- 211
Bruce12 - 73
Bruce42-424
Bruce43-379
Bruce45-233
Bruce55-2066
Bruce18-339
Bruce19-324
Bruce21-329
Bruce04-1543
Bruce07-1250
Bruce09-588
Bruce16-548
Bruce30-1505

Strain name BaseView Strain Host Isolated_in Species- Contact Group Year
biovar

8416 8416 Human Inner Mongolia, 2012 4 5 4 12 2 2 3 2 6 42 8 6 6 7 3 3

China
2013Jiang#095 0 Brucella2013 Human Inner Mongolia, B. abortus Buyun Cui 2012 30 4 5 4 12 2 2 3 2 6 42 8 6 6 7 3 3
China biovar 3

Frontiers in Microbiology | www.frontiersin.org

2013Garofolo_ 3 Brucella_ 9263 Bufalo Albanella,Italy B. abortus_ Giuliano 2011 36 72 4 5 3 12 2 2 3 1 6 42 8 5 6 7 3 3
9263 ITALIA_1 biovar3 Garofolo
2013Jiang#108 3 Brucella2013 NM1068 Cattle Inner Mongolia, B. abortus Buyun Cui 1985 36 4 5 3 12 2 2 3 1 6 42 8 6 7 7 3 3
China biovar 3
2013Jiang#105 3 Brucella2013 NM1065 Cattle Inner Mongolia, B. abortus Buyun Cui 1985 117 4 5 3 12 2 2 3 2 6 42 8 6 4 4 3 3
China biovar 3
2013Jiang#131 3 Brucella2013 NM1158 Cattle Inner Mongolia, B. abortus Buyun Cui 1988 36 4 5 3 12 2 2 3 1 6 42 8 6 4 7 3 3
China biovar 3
2013Jiang#140 3 Brucella2013 NM1175 Cattle Inner Mongolia, B. abortus Buyun Cui 1990 36 4 5 3 12 2 2 3 1 6 42 8 6 4 7 3 3
China biovar 3
2013Garofolo_ 3 Brucella_ 3636 Cattle Monte San B. abortus_ Giuliano 2011 36 72 4 5 3 12 2 2 3 1 6 42 8 4 6 7 3 3
3636 ITALIA_1 Giacomo,Italy biovar3 Garofolo

5
2013Garofolo_ 3 Brucella_ 3916 Cattle Monte San B. abortus_ Giuliano 2011 36 72 4 5 3 12 2 2 3 1 6 42 8 4 6 7 3 3
3916 ITALIA_1 Giacomo,Italy biovar3 Garofolo
2013Garofolo_ 3 Brucella_ 3920 Cattle Teggiano,Italy B. abortus_ Giuliano 2011 36 72 4 5 3 12 2 2 3 1 6 42 8 4 6 7 3 3
3920 ITALIA_1 biovar3 Garofolo
2013Garofolo_ 3 Brucella_ 4363 Cattle Laurenzana,Italy B. abortus_ Giuliano 2011 36 72 4 5 3 12 2 2 3 1 6 42 8 4 6 7 3 3
4363 ITALIA_1 biovar3 Garofolo
MLVA16

2013Garofolo_ 3 Brucella_ 12183 Cattle Corleto B. abortus_ Giuliano 2011 36 72 4 5 3 12 2 2 3 1 6 42 8 4 6 7 3 3

12183 ITALIA_1 Monforte,Italy biovar3 Garofolo
2013Garofolo_ 3 Brucella_ 12185 Cattle San Rufo,Italy B. abortus_ Giuliano 2011 36 72 4 5 3 12 2 2 3 1 6 42 8 4 6 7 3 3
12185 ITALIA_1 biovar3 Garofolo
2013Garofolo_ 3 Brucella_ 21571 Cattle Monte San B. abortus_ Giuliano 2011 36 72 4 5 3 12 2 2 3 1 6 42 8 4 6 7 3 3
21571 ITALIA_1 Giacomo,Italy biovar3 Garofolo
2013Garofolo_ 3 Brucella_ 21675 Cattle Monte San B. abortus_ Giuliano 2011 36 72 4 5 3 12 2 2 3 1 6 42 8 4 6 7 3 3
21675 ITALIA_1 Giacomo,Italy biovar3 Garofolo
2013Garofolo_ 3 Brucella_ 22839 Bufalo Albanella,Italy B. abortus_ Giuliano 2011 36 72 4 5 3 12 2 2 3 1 6 42 8 4 6 7 3 3
22839 ITALIA_1 biovar3 Garofolo
2013Garofolo_ 3 Brucella_ 22842 Cattle Teggiano,Italy B. abortus_ Giuliano 2011 36 72 4 5 3 12 2 2 3 1 6 42 8 4 6 7 3 3
22842 ITALIA_1 biovar3 Garofolo
2013Garofolo_ 3 Brucella_ 5362 Bufalo Monte San B. abortus_ Giuliano 2011 36 72 4 5 3 12 2 2 3 1 6 42 8 4 6 7 3 3
5362 ITALIA_1 Giacomo,Italy biovar3 Garofolo
2013Garofolo_ 3 Brucella_ 8980 Cattle Teggiano,Italy B. abortus_ Giuliano 2011 36 72 4 5 3 12 2 2 3 1 6 42 8 4 6 7 3 3
8980 ITALIA_1 biovar3 Garofolo

(Continued)

December 2015 | Volume 6 | Article 1375

Identification of an Unusual B. abortus SPR Strain
 TABLE 2 | Continued
Kang et al.

Distance
MLVA8
MLVA11
MLVA16
Bruce06-1322
Bruce08-1134
Bruce11-211
Bruce12-73
Bruce42-424
Bruce43-379
Bruce45-233
Bruce55-2066
Bruce18-339
Bruce19-324
Bruce21-329
Bruce04-1543
Bruce07-1250
Bruce09-588
Bruce16-548
Bruce30-1505

Strain name BaseView Strain Host Isolated_in Species- Contact Group Year
biovar

2013Garofolo_ 3 Brucella_ 8984 Cattle Sassano,Italy B. abortus_ Giuliano 2011 36 72 4 5 3 12 2 2 3 1 6 42 8 4 6 7 3 3

8984 ITALIA_1 biovar3 Garofolo
2013Jiang#093 3 Brucella2013 2011166 Human Chongqing, B. abortus Buyun Cui 2011 36 4 5 3 12 2 2 3 1 6 42 8 4 6 7 3 3
China biovar 3

Frontiers in Microbiology | www.frontiersin.org

2013Jiang#094 3 Brucella2013 YLQ Human Zhejiang, China B. abortus Buyun Cui 2006 36 4 5 3 12 2 2 3 1 6 42 8 4 6 7 3 3
biovar 3
2013Jiang#104 4 Brucella2013 NM1061 Cattle Inner Mongolia, B. abortus Buyun Cui 1984 112 4 5 3 12 2 2 3 3 6 42 8 5 4 7 3 3
China biovar 3
2013Jiang#092 4 Brucella2013 2011165\’ Human Chongqing, B. abortus Buyun Cui 2011 36 4 5 3 12 2 2 3 1 6 42 8 4 7 7 3 3
China biovar 3
2012 4 Brucella2012 LNIV- Alentejo, B. abortus 3 Cristina B. abortus 2006 36 72 4 5 3 12 2 2 3 1 6 42 8 4 7 7 3 3
Ferreira#146 328Ba3-06 Portugal Ferreira
2013Garofolo_ 4 Brucella_ 3272 Cattle Apricena,Italy B. abortus_ Giuliano 2011 4 5 3 12 2 2 3 3 6 42 8 7 6 3 3 3
3272 ITALIA_1 biovar3 Garofolo
2013Garofolo_ 4 Brucella_ 18081 Cattle Apricena,Italy B. abortus_ Giuliano 2011 4 5 3 12 2 2 3 3 6 42 8 7 6 3 3 3

6
18081 ITALIA_1 biovar3 Garofolo
2006 4 Brucella2012 BCCN#99- Cattle Mongolia B. abortus 7 Gilles B. abortus 1999 36 72 4 5 3 12 2 2 3 1 6 42 8 5 6 4 3 3
LeFlèche#119 98 Vergnaud
2013Jiang#089 4 Brucella2013 Cattle Hebei, China B. abortus Buyun Cui 2011 36 4 5 3 12 2 2 3 1 6 42 8 4 5 7 3 3
biovar 3
2013Jiang#090 4 Brucella2013 Cattle Hebei, China B. abortus Buyun Cui 2011 36 4 5 3 12 2 2 3 1 6 42 8 4 5 7 3 3
biovar 3
2006 4 Brucella2012 BCCN#94- Cattle Limoges, B. abortus 3 Gilles B. abortus 1994 36 72 4 5 3 12 2 2 3 1 6 42 8 4 5 7 3 3
LeFlèche#112 18 France Vergnaud
2013Jiang#100 4 Brucella2013 NM1051 Cattle Inner Mongolia, B. abortus Buyun Cui 1984 36 4 5 3 12 2 2 3 1 6 42 8 6 4 8 3 3
China biovar 3
2006 4 Brucella2012 REF 292 Cattle England B. abortus 4 Gilles B. abortus 30 78 4 5 4 12 2 2 3 2 6 42 8 3 4 3 3 5
LeFlèche#005 Vergnaud
2009Her#004 4 Brucella2012 KRef04 Cattle England B. abortus 4 Moon Her B. abortus 30 78 4 5 4 12 2 2 3 2 6 42 8 3 4 3 3 5
2012 4 Brucella2012 REF 292 B. abortus 4 Cristina B. abortus 30 78 4 5 4 12 2 2 3 2 6 42 8 3 4 3 3 5
Ferreira#213 Ferreira
2013Jiang#127 4 Brucella2013 NM1147 Cattle Inner Mongolia, B. abortus Buyun Cui 1988 117 4 5 3 12 2 2 3 2 6 42 8 5 4 3 3 3
China biovar 3
2013Jiang#130 4 Brucella2013 NM1156 Sheep Inner Mongolia, B. abortus Buyun Cui 1988 36 4 5 3 12 2 2 3 1 6 42 8 6 4 4 3 3
China biovar 3
2013Jiang#125 4 Brucella2013 NM1140 Cattle Inner Mongolia, B. abortus Buyun Cui 1988 36 4 5 3 12 2 2 3 1 6 42 8 5 4 7 3 3
China biovar 3

(Continued)

December 2015 | Volume 6 | Article 1375

Identification of an Unusual B. abortus SPR Strain
 Kang et al.

TABLE 2 | Continued

Frontiers in Microbiology | www.frontiersin.org

Distance
MLVA8
MLVA11
MLVA16
Br uc e06- 1322
Br uc e08- 1134
Bruce11-211
Bruce12-73
Bruce42-424
Bruce43-379
Bruce45-233
Bruce55-2066
Bruce18-339
Bruce19-324
Bruce21-329
Bruce04-1543
Bruce07-1250
Bruce09-588
Bruce16-548
Bruce30-1505

Strain name BaseView Strain Host Isolated_in Species- Contact Group Year
biovar

2013Jiang#126 4 Brucella2013 NM1146 Cattle Inner Mongolia, B. abortus Buyun Cui 1988 36 4 5 3 12 2 2 3 1 6 42 8 5 4 7 3 3
China biovar 3
2013Jiang#128 4 Brucella2013 NM1148 Cattle Inner Mongolia, B. abortus Buyun Cui 1988 36 4 5 3 12 2 2 3 1 6 42 8 5 4 7 3 3
China biovar 3
2013Jiang#141 4 Brucella2013 NM1176 Cattle Inner Mongolia, B. abortus Buyun Cui 1990 36 4 5 3 12 2 2 3 1 6 42 8 5 4 7 3 3
China biovar 3
2013Jiang#150 4 Brucella2013 NM1215 Cattle Inner Mongolia, B. abortus Buyun Cui 1994 36 4 5 3 12 2 2 3 1 6 42 8 5 4 7 3 3
China biovar 3

7
2013Jiang#151 4 Brucella2013 NM1218 Cattle Inner Mongolia, B. abortus Buyun Cui 1995 36 4 5 3 12 2 2 3 1 6 42 8 5 4 7 3 3
China biovar 3
2013Jiang#146 4 Brucella2013 NM1185 Cattle Inner Mongolia, B. abortus Buyun Cui 1990 36 4 5 3 12 2 2 3 1 6 42 8 6 4 5 3 3
China biovar 3
2013Jiang#152 4 Brucella2013 NM1219 Cattle Inner Mongolia, B. abortus Buyun Cui 1995 36 4 5 3 12 2 2 3 1 6 42 8 5 5 7 3 3
China biovar 3
2013Jiang#113 4 Brucella2013 NM1075 Cattle Inner Mongolia, B. abortus Buyun Cui 1985 117 4 5 3 12 2 2 3 2 8 42 8 5 6 3 3 3
China biovar 3
2006 4 Brucella2012 BfR 95 Mouse ? B. abortus 1 Gilles B. abortus 28 82 4 5 4 12 2 2 3 3 6 42 8 3 6 3 3 5
LeFlèche#135 Vergnaud
2009Her#011 4 Brucella2012 KRef15 Cattle USA B. abortus 1 Moon Her B. abortus 28 82 4 5 4 12 2 2 3 3 6 42 8 3 6 3 3 5
2013Jiang#083 4 Brucella2013 Cattle Xinjiang, China B. abortus Buyun Cui 2011 36 4 5 3 12 2 2 3 1 6 42 8 4 6 5 3 3
biovar 3
2013Garofolo_ 4 Brucella_ 3921 Cattle San Gregorio B. abortus_ Giuliano 2011 36 72 4 5 3 12 2 2 3 1 6 42 8 4 6 6 3 3
3921 ITALIA_1 Magno,Italy biovar3 Garofolo
2013Garofolo_ 4 Brucella_ 5007 Cattle San Gregorio B. abortus_ Giuliano 2011 36 72 4 5 3 12 2 2 3 1 6 42 8 4 6 6 3 3
5007 ITALIA_1 Magno,Italy biovar3 Garofolo

December 2015 | Volume 6 | Article 1375

Identification of an Unusual B. abortus SPR Strain
Kang et al.
phenotypic and molecular typing method. However, it shared
the same phage lysis profiles to that of B. melitensis biovar 1. In
conclusion, strain 8416 is the only SPR strain isolated from the
infected human thus far with a similar phage lysis pattern with
B. melitensis 16 M. However, despite same resistant to phage Tb,
we could not comprehensively compare with phage lysis profiles
of the three reported SPR strains due to different Brucella phages
tested among them.
Currently, MLVA has been mainly used for tracking the
variances of the bacterial genus with a high homology, such as
Brucella genus (Haguenoer et al., 2011). The MLVA-16 (panel
1, 2A and 2B) assay was widely used for molecular typing of
a larger collection of isolates at both species and biovars level.
The panel 1 comprised eight minisatellite markers for species
identification (Le Fleche et al., 2006) and the panel 2 markers were
found with a higher biovar discriminatory power. Surprisingly,
the MLVA-16 typing results showed that strain 8416 was clustered
into the Chinese B. abortus biovar 3 strains (Jiang et al., 2013)
with four variable loci (bruce04, 07, 11, and 55). Actually, among
the four known panel 1 genotypes (28, 30, 112, 116), strain 8416
(genotype 30) was distinct from other 65 Chinese B. abortus
biovar 3 strains isolated previously from different geographic
REFERENCES
Adone, R., and Pasquali, P. (2013). Epidemiosurveillance of brucellosis. Rev. Sci.
Tech. 32, 199–205.
Alton, G. G., Jones, L. M., and Pietz, D. E. (1975). Laboratory techniques in
brucellosis. Monogr. Ser. No.55 World Health Organ. 1975, 1–163.
Bounaadja, L., Albert, D., Chenais, B., Henault, S., Zygmunt, M. S., Poliak, S.,
et al. (2009). Real-time PCR for identification of Brucella spp.: a comparative
study of IS711, bcsp31 and per target genes. Vet. Microbiol. 137, 156–164. doi:
10.1016/j.vetmic.2008.12.023
Bricker, B. J., and Halling, S. M. (1994). Differentiation of Brucella abortus bv. 1, 2,
and 4, Brucella melitensis, Brucella ovis, and Brucella suis bv. 1 by PCR. J. Clin.
Microbiol. 32, 2660–2666.
Chen, Y., Ke, Y., Wang, Y., Yuan, X., Zhou, X., Jiang, H., et al. (2013).
Changes of predominant species/biovars and sequence types of Brucella isolates,
Inner Mongolia, China. BMC Infect. Dis. 13:514. doi: 10.1186/1471-2334-1
3-514
Corbel, M. J., and Morris, J. A. (1974). Studies on a smooth phage-resistant
variant of Brucella abortus I. Immunological properties. Br. J. Exp. Pathol. 55,
78–87.
Corbel, M. J., and Morris, J. A. (1975). Studies on a smooth phage resistant
variant of Brucella abortus II. Mechanism of phage resistance. Br. J. Exp. Pathol.
56, 1–7.
Corbel, M. J., and Thomas, E. L. (1976). The in vivo activity of a smooth
phage-resistant variant of Brucella abortus strain 19. Br. Vet. J. 132,
121–123.
Grissa, I., Bouchon, P., Pourcel, C., and Vergnaud, G. (2008). On-line resources for
bacterial micro-evolution studies using MLVA or CRISPR typing. Biochimie 90,
660–668. doi: 10.1016/j.biochi.2007.07.014
Haguenoer, E., Baty, G., Pourcel, C., Lartigue, M. F., Domelier, A. S., Rosenau, A.,
et al. (2011). A multi locus variable number of tandem repeat analysis (MLVA)
scheme for Streptococcus agalactiae genotyping. BMC Microbiol. 11:171. doi:
10.1186/1471-2180-11-171
Harrington, R. Jr., Bond, D. R., and Brown, G. M. (1977). Smooth phage-resistant
Brucella abortus from bovine tissue. J. Clin. Microbiol. 5, 663–664.
Frontiers in Microbiology | www.frontiersin.org
Identification of an Unusual B. abortus SPR Strain
origins, suggesting that more B. abortus strains phenotypically
identified as biovar 3 are required for the comparison. The MLVA
assay confirmed that B. abortus biovar 3 is a heterogeneous group
(Le Fleche et al., 2006), and in agreement with the B. abortus
biovar 3 divided into two sub-biovar 3a and 3b (Huber et al.,
2009).
In this study, an atypical B. abortus strain displaying a phage
lysis profile similar to B. melitensis biovars 1 was identified.
Most importantly, the lysis pattern by bacteriophages observed
in this newly uncovered B. abortus SPR strain. Although phage
typing in general can successfully classify Brucella species, our
research calls for attention as to conclusions on SPR strains.
Further investigation focusing on the strain 8416’s whole genomic
variations associated with phage resistance is needed.
ACKNOWLEDGMENTS
This study was supported by the National Nature Science
Foundation (81360444, 81460319) and the Inner Mongolia
Autonomous Region of Nature Science Foundation
(2013MS1105).
Huber, B., Scholz, H. C., Lucero, N., and Busse, H. J. (2009). Development of a PCR
assay for typing and subtyping of Brucella species. Int. J. Med. Microbiol. 299,
563–573. doi: 10.1016/j.ijmm.2009.05.002
Jiang, H., Wang, H., Xu, L., Hu, G., Ma, J., Xiao, P., et al. (2013). MLVA genotyping
of Brucella melitensis and Brucella abortus isolates from different animal species
and humans and identification of Brucella suis vaccine strain S2 from cattle in
China. PLoS ONE 8:e76332. doi: 10.1371/journal.pone.0076332
Jones, L. M., McDuff, C. R., and Wilson, J. B. (1962). Phenotypic alterations in the
colonial morphology of Brucella abortus due to a bacteriophage carrier state.
J. Bacteriol. 83, 860–866.
Le Fleche, P., Jacques, I., Grayon, M., Al Dahouk, S., Bouchon, P., Denoeud, F.,
et al. (2006). Evaluation and selection of tandem repeat loci for a Brucella MLVA
typing assay. BMC Microbiol. 6:9. doi: 10.1186/1471-2180-6-9
Ocampo-Sosa, A. A., Aguero-Balbin, J., and Garcia-Lobo, J. M. (2005).
Development of a new PCR assay to identify Brucella abortus biovars 5, 6
and 9 and the new subgroup 3b of biovar 3. Vet. Microbiol. 110, 41–51. doi:
10.1016/j.vetmic.2005.06.007
Van der Henst, C., De Barsy, M., Zorreguieta, A., Letesson, J. J., and De Bolle, X.
(2013). The Brucella pathogens are polarized bacteria. Microbes Infect. 15,
998–1004. doi: 10.1016/j.micinf.2013.10.008
Zhang, W. Y., Guo, W. D., Sun, S. H., Jiang, J. F., Sun, H. L., Li, S. L., et al. (2010).
Human brucellosis, Inner Mongolia, China. Emerg. Infect. Dis. 16, 2001–2003.
doi: 10.3201/eid1612.091081
Conflict of Interest Statement: The authors declare that the research was
conducted in the absence of any commercial or financial relationships that could
be construed as a potential conflict of interest.
Copyright © 2015 Kang, Li, Piao, Tian, Jiang, Jia, Lin, Cui, Chang, Guo and Zhu.
This is an open-access article distributed under the terms of the Creative Commons
Attribution License (CC BY). The use, distribution or reproduction in other forums
is permitted, provided the original author(s) or licensor are credited and that the
original publication in this journal is cited, in accordance with accepted academic
practice. No use, distribution or reproduction is permitted which does not comply
with these terms.
8 December 2015 | Volume 6 | Article 1375
 REVIEW ARTICLE
Cues and regulatory pathways involved in natural competence
and transformation in pathogenic and environmental
Gram-negative bacteria
Patrick Seitz & Melanie Blokesch
Global Health Institute, School of Life Sciences, Ecole Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland
Correspondence: Melanie Blokesch, EPFL-
SV-UPBLO/Station 19, Ecole Polytechnique
Fédérale de Lausanne (EPFL), CH-1015
Lausanne, Switzerland. Tel.: +41 21 693
0653; fax: +41 21 693 7210; e-mail:
melanie.blokesch@epfl.ch
Received 7 May 2012; revised 27 July 2012;
accepted 21 August 2012. Final version
published online 20 September 2012.
DOI: 10.1111/j.1574-6976.2012.00353.x
Editor: Hauke Hennecke
Keywords
horizontal gene transfer; DNA uptake;
carbon catabolite repression; quorum
sensing; DNA repair; type IV pilus.
MICROBIOLOGY REVIEWS
Introduction
In 2011, the spread of food-borne gastroenteritis in
Europe, most notably in Germany, was widely reported.
The pathogen caused severe bloody diarrhoea, often
accompanied by haemolytic uremic syndrome. The causa-
tive agent of the outbreak was Escherichia coli serotype
O104:H4. The strain was unique in that it combined the
‘potentials of two different pathogens: Shiga-toxin-
producing E. coli and enteroaggregative E. coli’ (Bielaszewska
et al., 2011). Multiple isolates of this outbreak were fully
sequenced for molecular epidemiological analysis (Rasko
et al., 2011; Grad et al., 2012). Such whole-genome
sequencing (WGS) studies are extremely helpful in analy-
sing single-nucleotide polymorphisms (SNPs) and allow
us to understand patterns of genetic diversity, which can
help in the identification of transmission routes. Further-
more, WGS sheds light on the acquisition of horizontally
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
Abstract
Bacterial genomics is flourishing, as whole-genome sequencing has become
affordable, readily available and rapid. As a result, it has become clear how fre-
quently horizontal gene transfer (HGT) occurs in bacteria. The potential impli-
cations are highly significant because HGT contributes to several processes,
including the spread of antibiotic-resistance cassettes, the distribution of
toxin-encoding phages and the transfer of pathogenicity islands. Three modes
of HGT are recognized in bacteria: conjugation, transduction and natural
transformation. In contrast to the first two mechanisms, natural competence
for transformation does not rely on mobile genetic elements but is driven
solely by a developmental programme in the acceptor bacterium. Once the
bacterium becomes competent, it is able to take up DNA from the environ-
ment and to incorporate the newly acquired DNA into its own chromosome.
The initiation and duration of competence differ significantly among bacteria.
In this review, we outline the latest data on representative naturally trans-
formable Gram-negative bacteria and how their competence windows differ.
We also summarize how environmental cues contribute to the initiation of
competence in a subset of naturally transformable Gram-negative bacteria
and how the complexity of the niche might dictate the fine-tuning of the
competence window.
acquired genetic elements. Horizontal gene transfer
(HGT) is important in rendering harmless bacteria into
major human pathogens, as exemplified by the commen-
sal bacterium E. coli mentioned above. Indeed, Rasko
et al. (2011) concluded from their WGS findings that
‘horizontal genetic exchange allowed for the emergence of
the highly virulent Shiga-toxin–producing enteroaggrega-
tive E. coli O104:H4 strain that caused the German out-
break’. Recent reviews on this topic concluded that there
is substantial evidence that HGT contributes to the acqui-
sition, integration and maintenance of pathogenicity
islands (PAIs) in bacterial genomes (Hacker et al., 1997,
2003; Dobrindt et al., 2004, 2010; Ahmed et al., 2008). In
general, it is widely accepted that HGT significantly
impacts the evolution of bacterial genomes and their
adaptation to new environments (a nonexhaustive list of
reviews on this topic published within this century would
be Eisen, 2000; Ochman et al., 2000; Boucher et al., 2001;
FEMS Microbiol Rev 37 (2013) 336–363
Regulation of natural competence in Gram-negative bacteria
Koonin et al., 2001; Boucher et al., 2003; Thomas & Niel-
sen, 2005; Davidsen et al., 2007; Cohan & Koeppel, 2008;
Koonin & Wolf, 2008; Ambur et al., 2009; Popa & Dagan,
2011; Stokes & Gillings, 2011; Wiedenbeck & Cohan, 2011).
HGT in bacteria occurs through three main mecha-
nisms: transduction, conjugation, and natural transforma-
tion. The latter mechanism is based on the uptake of free
DNA from the environment and therefore does not rely
on mobile genetic elements such as phages and plasmids;
instead, it is solely encoded by the acceptor bacterium.
Natural competence is the developmental state of the bac-
terium in which it is able to take up external DNA and
to recombine this DNA into the chromosome, thereby
undergoing natural transformation. Natural competence
and transformation are common to a wide variety of bac-
terial species and have been extensively reviewed (Lorenz
& Wackernagel, 1994; Dubnau, 1999; Chen & Dubnau,
2004). The composition and dynamics of the DNA-
uptake complexes (Box 1) have also been the focus of
several excellent recent reviews (Averhoff & Friedrich,
2003; Allemand & Maier, 2009; Claverys et al., 2009; Bur-
ton & Dubnau, 2010; Allemand et al., 2012). However,
less is known about the initiation of competence, particu-
larly in Gram-negative bacteria. It has become clear in
recent years that there are major differences in the regula-
tory network involved in competence induction in
Gram-negative bacteria. Thus, the present review aims to
provide an overview of how environmental signals drive
natural competence and transformation in these organ-
isms. Differences and similarities in the conditions that
induce competence in Gram-positive bacteria (a topic
Box 1 . A hypothetical mode for DNA update in V. cholerae
FEMS Microbiol Rev 37 (2013) 336–363
337
that has been recently reviewed by Claverys et al., 2006)
are highlighted throughout the text.
Finally, although this review emphasizes on pathogens
(with or without a niche outside of humans), a plethora
of nonpathogenic bacteria are known to be naturally
transformable. The rationale behind our focus on patho-
gens is based on two factors: (1) Research on HGT has
long been influenced by medically relevant questions,
such as how virulence factors are acquired by pathogens
and how antibiotic-resistance genes spread among micro-
organisms. In fact, the discovery of natural transforma-
tion in bacteria was based on the seminal work by
Frederick Griffith. Griffith was interested in understand-
ing the difference between virulent and nonvirulent
strains of Streptococcus pneumoniae and how such strains
can interconvert (Griffith, 1928). A more recent example
would be the opportunistic pathogen Porphyromonas
gingivalis. Porphyromonas gingivalis belongs to the phylum
bacteroidetes and contributes to periodontal disease upon
stable colonization of the human oral cavity. Recent data
demonstrated that natural competence and transforma-
tion is the major driving force behind DNA exchange in
this organism (Tribble et al., 2012). The resulting genetic
variability is most likely the reason for P. gingivalis’
survival in the human host and its evasion from human
immune defences (Tribble et al., 2012). (2) The second
reason that pathogenic bacteria are sometimes considered
more appropriate for the study of HGT is the somewhat
biased sequencing effort. Whereas often only one or very
few whole-genome sequences of nonpathogenic bacteria
are available, a trend towards sequencing many different
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
338
isolates of human pathogens has recently emerged
(Tettelin et al., 2005; Croucher et al., 2011; Mutreja et al.,
2011). The use of information from multiple isolates
facilitates the identification of horizontally acquired
regions and the underlying regulatory circuits driving
HGT.
Caught in competence – constitutively
competent Gram-negative bacteria
Helicobacter pylori responds to DNA damage
Helicobacter pylori is an extremely successful human path-
ogen. This bacterium colonizes the gastric epithelia of
more than half of the world’s population (Suerbaum &
Michetti, 2002) and has probably done so since humans
migrated out of Africa roughly 60 000 years ago (Falush
et al., 2003; Linz et al., 2007). Helicobacter pylori is also
one of the most diverse bacterial species (Achtman et al.,
1999). This bacterium’s diversity has been studied for
almost two decades, and it is clear that every person
infected by H. pylori carries a unique strain (Langenberg
et al., 1986; Majewski & Goodwin, 1988; Achtman et al.,
1999). Extremely frequent recombination events are key
to this extraordinary genetic diversity (Suerbaum et al.,
1998; Kersulyte et al., 1999; Falush et al., 2001; Morelli
et al., 2010). This high recombination frequency and
frequent horizontal flow of genetic material in H. pylori is
directly linked with its ability to undergo natural trans-
formation, the topic of this review. However, in terms of
natural competence, H. pylori differs from other Gram-
negative bacteria in two important respects: the mecha-
nism of the DNA-uptake process and the constitutive
competence state of the bacterium.
Most naturally competent bacteria rely on type IV
pilus-related DNA-uptake machineries (Hobbs & Mattick,
1993 and for recent review see Chen & Dubnau, 2004;
Allemand & Maier, 2009; Burton & Dubnau, 2010; Box
1). In H. pylori, however, the components involved in the
translocation of transforming DNA resemble type IV
secretion systems (T4SS; Hofreuter et al., 1998, 2000;
Karnholz et al., 2006), such as the archetypical VirB/
VirD4 system of Agrobacterium tumefaciens (for recent
review see Alvarez-Martinez & Christie, 2009). Indeed,
many bacteria use similar systems to export both protein
and DNA substrates. In addition to this competence-
related T4SS, known as the comB system, many H. pylori
strains harbour a ‘classical’ T4SS encoded on a PAI. This
system is used to inject the virulence factor CagA into
gastric epithelial cells (Odenbreit et al., 2000). These
two T4SS are independent of each other. Moreover, the
Cag-PAI is fully dispensable in natural transformation
(Hofreuter et al., 2000; Israel et al., 2000). Interestingly
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
P. Seitz & M. Blokesch
enough, another naturally competent e-proteobacterium,
Campylobacter jejuni, does not seem to share this depen-
dency on a transformation-related T4SS; instead DNA
uptake and transformation is dependent on genes encod-
ing components of a putative type II secretion or type IV
pilus-like system (Wiesner et al., 2003). More information
on natural competence of C. jejuni has been recently
reviewed by Young et al. (2007). And although DNA
uptake in H. pylori is exceptional, recent studies indicate
that certain steps might be conserved in DNA transport.
Whereas the uptake of DNA into the periplasm by a
T4SS is specific to H. pylori, transport across the inner
membrane involves homologous proteins found in both
naturally competent Gram-negative and Gram-positive
bacteria (Stingl et al., 2010; Box 1).
DNA uptake is not the only ‘exception that proves the
rule’ in H. pylori. An even more striking difference to
most other naturally transformable bacteria occurs at the
regulatory level of competence. In contrast to those bacte-
ria, H. pylori does not limit its competence window in
response to environmental stimuli. Instead, this bacterium
is naturally competent during all growth phases (Israel
et al., 2000; Baltrus & Guillemin, 2006). Because transfor-
mation frequencies specifically peak within certain growth
phases during in vitro growth (Baltrus & Guillemin,
2006), a regulatory system must exist to at least fine-tune
DNA uptake. The mediators involved in this process
remain unknown. The only stimulus experimentally pro-
ven to increase natural transformation in H. pylori is
Fig. 1. The most prominent environmental cues involved in
competence induction or the fine-tuning of competence. These signals
include genotoxic stresses causing DNA damage, such as UV light or
certain antibiotics (in red); bacterial cell-cell communication systems
(e.g. quorum-sensing; in green); the starvation of preferred carbon
sources, leading to the accumulation of the intracellular secondary
messenger cAMP (carbon catabolite repression; in blue); and the
presence or absence of certain carbon sources (such as the GlcNAc
polymer chitin for Vibrio cholerae). Details are provided in the text.
FEMS Microbiol Rev 37 (2013) 336–363
Regulation of natural competence in Gram-negative bacteria
DNA damage (Dorer et al., 2010; Fig. 1). Indeed, Dorer
et al. (2010) showed that mutations in a DNA repair pro-
tein, as well as treatment with the DNA-damaging agent
ciprofloxacin, led to the upregulation of transcription of
many genes involved in DNA uptake and to increased
recombination. Consequently, transformation frequencies
increased four to fivefold. Among the upregulated genes
was a homologue of a T4 phage-derived lysozyme that
had previously been characterized for its lytic activity
(Marsich et al., 2002). Therefore, DNA damage might
lead not only to increased DNA uptake but also to the
enhanced lysis of neighbouring cells and ultimately to
high levels of natural transformation (for recent review
see Dorer et al., 2011). The induction of competence
upon DNA damage, coupled with fratricide, has also been
observed in the Gram-positive bacterium S. pneumoniae
(Guiral et al., 2005; Havarstein et al., 2006; Prudhomme
et al., 2006). Although the two bacterial species are only
distantly related, they share another characteristic:
S. pneumoniae and H. pylori both lack a bona fide SOS
response system. Competence induction in response to
DNA damage might thus represent a product of conver-
gent evolution in distantly related bacteria that lack an
SOS response system (Dorer et al., 2010). This same con-
dition holds for the Gram-negative pathogen Legionella
pneumophila. In this organism natural competence is also
induced upon antibiotics- and UV-induced DNA damage
(Charpentier et al., 2011).
Some naturally competent bacteria are biased towards
the uptake of genetic material derived from closely related
neighbours. Various strategies have evolved to this end
(Fig. 2). Helicobacter pylori is believed to be able to take
up any kind of DNA, independent of its source, but
recent results indicated that incoming DNA can be con-
trolled by restriction-modification (R-M) systems (Aras
et al., 2002; Humbert & Salama, 2008; Humbert et al.,
2011). R-M systems are based on restriction endonucleas-
es (REases), which cleave DNA at specific sequence
motifs, and corresponding DNA methyltransferases,
which inhibit the DNA cleavage of self DNA by methyla-
tion of the respective recognition sites. This mechanism
recognizes and degrades foreign genetic material (Fig. 2).
Indeed, R-M systems were traditionally associated with
protection against bacteriophages or conjugative plasmids
(for review see Kessler & Manta, 1990). It is not entirely
clear how the REases destroy unmethylated DNA to
reduce transformation frequencies in H. pylori (Humbert
et al., 2011) as current models suggest that only single-
stranded DNA enters the cytoplasm (as reviewed by Chen
& Dubnau, 2004; Allemand & Maier, 2009; Burton &
Dubnau, 2010; Allemand et al., 2012; Box 1). Thus, it
was concluded in a recent review: ‘Because restriction
enzymes prefer double-stranded DNA, they likely act
FEMS Microbiol Rev 37 (2013) 336–363
339
prior to unwinding and perhaps extracellularly’ (Dorer
et al., 2011). The idea that such foreign species-derived
DNA degradation could occur outside the cells might
very well be possible as DNA damage-induced lysis of
neighbouring cells would not only release these cells’
DNA but also their REases.
Nuclease-mediated degradation of extracellular DNA
has also been described for nontransformable variants of
C. jejuni (Gaasbeek et al., 2009, 2010). More specifically,
these authors showed that acquisition of prophage-
encoded endonucleases lead to an inhibition of natural
transformability of C. jejuni (Gaasbeek et al., 2009, 2010).
This is in excellent agreement with an earlier publication
on another naturally competent Gram-negative bacterium,
Vibrio cholerae: in this study Blokesch and Schoolnik dem-
onstrated that an extracellular nuclease Dns degrades
external and potentially transforming DNA thus inhibiting
transformation of this organism (Blokesch & Schoolnik,
2008). However, repression of nuclease production at high
cell density occurs via a quorum-sensing regulatory circuit
and thus still allows natural transformation to occur in
V. cholerae (see below; Blokesch & Schoolnik, 2008; Lo
Scrudato & Blokesch, 2012).
The competence-induced uptake and secretion
of DNA – the unique case of Neisseria
Neisseria spp. have been known to be naturally competent
for more than half a century (Alexander & Redman,
1953; Catlin & Cunningham, 1961). Recent research on
the natural competence and transformation of Neisseria
has primarily focused on the human pathogens Neisseria
meningitidis and Neisseria gonorrhoeae. These species are
very similar with respect to competence regulation and to
their DNA-uptake mechanisms (Koomey, 1998). Thus,
many findings derived from one organism can readily be
applied to the other species.
Like H. pylori, Neisseriae are competent for natural
transformation during all growth phases (Sparling, 1966).
Environmental stimuli or secreted factor responsible for
the induction or enhancement of competence and trans-
formation have not been identified to date, except for a
minor effect of growth temperature (Sparling, 1966). No
further studies have been conducted on growth tempera-
ture, most likely because it only reflects elevated metabolic
activity. Natural transformation peaked at 37 °C for
N. gonorrhoeae (Sparling, 1966), which corresponds to the
constant ambient temperature of its niche, which is the
mucosal tissue of the human urogenital tract. As for
H. pylori, no environmental reservoir is known for
N. gonorrhoeae. Within the human host, the immune sys-
tem puts Neisseriae under strong selective pressure. Thus,
the extensive horizontal flow of genetic material mediated
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
340 P. Seitz & M. Blokesch

Fig. 2. Different strategies have evolved to enhance the probability of integrating species-specific DNA. Certain Gram-negative bacteria, such as
Haemophilus influenzae and Neisseria gonorrhoeae, are notably fastidious about the kind of DNA that they absorb. To sort species-specific from
nonspecies-specific DNA, these organisms have evolved specific DNA uptake sequences (DUS) that are overrepresented in their genomes and
recognized by naturally competent members of the same species through as-yet-unknown receptor proteins. These receptors are most likely
surface-exposed and initiate the ingestion of species-specific DNA into the cytoplasm of the bacterium (I). An alternative strategy to increase the
chances of taking up species-specific DNA is to link the regulation of natural competence and transformation to species-specific autoinducer
synthesis and recognition (II). The involvement of a ‘competence pheromone’ is a common strategy of Gram-positive bacteria and has also been
recently suggested for the Gram-negative bacterium Vibrio cholerae. Other bacteria, such as Helicobacter pylori, have evolved specific restriction-
modification (R-M) systems, which assure that species-nonspecific DNA is degraded and thus cannot recombine with the chromosome. The exact
mechanism is so far unknown for this mode of restricting nonrelated DNA. One possible mechanism could be that the restriction enzymes
(REases) are released from lysed bacteria (together with the DNA) and degrade species-nonspecific dsDNA extracellularly. Finally, as in the Gram-
positive bacterium Streptococcus pneumoniae, certain naturally competent Gram-negative bacteria initiate the donation of DNA from their
siblings. This process may be mediated either by autolysis, by the intentional killing of a subpopulation (e.g. the lysis of neighbouring cells) or by
the active secretion of DNA through a T4SS, as in a subset of Neisseria gonorrhoeae strains.

by natural transformation might provide an important
means of adaptation. Indeed, natural transformation
might contribute significantly to the high genomic diver-
sity of Neisseria (Viscidi & Demma, 2003; Hanage et al.,
2005; Maiden, 2008). Researchers provided evidence to
support extensive interspecies recombination and inter-
generic transfer of DNA in the late 1990s (Feil et al., 1996,
1999; Kroll et al., 1998). Only recently have genotyping
techniques made it possible to determine the extent of
diversity in Neisseria (for recent review see Maiden, 2008).
As described earlier, H. pylori actively secretes a lyso-
zyme-like protein to increase the concentration of naked
DNA in the environment (Dorer et al., 2010). Neisseria
gonorrhoeae facilitates the release of DNA by two different
mechanisms. In this organism, DNA is either donated to
siblings via autolysis or actively exported via a T4SS
(Hamilton et al., 2005; Hamilton & Dillard, 2006; Fig. 2).
Although the T4SS involved in DNA secretion is found
only in a subset of N. gonorrhoeae strains (Dillard &
Seifert, 2001) and does not contribute to DNA secretion
in N. meningitidis (Woodhams et al., 2012), all investi-
gated Neisseria strains can undergo autolysis. Little is
known about the regulation of either of these processes
or whether they are driven by environmental stimuli;
however, both mechanisms promote genetic transfer
within bacterial communities and might be beneficial for
genome maintenance (Fig. 2).
Whereas H. pylori limits interspecies transformation via
intracellular R-M systems, Neisseria spp. discriminate
between self and foreign DNA at the level of the

ª 2012 Federation of European Microbiological Societies FEMS Microbiol Rev 37 (2013) 336–363
Published by Blackwell Publishing Ltd. All rights reserved
Regulation of natural competence in Gram-negative bacteria
DNA-uptake machinery. In fact, the finding that N. menin-
gitidis and N. gonorrhoeae preferentially take up DNA from
closely related species is one of the strongest arguments
supporting the ‘DNA for repair’ hypothesis. More precisely,
for Neisseria spp., as well as for Haemophilus influenzae,
species-specific uptake of genetic material is dependent on
so-called DNA-uptake sequences (DUS; Goodman & Scoc-
ca, 1988; Elkins et al., 1991), which are conserved short
DNA sequence motifs (Danner et al., 1980; Fitzmaurice
et al., 1984; Elkins et al., 1991; Aas et al., 2002). DUS
motifs may serve as a binding target for a still unidentified
and hypothetical receptor protein localized on the outside
of DUS-containing Gram-negative bacteria (Fig. 2).
Furthermore, DUS motifs are highly overrepresented in the
bacterial genomes. Early estimates of roughly one DUS
motif per kb of DNA (Goodman & Scocca, 1991) were vali-
dated when the first two Neisseria genomes were made
publicly available [GenBank numbers AE002098 (Tettelin
et al., 2000) and AE004969]. Interestingly, further analysis
revealed a strong bias in DUS frequencies towards genome
maintenance genes (Davidsen et al., 2004). DUS motifs
provide yet another mechanism to support species-specificity
in transformation (Fig. 2) and may play a role in ‘recom-
binational repair’ (Michod et al., 2008).
Recent findings, however, call into question both the
nature and function of DUS (Ambur et al., 2007; Duffin
& Seifert, 2010). DNA uptake can occur in the absence of
DUS, and the effect of DUS on transformation frequen-
cies varies significantly among strains of N. gonorrhoeae
(Duffin & Seifert, 2010). Because the effect of DUS on
natural transformation efficiency was not proportional to
its effect on DNA uptake, the authors proposed an addi-
tional role for DUS downstream of the DNA-uptake pro-
cess (Duffin & Seifert, 2010). In addition, Duffin &
Seifert (2012) demonstrated that DUS sequences can
enhance the somewhat less efficient transformation by
single-stranded DNA and that this phenomenon shows
strand preference. The identification of a human gene
fragment in the genome of N. gonorrhoeae provides fur-
ther evidence for DUS-independent DNA uptake (Ander-
son & Seifert, 2011). This finding highlights the
evolutionary potential of N. gonorrhoeae and the impor-
tance of HGT in host-pathogen interactions.
Competence induction based on
starvation signals
Haemophilus influenzae requires a nutrient
downshift to induce competence
Haemophilus influenzae is yet another Gram-negative bac-
terium in which natural competence and transformation
has been studied for many years. As in the Neisseriaceae
FEMS Microbiol Rev 37 (2013) 336–363
341
and H. pylori, the primary ecological niche of this organism
is the human body, specifically, the upper respiratory tract.
Haemophilus influenzae also causes ear infections and men-
ingitis in children and respiratory disease in the elderly, as
well as in patients with cystic fibrosis and AIDS.
The environmental signal that triggers the onset of natu-
ral competence in H. influenzae is not known; however,
because ‘the human respiratory tract is a hostile environ-
ment for bacteria’, and ‘nutrients and energy sources are
limited’ (Murphy & Brauer, 2011), it is not surprising that
a link between starvation for preferred carbon sources and
competence induction has been discovered in vitro for this
organism. Indeed, in the laboratory, H. influenzae becomes
naturally transformable when the culture reaches stationary
phase (Redfield, 1991) or when the cells are transferred
from rich to starvation medium (Herriott et al., 1970). The
addition of the secondary messenger cyclic adenosine
monophosphate (cAMP) to exponentially growing bacteria
also induces competence (Wise et al., 1973). As a conse-
quence it was shown that cAMP receptor protein (CRP)
and adenylate cyclase (CyaA) are essential for the develop-
ment of natural transformation in this organism (Chandler,
1992; Dorocicz et al., 1993).
Recent work by Rosemary Redfield’s group and others
have provided further insight into the induction of
competence in H. influenzae. These researchers identified
the gene sxy (Redfield, 1991; homologue in other compe-
tent bacteria is tfoX), which is essential for natural
competence. Furthermore, Sxy induced constitutive com-
petence upon overexpression from a multi-copy plasmid
(Williams et al., 1994). The expression of sxy/tfoX as an
early competence gene was increased upon the addition
of cAMP, a phenomenon consistent with the increase in
transformability (Zulty & Barcak, 1995). Sxy may act as a
positive regulator of late competence-specific genes;
however, the mechanism of this positive regulation has
long been debated because Sxy lacks recognizable DNA-
binding motifs (Macfadyen, 2000).
The competence regulon in H. influenzae was identified
in 2005 using microarray expression profiling (Redfield
et al., 2005). The regulon consisted of 25 genes in 13
transcriptional units, each containing a characteristic
promoter-associated 22-bp element, the competence regu-
latory element (CRE; Karudapuram & Barcak, 1997; Mac-
fadyen, 2000; Redfield et al., 2005). In these studies,
competence-induced transcription was again strongly
dependent on cAMP and the competence regulator Sxy.
Based on this information, it was proposed that Sxy
might act as an accessory factor, directing CRP to CRE
sites and/or stabilizing the transcription-initiation com-
plex (Macfadyen, 2000; Redfield et al., 2005). The CRE
elements were later renamed as CRP-S sites, describing a
binding site for CRP; these sites, unlike canonical
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
342
CRP-binding sites (CPR-N), are dependent on Sxy (Cam-
eron & Redfield, 2006, 2008). What is the exact nature of
the link between cAMP/CRP, Sxy and competence induc-
tion? Although there is no definitive answer at present,
the current model (based on work by the Redfield labora-
tory) is as follows: (1) CRP and cAMP are required to
strongly induce the transcription of sxy (Cameron et al.,
2008) via the binding of the cAMP/CRP complex to a
CRP-binding site in the sxy promoter region (Zulty &
Barcak, 1995; Cameron et al., 2008). (2) A stem loop
structure is formed within the 5′ UTR of the sxy mRNA,
which is involved in controlling both the amount and
translation efficiency of the sxy mRNA (Cameron et al.,
2008). The authors of this study suggested that the latter
was caused by the sequestration of the SD site by the
mRNA stem loop structure, which limited the binding of
ribosomes. Thus, the increase in Sxy protein following
the transfer of the H. influenzae cells to starvation medium
might occur because the secondary structure of the sxy
mRNA has ‘the potential to play a sensory role’ (Cameron
et al., 2008). Such a sensory function might be based on
the speed of transcription; ribosomes may bind before the
5′ UTR stem loop is properly folded, ‘thus making the
progress of RNA polymerase a potential transducer of
nutritional signals’ (Cameron et al., 2008). Alternatively,
the 5′ UTR of the sxy mRNA might be directly involved in
regulation, perhaps by a riboswitch-like mechanism. This
hypothesis is supported by the recently described ribo-
switch, which was identified in association with a sxy
homologue in V. cholerae (tfoXGEMM, Sudarsan et al.,
2008; details below). In conclusion, cAMP and CRP, as the
major players in carbon-catabolite repression (CCR), are
important in the natural transformation of H. influenzae
because they are involved in the transcription of both
early- (sxy) and late- (e.g. comA) competence genes. This
feature couples the nutritional state of the bacterium to
the induction of competence.
Apart from its cAMP-dependent competence induction,
H. influenzae has a preference for species-specific DNA.
The species-specificity of the DNA is discriminated at the
level of the DNA-uptake process through the recognition
of specific DUS sequences, as described above for Neisseria
species (Danner et al., 1980; Fitzmaurice et al., 1984;
Smith et al., 1999).
Environmental cues involved in
competence induction in bacteria
commonly found outside a host
Natural transformation in Pseudomonadaceae
In contrast to the highly adapted human pathogens
discussed earlier, the pseudomonads are an extremely
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
P. Seitz & M. Blokesch
versatile group of environmental bacteria that live pre-
dominantly in soil, sediment, and water, occurring less
commonly as opportunistic human pathogens (for recent
review on Pseudomonas taxonomy and isolation of type
strains see Peix et al., 2009). Many Pseudomonas spp.
undergo natural transformation, including P. mendocina,
P. alcaligenes, P. pseudoalcaligenes, P. fluorescencs and
P. stutzeri (Carlson et al., 1983; Demaneche et al., 2001);
to date, there is no evidence for natural competence of
P. aeruginosa. Thus, most of the research on natural com-
petence has focused on the ubiquitous soil bacterium
P. stutzeri. This bacterium has a notably complex metabo-
lism, which allows it not only to grow on a wide variety
of carbon sources but also to degrade environmental
pollutants and to fix nitrogen (as recently reviewed by
Lalucat et al., 2006).
Most of the research on the regulation and mechanism
of natural transformation in P. stutzeri was performed in
Wilfried Wackernagel’s laboratory, and a seminal review,
‘Bacterial Gene Transfer by Natural Genetic Transforma-
tion in the Environment’, was published by Lorenz &
Wackernagel (1994). This group showed, for example,
that as with most other Gram-negative bacteria, the trans-
formability of P. stutzeri correlates with the expression of
type IV pilus structural or assembly genes (Graupner
et al., 2000), indicating that DNA uptake in P. stutzeri
might be similar to the mechanism proposed for other
Gram-negative bacteria (as reviewed by Chen & Dubnau,
2004 and described in Box 1). There are, however, several
interesting differences. For example and in contrast to
H. influenzae, transformation by single-stranded DNA
(ssDNA) is efficient in P. stutzeri, reaching up to 5%
compared with double-stranded DNA (dsDNA; Meier
et al., 2002). Notably, transformation with ssDNA was
dependent on the same components identified for
dsDNA, indicating that both substrates follow identical
routes into the cell (Meier et al., 2002). Furthermore, and
again in contrast to N. gonorrhoeae or H. influenzae,
DNA uptake was not dependent on the presence of
DNA-uptake motifs. However, early competition experi-
ments with either homologous or heterologous DNA sug-
gested that P. stutzeri might still discriminate between self
and foreign DNA at an early step of transformation
through an unknown mechanism (Carlson et al., 1983;
Lorenz & Wackernagel, 1990). It was also shown that
homology-facilitated illegitimate recombination is much
less efficient in P. stutzeri than in other naturally trans-
formable bacteria, such as S. pneumoniae or Acinetobacter
(Meier & Wackernagel, 2003). Finally, DNA restriction
has also been proposed as a barrier to natural transforma-
tion in P. stutzeri (Berndt et al., 2003). All of these fac-
tors result in a limited exchange of heterologous DNA
(Fig. 2), which has important implications for the spread
FEMS Microbiol Rev 37 (2013) 336–363
Regulation of natural competence in Gram-negative bacteria
of alleles in populations of different Pseudomonas species
and strains.
Transformation levels not only varied significantly
among strains of P. stutzeri (Lorenz & Sikorski, 2000;
Sikorski et al., 2002), but they also varied across growth
phases (Carlson et al., 1983; Lorenz & Wackernagel,
1990). Specifically, transformation frequencies for growth
in rich media peaked at the onset of the stationary phase,
when nutrients start to become limited (Lorenz & Wack-
ernagel, 1990). Thus, to mimic the natural habitat of this
bacterium and to study nutrient situation and its effect
on competence induction in P. stutzeri, Lorenz & Wack-
ernagel (1991) tested growth and transformation on soil-
extract-containing medium. These authors demonstrated
that nutrients in those extracts were limited, allowing for
the growth of only one or two generations. By supple-
menting the soil media with different combinations of
carbon sources, phosphate and ammonium, these
researchers determined the effect of each nutrient on
growth and transformation. Limiting N, C or P stimu-
lated transformation; this effect was most pronounced in
soil extracts spiked with pyruvate and phosphate but
lacking additional ammonium (Lorenz & Wackernagel,
1991). The underlying regulatory mechanism remains
unknown.
Soil microcosm experiments demonstrated that condi-
tions supporting the efficient transformation of P. stutzeri
are readily encountered in the environment (Sikorski
et al., 1998). The same phenomenon was observed in
P. fluorescens. Interestingly, transformation frequencies
were significantly higher in nonsterile soil samples than in
gamma-irradiated microcosms, possibly ‘due to the pres-
ence of an organic compound in soil that is in part
destroyed by soil sterilization’ (Demaneche et al., 2001).
Even more strikingly, none of the in vitro conditions tested
induced competence in P. fluorescens (Demaneche et al.,
2001), illustrating once more that most of the environ-
mental signals that foster natural competence and transfor-
mation have not yet been identified. Thus, many bacteria
used in the laboratory might have the potential for compe-
tence, although the trigger has not yet been elucidated.
Competence of Acinetobacter – is a nutrient
boost required?
Another well-studied, naturally competent, Gram-negative
bacterium is Acinetobacter. Acinetobacter spp. belong to
the c-proteobacteria, order Pseudomonadales. Acinetobac-
ter strains are abundant in soil and water (Warskow &
Juni, 1972), where they have been determined to make
up at least 0.001% of the total culturable aerobic bacterial
population (Baumann, 1968). Acinetobacter strains, how-
ever, are also frequently isolated from patients and have
FEMS Microbiol Rev 37 (2013) 336–363
343
recently ‘emerged from an organism of questionable path-
ogenicity to an infectious agent of importance to hospi-
tals worldwide’ (Munoz-Price & Weinstein, 2008). More
importantly, this bacterium readily acquires resistance to
antibiotics, another trait linked with HGT.
With respect to natural competence and transforma-
tion, most studies have been performed on Acinetobacter
baylyi BD413 (Carr et al., 2003; representative strains of
this species were earlier named Acinetobacter calcoaceticus
BD4, A. calcoaceticus BD413, A. calcoaceticus ADP1, and
Acinetobacter spp. strain ADP1; Young et al., 2005). The
transformation efficiency of A. baylyi BD413 (formerly
named A. calcoaceticus BD413) was shown to be notably
high, ranging from 0.1% to 0.7% of all cells in an auto-
trophy/prototrophy transformation assay (Juni & Janik,
1969). Furthermore, this bacterium is not fastidious about
the source of DNA; no species-specificity was observed
with respect to DNA uptake (Lorenz et al., 1992; Palmen
et al., 1993; Palmen & Hellingwerf, 1997).
Acinetobacter baylyi BD413 is always transformable,
though with large variations in efficiency throughout the
growth cycle (Cruze et al., 1979). Reports on the time of
peak transformability are conflicting. The initial study
suggested that ‘a peak of competency is found as the cul-
ture enters the stationary phase’ (Juni, 1978) and that
transformation was less efficient in the exponential phase
(Juni & Janik, 1969). This finding was confirmed by
others who found that the appearance of transformants
was maximal at the end of the exponential growth phase
for the soil bacteria Bacillus subtilis and A. baylyi BD413
(Lorenz et al., 1991). These authors also demonstrated
that there was no active DNA release by competent
Acinetobacter cells. Other groups provided evidence that
the highest transformation frequencies occurred early in
the exponential phase, which was consistent with rapid
growth (Cruze et al., 1979). The latter finding was also
confirmed by several groups (Palmen et al., 1993, 1994;
Porstendörfer et al., 2000), who showed that natural
transformation was maximally induced after the dilution
of an overnight culture into fresh medium or ‘after an
increase in nutrient availability, preceded by cessation of
growth’ (Palmen et al., 1994). In an earlier review, these
contradictory findings were resolved with the conclusion
that competence ‘reaches its maximum during early to
late log phase in A. calcoaceticus’ (Lorenz & Wackernagel,
1994). Based on the results of a study on early compe-
tence induction, Palmen et al. (1994) concluded that ‘the
biological function of natural transformation in A. calco-
aceticus is not to provide the cell with nutrients’ and that
‘the regulation of competence development as a function
of the stringency of carbon, nitrogen, and phosphate limi-
tation was just opposite to what one would expect if the
‘nutrient supply hypothesis’ would apply’.
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
344
Beate Averhoff and collaborators demonstrated that the
transcription of at least two of the competence genes,
comP and comB, did not correlate with the transforma-
tion pattern (Herzberg et al., 2000; Porstendörfer et al.,
2000). Instead comP expression peaked in the late station-
ary phase, a finding that was also confirmed at the
protein level (Porstendörfer et al., 2000). ComB transcrip-
tional activation also turned out as growth phase depen-
dent: expression slightly peaked after dilution of the
bacteria in fresh medium followed by an immediate
decreased of comB transcription. Within mid-exponential
phase expression of comB resumed up to a constant and
high level of comB transcription in late stationary phase
(Herzberg et al., 2000). Based on these data Herzberg
et al. (2000) concluded that the DNA machinery is
already present in late stationary phase; diluting the cells
into fresh medium then provides the required energy to
allow efficient DNA uptake and consequently transforma-
tion. This conclusion is consistent with other reports on
competence induction and transformation (Palmen et al.,
1994; Porstendörfer et al., 2000) and the finding that
transformation of A. baylyi BD413 is not inhibited by the
protein synthesis inhibitor chloramphenicol (Palmen &
Hellingwerf, 1997). The group of B. Averhoff has also
conducted work on the natural transformation of an unu-
sual Gram-negative bacterium, the thermophile Thermus
thermophilus. This bacterium is a record-holder with
respect to natural transformability (Koyama et al., 1986)
and a detailed review on T. thermophilus transformation
was recently published (Averhoff, 2009). The impact and
basis of natural competence and transformation for sur-
vival of Thermus in and adaptation to hot environments
has recently been analysed (Omelchenko et al., 2005;
Averhoff & Müller, 2010).
Based on these in vitro studies described earlier, it is dif-
ficult to determine the environmental signals that could
trigger competence in Acinetobacter. Nielsen et al. (1997)
performed soil microcosm studies (‘in situ’) and observed
the induction of competence by nutrient upshifts, as well as
a transformation-enhancing effect of phosphate. These
authors concluded that ‘poorly transformable A. calcoaceti-
cus cells can be induced to undergo natural transformation
with chromosomal DNA in soil’ and that ‘their level of
competence is influenced by their metabolic state’ (Nielsen
et al., 1997). Such transformability was maintained for sev-
eral hours after induction. A recent in situ study confirmed
that A. baylyi entered competence soon after the inocula-
tion of leaves, as evaluated by the re-isolation of bacteria
from the leaves and in vitro provision of transforming
DNA (Pontiroli et al., 2009). Furthermore, these authors
directly visualized transformants in situ using gfp reporter
fusion and transforming DNA derived from transgenic
plants. They concluded that the phytosphere might consti-
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
P. Seitz & M. Blokesch
tute a hotspot for host/pathogen HGT (Pontiroli et al.,
2009). In summary, the induction of transformability of
Acinetobacter by an upshift in nutrients may behave in a
manner contrary to what was observed for many other nat-
urally transformable bacteria in which competence and
transformation are often linked with starvation signals.
Natural transformation in aquatic Vibrio
spp., with a focus on V. cholerae
Vibrio cholerae is a human pathogen and the causative
agent of cholera; however, this bacterium’s main niches
are aquatic environments, such as rivers, estuaries, and
coastal regions. Within these environments, V. cholerae is
often found in association with chitinous zooplankton or
their molts (Colwell, 1996; Lipp et al., 2002; Pruzzo et al.,
2008). This association results from the ability of
V. cholerae and other aquatic Vibrio species to degrade
the chitinous exoskeletons of zooplankton and use them
as a carbon and nitrogen source (for review see Keyhani
& Roseman, 1999). Because the abundance of plankton
changes significantly between plankton blooms, Vibrio
species regulate gene expression in accordance with nutri-
ent availability. Whether such environmental changes also
influence competence induction is discussed below.
HGT within the species V. cholerae
Comparative genomic hybridization experiments per-
formed by John Mekalanos’s group revealed a high degree
of conservation among V. cholerae strains (Dziejman
et al., 2002). This study was based on microarray hybrid-
ization experiments and compared the strains’ genomes
to the first sequenced strain of V. cholerae, which was a
pandemic O1 El Tor strain (N16961; Heidelberg et al.,
2000). Only presence/absence data about genes and ope-
rons were collected; no information about novel genetic
material was retrievable with this experimental approach.
The authors of this study identified a limited number of
genomic islands that were specific to the seventh pan-
demic strain and absent in earlier isolates, such as the
representative strains of the classical biotype (the causa-
tive agent of the sixth and, likely earlier, cholera pandem-
ics; Dziejman et al., 2002). Furthermore, this initial study
focused mainly on patient isolates and only investigated
representatives of the serogroups O1 and O139. A follow-
up study by Dziejman et al. broadened this analysis to
include pathogenic non-O1 and non-O139 V. cholerae
isolates. These strains were ‘quite divergent’ from the
V. cholerae representatives belonging to the O1 and O139
serogroups (Dziejman et al., 2005).
Technologically advanced, large-scale WGS studies
made it possible to generate a clearer picture of SNPs and
FEMS Microbiol Rev 37 (2013) 336–363
Regulation of natural competence in Gram-negative bacteria
horizontally acquired genetic material; these data enabled
the construction of phylogenetic trees from which to
derive information about the evolutionary relationships
among strains. For example, recent WGS studies on
V. cholerae suggested that several transcontinental trans-
mission events have occurred (Mutreja et al., 2011), sig-
nificantly contributing to the understanding of the recent
emergence of this pathogen in an enormous cholera out-
break in Haiti (Chin et al., 2010; Hendriksen et al., 2011;
Mutreja et al., 2011). Rita Colwell and collaborators also
used WGS to understand the genetic diversity of 23 dif-
ferent V. cholerae strains isolated over the past 98 years.
These authors identified 73 genomic islands (containing
five or more ORFs) in their study and thus provided
strong evidence for ‘extensive lateral gene transfer in
V. cholerae’ (Chun et al., 2009).
Natural competence has only recently been described
in V. cholerae (Meibom et al., 2005). The authors demon-
strated that during growth on chitinous surfaces – a com-
mon environmental niche for this bacterium (Lipp et al.,
2002), V. cholerae can take up naked DNA and recom-
bine it into its genome (Meibom et al., 2005). As a mech-
anism of HGT, natural competence enables V. cholerae to
acquire new genes, including those that specify patho-
genic features. Indeed, Blokesch and Schoolnik demon-
strated experimentally that the O1-to-O139 serogroup
conversion could occur in a single step through a natural
transformation-mediated exchange of the large O-antigen
cluster (> 32 kb exchanged for > 42 kb; Blokesch &
Schoolnik, 2007). The transformants displayed all pheno-
types associated with the acquired gene cluster, including
the changed O-antigen chain of the LPS and the capsular
material surrounding the cells (Blokesch & Schoolnik,
2007). A stretch of homologous DNA located within the
O-antigen cluster of many different V. cholerae strains
(the IS1358 element) enabled transformants with only
partial exchange of the cluster (only the upstream or
downstream region of IS1358) to be obtained (Blokesch
& Schoolnik, 2007). In this way, natural transformation
may contribute to a ‘reshuffling’ of O-antigen genes
and thus to the creation of new serogroups of V. cholerae.
That serogroup conversion of V. cholerae occurs fre-
quently in nature was also supported by comparative
genomics (Chun et al., 2009). A study notably similar to
that conducted by Blokesch & Schoolnik (2007) was per-
formed 4 years later, and the authors demonstrated a
carbotype conversion of Vibrio vulnificus based on chitin-
induced transformation (Neiman et al., 2011). In this
case, the transformation event led to the exchange of a
capsular polysaccharide locus in this organism (Neiman
et al., 2011). Udden et al. (2008) used a similar experi-
mental approach to show that chitin-induced natural
competence of V. cholerae may be responsible for the
FEMS Microbiol Rev 37 (2013) 336–363
345
spread of cholera-toxin prophages, most notably from
non-O1/non-O139 strains. These authors provided evi-
dence that strains that were unable to produce infectious
cholera-toxin phage particles could still undergo horizontal
transfer of the prophage using natural transformation
(Udden et al., 2008). Apart from these virulence-associated
gene clusters, the transfer of genes involved in sugar
metabolism via chitin-induced natural transformation has
also been demonstrated for V. cholerae isolates from the
Californian coast (Miller et al., 2007). Thus, natural com-
petence and transformation of V. cholerae provide a
newly recognized mechanism by which this organism can
effectively acquire new genes, thereby adapting to chang-
ing environmental conditions. Ways in which the natural
competence programme is regulated in V. cholerae are
summarized below.
Competence induction in some bacteria
involves more than just one signal – the three
interlinked regulatory pathways of V. cholerae
Sensing the carbon source chitin and colonizing
chitinous surfaces
The resolution of the genome sequence of V. cholerae
(Heidelberg et al., 2000) provided the starting point for
the investigation of regulatory circuits using microarray
expression profiling. This approach enabled the identifica-
tion of genes that were specifically activated during
V. cholerae’s association with chitinous crab-shell frag-
ments or purified soluble chitin oligosaccharides, the
so-called chitin-utilization programme (Meibom et al.,
2004). Genes that specifically responded to chitin oligo-
saccharides included those genes predicted to encode
components for the biogenesis and function of a type IV
pilus (Fullner & Mekalanos, 1999; Meibom et al., 2004),
although this pilus has not yet been visualized. Compo-
nents of type IV pilus complexes in naturally competent
bacteria are hypothesized to participate in the transport
of DNA through the outer membrane and the peptidogly-
can (reviewed by Chen & Dubnau, 2004; Box 1), but ‘it
is currently unclear how the components interact to gen-
erate movement of the incoming DNA through the
cell envelope’ (Allemand & Maier, 2009). The chitin-
dependent induction of type IV pilus-encoding genes
cohered with natural competence and transformation,
demonstrating for the first time that V. cholerae is a natu-
rally transformable bacterium (Meibom et al., 2005).
But how is the chitinous surface sensed, and how is the
signal transmitted to induce gene-specific expression? In
this context, the induction of genes involved in chitin
degradation is strictly dependent on a unique two-
component signalling system composed of a chitin hybrid
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
346 P. Seitz & M. Blokesch

Fig. 3. A complex, threefold regulatory network is involved in the natural transformation of Vibrio cholerae. As shown in Fig. 1, the induction of
natural competence and transformation in V. cholerae is linked with three of the common environmental cues: (1) the chitinous surface as the
sole carbon source for this bacterium and other Vibrio spp. (brown pathway); (2) CCR and the induction of competence after starvation for
preferred PTS-transported sugars (blue pathways); and (3) cell-to-cell communication, such as that exerted by the quorum-sensing systems (green
pathway). Left – the chitin pathway. Extracellular chitinase (e.g. ChiA-1 and ChiA-2 directly secreted by V. cholerae or chitinases from other
chitinolytic aquatic bacteria, which are abundant in this environment) degrades the insoluble N-acetylglucosamine (GlcNAc) polymer chitin and
releases diacetylchitobiose units as the major product (together with some triacetylchitotriose). These chitin oligomers are recognized by a sensory
histidine kinase, ChiS, and a signal transduction pathway subsequently leads to the induction of tfoX and tfoR expression. tfoR encodes for a
small regulatory RNA, which enhances the translation of the tfoX mRNA by freeing the Shine-Dalgarno (SD) site, as well as the start codon
(AUG) from a secondary stem loop structure formed by the 5′ UTR of the tfoX mRNA. TfoX, as the major regulator of transformation, is involved
in the induction of competence genes. This induction most likely involves pairing with the cAMP-CRP complex (grey double arrow), as TfoX itself
does not contain any obvious DNA-binding domain. Middle –CCR. In the absence of preferred PTS-transported sugars (such as glucose or the
chitin monomer GlcNAc), the PTS systems are not saturated, and the phosphoryl group is kept by enzyme II (EII). EII–P is a direct activator of the
cAMP-producing enzyme CyaA. Accumulating cAMP, together with its binding protein (CRP), leads to the induction of the competence genes.
This complex might interact with TfoX to allow for the binding of cAMP-CRP to competence-specific CRP-S sites. Right – high cell density, as
measured by QS, is required for competence induction and transformation. V. cholerae produces two main autoinducers: CAI-1 and autoinducer
2 (AI-2). At high cell density (HCD), the binding of the autoinducer to the respective receptors initiates a signalling cascade (simplified), which
conclusively results in the production of the major regulator of QS, HapR. HapR acts as repressor of the extracellular nuclease gene dns and as a
positive regulator of the competence genes comEA and comEC. The encoded proteins of all three genes are expected to be in direct contact
with transforming DNA; these proteins either induce the degradation of external DNA at low cell density (Dns) or are part of the DNA-uptake
machinery (ComEA and ComEC; see Box 1). Thus, QS fine-tunes the fate of the surrounding DNA. References on which the scheme is based on
are mentioned in the text.

sensor/kinase ChiS (Li & Roseman, 2004) and a thus far
unidentified response regulator tentatively named ChiR
(Fig. 3). Li & Roseman (2004) speculated that ChiS works
in conjunction with an N,N′-diacetylchitobiose ABC-type
transporter and that ChiS is activated upon the transport
of the chitin degradation product (GlcNAc)2. The con-
nection between ChiS and the expression of the type IV
pilus-encoding genes became obvious when the latter
genes were not upregulated in a V. cholerae chiS mutant
during growth on chitin (Meibom et al., 2004). Notewor-
thy, a chiS mutant is not naturally transformable under
chitin-dependent competence-inducing conditions (M. Blok-
esch, unpublished data).
The V. cholerae chitin utilization programme also indi-
cated that a gene encoding a protein with similarity to
the transformation regulators of other naturally compe-
tent bacteria (TfoX/Sxy) was induced during growth on
chitin (Meibom et al., 2004). TfoX was later shown to be
essential for natural transformation of V. cholerae (Mei-
bom et al., 2005). There is substantial evidence that TfoX
is downstream of ChiS in the chitin-dependent signalling
cascade (Fig. 3). Expression of tfoX in trans or in cis

ª 2012 Federation of European Microbiological Societies FEMS Microbiol Rev 37 (2013) 336–363
Published by Blackwell Publishing Ltd. All rights reserved
Regulation of natural competence in Gram-negative bacteria
(Meibom et al., 2005; Lo Scrudato & Blokesch, 2012)
relieved the requirement for chitin to induce the pilus-
specific genes and other genes involved in DNA uptake
(Box 1), henceforth referred to as competence genes. Fur-
thermore, the transformation-negative phenotype of a
chiS mutant can be overcome by chitin-independent and
low-level expression of tfoX (Lo Scrudato & Blokesch,
2012), which restores natural transformability in this
strain (M. Blokesch, unpublished data).
Haruo Watanabe’s group demonstrated another link
between chitin-dependent transformation and TfoX pro-
duction in V. cholerae. These researchers first confirmed
the transcriptional data of Meibom et al. (2004): in con-
trast to (GlcNAc)n oligomers (n  2), the chitin mono-
mer GlcNAc is not sufficient to induce tfoX expression
(Yamamoto et al., 2010). In addition, Watanabe et al.
discovered a translational component of chitin-induced
TfoX production (Yamamoto et al., 2010). The latter
finding was supported by the discovery of a chitin-
induced small RNA (named TfoR), which positively con-
tributes to the translation of the tfoX mRNA (Yamamoto
et al., 2011; Fig. 3). These authors showed in vitro that
the 102-nucleotide sRNA TfoR activates the translation of
the tfoX mRNA, most likely by freeing the SD motif and
the start codon from a secondary structure formed by the
5′ UTR of the tfoX mRNA in the absence of TfoR. This
mechanism mimics the analogous mechanism in H. influ-
enzae (Cameron et al., 2008): a 5′ stem loop structure
formed within the tfoX/sxy mRNA is significant in initiat-
ing translation, but the release of this stem loop by base
pairing with the sRNA TfoR seems to be unique to
V. cholerae and, possibly, other Vibrio spp. (Yamamoto
et al., 2011).
CCR and the requirement for cAMP and CRP
in V. cholerae
When chitin-induced natural competence was first
described, the authors demonstrated that glucose
prevented natural transformation. They hypothesized
therefore that CCR might play a role in competence
induction (Meibom et al., 2005; Figs 1 and 3). Vibrio
cholerae takes advantage of the abundance of zooplankton
in its natural environment. Zooplankton have chitinous
exoskeletons, which provide a nutritious surface that can
serve as the sole carbon source for V. cholerae and can
induce natural competence (Fig. 1). Therefore, the link
between CCR and chitin-induced natural competence was
further investigated (Blokesch, 2012). In this study, the
hypothesis that CCR is involved in the regulation of
natural competence was further strengthened by showing
that not only glucose but also other phosphoenolpyr-
uvate : carbohydrate phosphotransferase system (PTS)-
FEMS Microbiol Rev 37 (2013) 336–363
347
transported sugars, including the chitin monomer GlcNAc,
interfere with natural transformation (Blokesch, 2012).
Such carbon sources are known to counteract against the
intracellular accumulation of the secondary messenger
cAMP (for review see Deutscher et al., 2006). The author
of this study also demonstrated that the lack of cAMP or
of the CRP changes at least three levels of chitin-depen-
dent natural competence induction in V. cholerae: chitin
surface colonization, chitin degradation/metabolism and
competence gene expression (Blokesch, 2012). Indeed,
with respect to chitin metabolism, V. cholerae strains defi-
cient in CCR showed significantly lower expression of
chiA-2 (Blokesch, 2012). ChiA-2 and chiA-1 encode two
redundant extracellular chitinases, and at least one of these
chitinases is required for growth on chitin (Meibom et al.,
2004). The lack of expression of such chitin degradation
enzymes might explain why V. cholerae mutants that are
defective for CCR, such as a crp minus strain, are unable
to grow with GlcNAc oligomers as the sole carbon source
and are unable to colonize the chitinous surface (Blokesch,
2012).
The expression of the competence genes also being
directly dependent on CCR was further proven by uncou-
pling natural competence induction from chitin surface
colonization and chitin metabolism (Blokesch, 2012;
Lo Scrudato & Blokesch, 2012). Using different chitin-
independent competence-inducing conditions and differ-
ent readouts (reverse transcription followed by PCR and
transcriptional reporter fusions for Blokesch, 2012 and Lo
Scrudato & Blokesch, 2012, respectively), a change in
competence gene expression became obvious when com-
paring wild-type V. cholerae with CCR-deficient strains.
Furthermore, a V. cholerae strain that lacked the cAMP-
degrading enzyme cAMP phosphodiesterase (cpdA)
showed enhanced frequencies of natural transformation
(Lo Scrudato & Blokesch, 2012). Based on these and
other findings, Lo Scrudato & Blokesch (2012) concluded
that an imbalanced intracellular cAMP pool affects com-
petence induction at the transcriptional level. How these
two regulatory circuits – chitin-dependent induction of
TfoX and CCR – are connected is not fully understood.
However, a current hypothesis is that the transformation
regulator TfoX can act only in conjunction with CRP.
This speculation is based on the suggested CRP-Sxy inter-
action as described earlier for H. influenzae. Indeed, based
on published expression data for V. cholerae grown on
chitin surfaces or tfoX overexpression (Meibom et al.,
2004, 2005), competence-specific CRP-S sites have also
been predicted in silico by Cameron and Redfield for
V. cholerae (Cameron & Redfield, 2006). It might be
interesting to note that cAMP-CRP also positively regu-
late the expression of the integron-specific integrase gene
in V. cholerae (Baharoglu et al., 2012), which might avoid
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
348
the need to homologously recombine the incoming trans-
forming DNA into the chromosome; instead, the newly
acquired DNA could be integrated into the gene-captur-
ing integron island (Cambray et al., 2010).
Bacterial communication is key for competence
induction in V. cholerae
The first suggestion of the involvement of quorum sensing
(QS) in natural competence and transformation (Fig. 3)
was based on the finding that V. cholerae was transform-
able in a more efficient manner after an extended growth
period on chitin surfaces, which is in accordance with
increased cell densities (Meibom et al., 2005). Blokesch &
Schoolnik (2008) followed up on this finding and
described the connection between QS and natural compe-
tence and transformation in further detail. They demon-
strated that the main regulator of QS, HapR, which is
only produced at high cell density, represses the gene
encoding the extracellular nuclease Dns (Blokesch &
Schoolnik, 2008; Fig. 3). This repression is an essential
step in the regulation of natural transformation because
the nuclease Dns degrades surrounding DNA, thereby
destroying any potential transforming material (Blokesch
& Schoolnik, 2008). Recently, other researchers also con-
firmed the HapR-dependent down-regulation of dns
(Seper et al., 2011).
HapR also acting as a positive regulator of at least a
subset of competence genes, such as the periplasmic
DNA-binding protein encoding gene comEA, became
obvious early on (Meibom et al., 2005; Blokesch &
Schoolnik, 2008). This speculation was initially based on
microarray expression data (Meibom et al., 2005) and
V. cholerae hapR/dns double knockout mutants being
transformable at an c. 100-fold lower frequency than a
dns single knockout mutant, even though the degradation
of extracellular DNA was fully abolished (Blokesch &
Schoolnik, 2008). The expression of comEA being indeed
regulated in a QS-dependent manner was nicely demon-
strated using transcriptional fusion with a luciferase-based
readout (Antonova & Hammer, 2011) and also more
recently by detecting comEA expression using transcrip-
tional fusions with fluorescent proteins (Lo Scrudato &
Blokesch, 2012). The latter study significantly extended
this finding by not only monitoring the expression of
comEA but also other competence genes, such as pilA
(Lo Scrudato & Blokesch, 2012; Fig. 3). Furthermore, Lo
Scrudato and Blokesch investigated a plethora of compe-
tence genes with respect to any potential QS-dependent
regulation. The results of this study were remarkable
because they provided good evidence for only a small
subset of competence genes of V. cholerae being coregu-
lated by TfoX and QS (Lo Scrudato & Blokesch, 2012).
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
P. Seitz & M. Blokesch
More importantly, the genes that are indeed linked with
QS (comEA and comEC) are all extremely relevant for
transformation because the encoded proteins are hypothe-
sized to directly interact with the transforming DNA (Box
1). Indeed, Dns degrades the DNA at low cell density,
whereas ComEA and ComEC are produced at high cell
density and present major components of the DNA-
uptake machinery (Box 1; Suckow et al., 2011), thereby
directly interacting with the incoming DNA in V. chole-
rae. Thus, in conclusion, QS acts as a switch from extra-
cellular DNA degradation to DNA uptake in V. cholerae
(Lo Scrudato & Blokesch, 2012).
Based on HapR being involved in the natural compe-
tence and transformation of V. cholerae, several research
groups became interested in the question of how the
autoinducers of V. cholerae contribute to this regulatory
circuit. Based on work by Bonnie Bassler and collabora-
tors, two main autoinducers have been identified and have
been allocated a specific role in V. cholerae. Cholera
autoinducer 1 (CAI-1), a (S)-3-hydroxytridecan-4-one
(Higgins et al., 2007), is thought to play a role as an intra-
species communication agent. By contrast, AI-2, a furano-
syl borate diester (Chen et al., 2002), serves as a ‘universal
signal’ in bacteria as this molecule is produced and sensed
by many bacteria and therefore allows interspecies com-
munication (Xavier & Bassler, 2003). Comparing V. cholerae
mutants devoid of the capacity to synthesize CAI-1, AI-2
or both autoinducers, the group of Melanie Blokesch
showed that in the absence of CAI-1, natural transforma-
tion is (almost) completely abolished in experiments
mimicking the natural chitinous environment (Suckow
et al., 2011) and that transformation was consistently
undetectable under homogenous competence-inducing
conditions (Lo Scrudato & Blokesch, 2012). Furthermore,
the authors of these studies were never able to detect any
transformants in the absence of both autoinducers (CAI-1
and AI-2), a phenotype that mirrors hapR minus strains
(Suckow et al., 2011; Lo Scrudato & Blokesch, 2012). Not
surprisingly no HapR protein was detectable in CAI-1/
AI-2 synthase-deficient cells (Lo Scrudato & Blokesch,
2012). This transformation-negative phenotype of autoin-
ducer-deficient V. cholerae cells was not restorable by
cross-feeding of solely AI-2 from co-cultured bacteria, in
contrast to the efficient restoration of transformation by
cross-fed CAI-1 (Suckow et al., 2011). Suckow et al.
therefore concluded that the intraspecies autoinducer
CAI-1 might be similar to competence pheromones
described in Gram-positive bacteria and that such a
regulation might increase the chances of taking up
species-specific DNA (Fig. 2). However, even though
another study illustrated the same tendency, it did not
show such a strict CAI-1-dependency. More specifically,
Antonova & Hammer (2011) concluded in their study
FEMS Microbiol Rev 37 (2013) 336–363
Regulation of natural competence in Gram-negative bacteria
that ‘Vibrio-specific CAI-1 appears to play a major role
and interspecies AI-2 a minor role, suggesting that
induction of DNA uptake may not be restricted exclu-
sively to a response to autoinducers produced by Vibrio
species, but that HGT may also be promoted by AI-2
derived from non-Vibrio members of a biofilm’. These
authors also showed that strains lacking both autoinducers
(CAI-1 and AI-2) were readily transformable and thereby
significantly above the detection level of a nontransform-
able hapR mutant strain (Antonova & Hammer, 2011).
Based on these data, it can be hypothesized that the dis-
crepancy in autoinducer dependency in these different
V. cholerae strains is not due to a different connection
between the QS regulator HapR and natural transforma-
tion but most likely reflects a difference in the QS circuit-
ries. Further studies will be required to confirm or refute
this hypothesis.
The induction of competence in V. cholerae
compared with the other naturally
transformable bacteria
Based on the results of CAI-1 dependency for natural
transformation in V. cholerae, CAI-1 can be considered a
competence pheromone (Suckow et al., 2011; Lo Scruda-
to & Blokesch, 2012). Indeed, this molecule might per-
form a function similar to a function proposed for
Gram-positive bacteria (for a seminal review on compe-
tence induction in Gram-positive bacteria see Claverys
et al., 2006). For example, competence induction in
S. pneumoniae is not solely dependent on the cell density
and thereby on the passive accumulation of the compe-
tence-stimulating peptide (CSP), a competence phero-
mone. Instead, the production of CSP varies with
changing environmental conditions (Claverys et al., 2000,
2006). Claverys et al. showed that DNA-damaging antibi-
otics induce the expression of the competence regulon in
a CSP-dependent manner in S. pneumoniae. These
authors argued that CSP serves as a stress signal and can
therefore be considered an alarmone rather than a quo-
rum-sensing effector (Prudhomme et al., 2006). Prud-
homme et al. (2006) therefore concluded that ‘CSP could
thus play a crucial role in generating genetic diversity
under stress conditions for a species that seems unable to
rely on inducible mutagenic repair (such as the SOS
response)’. The hypothesis that CAI-1 in V. cholerae is
similar to competence pheromones is supported by the
fact that the synthesis of CAI-1 by the synthase CqsA is
also not constitutive in V. cholerae. In fact, Liang et al.
(2007, 2008) demonstrated that CCR is involved in the
post-transcriptional regulation of cqsA expression in
V. cholerae, providing another connection among the
three regulatory pathways that drive natural competence
FEMS Microbiol Rev 37 (2013) 336–363
349
and transformation in V. cholerae. This process also con-
nects a process indicative of preferred carbon source star-
vation, or CCR, to the induction of competence. This
link might reflect a similarity to the function of natural
transformation in S. pneumoniae ‘as a rescue process’
(Prudhomme et al., 2006).
In a recent review by Ng and Bassler, the authors wrote
‘One oddity is that homologues of cqsA and cqsS, called
lqsA and lqsS respectively, exist in the distantly related bac-
terium L. pneumophila. LqsA produces 3-hydroxypentad-
ecan-4-one; a molecule with a longer hydrocarbon chain
than CAI-1 (Spirig et al., 2008)’ (Ng & Bassler, 2009).
Interestingly, a recent analysis of six fully sequenced
L. pneumophila genomes suggested that the genomes were
highly dynamic, as a result of extensive HGT and recombi-
nation (Gomez-Valero et al., 2011). Furthermore, ‘L. pneu-
mophila ranks as the prokaryote with the widest variety of
eukaryotic-like proteins’, all of which may have been
acquired from the host by HGT (Cazalet et al., 2004). At
this point, it is tempting to speculate that a-hydroxyketone
signalling molecules are commonly involved in the regula-
tion of natural competence. Thus, it will be interesting to
see whether the Legionella autoinducer-1 (LAI-1; Spirig
et al., 2008) is also required for natural competence and
transformation in this bacterium. It had long been appreci-
ated that L. pneumophila is naturally transformable (Stone
& Kwaik, 1999), and recent studies have provided evidence
that competence might fulfil a similar function in this
organism as in S. pneumoniae, replacing the absent SOS
response (Charpentier et al., 2011).
The induction of natural competence
in noncholera Vibrio species
Studies in the 1990s had already suggested that ‘marine
Vibrio spp.’ were naturally transformable (Frischer et al.,
1990, 1993; Jeffrey et al., 1990; Paul et al., 1992). These
aquatic isolates were assigned to the genus Vibrio and
were most similar to V. campbelli based on the biochemi-
cal features described in Bergey’s manual; however, fol-
low-up studies on this phenomenon provided evidence
that the isolates belonged to the genus Pseudomonas, a
finding based on 16S-rRNA gene analysis (Frischer et al.,
1996). Thus, the natural transformability of Vibrio spp.
had not been demonstrated before Meibom et al. (2005)
described chitin-induced natural transformation in
V. cholerae.
Following the first studies on natural competence in
V. cholerae (Meibom et al., 2005; Blokesch & Schoolnik,
2007, 2008), other researchers have found that V. cholerae
is not the only representative of the genus Vibrio to be
naturally transformable; natural transformation has also
been demonstrated in V. vulnificus (Gulig et al., 2009),
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
350
Vibrio parahaemolyticus (Chen et al., 2010) and Vibrio
fischeri (Pollack-Berti et al., 2010). The first two studies
used a modified version of the natural transformation
protocol initially described for V. cholerae (Meibom et al.,
2005; Blokesch & Schoolnik, 2007) to modify the bacteria
genetically (Gulig et al., 2009; Chen et al., 2010). Because
this protocol is based on competence induction on chitin-
ous surfaces, a common theme of chitin-based induction
seems clear. Indeed, it was experimentally shown that
many other species of the genus Vibrio grow on chitin
and that most of them contain proteins homologous to
those identified in the V. cholerae chitin utilization pro-
gramme (Meibom et al., 2004 and as summarized in
Hunt et al., 2008). Thus, the chitinous surface may be a
common niche for Vibrio spp., and chitin-induced natu-
ral competence may be widespread.
This finding was further emphasized by the group of
Ned Ruby, who described the induction of natural com-
petence in the symbiotic Vibrio, V. fischeri (Pollack-Berti
et al., 2010). These researchers provided evidence that as
in V. cholerae, natural competence in V. fischeri is
induced in the presence of chitin oligomers and in a
TfoX-dependent manner. Interestingly, Pollack-Berti et al.
identified a second TfoX-like protein, named TfoY, in
V. fischeri. Indeed, all fully sequenced Vibrionaceae con-
tain these two paralogues of TfoX (Pollack-Berti et al.,
2010). The authors of this study also demonstrated that
the two paralogs were unable to compensate fully for the
loss of the counterpart and accordingly that they influ-
ence natural transformation in distinct manners. Interest-
ingly, the corresponding TfoY counterpart of V. cholerae
had been discovered earlier during a search for RNA
motifs and riboswitches in bacteria because it contains a
Genes for the Environment, for Membranes, and for
Motility (GEMM) motif (Weinberg et al., 2007; Sudarsan
et al., 2008). In this context, Weinberg et al. described
that based on the conserved domain database (CDD),
V. cholerae contains two genes belonging to the COG3070
family (encompassing TfoX-domains). As TfoY (named
tfoXGEMM in Weinberg et al., 2007) but not tfoX
contained this structured GEMM motif, the authors con-
cluded ‘in V. cholerae and related bacteria, GEMM might
participate in chitin-induced competence, or even regu-
late competence in environments not containing elevated
chitin concentrations’. In line with the latter hypothesis,
TfoY does not play an obvious role in chitin-induced nat-
ural competence of V. cholerae. Indeed, a V. cholerae
knockout strain of tfoY is fully transformable during
growth on chitinous surfaces and artificial expression of
tfoY does not render V. cholerae naturally competent
(M. Blokesch, unpublished data).
In conclusion, the regulation of natural competence
and transformation in Vibrio spp. seems to be consistent
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
P. Seitz & M. Blokesch
with respect to chitin-dependent induction and the
requirement for TfoX. The QS circuits, by contrast, differ
among these organisms, and not all Vibrio spp. contain a
CqsA/CqsS system (for recent review see Milton, 2006).
Additionally, not all Vibrio spp. that do contain a CqsA/
CqsS system produce the same autoinducer. For example,
work by Bonnie Bassler’s group demonstrated that the
Vh-CAI-1 autoinducer of Vibrio harveyi differs slightly
from its counterpart in V. cholerae in that the former
contains an 8-carbon tail instead of a 10-carbon tail
(Ng et al., 2011). An 8-carbon tail CAI-1 was also present
in spent supernatants of other Vibrio spp., including
V. parahaemolyticus, Vibrio alginolyticus, Vibrio anguilla-
rum and Vibrio furnissii (Ng et al., 2011). The authors of
this study concluded that ‘different Vibrio species display
unique production and detection profiles for the CAI-1
family of molecules’. One could hypothesize that ‘the
forces that drove these two species [= V. cholerae and
V. harveyi] to evolve different signalling specificities’ (Ng
et al., 2011) might be linked with natural transformation
and species-specificity of the system (Fig. 2). Further
studies will be required to elucidate the connection
between QS and the natural transformability in nonchol-
era Vibrios and especially in those Vibrios that do not
contain CqsA/CqsS systems.
The induction of natural transformation
in plant pathogenic bacteria
Ralstonia solanacearum – a large host range
plant pathogen takes up DNA
Ralstonia solanacearum is a representative bacterium of
the b-proteobacteria class, family Burkholderiaceae. Most
importantly, R. solanacearum is a major plant pathogen
‘because of its aggressiveness, large host range, broad
geographical distribution and long persistence in soil and
water environments’ as recently reviewed (Genin, 2010).
After the genome of R. solanacearum was sequenced
(Salanoubat et al., 2002 and reviewed in Genin & Boucher,
2004), analysis revealed some interesting features. First, as
in V. cholerae (Heidelberg et al., 2000), the genome of
R. solanacearum is split into two chromosomes of
unequal sizes (3.7 and 2.1 megabases). Second, the gen-
ome displayed a mosaic structure with respect to the
G+C content throughout several kilobases of both chro-
mosomes (7% of the total genomes differed). Together
with the fact that the encoded genes differed in codon
usage from the rest of the bacterium, this difference was
indicative of HGT (Genin & Boucher, 2004). The authors
of this review concluded ‘that [these regions] could play
an important role in the rapid adaptation of the bacte-
rium to the change of ecological niche’ (Genin & Bou-
FEMS Microbiol Rev 37 (2013) 336–363
Regulation of natural competence in Gram-negative bacteria
cher, 2004). Because R. solanacearum is naturally trans-
formable, this mode of HGT might have contributed sig-
nificantly to the mosaic structure of the genome. Indeed,
two recent studies provided evidence that large regions of
DNA (from 30 to 90 kb and up to 80 kb, respectively)
can be integrated into the recipient’s genome (Coupat
et al., 2008; Coupat-Goutaland et al., 2011).
With respect to the regulation of natural competence
in R. solanacearum, transformation frequencies peaked in
the middle of the exponential phase and dropped quickly
afterwards (Bertolla et al., 1997). Furthermore, as is the
case for many other naturally competent bacteria, growth
on minimal medium reflecting limiting growth conditions
permitted higher transformation efficiencies than those
observed in rich medium (Bertolla et al., 1997). Bertolla
et al. (1999) also demonstrated that natural transforma-
tion occurs in situ (e.g. in planta). During cell-host inter-
action, natural transformation was most efficient while
R. solanacearum cells were multiplying within the plant
vessels (Bertolla et al., 1999).
Ralstonia solanacearum also harbours more than one
quorum-sensing system (Whitehead et al., 2001). Specifi-
cally, this species contains a LysR-type regulator, PhcA,
which regulates the production of extracellular polysac-
charide (EPS) and extracellular enzymes (Brumbley et al.,
1993). PhcA is subject to regulation by the two-component
system, which responds to the autoinducer 3-hydroxypalmitic
acid methyl ester (3OH PAME; Clough et al., 1997;
Flavier et al., 1997). This autoinducer is most likely syn-
thesized by the synthase PhcB, which exhibits homology
to small-molecule SAM-dependent methyltransferases
(Flavier et al., 1997). Interestingly, the CAI-1 synthase
CqsA of V. cholerae also uses SAM as a substrate (as do
many other autoinducer synthases; Wei et al., 2011),
and, at first glance, CAI-1 does show similarities to the
3OH PAME of R. solanacearum. Evidence that, as in
V. cholerae, this quorum-sensing system is also involved
in the regulation of natural transformation was generated
by a recent study by Kang et al., who demonstrated that
the gene encoding the major pilin subunit of a type
IV pilus, pilA, is essential for natural transformation of
R. solanacearum and is regulated in a cell density-dependent
manner. Specifically, these authors demonstrated that the
expression of pilA decreases at high cell density and that
this downregulation is dependent on the major regulator
of QS, PhcA (Kang et al., 2002).
Another plant pathogen, Xylella fastidiosa,
was only recently discovered to be naturally
transformable
Another plant pathogen, X. fastidiosa, has also been iden-
tified as being naturally transformable (Kung & Almeida,
FEMS Microbiol Rev 37 (2013) 336–363
351
2011). This bacterium belongs to the class c-proteobacte-
ria, family Xanthomonadaceae. The lifestyle of X. fastidi-
osa has recently been reviewed and the authors
mentioned that ‘strains of X. fastidiosa have been associ-
ated with a large number of diseases, many causing great
economic losses’ (Chatterjee et al., 2008). This bacterium
causes disease in many plants, including grapes and citrus
fruits. Transmission of this bacterium is mediated by
insects, such as sharpshooters (Chatterjee et al., 2008).
The ability to exchange genes horizontally may benefit
X. fastidiosa bacteria and ‘potentially permit them to
explore a wider variety of host plants’ (Kung & Almeida,
2011).
The information regarding when and how natural
competence in Xylella is induced is still limited; however,
Kung & Almeida (2011) demonstrated that cell growth
affects the number of transformants, and transformants
were obtained more efficiently if the transforming DNA
was provided upon entry into exponential phase, with the
efficiency declining significantly afterwards. These authors
also demonstrated that methylated plasmid DNA served
better as transforming DNA than its unmethylated coun-
terpart, suggesting that R-M systems might play a role in
the transformation of X. fastidiosa (Kung & Almeida,
2011) comparable with what was demonstrated for
H. pylori (as described above; Humbert & Salama, 2008;
Humbert et al., 2011). Finally, the authors observed a
medium-specificity of natural transformation because
transformation occurred only in nutrient-limited medium
(Kung & Almeida, 2011). Thus, the induction of natural
competence seems to be linked with the nutritional status
of the cell in X. fastidiosa. Chitin has been proposed as a
carbon source for X. fastidiosa upon entry into insect vec-
tors (e.g. leafhoppers), and chitinolytic activity has been
demonstrated in vitro (Killiny et al., 2010). Thus, it is
tempting to speculate that a co-regulation between chitin
utilization and CCR, as described earlier for V. cholerae,
might also be involved in competence induction in
X. fastidiosa.
Evolution in an evolutionary system –
what benefits does natural
transformation bring to an organism?
The question of why natural transformation exists
remains the subject of ongoing debate. Three hypotheses
are frequently discussed: ‘DNA for food’, ‘DNA for
repair’, and ‘DNA for evolution’. These hypotheses are
not necessarily mutually exclusive, but they are not
equally supported by the data. This topic has recently
been extensively reviewed elsewhere (Michod et al., 2008;
Vos, 2009); however, due to the strong link between the
induction of natural competence by environmental signals
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
352
and the role transformation may play in bacteria, we
summarize some of the major arguments below.
DNA for food
The role of DNA as a source of nutrients was initially
proposed by Redfield (2001). She wrote, ‘the simplest
model thus might be that the nutrients that transforma-
tion reliably provides pay the evolutionary bills (and have
been responsible for the evolution of its regulation), and
as a bonus the cell gets the occasional benefits of recom-
bination and repair’ (Redfield, 1993b). Among others, the
main arguments put forward by Redfield for a role of
transformation as ‘Genes for Breakfast’ were as follows:
(1) DNA as food provides a direct short-term advantage,
whereas a role of novel genes in evolution might be
selected only in the long run. (2) H. influenzae and
B. subtilis react to nutritional limitations when inducing
competence (Redfield, 1993a, b). Notably, this is not the
case for S. pneumoniae. In this species, nutrient starvation
has never been observed to induce competence (Claverys
et al., 2006); instead, during growth in rich medium the
bacteria acquire competence in early log phase and main-
tain it only over a short period of time or differentially
said ‘competence is induced in times of feast rather than
famine’ (Claverys & Havarstein, 2007). (3) Neither H.
influenzae nor B. subtilis induce competence upon DNA
damage. (4) Unrelated DNA can be used as a nutrient
source, especially because it could not recombine homol-
ogously into the chromosome. (5) The addition of ribo-
nucleotide monophosphates, most notably AMP and
GMP, to competence-inducing starvation medium
reduced the natural transformation in H. influenzae by
more than two orders of magnitude and significantly
reduced the expression of competence genes in this
organism (MacFadyen et al., 2001). These authors argued
that the depletion of purines within the cell induces com-
petence and that the incoming DNA could subsequently
replenish the purine pool. Interestingly, this effect was
not observable for desoxyribonucleotide monophosphates,
triphosphates, or the free bases (MacFadyen et al., 2001).
(6) The poor-quality DNA derived from dead cells might
not be suitable for transformation-mediated evolution
(Redfield, 1988; Redfield et al., 1997). But certain competent
bacteria kill their (non-or not yet competent) siblings
within the population, whereas other bacteria actively
donate DNA through a T4SS (Hamilton et al., 2005;
Hamilton & Dillard, 2006) or through a currently
unknown mechanism (Stewart et al., 1983), suggesting
that not all transforming DNA is of poor quality.
Recent studies on E. coli support Redfield’s work.
Finkel & Kolter (2001) provided evidence that E. coli can
grow with DNA as the sole source of carbon and energy;
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
P. Seitz & M. Blokesch
however, this ability was disrupted in E. coli mutants
lacking genes that encode for potential competence-
related proteins. The authors of this and a follow-up
study identified eight genes encoding proteins that are
between 12% and 74% identical to the H. influenzae
counterparts involved in natural competence (Finkel &
Kolter, 2001; Palchevskiy & Finkel, 2006). In the absence
of these genes, most of these strains encountered a sta-
tionary phase competition defect during co-culture with
the parental wild-type bacteria. The authors of this study
concluded that taking up DNA for nutritional purposes,
‘particularly when that DNA is heterologous and less
likely to recombine onto the chromosome’ (Finkel &
Kolter, 2001; Palchevskiy & Finkel, 2006), might confer a
significant advantage even over the acquisition of a bene-
ficial gene by HGT. However, because most of the
homologous proteins identified in E. coli and other proteo-
bacteria (Palchevskiy & Finkel, 2006) resemble the type
IV pilus part of the DNA-uptake machinery, the question
arises as to whether such DNA indeed reaches the cyto-
plasm as linear ssDNA, as occurs in naturally competent
bacteria. Alternatively, the type IV pilus-like structure
may assist in recruiting free dsDNA into the periplasm
and thus facilitate transport across the outer membrane
through the secretin PilQ/HofQ. DNA might be further
degraded in the periplasm into nucleosides, which are
subsequently taken up into the cytoplasm by specific
nucleoside transporters to serve as a source of carbon and
energy. Thus, whereas the first part of this process would
resemble natural competence-induced DNA uptake and
involve type IV pilus-like protein components, DNA
transport across the inner membrane with concomitant
degradation of one strand might not be identical in
‘nutritional competence’ (Palchevskiy & Finkel, 2006).
Indeed, DNA uptake is a 2-step process in naturally
competent H. pylori (Stingl et al., 2010); however, as dis-
cussed earlier, H. pylori uses a T4SS, not a type IV
pilus-like structure, to shuffle the DNA across the outer
membrane. But there are good indications that this 2-step
DNA-uptake process also holds true for other Gram-
negative bacteria, as demonstrated in a recent review
(Kru ger & Stingl, 2011). Furthermore, V. cholerae regu-
lates the competence genes required for DNA movement
across the outer membrane differentially from the compe-
tence genes whose products are involved in DNA translo-
cation across the periplasmic space and the inner
membrane (Lo Scrudato & Blokesch, 2012).
Other findings also oppose the ‘DNA for food’ hypoth-
esis. For example, David Dubnau stated in a review from
1999 that B. subtilis possesses a powerful extracellular
nuclease and adequate uptake systems for DNA degrada-
tion products (Dubnau, 1999). Extracellular nucleases
have also been described for Vibrio cholerae (Newland
FEMS Microbiol Rev 37 (2013) 336–363
Regulation of natural competence in Gram-negative bacteria
et al., 1985; Focareta & Manning, 1987). Most notably,
the main nuclease in this organism, Dns, is oppositely
regulated from the competence genes that are directly
involved in DNA uptake (Blokesch & Schoolnik, 2008;
Lo Scrudato & Blokesch, 2012), and Dns is a major
inhibitor of natural transformation in V. cholerae because
it degrades the transforming material around the cell
(Blokesch & Schoolnik, 2008). A role of this nuclease in
the utilization of DNA as a nutrient source has been sug-
gested (Blokesch & Schoolnik, 2008) and was experimen-
tally supported with respect to the utilization of DNA as
a phosphate source (Seper et al., 2011). Interestingly, the
crystal structure of a concentrative nucleoside transporter
of V. cholerae (NupC) has recently been solved. This pro-
tein uses a sodium-ion gradient for nucleoside transport
across the inner membrane (Johnson et al., 2012) and
may be involved in the uptake of nucleotides released by
the extracellular nuclease Dns.
Another important point that undermines the DNA for
food hypothesis is the energy associated with DNA uptake
itself. As mentioned earlier and discussed in detail else-
where (Chen & Dubnau, 2004; Allemand & Maier, 2009;
Burton & Dubnau, 2010; Allemand et al., 2012), the
DNA-uptake machinery is most likely a multiprotein
complex (Box 1). Although the composition and mode of
action of this complex has not fully been elucidated, the
complex is assumed to resemble type IV pili (with the
exception of H. pylori, as discussed earlier). Such type IV
pilus-like structures (or shortened forms, also known as
pseudopili; Pugsley, 1993; Chen & Dubnau, 2004) allow
for the transport of DNA across the peptidoglycan layer
and/or the outer membrane of Gram-positive and Gram-
negative bacteria, respectively. An inner membrane chan-
nel subsequently allows translocation of the DNA across
the inner membrane; this structure is probably conserved
across all naturally transformable bacteria (Draskovic &
Dubnau, 2005; Stingl et al., 2010; Suckow et al., 2011).
Consistent with the resemblance of the components, the
forces generated by type IV pilus retraction and DNA
uptake were in the same range for both systems and rep-
resent some of the strongest linear motors characterized
to date (Merz et al., 2000; Maier et al., 2002, 2004). As
recently reviewed by Berenike Maier and others (Maier,
2005; Allemand & Maier, 2009; Allemand et al., 2012)
such ‘directed DNA translocation is often energetically
unfavourable and requires an active process that uses
energy, namely the action of molecular motors’ (Alle-
mand et al., 2012). Thus, the question arises as to
whether the use of transforming DNA as an energy
source would be able to supply enough energy to
compensate for the consuming uptake process and still
provide sufficient extra energy to be more cost-effective
than the de novo synthesis of nucleotides. Furthermore,
FEMS Microbiol Rev 37 (2013) 336–363
353
incoming ssDNA is protected against degradation in
naturally competent bacteria by competence-specific
proteins such as DprA (Mortier-Barriere et al., 2007).
This mechanism is also more consistent with a role of the
transforming DNA in DNA-repair processes or in the
donation of new alleles/genes.
DNA for repair
Arguments for the repair hypothesis are principally based
on the following facts:
The ability to take up DNA in some Gram-negative,
naturally transformable bacteria is highly biased towards
genetic material from the same or closely related species.
Various strategies have evolved to this end (Fig. 2). For
example, N. gonorrhoeae and H. influenzae discriminate
between self and foreign DNA through the recognition of
DUS that are overrepresented in their own genomes
(Danner et al., 1980; Fitzmaurice et al., 1984; Elkins
et al., 1991). Vibrio cholerae, in contrast, does not dis-
criminate between self and foreign DNA at the level of
the DNA uptake (Suckow et al., 2011). Because compe-
tence induction in this organism is tightly linked with an
accumulation of the species-specific autoinducer CAI-1
(Suckow et al., 2011; Lo Scrudato & Blokesch, 2012), spe-
cies-specific DNA is highly likely to reach the cytosol. In
contrast, H. pylori does not display any preference for
species-specific DNA. This assumption is based on the
fact that competence is constitutive in this organism and
that DUS-dependent sorting does not occur at the level
of the DNA-uptake machinery. However, recent data has
indicated that a mechanism based on R-M systems might
control DNA uptake in H. pylori, as explained above
(Aras et al., 2002; Humbert & Salama, 2008; Humbert
et al., 2011). This system may also ensure the species-
specificity of transforming DNA by recognizing and
degrading foreign genetic material, and protecting the
genome from foreign DNA (Fig. 2; Aras et al., 2002;
Humbert & Salama, 2008; Humbert et al., 2011). Another
source for species-specific DNA may be fratricide. More
precisely, bacterial fratricide is associated with natural
competence of S. pneumoniae (Guiral et al., 2005; Havar-
stein et al., 2006 and recent review by Claverys & Havar-
stein, 2007) and probably also in H. pylori (Dorer et al.,
2010; Fig. 2). In this context, bacterial cells of the same
population are killed to provide transforming DNA.
Based on several examples of the intentional killing of
bacterial siblings, Gilmore & Haas (2005) concluded that
‘the selective lysis of siblings by a subpopulation of bacte-
rial cells appears to be a highly evolved and complex pro-
cess’.
Based on these important points, the idea that natural
transformation serves as a mechanism of DNA repair
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
354
seems sound. In this scenario, the uptake of DNA from
closely related organisms would facilitate the maintenance
of genomic integrity. Evidence from experiments on the
naturally competent Gram-positive bacterium B. subtilis
supports this hypothesis (Michod et al., 1988; Wojcie-
chowski et al., 1989), but recent findings on the two
Gram-negative bacteria, H. pylori and L. pneumophila,
with nonsense mutations in their DNA-uptake systems,
did not support the repair hypothesis (Dorer et al.,
2010; Charpentier et al., 2011). In these studies the
bacterial strains incapable of DNA uptake showed no
increased sensitivity to genotoxic agents (Dorer et al.,
2010; Charpentier et al., 2011).
DNA for evolution
Finally, natural transformation might enable rapid evolu-
tion when diversity may be beneficial; these circumstances
include such stresses as high population densities, DNA
damage, abundance or a lack of certain carbon sources
and/or starvation. All of these conditions are known to
induce natural competence in at least a subset of natu-
rally transformable bacteria, as described earlier and illus-
trated in Fig. 1. A recent study on the long-term in vitro
passage of H. pylori (c. 1000 generations) supported an
evolutionary advantage of natural transformation because
competent bacteria increased their fitness faster than
those unable to take up external DNA (Baltrus et al.,
2008). Another recent study investigated the occurrence
of multi-drug-resistant (MDR) strains based on recombi-
nation following DNA uptake as part of the natural trans-
formation process of A. baylyi (Perron et al., 2012). The
authors demonstrated that in the presence of recombina-
tion, resistance genes were readily exchanged, and MDR
strains were obtained within fewer generations (Perron
et al., 2012).
However, as noted by Perron et al. and as described in
depth in a recent review on this topic (MacLean et al.,
2010), it is important that ‘evolution experiments offer a
useful approach to uncover the factors determining the
evolution of resistance, but most experiments have stud-
ied clonal populations without any contribution of
recombination’. Furthermore, the ‘benefits of recombina-
tion are context-dependent’, and experimental setups are
crucial to the outcome of such experiments. Such context
dependency was also highlighted in a recent study by
Engelmoer & Rozen (2011). These authors emphasized
once more the biased setup of most experimental studies,
which only examine benefits dependent on the acquisition
of DNA as part of natural transformation. However,
because natural competence is a developmental programme
and often induces other processes apart from DNA uptake
(e.g. fratricide in S. pneumoniae or competence-dependent
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
P. Seitz & M. Blokesch
growth arrest, as seen in B. subtilis), benefits arising from
these processes might have been overlooked in the past
(Engelmoer & Rozen, 2011). In this study, the authors
investigated the natural competence of S. pneumoniae and
confirmed that transformation is beneficial in a ‘DNA for
repair’ scenario upon treating cells with DNA-damaging
agents (Engelmoer & Rozen, 2011), but they also provided
evidence that competence is beneficial in withstanding
other kinds of stresses and that these benefits do not rely
on transformation (Engelmoer & Rozen, 2011). The
authors concluded that their findings were in line with
‘Claverys’ hypothesis (Claverys et al., 2006) that compe-
tence but not necessarily transformation may act as a
general process to relieve stress’ (Engelmoer & Rozen,
2011).
In summary, we conclude that there are many benefits
of natural competence and transformation, and there
might be no single reason that transformation is main-
tained in bacteria. The fact that numerous bacteria are
known or predicted to be naturally transformable is a
strong indication of the importance of this mode of
HGT.
Concluding remarks
Regulation of the competence window may
have co-evolved as an ability to respond to
environmental cues
Summarizing the information on competence induction
presented here, we conclude that natural competence
occurs constitutively in certain Gram-negative bacteria
but is tightly regulated and transient in others. But still
the question of ‘who’s competent and when’, which was
initially asked by Solomon & Grossman (1996), cannot
yet be fully answered. Interestingly, several recent studies
examine how natural competence and transformation are
evolutionarily maintained over time (Johnsen et al., 2009;
Maughan & Redfield, 2009). The question of why some
naturally competent bacteria are competent only in
a brief, finely tuned time window has not yet been
Fig. 4. The direct correlation between the variation encountered in
an organism’s niche and the complexity of the regulatory network
driving competence induction. For details, see the text.
FEMS Microbiol Rev 37 (2013) 336–363
Regulation of natural competence in Gram-negative bacteria
conclusively addressed. An examination of the
environments of naturally competent bacteria might yield
evidence as to why some of these organisms have tighter
competence windows than others (Fig. 4). For example,
H. pylori, as a gastric pathogen, solely colonizes the epi-
thelium of the human stomach, and no other reservoir
has been reported. Helicobacter pylori does not need to
cope with sudden and/or dramatic changes in its environ-
ment (Fig. 4). Instead, such parameters as pH or nutrient
availability are likely to fluctuate only within a narrow
range. However, during long-term infection, environmen-
tal alterations might include the ageing of the host and
concomitant changes in immune status or a secondary
infection with another H. pylori strain. In this context,
natural transformation can lead to the rapid spread of
beneficial mutations and thus provides a means for adap-
tion. Because H. pylori is constitutively competent, it
maintains high genetic plasticity during infection. There-
fore, the bacteria are constantly able to fine-tune interac-
tions with their human hosts, as suggested in a recent
review on the evolution and phenotypic diversification of
H. pylori (Suerbaum & Josenhans, 2007).
The dynamics might differ significantly in naturally
transformable bacteria that frequently encounter dramatic
fluctuations in their environment (Fig. 4). A good exam-
ple is V. cholerae, an aquatic and naturally transformable
bacterium. In its environmental niche, this organism can
be found as free-living bacterium in a planktonic state or
as part of a sessile biofilm community (Yildiz & Visick,
2009). Furthermore, V. cholerae is well known for its abil-
ity to interact with phyto- and zooplankton, of which the
mucilaginous sheath and the exoskeleton, respectively,
can serve as carbon and energy sources (reviewed by Lipp
et al., 2002). The abundance of such carbon sources,
however, can fluctuate significantly, and climate changes
and human behaviour are thought to contribute signifi-
cantly to plankton blooms and V. cholerae abundance
(Colwell, 1996; Lobitz et al., 2000; Lipp et al., 2002; Pru-
zzo et al., 2008; Fig. 4). Furthermore, selection pressure
from environmental cholera bacteriophages, which are
also thought to play a significant role in cholera seasonal-
ity and the emergence of new clones (Faruque et al.,
2005), as well as pressure from protozoan grazers (Matz
et al., 2005; Erken et al., 2011), may vary in time and
space. Thus, the adjustment between famine and feast
and between growth and growth arrest including the
ability to enter the state of natural competence must be
fine-tuned (Fig. 4). Further studies will be required to
address the potential disadvantages of constitutive compe-
tence and transformation under changing environmental
conditions; such weaknesses could explain why the com-
petence window is adjusted according to the bacterium’s
lifestyle.
FEMS Microbiol Rev 37 (2013) 336–363
355
Acknowledgements
We apologize to those researchers whose work we could
not include due to space limitations. This work was sup-
ported by Swiss National Science Foundation (SNSF)
grant 31003A_127029.
References
Aas FE, Wolfgang M, Frye S, Dunham S, Lovold C & Koomey M
(2002) Competence for natural transformation in Neisseria
gonorrhoeae: components of DNA binding and uptake linked
to type IV pilus expression. Mol Microbiol 46: 749–760.
Achtman M, Azuma T, Berg DE, Ito Y, Morelli G, Pan ZJ,
Suerbaum S, Thompson SA, van der Ende A & van Doorn LJ
(1999) Recombination and clonal groupings within
Helicobacter pylori from different geographical regions. Mol
Microbiol 32: 459–470.
Ahmed N, Dobrindt U, Hacker J & Hasnain SE (2008)
Genomic fluidity and pathogenic bacteria: applications in
diagnostics, epidemiology and intervention. Nat Rev
Microbiol 6: 387–394.
Alexander HE & Redman W (1953) Transformation of type
specificity of meningococci; change in heritable type
induced by type-specific extracts containing
desoxyribonucleic acid. J Exp Med 97: 797–806.
Allemand JF & Maier B (2009) Bacterial translocation motors
investigated by single molecule techniques. FEMS Microbiol
Rev 33: 593–610.
Allemand JF, Maier B & Smith DE (2012) Molecular motors
for DNA translocation in prokaryotes. Curr Opin Biotechnol
23: 503–509.
Alvarez-Martinez CE & Christie PJ (2009) Biological diversity
of prokaryotic type IV secretion systems. Microbiol Mol Biol
Rev 73: 775–808.
Ambur OH, Frye SA & Tonjum T (2007) New functional
identity for the DNA uptake sequence in transformation
and its presence in transcriptional terminators. J Bacteriol
189: 2077–2085.
Ambur OH, Davidsen T, Frye SA, Balasingham SV, Lagesen K,
Rognes T & Tonjum T (2009) Genome dynamics in major
bacterial pathogens. FEMS Microbiol Rev 33: 453–470.
Anderson MT & Seifert HS (2011) Opportunity and means:
horizontal gene transfer from the human host to a bacterial
pathogen. mBio 2: e00005–e00011.
Antonova ES & Hammer BK (2011) Quorum-sensing
autoinducer molecules produced by members of a
multispecies biofilm promote horizontal gene transfer to
Vibrio cholerae. FEMS Microbiol Lett 322: 68–76.
Aras RA, Small AJ, Ando T & Blaser MJ (2002) Helicobacter
pylori interstrain restriction-modification diversity prevents
genome subversion by chromosomal DNA from competing
strains. Nucleic Acids Res 30: 5391–5397.
Averhoff B (2009) Shuffling genes around in hot
environments: the unique DNA transporter of Thermus
thermophilus. FEMS Microbiol Rev 33: 611–626.
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
356
Averhoff B & Friedrich A (2003) Type IV pili-related natural
transformation systems: DNA transport in mesophilic and
thermophilic bacteria. Arch Microbiol 180: 385–393.
Averhoff B & Müller V (2010) Exploring research frontiers in
microbiology: recent advances in halophilic and thermophilic
extremophiles. Res Microbiol 161: 506–514.
Baharoglu Z, Krin E & Mazel D (2012) Connecting
environment and genome plasticity in the characterization
of transformation-induced SOS regulation and carbon
catabolite control of the Vibrio cholerae integron integrase.
J Bacteriol 194: 1659–1667.
Baltrus DA & Guillemin K (2006) Multiple phases of
competence occur during the Helicobacter pylori growth
cycle. FEMS Microbiol Lett 255: 148–155.
Baltrus DA, Guillemin K & Phillips PC (2008) Natural
transformation increases the rate of adaptation in the
human pathogen Helicobacter pylori. Evolution 62: 39–49.
Baumann P (1968) Isolation of Acinetobacter from soil and
water. J Bacteriol 96: 39–42.
Berndt C, Meier P & Wackernagel W (2003) DNA restriction
is a barrier to natural transformation in Pseudomonas
stutzeri JM300. Microbiology 149: 895–901.
Bertolla F, Van Gijsegem F, Nesme X & Simonet P
(1997) Conditions for natural transformation of
Ralstonia solanacearum. Appl Environ Microbiol 63:
4965–4968.
Bertolla F, Frostesard A, Brito B, Nesme X & Simonet P
(1999) During infection of its host, the plant pathogen
Ralstonia solanacearum naturally develops a state of
competence and exchanges genetic material. Mol Plant
Microbe Interact 12: 467–472.
Bielaszewska M, Mellmann A, Zhang W, Kock R, Fruth A,
Bauwens A, Peters G & Karch H (2011) Characterisation
of the Escherichia coli strain associated with an outbreak
of haemolytic uraemic syndrome in Germany, 2011:
a microbiological study. Lancet Infect Dis 11:
671–676.
Blokesch M (2012) Chitin colonization, chitin degradation and
chitin-induced natural competence of Vibrio cholerae are
subject to catabolite repression. Environ Microbiol 14:
1898–1912.
Blokesch M & Schoolnik GK (2007) Serogroup conversion of
Vibrio cholerae in aquatic reservoirs. PLoS Pathog 3: e81.
Blokesch M & Schoolnik GK (2008) The extracellular nuclease
Dns and its role in natural transformation of Vibrio
cholerae. J Bacteriol 190: 7232–7240.
Boucher Y, Nesbo CL & Doolittle WF (2001) Microbial genomes:
dealing with diversity. Curr Opin Microbiol 4: 285–289.
Boucher Y, Douady CJ, Papke RT, Walsh DA, Boudreau ME,
Nesbo CL, Case RJ & Doolittle WF (2003) Lateral gene
transfer and the origins of prokaryotic groups. Annu Rev
Genet 37: 283–328.
Brumbley SM, Carney BF & Denny TP (1993) Phenotype
conversion in Pseudomonas solanacearum due to
spontaneous inactivation of PhcA, a putative LysR
transcriptional regulator. J Bacteriol 175: 5477–5487.
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
P. Seitz & M. Blokesch
Burton B & Dubnau D (2010) Membrane-associated DNA
transport machines. Cold Spring Harb Perspect Biol 2:
a000406.
Cambray G, Guerout AM & Mazel D (2010) Integrons. Annu
Rev Genet 44: 141–166.
Cameron AD & Redfield RJ (2006) Non-canonical CRP sites
control competence regulons in Escherichia coli and many
other gamma-proteobacteria. Nucleic Acids Res 34: 6001–6014.
Cameron AD & Redfield RJ (2008) CRP binding and
transcription activation at CRP-S sites. J Mol Biol 383:
313–323.
Cameron AD, Volar M, Bannister LA & Redfield RJ (2008)
RNA secondary structure regulates the translation of sxy
and competence development in Haemophilus influenzae.
Nucleic Acids Res 36: 10–20.
Carlson CA, Pierson LS, Rosen JJ & Ingraham JL (1983)
Pseudomonas stutzeri and related species undergo natural
transformation. J Bacteriol 153: 93–99.
Carr EL, Kampfer P, Patel BK, Gurtler V & Seviour RJ (2003)
Seven novel species of Acinetobacter isolated from activated
sludge. Int J Syst Evol Microbiol 53: 953–963.
Catlin BW & Cunningham LS (1961) Transforming activities
and base contents of deoxyribonucleate preparations from
various Neisseriae. J Gen Microbiol 26: 303–312.
Cazalet C, Rusniok C, Bruggemann H et al. (2004) Evidence in
the Legionella pneumophila genome for exploitation of host
cell functions and high genome plasticity. Nat Genet 36:
1165–1173.
Chandler MS (1992) The gene encoding cAMP receptor
protein is required for competence development in
Haemophilus influenzae Rd. P Natl Acad Sci USA 89:
1626–1630.
Charpentier X, Kay E, Schneider D & Shuman HA (2011)
Antibiotics and UV radiation induce competence for natural
transformation in Legionella pneumophila. J Bacteriol 193:
1114–1121.
Chatterjee S, Almeida RP & Lindow S (2008) Living in two
worlds: the plant and insect lifestyles of Xylella fastidiosa.
Annu Rev Phytopathol 46: 243–271.
Chen I & Dubnau D (2004) DNA uptake during bacterial
transformation. Nat Rev Microbiol 2: 241–249.
Chen X, Schauder S, Potier N, Van Dorsselaer A, Pelczer I,
Bassler BL & Hughson FM (2002) Structural identification
of a bacterial quorum-sensing signal containing boron.
Nature 415: 545–549.
Chen Y, Dai J, Morris JG Jr & Johnson JA (2010) Genetic
analysis of the capsule polysaccharide (K antigen) and
exopolysaccharide genes in pandemic Vibrio
parahaemolyticus O3:K6. BMC Microbiol 10: 274.
Chin CS, Sorenson J, Harris JB et al. (2010) The origin of
the Haitian cholera outbreak strain. N Engl J Med 364:
33–42.
Chun J, Grim CJ, Hasan NA et al. (2009) Comparative
genomics reveals mechanism for short-term and long-term
clonal transitions in pandemic Vibrio cholerae. P Natl Acad
Sci USA 106: 15442–15447.
FEMS Microbiol Rev 37 (2013) 336–363
Regulation of natural competence in Gram-negative bacteria
Claverys JP & Havarstein LS (2007) Cannibalism and
fratricide: mechanisms and raisons d’etre. Nat Rev Microbiol
5: 219–229.
Claverys JP, Prudhomme M, Mortier-Barriere I & Martin B
(2000) Adaptation to the environment: Streptococcus
pneumoniae, a paradigm for recombination-mediated
genetic plasticity? Mol Microbiol 35: 251–259.
Claverys JP, Prudhomme M & Martin B (2006) Induction of
competence regulons as a general response to stress in
gram-positive bacteria. Annu Rev Microbiol 60: 451–475.
Claverys JP, Martin B & Polard P (2009) The genetic
transformation machinery: composition, localization, and
mechanism. FEMS Microbiol Rev 33: 643–656.
Clough SJ, Lee KE, Schell MA & Denny TP (1997) A two-
component system in Ralstonia (Pseudomonas) solanacearum
modulates production of PhcA-regulated virulence factors in
response to 3-hydroxypalmitic acid methyl ester. J Bacteriol
179: 3639–3648.
Cohan FM & Koeppel AF (2008) The origins of ecological
diversity in prokaryotes. Curr Biol 18: R1024–R1034.
Colwell RR (1996) Global climate and infectious disease: the
cholera paradigm. Science 274: 2025–2031.
Coupat B, Chaumeille-Dole F, Fall S, Prior P, Simonet P, Nesme
X & Bertolla F (2008) Natural transformation in the Ralstonia
solanacearum species complex: number and size of DNA that
can be transferred. FEMS Microbiol Ecol 66: 14–24.
Coupat-Goutaland B, Bernillon D, Guidot A, Prior P, Nesme X
& Bertolla F (2011) Ralstonia solanacearum virulence
increased following large interstrain gene transfers by natural
transformation. Mol Plant Microbe Interact 24: 497–505.
Croucher NJ, Harris SR, Fraser C et al. (2011) Rapid
pneumococcal evolution in response to clinical
interventions. Science 331: 430–434.
Cruze JA, Singer JT & Finnerty WR (1979) Conditions for
quantitative transformation in Acinetobacter calcoaceticus.
Curr Microbiol 3: 129–132.
Danner DB, Deich RA, Sisco KL & Smith HO (1980) An
eleven-base-pair sequence determines the specificity of DNA
uptake in Haemophilus transformation. Gene 11: 311–318.
Davidsen T, Rodland EA, Lagesen K, Seeberg E, Rognes T &
Tonjum T (2004) Biased distribution of DNA uptake
sequences towards genome maintenance genes. Nucleic Acids
Res 32: 1050–1058.
Davidsen T, Koomey M & Tonjum T (2007) Microbial
genome dynamics in CNS pathogenesis. Neuroscience 145:
1375–1387.
Demaneche S, Kay E, Gourbiere F & Simonet P (2001)
Natural transformation of Pseudomonas fluorescens and
Agrobacterium tumefaciens in soil. Appl Environ Microbiol
67: 2617–2621.
Deutscher J, Francke C & Postma PW (2006) How
phosphotransferase system-related protein phosphorylation
regulates carbohydrate metabolism in bacteria. Microbiol
Mol Biol Rev 70: 939–1031.
Dillard JP & Seifert HS (2001) A variable genetic island
specific for Neisseria gonorrhoeae is involved in providing
FEMS Microbiol Rev 37 (2013) 336–363
357
DNA for natural transformation and is found more
often in disseminated infection isolates. Mol Microbiol 41:
263–277.
Dobrindt U, Hochhut B, Hentschel U & Hacker J (2004)
Genomic islands in pathogenic and environmental
microorganisms. Nat Rev Microbiol 2: 414–424.
Dobrindt U, Chowdary MG, Krumbholz G & Hacker J (2010)
Genome dynamics and its impact on evolution of
Escherichia coli. Med Microbiol Immunol 199: 145–154.
Dorer MS, Fero J & Salama NR (2010) DNA damage triggers
genetic exchange in Helicobacter pylori. PLoS Pathog 6:
e1001026.
Dorer MS, Sessler TH & Salama NR (2011) Recombination
and DNA repair in Helicobacter pylori. Annu Rev Microbiol
65: 329–348.
Dorocicz IR, Williams PM & Redfield RJ (1993) The
Haemophilus influenzae adenylate cyclase gene: cloning,
sequence, and essential role in competence. J Bacteriol 175:
7142–7149.
Draskovic I & Dubnau D (2005) Biogenesis of a putative
channel protein, ComEC, required for DNA uptake:
membrane topology, oligomerization and formation of
disulphide bonds. Mol Microbiol 55: 881–896.
Dubnau D (1999) DNA uptake in bacteria. Annu Rev
Microbiol 53: 217–244.
Duffin PM & Seifert HS (2010) DNA uptake sequence-
mediated enhancement of transformation in Neisseria
gonorrhoeae is strain dependent. J Bacteriol 192:
4436–4444.
Duffin PM & Seifert HS (2012) Genetic transformation of
Neisseria gonorrhoeae shows a strand preference. FEMS
Microbiol Lett 334: 44–48.
Dziejman M, Balon E, Boyd D, Fraser CM, Heidelberg JF &
Mekalanos JJ (2002) Comparative genomic analysis of
Vibrio cholerae: genes that correlate with cholera
endemic and pandemic disease. P Natl Acad Sci USA 99:
1556–1561.
Dziejman M, Serruto D, Tam VC et al. (2005) Genomic
characterization of non-O1, non-O139 Vibrio cholerae
reveals genes for a type III secretion system. P Natl Acad Sci
USA 102: 3465–3470.
Eisen JA (2000) Horizontal gene transfer among microbial
genomes: new insights from complete genome analysis. Curr
Opin Genet Dev 10: 606–611.
Elkins C, Thomas CE, Seifert HS & Sparling PF (1991)
Species-specific uptake of DNA by gonococci is mediated by
a 10-base-pair sequence. J Bacteriol 173: 3911–3913.
Engelmoer DJ & Rozen DE (2011) Competence increases
survival during stress in Streptococcus pneumoniae. Evolution
65: 3475–3485.
Erken M, Weitere M, Kjelleberg S & McDougald D (2011) In
situ grazing resistance of Vibrio cholerae in the marine
environment. FEMS Microbiol Ecol 76: 504–512.
Falush D, Kraft C, Taylor NS, Correa P, Fox JG, Achtman M
& Suerbaum S (2001) Recombination and mutation during
long-term gastric colonization by Helicobacter pylori:
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
358
estimates of clock rates, recombination size, and minimal
age. P Natl Acad Sci USA 98: 15056–15061.
Falush D, Wirth T, Linz B et al. (2003) Traces of human
migrations in Helicobacter pylori populations. Science 299:
1582–1585.
Faruque SM, Naser IB, Islam MJ, Faruque AS, Ghosh AN,
Nair GB, Sack DA & Mekalanos JJ (2005) Seasonal
epidemics of cholera inversely correlate with the prevalence
of environmental cholera phages. Proc Natl Acad Sci USA
102: 1702–1707.
Feil E, Zhou J, Maynard Smith J & Spratt BG (1996) A
comparison of the nucleotide sequences of the adk and recA
genes of pathogenic and commensal Neisseria species:
evidence for extensive interspecies recombination within
adk. J Mol Evol 43: 631–640.
Feil EJ, Maiden MCJ, Achtman M & Spratt BG (1999) The
relative contributions of recombination and mutation to the
divergence of clones of Neisseria meningitidis. Mol Biol Evol
16: 1496–1502.
Finkel SE & Kolter R (2001) DNA as a nutrient: novel role for
bacterial competence gene homologs. J Bacteriol 183: 6288–
6293.
Fitzmaurice WP, Benjamin RC, Huang PC & Scocca JJ (1984)
Characterization of recognition sites on bacteriophage
HP1c1 DNA which interact with the DNA uptake system
of Haemophilus influenzae Rd. Gene 31: 187
–196.
Flavier AB, Clough SJ, Schell MA & Denny TP (1997)
Identification of 3-hydroxypalmitic acid methyl ester as a
novel autoregulator controlling virulence in Ralstonia
solanacearum. Mol Microbiol 26: 251–259.
Focareta T & Manning PA (1987) Extracellular proteins of
Vibrio cholerae: molecular cloning, nucleotide sequence and
characterization of the deoxyribonuclease (DNase) together
with its periplasmic localization in Escherichia coli K-12.
Gene 53: 31–40.
Frischer ME, Thurmond JM & Paul JH (1990) Natural
plasmid transformation in a high-frequency-of-
transformation marine Vibrio strain. Appl Environ Microbiol
56: 3439–3444.
Frischer ME, Thurmond JM & Paul JH (1993) Factors
affecting competence in a high-frequency of transformation
marine Vibrio. J Gen Microbiol 139: 753–761.
Frischer ME, Williams HG, Bennison B, Drake GR, Balkwill
DL & Paul JH (1996) The naturally transformable marine
bacterium WJT-1C formally identified as “Vibrio” is a
pseudomonad. Curr Microbiol 33: 287–291.
Fullner KJ & Mekalanos JJ (1999) Genetic characterization of a
new type IV-A pilus gene cluster found in both classical
and El Tor biotypes of Vibrio cholerae. Infect Immun 67:
1393–1404.
Gaasbeek EJ, Wagenaar JA, Guilhabert MR, Wosten MM, van
Putten JP, van der Graaf-van Bloois L, Parker CT & van der
Wal FJ (2009) A DNase encoded by integrated element
CJIE1 inhibits natural transformation of Campylobacter
jejuni. J Bacteriol 191: 2296–2306.
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
P. Seitz & M. Blokesch
Gaasbeek EJ, Wagenaar JA, Guilhabert MR, van Putten JP,
Parker CT & van der Wal FJ (2010) Nucleases encoded
by the integrated elements CJIE2 and CJIE4 inhibit
natural transformation of Campylobacter jejuni. J Bacteriol
192: 936–941.
Genin S (2010) Molecular traits controlling host range and
adaptation to plants in Ralstonia solanacearum. New Phytol
187: 920–928.
Genin S & Boucher C (2004) Lessons learned from the
genome analysis of Ralstonia solanacearum. Annu Rev
Phytopathol 42: 107–134.
Gilmore MS & Haas W (2005) The selective advantage of
microbial fratricide. P Natl Acad Sci USA 102: 8401–
8402.
Gomez-Valero L, Rusniok C, Jarraud S, Vacherie B, Rouy Z,
Barbe V, Medigue C, Etienne J & Buchrieser C (2011)
Extensive recombination events and horizontal gene transfer
shaped the Legionella pneumophila genomes. BMC Genomics
12: 536.
Goodman SD & Scocca JJ (1988) Identification and
arrangement of the DNA sequence recognized in specific
transformation of Neisseria gonorrhoeae. P Natl Acad Sci
USA 85: 6982–6986.
Goodman SD & Scocca JJ (1991) Factors influencing the
specific interaction of Neisseria gonorrhoeae with
transforming DNA. J Bacteriol 173: 5921–5923.
Grad YH, Lipsitch M, Feldgarden M et al. (2012) Genomic
epidemiology of the Escherichia coli O104:H4 outbreaks in
Europe, 2011. P Natl Acad Sci USA 109: 3065–3070.
Graupner S, Frey V, Hashemi R, Lorenz MG, Brandes G &
Wackernagel W (2000) Type IV pilus genes pilA and pilC of
Pseudomonas stutzeri are required for natural genetic
transformation, and pilA can be replaced by corresponding
genes from nontransformable species. J Bacteriol 182: 2184–
2190.
Griffith F (1928) The significance of pneumococcal types. J
Hyg 27: 113–159.
Guiral S, Mitchell TJ, Martin B & Claverys JP (2005)
Competence-programmed predation of noncompetent cells
in the human pathogen Streptococcus pneumoniae: genetic
requirements. P Natl Acad Sci USA 102: 8710–8715.
Gulig PA, Tucker MS, Thiaville PC, Joseph JL & Brown RN
(2009) USER friendly cloning coupled with chitin-based
natural transformation enables rapid mutagenesis of Vibrio
vulnificus. Appl Environ Microbiol 75: 4936–4949.
Hacker J, Blum-Oehler G, Mühldorfer I & Tschäpe H (1997)
Pathogenicity islands of virulent bacteria: structure,
function and impact on microbial evolution. Mol Microbiol
23: 1089–1097.
Hacker J, Hentschel U & Dobrindt U (2003) Prokaryotic
chromosomes and disease. Science 301: 790–793.
Hamilton HL & Dillard JP (2006) Natural transformation of
Neisseria gonorrhoeae: from DNA donation to homologous
recombination. Mol Microbiol 59: 376–385.
Hamilton HL, Dominguez NM, Schwartz KJ, Hackett KT &
Dillard JP (2005) Neisseria gonorrhoeae secretes
FEMS Microbiol Rev 37 (2013) 336–363
Regulation of natural competence in Gram-negative bacteria
chromosomal DNA via a novel type IV secretion system.
Mol Microbiol 55: 1704–1721.
Hanage WP, Fraser C & Spratt BG (2005) Fuzzy species
among recombinogenic bacteria. BMC Biol 3: 6.
Havarstein LS, Martin B, Johnsborg O, Granadel C & Claverys
JP (2006) New insights into the pneumococcal fratricide:
relationship to clumping and identification of a novel
immunity factor. Mol Microbiol 59: 1297–1307.
Heidelberg JF, Eisen JA, Nelson WC et al. (2000) DNA
sequence of both chromosomes of the cholera pathogen
Vibrio cholerae. Nature 406: 477–483.
Hendriksen RS, Price LB, Schupp JM et al. (2011) Population
genetics of Vibrio cholerae from Nepal in 2010: evidence on
the origin of the Haitian outbreak. mBio 2: e00157–11.
Herriott RM, Meyer EM & Vogt M (1970) Defined nongrowth
media for stage II development of competence in
Haemophilus influenzae. J Bacteriol 101: 517–524.
Herzberg C, Friedrich A & Averhoff B (2000) comB, a novel
competence gene required for natural transformation of
Acinetobacter sp. BD413: identification, characterization, and
analysis of growth-phase-dependent regulation. Arch
Microbiol 173: 220–228.
Higgins DA, Pomianek ME, Kraml CM, Taylor RK,
Semmelhack MF & Bassler BL (2007) The major Vibrio
cholerae autoinducer and its role in virulence factor
production. Nature 450: 883–886.
Hobbs M & Mattick JS (1993) Common components in the
assembly of type 4 fimbriae, DNA transfer systems,
filamentous phage and protein-secretion apparatus: a
general system for the formation of surface-associated
protein complexes. Mol Microbiol 10: 233–243.
Hofreuter D, Odenbreit S, Henke G & Haas R (1998) Natural
competence for DNA transformation in Helicobacter pylori:
identification and genetic characterization of the comB
locus. Mol Microbiol 28: 1027–1038.
Hofreuter D, Odenbreit S, Puls J, Schwan D & Haas R (2000)
Genetic competence in Helicobacter pylori: mechanisms and
biological implications. Res Microbiol 151: 487–491.
Humbert O & Salama NR (2008) The Helicobacter pylori
HpyAXII restriction-modification system limits exogenous
DNA uptake by targeting GTAC sites but shows asymmetric
conservation of the DNA methyltransferase and restriction
endonuclease components. Nucleic Acids Res 36: 6893–6906.
Humbert O, Dorer MS & Salama NR (2011) Characterization
of Helicobacter pylori factors that control transformation
frequency and integration length during inter-strain DNA
recombination. Mol Microbiol 79: 387–401.
Hunt DE, Gevers D, Vahora NM & Polz MF (2008)
Conservation of the chitin utilization pathway in the
Vibrionaceae. Appl Environ Microbiol 74: 44–51.
Israel DA, Lou AS & Blaser MJ (2000) Characteristics of
Helicobacter pylori natural transformation. FEMS Microbiol
Lett 186: 275–280.
Jeffrey WH, Paul JH & Stewart GJ (1990) Natural
transformation of a marine Vibrio species by plasmid DNA.
Microb Ecol 19: 259–268.
FEMS Microbiol Rev 37 (2013) 336–363
359
Johnsen PJ, Dubnau D & Levin BR (2009) Episodic selection
and the maintenance of competence and natural
transformation in Bacillus subtilis. Genetics 181: 1521–1533.
Johnson ZL, Cheong CG & Lee SY (2012) Crystal structure of
a concentrative nucleoside transporter from Vibrio cholerae
at 2.4 A. Nature 483: 489–493.
Juni E (1978) Genetics and physiology of Acinetobacter. Annu
Rev Microbiol 32: 349–371.
Juni E & Janik A (1969) Transformation of Acinetobacter calco-
aceticus (Bacterium anitratum). J Bacteriol 98: 281–288.
Kang Y, Liu H, Genin S, Schell MA & Denny TP (2002)
Ralstonia solanacearum requires type 4 pili to adhere to
multiple surfaces and for natural transformation and
virulence. Mol Microbiol 46: 427–437.
Karnholz A, Hoefler C, Odenbreit S, Fischer W, Hofreuter D
& Haas R (2006) Functional and topological
characterization of novel components of the comB DNA
transformation competence system in Helicobacter pylori.
J Bacteriol 188: 882–893.
Karudapuram S & Barcak GJ (1997) The Haemophilus
influenzae dprABC genes constitute a competence-inducible
operon that requires the product of the tfoX (sxy) gene for
transcriptional activation. J Bacteriol 179: 4815–4820.
Kersulyte D, Chalkauskas H & Berg DE (1999) Emergence of
recombinant strains of Helicobacter pylori during human
infection. Mol Microbiol 31: 31–43.
Kessler C & Manta V (1990) Specificity of restriction
endonucleases and DNA modification methyltransferases a
review. Gene 92: 1–248.
Keyhani NO & Roseman S (1999) Physiological aspects of
chitin catabolism in marine bacteria. Biochim Biophys Acta
1473: 108–122.
Killiny N, Prado SS & Almeida RP (2010) Chitin utilization by
the insect-transmitted bacterium Xylella fastidiosa. Appl
Environ Microbiol 76: 6134–6140.
Koomey M (1998) Competence for natural transformation in
Neisseria gonorrhoeae: a model system for studies of
horizontal gene transfer. APMIS Suppl 84: 56–61.
Koonin EV & Wolf YI (2008) Genomics of bacteria and
archaea: the emerging dynamic view of the prokaryotic
world. Nucleic Acids Res 36: 6688–6719.
Koonin EV, Makarova KS & Aravind L (2001) Horizontal gene
transfer in prokaryotes: quantification and classification.
Annu Rev Microbiol 55: 709–742.
Koyama Y, Hoshino T, Tomizuka N & Furukawa K (1986)
Genetic transformation of the extreme thermophile Thermus
thermophilus and of other Thermus spp. J Bacteriol 166:
338–340.
Kroll JS, Wilks KE, Farrant JL & Langford PR (1998)
Natural genetic exchange between Haemophilus and
Neisseria: intergeneric transfer of chromosomal genes
between major human pathogens. P Natl Acad Sci USA
95: 12381–12385.
Kru ger NJ & Stingl K (2011) Two steps away from novelty–
principles of bacterial DNA uptake. Mol Microbiol 80: 860–
867.
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
360
Kung SH & Almeida RP (2011) Natural competence and
recombination in the plant pathogen Xylella fastidiosa. Appl
Environ Microbiol 77: 5278–5284.
Lalucat J, Bennasar A, Bosch R, Garcia-Valdes E & Palleroni
NJ (2006) Biology of Pseudomonas stutzeri. Microbiol Mol
Biol Rev 70: 510–547.
Langenberg W, Rauws EAJ, Widjojokusumo A, Tytgat GNJ &
Zanen HC (1986) Identification of Campylobacter pyloridis
isolates by restriction endonuclease DNA analysis. J Clin
Microbiol 24: 414–417.
Li X & Roseman S (2004) The chitinolytic cascade in Vibrios is
regulated by chitin oligosaccharides and a two-component
chitin catabolic sensor/kinase. P Natl Acad Sci USA 101:
627–631.
Liang W, Pascual-Montano A, Silva AJ & Benitez JA (2007)
The cyclic AMP receptor protein modulates quorum
sensing, motility and multiple genes that affect intestinal
colonization in Vibrio cholerae. Microbiology 153: 2964–
2975.
Liang W, Sultan SZ, Silva AJ & Benitez JA (2008) Cyclic
AMP post-transcriptionally regulates the biosynthesis of a
major bacterial autoinducer to modulate the cell density
required to activate quorum sensing. FEBS Lett 582:
3744–3750.
Linz B, Balloux F, Moodley Y et al. (2007) An African origin
for the intimate association between humans and
Helicobacter pylori. Nature 445: 915–918.
Lipp EK, Huq A & Colwell RR (2002) Effects of global climate
on infectious disease: the cholera model. Clin Microbiol Rev
15: 757–770.
Lo Scrudato M & Blokesch M (2012) The regulatory network
of natural competence and transformation of Vibrio
cholerae. PLoS Genet 8: e1002778.
Lobitz B, Beck L, Huq A, Wood B, Fuchs G, Faruque AS &
Colwell R (2000) Climate and infectious disease: use of
remote sensing for detection of Vibrio cholerae by indirect
measurement. P Natl Acad Sci USA 97: 1438–1443.
Lorenz MG & Sikorski J (2000) The potential for intraspecific
horizontal gene exchange by natural genetic transformation:
sexual isolation among genomovars of Pseudomonas stutzeri.
Microbiology 146(Pt 12): 3081–3090.
Lorenz MG & Wackernagel W (1990) Natural genetic
transformation of Pseudomonas stutzeri by sand-adsorbed
DNA. Arch Microbiol 154: 380–385.
Lorenz MG & Wackernagel W (1991) High frequency of
natural genetic transformation of Pseudomonas stutzeri in
soil extract supplemented with a carbon/energy and
phosphorus source. Appl Environ Microbiol 57: 1246–
1251.
Lorenz MG & Wackernagel W (1994) Bacterial gene transfer
by natural genetic transformation in the environment.
Microbiol Rev 58: 563–602.
Lorenz MG, Gerjets D & Wackernagel W (1991) Release of
transforming plasmid and chromosomal DNA from two
cultured soil bacteria. Arch Microbiol 156: 319–326.
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
P. Seitz & M. Blokesch
Lorenz MG, Reipschlager K & Wackernagel W (1992) Plasmid
transformation of naturally competent Acinetobacter
calcoaceticus in non-sterile soil extract and groundwater.
Arch Microbiol 157: 355–360.
Macfadyen LP (2000) Regulation of competence development
in Haemophilus influenzae. J Theor Biol 207: 349–359.
MacFadyen LP, Chen D, Vo HC, Liao D, Sinotte R & Redfield
RJ (2001) Competence development by Haemophilus
influenzae is regulated by the availability of nucleic acid
precursors. Mol Microbiol 40: 700–707.
MacLean RC, Hall AR, Perron GG & Buckling A (2010) The
population genetics of antibiotic resistance: integrating
molecular mechanisms and treatment contexts. Nat Rev
Genet 11: 405–414.
Maiden MC (2008) Population genomics: diversity and
virulence in the Neisseria. Curr Opin Microbiol 11: 467–
471.
Maier B (2005) Using laser tweezers to measure twitching
motility in Neisseria. Curr Opin Microbiol 8: 344–349.
Maier B, Potter L, So M, Long CD, Seifert HS & Sheetz MP
(2002) Single pilus motor forces exceed 100 pN. P Natl
Acad Sci USA 99: 16012–16017.
Maier B, Chen I, Dubnau D & Sheetz MP (2004) DNA
transport into Bacillus subtilis requires proton motive force
to generate large molecular forces. Nat Struct Mol Biol 11:
643–649.
Majewski SI & Goodwin CS (1988) Restriction endonuclease
analysis of the genome of Campylobacter pylori with a rapid
extraction method: evidence for considerable genomic
variation. J Infect Dis 157: 465–471.
Marsich E, Zuccato P, Rizzi S, Vetere A, Tonin E & Paoletti S
(2002) Helicobacter pylori expresses an autolytic enzyme:
gene identification, cloning, and theoretical protein
structure. J Bacteriol 184: 6270–6279.
Matz C, McDougald D, Moreno AM, Yung PY, Yildiz FH &
Kjelleberg S (2005) Biofilm formation and phenotypic
variation enhance predation-driven persistence of Vibrio
cholerae. P Natl Acad Sci USA 102: 16819–16824.
Maughan H & Redfield RJ (2009) Tracing the evolution of
competence in Haemophilus influenzae. PLoS ONE 4:
e5854.
Meibom KL, Li XB, Nielsen AT, Wu CY, Roseman S &
Schoolnik GK (2004) The Vibrio cholerae chitin utilization
program. P Natl Acad Sci USA 101: 2524–2529.
Meibom KL, Blokesch M, Dolganov NA, Wu C-Y & Schoolnik
GK (2005) Chitin induces natural competence in Vibrio
cholerae. Science 310: 1824–1827.
Meier P & Wackernagel W (2003) Mechanisms of homology-
facilitated illegitimate recombination for foreign DNA
acquisition in transformable Pseudomonas stutzeri. Mol
Microbiol 48: 1107–1118.
Meier P, Berndt C, Weger N & Wackernagel W (2002) Natural
transformation of Pseudomonas stutzeri by single-stranded
DNA requires type IV pili, competence state and comA.
FEMS Microbiol Lett 207: 75–80.
FEMS Microbiol Rev 37 (2013) 336–363
Regulation of natural competence in Gram-negative bacteria
Merz AJ, So M & Sheetz MP (2000) Pilus retraction powers
bacterial twitching motility. Nature 407: 98–102.
Michod RE, Wojciechowski MF & Hoelzer MA (1988) DNA
repair and the evolution of transformation in the bacterium
Bacillus subtilis. Genetics 118: 31–39.
Michod RE, Bernstein H & Nedelcu AM (2008) Adaptive
value of sex in microbial pathogens. Infect Genet Evol 8:
267–285.
Miller MC, Keymer DP, Avelar A, Boehm AB & Schoolnik GK
(2007) Detection and transformation of genome segments
that differ within a coastal population of Vibrio cholerae
strains. Appl Environ Microbiol 73: 3695–3704.
Milton DL (2006) Quorum sensing in Vibrios: complexity for
diversification. Int J Med Microbiol 296: 61–71.
Morelli G, Didelot X, Kusecek B, Schwarz S, Bahlawane C,
Falush D, Suerbaum S & Achtman M (2010)
Microevolution of Helicobacter pylori during prolonged
infection of single hosts and within families. PLoS Genet 6:
e1001036.
Mortier-Barriere I, Velten M, Dupaigne P et al. (2007) A key
presynaptic role in transformation for a widespread bacterial
protein: DprA conveys incoming ssDNA to RecA. Cell 130:
824–836.
Munoz-Price LS & Weinstein RA (2008) Acinetobacter
infection. N Engl J Med 358: 1271–1281.
Murphy TF & Brauer AL (2011) Expression of urease by
Haemophilus influenzae during human respiratory tract
infection and role in survival in an acid environment. BMC
Microbiol 11: 183.
Mutreja A, Kim DW, Thomson NR et al. (2011) Evidence for
several waves of global transmission in the seventh cholera
pandemic. Nature 477: 462–465.
Neiman J, Guo Y & Rowe-Magnus DA (2011) Chitin-induced
carbotype conversion in Vibrio vulnificus. Infect Immun 79:
3195–3203.
Newland JW, Green BA, Foulds J & Holmes RK (1985)
Cloning of extracellular DNase and construction of a
DNase-negative strain of Vibrio cholerae. Infect Immun 47:
691–696.
Ng WL & Bassler BL (2009) Bacterial quorum-sensing network
architectures. Annu Rev Genet 43: 197–222.
Ng WL, Perez LJ, Wei Y, Kraml C, Semmelhack MF & Bassler
BL (2011) Signal production and detection specificity in
Vibrio CqsA/CqsS quorum-sensing systems. Mol Microbiol
79: 1407–1417.
Nielsen KM, Bones AM & Van Elsas JD (1997) Induced
natural transformation of Acinetobacter calcoaceticus in soil
microcosms. Appl Environ Microbiol 63: 3972–3977.
Ochman H, Lawrence JG & Groisman EA (2000) Lateral gene
transfer and the nature of bacterial innovation. Nature 405:
299–304.
Odenbreit S, Puls J, Sedlmaier B, Gerland E, Fischer W &
Haas R (2000) Translocation of Helicobacter pylori CagA
into gastric epithelial cells by type IV secretion. Science 287:
1497–1500.
FEMS Microbiol Rev 37 (2013) 336–363
361
Omelchenko MV, Wolf YI, Gaidamakova EK, Matrosova VY,
Vasilenko A, Zhai M, Daly MJ, Koonin EV & Makarova KS
(2005) Comparative genomics of Thermus thermophilus and
Deinococcus radiodurans: divergent routes of adaptation to
thermophily and radiation resistance. BMC Evol Biol 5: 57.
Palchevskiy V & Finkel SE (2006) Escherichia coli competence
gene homologs are essential for competitive fitness and the
use of DNA as a nutrient. J Bacteriol 188: 3902–3910.
Palmen R & Hellingwerf KJ (1997) Uptake and processing of
DNA by Acinetobacter calcoaceticus–a review. Gene 192:
179–190.
Palmen R, Vosman B, Buijsman P, Breek CK & Hellingwerf KJ
(1993) Physiological characterization of natural
transformation in Acinetobacter calcoaceticus. J Gen Microbiol
139: 295–305.
Palmen R, Buijsman P & Hellingwerf KJ (1994) Physiological
regulation of competence induction for natural
transformation in Acinetobacter calcoaceticus. Arch Microbiol
162: 344–351.
Paul JH, Thurmond JM, Frischer ME & Cannon JP (1992)
Intergeneric natural plasmid transformation between E. coli
and a marine Vibrio species. Mol Ecol 1: 37–46.
Peix A, Ramirez-Bahena MH & Velazquez E (2009) Historical
evolution and current status of the taxonomy of genus
Pseudomonas. Infect Genet Evol 9: 1132–1147.
Perron GG, Lee AE, Wang Y, Huang WE & Barraclough TG
(2012) Bacterial recombination promotes the evolution of
multi-drug-resistance in functionally diverse populations.
Proc Biol Sci 279: 1477–1484.
Pollack-Berti A, Wollenberg MS & Ruby EG (2010) Natural
transformation of Vibrio fischeri requires tfoX and tfoY.
Environ Microbiol 12: 2302–2311.
Pontiroli A, Rizzi A, Simonet P, Daffonchio D, Vogel TM &
Monier JM (2009) Visual evidence of horizontal gene
transfer between plants and bacteria in the phytosphere of
transplastomic tobacco. Appl Environ Microbiol 75: 3314–
3322.
Popa O & Dagan T (2011) Trends and barriers to lateral gene
transfer in prokaryotes. Curr Opin Microbiol 14: 615–623.
Porstendörfer D, Gohl O, Mayer F & Averhoff B (2000)
ComP, a pilin-like protein essential for natural competence
in Acinetobacter sp. Strain BD413: regulation, modification,
and cellular localization. J Bacteriol 182: 3673–3680.
Prudhomme M, Attaiech L, Sanchez G, Martin B & Claverys
JP (2006) Antibiotic stress induces genetic transformability
in the human pathogen Streptococcus pneumoniae. Science
313: 89–92.
Pruzzo C, Vezzulli L & Colwell RR (2008) Global impact of
Vibrio cholerae interactions with chitin. Environ Microbiol
10: 1400–1410.
Pugsley AP (1993) The complete general secretory pathway in
gram-negative bacteria. Microbiol Rev 57: 50–108.
Rasko DA, Webster DR, Sahl JW et al. (2011) Origins of the
E. coli strain causing an outbreak of hemolytic-uremic
syndrome in Germany. N Engl J Med 365: 709–717.
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
362
Redfield RJ (1988) Evolution of bacterial transformation: is sex
with dead cells ever better than no sex at all? Genetics 119:
213–221.
Redfield RJ (1991) sxy-1, a Haemophilus influenzae mutation
causing greatly enhanced spontaneous competence.
J Bacteriol 173: 5612–5618.
Redfield RJ (1993a) Evolution of natural transformation:
testing the DNA repair hypothesis in Bacillus subtilis and
Haemophilus influenzae. Genetics 133: 755–761.
Redfield RJ (1993b) Genes for breakfast: the have-your-cake-
and-eat-it-too of bacterial transformation. J Hered 84: 400–
404.
Redfield RJ (2001) Do bacteria have sex? Nat Rev Genet 2:
634–639.
Redfield RJ, Schrag MR & Dean AM (1997) The evolution of
bacterial transformation: sex with poor relations. Genetics
146: 27–38.
Redfield RJ, Cameron AD, Qian Q, Hinds J, Ali TR, Kroll JS
& Langford PR (2005) A novel CRP-dependent regulon
controls expression of competence genes in Haemophilus
influenzae. J Mol Biol 347: 735–747.
Salanoubat M, Genin S, Artiguenave F et al. (2002) Genome
sequence of the plant pathogen Ralstonia solanacearum.
Nature 415: 497–502.
Seper A, Fengler VH, Roier S, Wolinski H, Kohlwein SD,
Bishop AL, Camilli A, Reidl J & Schild S (2011)
Extracellular nucleases and extracellular DNA play
important roles in Vibrio cholerae biofilm formation. Mol
Microbiol 82: 1015–1037.
Sikorski J, Graupner S, Lorenz MG & Wackernagel W (1998)
Natural genetic transformation of Pseudomonas stutzeri in a
non-sterile soil. Microbiology 144: 569–576.
Sikorski J, Teschner N & Wackernagel W (2002) Highly
different levels of natural transformation are associated with
genomic subgroups within a local population of
Pseudomonas stutzeri from soil. Appl Environ Microbiol 68:
865–873.
Smith HO, Gwinn ML & Salzberg SL (1999) DNA uptake
signal sequences in naturally transformable bacteria. Res
Microbiol 150: 603–616.
Solomon JM & Grossman AD (1996) Who’s competent and
when: regulation of natural genetic competence in bacteria.
Trends Genet 12: 150–155.
Sparling PF (1966) Genetic transformation of Neisseria
gonorrhoeae to streptomycin resistance. J Bacteriol 92: 1364–
1371.
Spirig T, Tiaden A, Kiefer P, Buchrieser C, Vorholt JA & Hilbi
H (2008) The Legionella autoinducer synthase LqsA
produces an alpha-hydroxyketone signaling molecule. J Biol
Chem 283: 18113–18123.
Stewart GJ, Carlson CA & Ingraham JL (1983) Evidence for an
active role of donor cells in natural transformation of
Pseudomonas stutzeri. J Bacteriol 156: 30–35.
Stingl K, Muller S, Scheidgen-Kleyboldt G, Clausen M & Maier
B (2010) Composite system mediates two-step DNA uptake
into Helicobacter pylori. P Natl Acad Sci USA 107: 1184–1189.
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
P. Seitz & M. Blokesch
Stokes HW & Gillings MR (2011) Gene flow, mobile genetic
elements and the recruitment of antibiotic resistance genes
into Gram-negative pathogens. FEMS Microbiol Rev 35:
790–819.
Stone BJ & Kwaik YA (1999) Natural competence for DNA
transformation by Legionella pneumophila and its association
with expression of type IV pili. J Bacteriol 181: 1395–1402.
Suckow G, Seitz P & Blokesch M (2011) Quorum sensing
contributes to natural transformation of Vibrio cholerae in a
species-specific manner. J Bacteriol 193: 4914–4924.
Sudarsan N, Lee ER, Weinberg Z, Moy RH, Kim JN, Link KH
& Breaker RR (2008) Riboswitches in eubacteria sense the
second messenger cyclic di-GMP. Science 321: 411–413.
Suerbaum S & Josenhans C (2007) Helicobacter pylori
evolution and phenotypic diversification in a changing host.
Nat Rev Microbiol 5: 441–452.
Suerbaum S & Michetti P (2002) Helicobacter pylori infection.
N Engl J Med 347: 1175–1186.
Suerbaum S, Smith JM, Bapumia K, Morelli G, Smith NH,
Kunstmann E, Dyrek I & Achtman M (1998) Free
recombination within Helicobacter pylori. Proc Natl Acad Sci
USA 95: 12619–12624.
Tettelin H, Saunders NJ, Heidelberg J et al. (2000) Complete
genome sequence of Neisseria meningitidis serogroup B
strain MC58. Science 287: 1809–1815.
Tettelin H, Masignani V, Cieslewicz MJ et al. (2005) Genome
analysis of multiple pathogenic isolates of Streptococcus
agalactiae: implications for the microbial “pan-genome”.
P Natl Acad Sci USA 102: 13950–13955.
Thomas CM & Nielsen KM (2005) Mechanisms of, and
barriers to, horizontal gene transfer between bacteria. Nat
Rev Microbiol 3: 711–721.
Tribble GD, Rigney TW, Dao DH, Wong CT, Kerr JE, Taylor BE,
Pacha S & Kaplan HB (2012) Natural competence is a
major mechanism for horizontal DNA transfer in the oral
pathogen Porphyromonas gingivalis. mBio 3: e00231–11.
Udden SMN, Zahid MSH, Biswas K, Ahmad QS, Cravioto A,
Nair GB, Mekalanos JJ & Faruque SM (2008) Acquisition of
classical CTX prophage from Vibrio cholerae O141 by El Tor
strains aided by lytic phages and chitin-induced
competence. Proc Natl Acad Sci USA 105: 11951–11956.
Viscidi RP & Demma JC (2003) Genetic diversity of
Neisseria gonorrhoeae housekeeping genes. J Clin Microbiol
41: 197–204.
Vos M (2009) Why do bacteria engage in homologous
recombination? Trends Microbiol 17: 226–232.
Warskow AL & Juni E (1972) Nutritional requirements of
Acinetobacter strains isolated from soil, water, and sewage.
J Bacteriol 112: 1014–1016.
Wei Y, Perez LJ, Ng WL, Semmelhack MF & Bassler BL (2011)
Mechanism of Vibrio cholerae autoinducer-1 biosynthesis.
ACS Chem Biol 6: 356–365.
Weinberg Z, Barrick JE, Yao Z et al. (2007) Identification of
22 candidate structured RNAs in bacteria using the
CMfinder comparative genomics pipeline. Nucleic Acids Res
35: 4809–4819.
FEMS Microbiol Rev 37 (2013) 336–363
Regulation of natural competence in Gram-negative bacteria
Whitehead NA, Barnard AM, Slater H, Simpson NJ &
Salmond GP (2001) Quorum-sensing in Gram-negative
bacteria. FEMS Microbiol Rev 25: 365–404.
Wiedenbeck J & Cohan FM (2011) Origins of bacterial
diversity through horizontal genetic transfer and
adaptation to new ecological niches. FEMS Microbiol Rev
35: 957–976.
Wiesner RS, Hendrixson DR & DiRita VJ (2003) Natural
transformation of Campylobacter jejuni requires components
of a type II secretion system. J Bacteriol 185: 5408–5418.
Williams PM, Bannister LA & Redfield RJ (1994) The
Haemophilus influenzae sxy-1 mutation is in a newly
identified gene essential for competence. J Bacteriol 176:
6789–6794.
Wise EM Jr, Alexander SP & Powers M (1973) Adenosine
3′:5′-cyclic monophosphate as a regulator of bacterial
transformation. P Natl Acad Sci USA 70: 471–474.
Wojciechowski MF, Hoelzer MA & Michod RE (1989)
DNA repair and the evolution of transformation in
Bacillus subtilis. II. Role of inducible repair. Genetics 121:
411–422.
Woodhams KL, Benet ZL, Blonsky SE, Hackett KT & Dillard
JP (2012) Prevalence and detailed mapping of the
gonococcal genetic island in Neisseria meningitidis. J
Bacteriol 194: 2275–2285.
Xavier KB & Bassler BL (2003) LuxS quorum sensing: more
than just a numbers game. Curr Opin Microbiol 6: 191–197.
FEMS Microbiol Rev 37 (2013) 336–363
363
Yamamoto S, Morita M, Izumiya H & Watanabe H (2010)
Chitin disaccharide (GlcNAc)2 induces natural competence
in Vibrio cholerae through transcriptional and translational
activation of a positive regulatory gene tfoXVC. Gene 457:
42–49.
Yamamoto S, Izumiya H, Mitobe J, Morita M, Arakawa E,
Ohnishi M & Watanabe H (2011) Identification of a chitin-
induced small rna that regulates translation of the tfoX gene,
encoding a positive regulator of natural competence in
Vibrio cholerae. J Bacteriol 193: 1953–1965.
Yildiz FH & Visick KL (2009) Vibrio biofilms: so much
the same yet so different. Trends Microbiol 17: 109–
118.
Young DM, Parke D & Ornston LN (2005) Opportunities for
genetic investigation afforded by Acinetobacter baylyi, a
nutritionally versatile bacterial species that is highly
competent for natural transformation. Annu Rev Microbiol
59: 519–551.
Young KT, Davis LM & Dirita VJ (2007) Campylobacter jejuni:
molecular biology and pathogenesis. Nat Rev Microbiol 5:
665–679.
Zulty JJ & Barcak GJ (1995) Identification of a DNA
transformation gene required for com101A+ expression and
supertransformer phenotype in Haemophilus influenzae.
P Natl Acad Sci USA 92: 3616–3620.
ª 2012 Federation of European Microbiological Societies
Published by Blackwell Publishing Ltd. All rights reserved
JOURNAL OF BACTERIOLOGY, Nov. 2001, p. 6288–6293 Vol. 183, No. 21
0021-9193/01/$04.00⫹0 DOI: 10.1128/JB.183.21.6288–6293.2001
Copyright © 2001, American Society for Microbiology. All Rights Reserved.

DNA as a Nutrient: Novel Role for Bacterial

Competence Gene Homologs
STEVEN E. FINKEL1* AND ROBERTO KOLTER2
Department of Biological Sciences, University of Southern California, Los Angeles,
California 90089-1340,1 and Department of Microbiology and Molecular Genetics,
Harvard Medical School, Boston, Massachusetts 021152
Received 5 April 2001/Accepted 7 August 2001

The uptake and stable maintenance of extracellular DNA, genetic transformation, is universally recognized
as a major force in microbial evolution. We show here that extracellular DNA, both homospecific and
heterospecific, can also serve as the sole source of carbon and energy supporting microbial growth. Mutants

Downloaded from http://jb.asm.org/ on August 12, 2014 by guest

unable to consume DNA suffer a significant loss of fitness during stationary-phase competition. In Escherichia
coli, the use of DNA as a nutrient depends on homologs of proteins involved in natural genetic competence and
transformation in Haemophilus influenzae and Neisseria gonorrhoeae. Homologs of these E. coli genes are present
in many members of the ␥ subclass of Proteobacteria, suggesting that the mechanisms for consumption of DNA
may have been widely conserved during evolution.

Horizontal gene transfer in bacteria can occur between or-
ganisms of the same or different species via one of three mech-
anisms: conjugation, transduction, or transformation (7). The
last mechanism relies on the cell being able to take up and
stably maintain extracellular DNA. Many bacteria are “natu-
rally competent” for genetic transformation and, at least under
some environmental conditions, can take up and integrate ex-
tracellular DNA. The mechanisms of DNA uptake in several
naturally competent gram-negative and gram-positive bacteria
have been extensively studied and reviewed (8, 9, 12, 14, 27, 29,
31). The process of transformation in naturally competent bac-
teria involves several steps. First, double-stranded DNA
(dsDNA) is bound to the surface of the cell and enters a
compartment where it becomes resistant to exogenous nucle-
ase. Next, one strand of the DNA enters the cytoplasm while
the other strand is degraded (9). Some bacteria, such as Hae-
mophilus influenzae and Neisseria gonorrhoeae, preferentially
take up homospecific DNA. Specificity of DNA uptake in these
organisms is determined by the presence of “uptake signal
sequences,” which are overrepresented conserved sequences
found throughout the genome (28). Finally, after a recombi-
nation event, the new DNA is integrated into the chromosome.
Natural competence and transformation have not been ob-
served to occur in many bacterial species, including Escherichia
coli. It has been proposed that natural competence, in addition
to playing a role in genetic recombination, might serve to allow
the use of extracellular DNA for a nutritional purpose (22, 24,
29, 30). That is, the uptake of DNA into the cell may have two
non-mutually exclusive functions: to provide DNA for genetic
transformation and to provide nutrients. While studying mech-
anisms of survival of E. coli during long-term starvation, we
identified a transposon insertion mutant that demonstrated an
inability to survive when competing with its wild-type parent
* Corresponding author. Mailing address: Department of Biological
Sciences, Program in Molecular Biology, SHS 172, University of
Southern California, Los Angeles, CA 90089-1340. Phone: (213) 821-
1498. Fax: (213) 740-8631. E-mail: sfinkel@usc.edu.
during extended stationary-phase incubation. The mutated
gene showed high homology with a putative competence gene
in H. influenzae. Since for some naturally transformable bac-
teria, genetic competence is induced during starvation, and
since no such natural induction of competence under standard
laboratory conditions has yet to be described for E. coli, we
chose to address whether the competitive disadvantage of this
mutant was due to an inability of that strain to compete for a
nutrient resource, namely, extracellular DNA.
MATERIALS AND METHODS
Transposon insertion mutagenesis and initial screen for stationary phase-
specific competition-defective mutants. All experiments were performed using
strains derived from E. coli K-12 strain ZK126 (W3110 ⌬lacU169 tna-2) (33).
ZK1142 is a nalidixic acid-resistant (Nalr) derivative of ZK126. Transposon
insertion mutagenesis of ZK126 using ␭NK1324 was performed as previously
described (16), resulting in a pool of mutant cells carrying a mini-Tn10d-Camr
insert conferring resistance to chloramphenicol. Using this transposon, a screen
for “stationary-phase-specific competition-defective” mutants was performed:
mutant candidates were identified after coculture at 37°C of individual transpo-
son insertion mutant strains with wild-type ZK1142 (Nalr) cells in 200 ␮l of
Luria-Bertani (LB) broth in 96-well microtiter plates. Both the Nalr allele (16)
and the presence of the chloramphenicol resistance (Camr) marker (S. Finkel
and R. Kolter, unpublished data) are neutral in the absence of drug selection.
Transposon insertions which resulted in the loss of mutant cells, as determined
by detecting Nalr cells and few or no Camr cells, after 5 days of competition were
then rescreened in 5-ml batch cultures (see below). Mutant candidates were
reconstructed by bacteriophage P1 transduction (19) and rescreened. The ZK126
hofQ::Tn10d-Camr mutant was kindly provided by L. Pratt and R. Kolter (un-
published data).
Batch culture competition assays. E. coli wild-type (ZK1142 Nalr) and mutant
(Camr) strains were each grown overnight in LB broth (reaching a density of
⬃5 ⫻ 109 CFU/ml.) Cultures were then inoculated 1:1,000,000 (vol/vol) into
fresh LB broth, either in coculture or alone. Cell titers were determined by serial
dilution on LB agar plates supplemented with nalidixic acid (20 ␮g/ml) or
chloramphenicol (30 ␮g/ml) as appropriate. The limit of detection of this titra-
tion method is ⬍100 CFU/ml.
DNA sequencing of transposon insertion sites. The DNA sequence of the
region flanking the transposon insertion was obtained using an arbitrary PCR-
based technique (4). The primers specific to the mini-Tn10d-Camr element were
primer 1L (CTGCCTCCCAGAGCCTG) and primer OUT 1L (CAGGCTCTC
CCCGTGGAGG).
Preparation of conditioned medium. Filter-sterilized conditioned medium was
prepared as follows. LB cultures (50 ml) were inoculated 1:1,000 (vol/vol) with

6288
VOL. 183, 2001 DNA AS A NUTRIENT 6289

cells from a fresh overnight culture of ZK126 and incubated for 5 days in 250-ml
Erlenmeyer flasks at 37°C with vigorous aeration. After 5 days, cells were pel-
leted and the supernatant was removed and filtered through a 0.2-␮m NYL filter
unit (Nalgene). It was essential that filters be rinsed with at least 100 ml of sterile
distilled water prior to use to ensure removal of trace contaminants on the filter
which are metabolizable by E. coli (data not shown). Supernatants treated with
DNase I were incubated at 37°C with 10 ␮g of DNase I (Sigma Chemical Co., St.
Louis, Mo.) per ml for 20 min prior to inoculation. Cultures were then inocu-
lated, and titers were determined after overnight incubation.
Preparation of minimal medium supplemented with purified DNA. M63 min-
imal medium (1⫻) was prepared as described (19) and supplemented with 1 mM
MgSO4 and 1 ␮g of vitamin B1 per ml. E. coli chromosomal DNA was prepared
as described previously (2). Isolated DNA was sonicated to an average length of
300 to 500 bp and extensively extracted with phenol, phenol-chloroform, chlo-
roform, and ethyl ether. DNA was then precipitated and reprecipitated with
ethanol and resuspended in sterile distilled water immediately before use. It was
essential to use freshly precipitated DNA, most likely because DNA stored for
long periods of time contained easily metabolized nucleotides or other break-
down products from the dsDNA. For the experiments in Fig. 3, chromosomal

Downloaded from http://jb.asm.org/ on August 12, 2014 by guest

DNA was added at a concentration of ⬃6 ␮g/ml. Cultures were inoculated and
titers were determined as described above.
DNA sequence analysis. DNA sequence similarity searches using the basic
BLAST, tBLASTn (1), and Microbial Genomes BLAST algorithms and open
reading frame searches using the ORF Finder program were performed at the
National Center for Biotechnology Information website (http://www.ncbi.nlm.ni-
h.gov/). Protein sequence alignments using the ClustalW 1.8, MAP, and PIMA
algorithms were performed at the Baylor College of Medicine Molecular Biology
website (http://www.hgsc.bcm.tmc.edu/SearchLauncher/).

RESULTS

Identification of the yhiR mutant. We performed a genetic

screen for stationary-phase-specific competition-defective mu-
tants; the analysis of one of those mutants is presented here.
The transposon insertion mutant exhibits no fitness loss during
the exponential phase of growth or upon entry into stationary
phase when cocultured with the wild-type parent. However,
after 2 days of coincubation in stationary phase, the mutant is
outcompeted by wild-type cells and is completely lost from the
culture after 10 to 12 days (Fig. 1A). Yet, when cultured
separately, the mutant and the wild-type strains show identical
patterns of exponential-phase growth and survival during long-
term stationary phase incubation (Fig. 1B).
Analysis of the DNA sequence flanking the insertion muta-
tion showed that the transposon inserted into yhiR, a gene of
previously unassigned function, located at 78.5 min on the E.
coli chromosome. Sequence comparisons identified this gene
as a homolog of the comJ (orfJ) gene of H. influenzae; the
predicted amino acid sequences of ComJ and YhiR are ⬃66%
identical and ⬃85% similar. comJ, located at chromosomal
nucleotide position 463327, was first identified as a gene adja-
cent to the H. influenzae competence locus, a cluster of genes
that mediate natural competence of this organism (32).
Though direct evidence that comJ plays a role in competence
is lacking, a deletion mutation removing part of comJ and an
adjacent gene causes a transformation-deficient phenotype in
H. influenzae (5). Since the yhiR gene appears to be present in
a single gene operon and transduction of the yhiR::Tn10d-
Camr mutation into fresh wild-type strains confers the compe-
tition-defective phenotype, we can assume that the effect of the
transposon insertion is due directly to the loss of YhiR activity.
Conditioned medium experiments indicate the use of DNA
as a nutrient. During long-term incubation in a rich medium
(LB broth), viable cell counts of E. coli ZK126 reach a plateau
of ⬃5 ⫻ 109 CFU per ml at the end of exponential phase, drop
FIG. 1. Survival patterns of yhiR and the wild-type (WT) parental
strain in the presence or absence of competition. LB cultures were
incubated for 12 days. (A) Cells grown in coculture; (B) cells grown
separately. The asterisk indicates no detectable counts (limit of detec-
tion of ⬍100 CFU/ml).
to ⬃5 ⫻ 107 CFU/ml after ⬃3 days (the death phase), and level
off at that density for many weeks (10, 33). We hypothesized
that during the death phase, DNA might be released from
dead cells into the medium and serve as a nutrient for the
minority of cells that are still alive. (It has already been shown
that amino acids are released for catabolic use, presumably
from dead cells, during this time [34].) To determine whether
DNA was available as a nutrient source and whether the yhiR
mutant is defective in the utilization of released DNA, we
prepared conditioned media by filter sterilization of 5-day-old
cultures of wild-type E. coli. When inoculated at ⬃5 ⫻ 103
CFU/ml, the wild-type cultures reach a density of ⬃8 ⫻ 106
CFU/ml after overnight incubation (Fig. 2). In contrast, yhiR
mutant cells reach a density of only ⬃6 ⫻ 105 CFU/ml, less
than 1/10 that of the wild type. Yet if the conditioned medium
is pretreated with DNase I for 20 min and then inoculated with
either wild-type or yhiR mutant cells, the growth yields are
identical to that of the wild type growing without added DNase
I. These results suggest that DNA is present in the conditioned
medium and is metabolized by wild-type E. coli but not by the
yhiR mutant.
6290 FINKEL AND KOLTER J. BACTERIOL.

Downloaded from http://jb.asm.org/ on August 12, 2014 by guest

FIG. 2. Growth yields for overnight cultures of wild-type (WT) or
yhiR cells incubated in conditioned medium. Culture medium was
prepared from 5-day-old LB cultures without or with DNase I treat-
ment prior to inoculation. Results are the averages of four experi- FIG. 3. Growth of E. coli utilizing DNA as the sole carbon source.
ments. Cells were grown in M63 minimal medium in the presence or absence
of purified E. coli chromosomal DNA as the sole carbon source.
Results are averages of three experiments. WT, wild type.
Use of defined minimal media supplemented with DNA.
While it is clear that prior to DNase I treatment the condi-
tioned medium contains a nutrient that only the wild-type cells Identification of other com gene homologs in E. coli. The
can utilize, we chose to directly address the question of experiments presented here demonstrate that E. coli can grow
whether this nutrient was DNA. Minimal medium which con- using DNA as a source of carbon and energy and that this
tained only inorganic salts, vitamin B1, and purified E. coli ability depends on a homolog of a gene in the competence
chromosomal DNA as the sole source of carbon and energy locus of another organism, H. influenzae. This prompted us to
was prepared. This medium was inoculated with wild-type or determine whether other H. influenzae competence gene ho-
yhiR mutant cells at ⬃5 ⫻ 103 CFU/ml and incubated for 4 mologs are present in the E. coli genome. In fact, E. coli has
days. As shown in Fig. 3, the growth yields of wild-type cells are homologs to eight H. influenzae genes believed to be involved
more than 50-fold greater than those of the yhiR mutant; in competence and transformation (Table 1). The H. influenzae
wild-type cells reach final densities of 2 ⫻ 105 to 5 ⫻ 105 com (11) locus includes seven genes (comABCDEFG) in a
CFU/ml. This directly demonstrates that E. coli has the ability putative operon plus comJ, which is transcribed divergently
to take up and utilize exogenous DNA as a carbon source. and separated from the cluster by the ponA gene, encoding
Similar results were obtained using heterologous DNAs, in- penicillin-binding protein 1a (Fig. 4). The eight E. coli com
cluding sonicated salmon sperm DNA or synthetic double- gene homologs show various degrees of amino acid sequence
stranded oligonucleotides (data not shown). However, when similarity to the H. influenzae com cluster gene products, rang-
minimal medium containing added DNA is first pretreated ing from 12 to 74% identity (Table 1). In E. coli, these genes
with DNase I, both wild-type and mutant cells grow similarly,
reaching final cell densities of ⬃2 ⫻ 109 CFU/ml (data not
shown). No growth of either wild-type or mutant strains when TABLE 1. Degree of similarity and location of E. coli and
cells were incubated in M63 minimal medium plus DNase I H. influenzae competence genes
alone, with no added DNA or other source of carbon, was H. influenzae E. coli % Identity
observed (data not shown). A similar result is observed when Gene Position (nt) Gene Position (min) (similarity)
wild-type and yhiR mutant strains are grown in minimal me-
dium with 100 mM concentrations of deoxynucleoside triphos- comA 460543 yrfD 75.8 19.8 (33.2)
comB 459748 yrfC 75.8 21.6 (59.2)
phates as the sole carbon source; both strains grow to the same comC 459245 yrfB 75.8 24.2 (49.7)
density (data not shown). Together, these results indicate that comD 458727 yrfA 75.8 11.8 (44.2)
yhiR mutants cannot catabolize dsDNA but can consume DNA comE 458304 hofQ 75.8 36.9 (68.9)
breakdown products. This defect in dsDNA consumption comF 456958 yhgH 76.4 39.5 (69.5)
causes the yhiR mutant to have a competitive disadvantage comG/orfG 456193 yhgI 76.4 73.7 (91.9)
comJ/orfJ 463327 yhiR 78.5 65.8 (85.1)
during coculture with its wild-type parent.
VOL. 183, 2001 DNA AS A NUTRIENT 6291

Downloaded from http://jb.asm.org/ on August 12, 2014 by guest

FIG. 4. Genetic maps of the H. influenzae com locus genes and their homologs. Dashed lines indicate the limits of gene clusters between H.
influenzae and E. coli. Arrows indicate direction of transcription, and arrow length is proportional to gene length (with the exception of ponA/mrcA
homologs, which are not to scale). Genes that are not contiguous are indicated by breaks between genes or clusters. Annotated sequence
information from the completed genomes of H. influenzae and E. coli was obtained from The Institute for Genomic Research (www.tigr.org) and
the University of Wisconsin (www.genetics.wisc.edu) websites, respectively. Except for the published pil locus sequences (6, 18, 25), unannotated
sequence data were obtained as follows: for S. enterica serovar Typhi, www.sanger.ac.uk/projects/S.typhi; for V. cholerae and S. putrefaciens,
www.tigr.org; for N. gonorrhoeae, dna1.chem.ou.edu/gono.html; for P. aeruginosa, www.pseudomonas.com; for K. pneumoniae, genome.wustl.edu/
gsc/Projects/bacterial/klebsiella/klebsiella.shtml.

are found in three loci: five genes (yrfDCBA and hofQ) in an tential advantage of taking up extracellular DNA solely for
apparent operon located at 75.8 min, two genes (yhgHI) in a nutritional purposes, particularly when that DNA is heterolo-
putative bicistronic operon at 76.4 min, and yhiR at 78.5 min gous and less likely to recombine onto the chromosome.
transcribed monocistronically. Whereas there is a chance to lose an essential function or
The identification in E. coli of homologs of the H. influenzae acquire a deleterious allele when taking in extracellular DNA
competence apparatus encouraged us to determine if another as genetic material, using DNA solely as a nutrient source
com gene homolog is essential for growth on DNA as the sole might pose little threat to the cell. Since so many organisms
carbon source. An insertional mutant with a mutation in the E. have developed mechanisms for natural competence and ge-
coli hofQ gene, a homolog of H. influenzae comE, was tested netic transformation, it seems that over evolutionary time, the
for its ability to compete with its wild-type parent and to utilize benefits of maintaining a system for horizontal genetic transfer
DNA as a sole carbon source. hofQ is located in a different outweigh the costs. However, having an additional, and not
operon from yhiR, over 2.5 min away. Under all conditions mutually exclusive, system for nutrient uptake could also pro-
tested it behaved identically to the yhiR mutant, showing a
stationary-phase competition defect during coculture with the
wild type (Fig. 5) and an inability to utilize extracellular DNA
as the sole source of carbon or energy (data not shown). Im-
portantly, both yhiR and hofQ mutants can be artificially in-
duced to competence, by treatment with calcium chloride or by
electroporation (3, 13, 15), as efficiently as the wild type (data
not shown). This indicates that the mechanism of DNA uptake
when DNA is used as a nutrient is distinct from the uptake
mechanism used during induction of artificial competence.

DISCUSSION

We have presented direct evidence supporting a model (9,

22, 24, 30) in which the DNA uptake function of a competence
system can be used for nutrient acquisition rather than (or in
addition to) obtaining and processing DNA for genetic trans-
formation. While the potential evolutionary fitness advantages FIG. 5. Survival patterns of the hofQ mutant and the wild-type
conferred by the acquisition of a beneficial gene by horizontal parental strain (WT) during competition in stationary phase. LB cul-
transfer are obvious (23), also as significant might be the po- tures were incubated for 12 days.
6292 FINKEL AND KOLTER
vide great benefit. It has been noted that because in naturally
transformable bacteria, such as B. subtilis, Streptococcus pneu-
moniae, H. influenzae, and N. gonorrhoeae, only a single strand
enters the cytoplasm during transformation (with the other
strand being degraded), this mechanism lacks efficiency as a
nutrient acquisition system (9). That is, if the mechanism for
DNA uptake, when the DNA is being used as a source of
nutrients rather than genetic information, is the same for “non-
competent” bacteria as for naturally transformable species,
then these bacteria would only be able to take up “half” of the
DNA as food. While this may be more of an issue for gram-
positive organisms, which do not have an outer membrane, we
feel that this may be less of a problem for gram-negative
organisms, particularly because current models of natural
transformation for these bacteria suggest that DNA degrada-
tion of the single strand which does not enter the cytoplasm
may take place in the periplasm (9), thus allowing the retention
and possible transport of the resulting nucleotides into the
cytoplasm for catabolic use. It has also been proposed that
bacteria may produce an extracellular nuclease which digests
dsDNA for catabolic purposes. However, we do not believe
that our data support this model in E. coli, since we would
expect the products of such a nuclease to help support the
growth of not only wild-type strains but also the yhiR or hofQ
mutants when grown in coculture with the wild-type.
It is interesting that the organization of these genes has been
conserved in many organisms. For example, the relative posi-
tion of the E. coli ponA homolog, mrcA, adjacent to the com-
ABCDE homologs, is similar to that of H. influenzae ponA (Fig.
4). In fact, this organization, with the ponA homolog tran-
scribed divergently from a five-gene cluster containing ho-
mologs of competence genes, is conserved in a wide variety of
gram-negative bacteria, including Salmonella enterica serovar
Typhi, Vibrio cholerae, Shewanella putrefaciens, N. gonorrhoeae,
Klebsiella pneumoniae, and Pseudomonas aeruginosa (Fig. 4).
What is particularly striking about these homologs is that the
genetic organization of the cluster is highly conserved but
levels of sequence conservation and gene sizes are quite vari-
able. This suggests that these genes were present in a common
ancestor and that this genetic organization has been conserved
over evolutionary time, but the functions encoded by these
genes may have diverged. As observed by A. Pugsley, homologs
of comE are part of a superfamily with roles in cellular pro-
cesses involving the movement of macromolecules across
membranes (21), including pilus biogenesis, protein secretion,
competence, and twitching motility; this prompted D. Dubnau,
in a review of mechanisms of DNA uptake in bacteria, to term
them PSTC proteins (9). For example, members of this family
play roles in competence in H. influenzae and N. gonorrhoeae,
type IV pilus biogenesis in N. gonorrhoeae and P. aeruginosa,
and twitching motility in P. aeruginosa (6, 18, 21, 25). It is
interesting that P. aeruginosa is not known to be naturally
competent.
Several investigators have noted that the E. coli yrfDBCA/
hofQ genes are homologs of genes involved in the production
of type IV pili in pseudomonads (18, 25, 26). In an effort to
address the lack of expression of type IV pili in K-12 E. coli,
Sauvonnet and coworkers (26) recently measured mRNA lev-
els expressed from the putative hofQ-containing operon (they
refer to the five-gene cluster as hofMNOPQ). No significant
J. BACTERIOL.
expression was observed in cells grown in LB medium during
exponential or early stationary phase, as determined by lacZ
fusions or reverse transcription-PCR techniques. However, the
fact that we observe a phenotype under conditions of compe-
tition in stationary-phase LB cultures and of outgrowth in
minimal media supplemented with DNA suggests that these
genes are, in fact, expressed. These differences may be due to
several factors, including the time points during stationary
phase when cells were harvested and the possibility that extra-
cellular DNA may act as an inducer of yrfD/hofM operon gene
expression.
Bacteria inhabit a wide variety of niches, and within many of
these environments extracellular DNA may be available. Esti-
mates of extracellular DNA concentrations in various marine
and aquatic environments range from 0.2 to 44 ␮g/liter (re-
viewed in reference 17). DNA has been shown to be quite
Downloaded from http://jb.asm.org/ on August 12, 2014 by guest
stable when complexed with various clays and soil minerals,
binding at concentrations in the microgram to milligram range
per gram of material (17). Extracellular DNA concentrations
within the mammalian host, including the gastrointestinal tract
and the mucosa in normal human lung, are estimated at hun-
dreds of micrograms per milliliter, reaching as much as 4
mg/ml in the lungs of cystic fibrosis patients (20). With an
abundance of DNA available in the environment and primarily
of heterologous origin, it is unlikely that much of the DNA
would be suitable for incorporation into the bacterial chromo-
some by homologous recombination, hence reducing its poten-
tial as a source of genetic diversity. Therefore, it is quite rea-
sonable that E. coli and other organisms would take advantage
of this rich nutrient source. Put simply, DNA is “good eating.”
It remains to be determined if the system in place for the
consumption of DNA arose first and then evolved into a sys-
tem for genetic transformation, or vice versa. Also remaining
to be demonstrated is the possibility that E. coli itself is capable
of natural genetic competence and transformation and that we
simply have not found the appropriate environmental condi-
tions to observe such a phenomenon in the laboratory. It is
possible that once the DNA is taken up, it can undergo two
fates: recombination-replication, to yield stable transforma-
tion, or digestion, to be used as a nutrient. Since the genes used
for this process, regardless of the final fate, appear to be
homologs of competence genes, we propose to name the E. coli
genes com genes after the designation of the H. influenzae
genes. The ability of the com genes to mediate utilization of
DNA as a nutrient appears to play a significant role in deter-
mining the relative fitness of organisms competing for very
limited nutrient resources in natural environments.
ACKNOWLEDGMENTS
We thank members of the Kolter lab and Leah Macfadyen for
helpful discussions; Erik Zinser, George O’Toole, Patrick Stragier,
Michael Farrell, and Miriam Susskind for comments on the manu-
script; and Vyacheslav Palchevskiy for technical assistance.
This work was supported by grants from the NIH and NSF to R.K.,
a postdoctoral fellowship from the Helen Hay Whitney Foundation to
S.E.F., and a grant from the USC/Norris Comprehensive Cancer Cen-
ter to S.E.F.
REFERENCES
1. Altschul, S. F., et al. 1997. Gapped BLAST and PSI-BLAST: a new gener-
ation of protein database search programs. Nucleic Acids Res. 25:3389–3402.
2. Ausubel, F. M., et al. (ed.). 1987. Current protocols in molecular biology.
VOL. 183, 2001
John Wiley and Sons, New York, N.Y.
3. Baur, B., K. Hanselmann, W. Schlimme, and B. Jenni. 1996. Genetic trans-
formation in freshwater: Escherichia coli is able to develop natural compe-
tence. Appl. Environ. Microbiol. 62:3673–3678.
4. Caetano-Annoles, G. 1993. Amplifying DNA with arbitrary oligonucleotide
primers. PCR Methods Appl. 3:85–92.
5. Dougherty, B. A., and H. O. Smith. 1999. Identification of Haemophilus
influenzae Rd transformation genes using cassette mutagenesis. Microbiol-
ogy 145:401–409.
6. Drake, S. L., and M. Koomey. 1995. The product of the pilQ gene is essential
for the biogenesis of type IV pili in Neisseria gonorrhoeae. Mol. Microbiol.
18:975–986.
7. Dreiseikelmann, B. 1994. Translocation of DNA across bacterial mem-
branes. Microbiol. Rev. 58:293–316.
8. Dubnau, D. 1991. Genetic competence in Bacillus subtilis. Microbiol. Rev.
55:395–424.
9. Dubnau, D. 1999. DNA uptake in bacteria. Annu. Rev. Microbiol. 53:217–
244.
10. Finkel, S. E., E. Zinser, and R. Kolter. 2000. Long-term survival and evolu-
tion in stationary phase, p. 231–238. In G. Storz and R. Hengge-Aronis (ed.),
Bacterial stress responses. ASM Press, Washington, D.C.
11. Fleischmann, R. D., et al. 1995. Whole-genome random sequencing and
assembly of Haemophilus influenzae Rd. Science 269:496–512.
12. Goodgal, S. H. 1982. DNA uptake in Haemophilus transformation. Annu.
Rev. Genet. 16:169–192.
13. Huang, R., and R. Reusch. 1995. Genetic competence in Escherichia coli
requires poly-beta-hydroxybutyrate/calcium polyphosphate membrane com-
plexes and certain divalent cations. J. Bacteriol. 177:486–489.
14. Kahn, M. E., and H. O. Smith. 1984. Transformation in Haemophilus: a
problem in membrane biology. J. Membr. Biol. 81:89–103.
15. Kimoto, H., and A. Taketo. 1997. Initial stage of DNA-electrotransfer into E.
coli cells. J. Biochem. 122:237–242.
16. Kleckner, N., J. Bender, and S. Gottesman. 1991. Uses of transposons with
emphasis on Tn10. Methods Enzymol. 204:139–180.
17. Lorenz, M. G., and W. Wackernagel. 1994. Bacterial gene transfer by natural
genetic transformation in the environment. Microbiol. Rev. 58:563–602.
18. Martin, R., A. A. Watson, T. F. McCaul, and J. S. Mattick. 1995. Charac-
terization of a five-gene cluster required for the biogenesis of type 4 fimbriae
in Pseudomonas aeruginosa. Mol. Microbiol. 16:497–508.
19. Miller, J. H. 1992. A short course in bacterial genetics. Cold Spring Harbor
DNA AS A NUTRIENT 6293
Laboratory Press, Cold Spring Harbor, N.Y.
20. Potter, J., S. Spector, L. Matthews, and J. Lemm. 1963. Studies on pulmo-
nary secretions. 3. The nucleic acids in whole pulmonary secretions from
patients with cystic fibrosis, bronchiectasis, and laryngectomy. Am. Rev.
Respir. Dis. 99:909–916.
21. Pugsley, A. P. 1993. The complete general secretory pathway in gram-neg-
ative bacteria. Microbiol. Rev. 57:50–108.
22. Redfield, R. 1993. Genes for breakfast: the have-your-cake-and-eat-it-too of
bacterial transformation. J. Hered. 84:400–404.
23. Redfield, R. J. 1988. Evolution of bacterial transformation: is sex with dead
cells ever better than no sex at all? Genetics 119:213–221.
24. Redfield, R. J., M. R. Schrag, and A. M. Dean. 1997. The evolution of
bacterial transformation: sex with poor relations. Genetics 146:27–38.
25. Ropp, P. A., and R. A. P. Nicholas. 1997. Cloning and characterization of the
ponA gene encoding penicillin-binding protein 1 from Neisseria gonorrhoeae
and Neisseria meningitidis. J. Bacteriol. 179:2783–2787.
26. Sauvonnet, N., P. Gounon, and A. P. Pugsley. 2000. PpdD type IV pilin of
Escherichia coli K-12 can be assembled into pili in Pseudomonas aeruginosa.
J. Bacteriol. 182:848–854.
27. Smith, H. O., D. B. Danner, and R. A. Deich. 1981. Genetic transformation.
Annu. Rev. Biochem. 50:41–68.
Downloaded from http://jb.asm.org/ on August 12, 2014 by guest
28. Smith, H. O., J.-F. Tomb, B. A. Dougherty, R. D. Fleischmann, and J. C.
Venter. 1995. Frequency and distribution of DNA uptake signal sequences in
the Haemophilus influenzae Rd genome. Science 269:538–540.
29. Solomon, J. M., and A. D. Grossman. 1996. Who’s competent and when:
regulation of natural genetic competence in bacteria. Trends Genet. 12:150–
155.
30. Stewart, G. J., and C. A. Carlson. 1986. The biology of natural transforma-
tion. Annu. Rev. Microbiol. 40:211–235.
31. Syvanen, M., and C. I. Kado. 1998. Horizontal gene transfer. Chapman &
Hall, London, United Kingdom.
32. Tomb, J.-F., H. el-Hajj, and H. O. Smith. 1991. Nucleotide sequence of a
cluster of genes involved in the transformation of Haemophilus influenzae
Rd. Gene 104:1–10.
33. Zambrano, M. M., D. A. Siegele, M. Almirón, A. Tormo, and R. Kolter. 1993.
Microbial competition: Escherichia coli mutants that take over stationary
phase cultures. Science 259:1757–1760.
34. Zinser, E. R., and R. Kolter. 1999. Mutations enhancing amino acid catab-
olism confer a growth advantage in stationary phase. J. Bacteriol. 181:5800–
5807.
 Blackwell Science, LtdOxford, UKMMIMolecular Microbiology0950-382X© 2005 The Authors; Journal compilation © 2005 Blackwell Publishing Ltd? 2005592376385Review ArticleNatural transformation of Neisseria gonorrhoeaeH. L. Hamilton
and J. P. Dillard

Molecular Microbiology (2006) 59(2), 376–385 doi:10.1111/j.1365-2958.2005.04964.x

First published online 9 November 2005

MicroReview

Natural transformation of Neisseria gonorrhoeae: from

DNA donation to homologous recombination

Holly L. Hamilton and Joseph P. Dillard* identification of additional transformation genes and
Department of Medical Microbiology and Immunology, provides insight into previous investigations of gono-
University of Wisconsin-Madison Medical School, coccal transformation. Here we review these recent
Madison, WI 53706, USA. developments and address the implications of natural
transformation in the evolution and pathogenesis
N. gonorrhoeae.
Summary

Gonococci undergo frequent and efficient natural

Overview
transformation. Transformation occurs so often that
the population structure is panmictic, with only one Bacteria use three major mechanisms for genetic
long-lived clone having been identified. This high exchange: conjugation, transduction and transformation.
degree of genetic exchange is likely necessary to Neisseria gonorrhoeae is one of at least 44 species of
generate antigenic diversity and allow the persistence bacteria that are naturally competent for genetic transfor-
of gonococcal infection within the human population. mation (Lorenz and Wackernagel, 1994; Sparling, 1966).
In addition to spreading different alleles of genes Phages able to infect N. gonorrhoeae are not known
for surface markers and allowing avoidance of the (Goldberg et al., 1978) and, although conjugative plas-
immune response, transformation facilitates the mids are present in some gonococcal isolates, trans-
spread of antibiotic resistance markers, a continuing formation is the only mechanism for mobilization of
problem for treatment of gonococcal infections. gonococcal chromosomal loci (Sox et al., 1978; Koomey,
Transforming DNA is donated by neighbouring 1998).
gonococci by two different mechanisms: autolysis or Natural transformation is the ability of bacteria to take
type IV secretion. All types of DNA are bound non- up and incorporate macromolecular DNA efficiently. It is
specifically to the cell surface. However, for DNA hypothesized that natural transformation evolved as a
uptake, Neisseria gonorrhoeae recognizes only DNA mechanism for using DNA as a food source (Finkel and
containing a 10-base sequence (GCCGTCTGAA) Kolter, 2001) or to aid in the repair of damaged chromo-
present frequently in the chromosome of neisserial somes (Solomon and Grossman, 1996). For gonococci, it
species. Type IV pilus components and several pilus- is clear that transformation has been harnessed as a
associated proteins are necessary for gonococcal powerful mechanism for generating genetic diversity,
DNA uptake. Incoming DNA is subject to restriction, spreading advantageous alleles and mediating some
making establishment of replicating plasmids difficult forms of antigenic variation (Hobbs et al., 1994; Fudyk
but not greatly affecting chromosomal transforma- et al., 1999; Snyder et al., 2004).
tion. Processing and integration of transforming DNA Neisseria spp. are unusual in that they do not regulate
into the chromosome involves enzymes required for competence like many other naturally competent bacteria,
homologous recombination. Recent research on DNA including Streptococcus pneumoniae, Bacillus subtilis and
donation mechanisms and extensive work on type IV Haemophilus influenzae; rather, the Neisseriae are
pilus biogenesis and recombination proteins have competent for natural transformation during all phases
greatly improved our understanding of natural trans- of growth (Biswas et al., 1977; Goodgal, 1982). The
formation in N. gonorrhoeae. The completion of the competence of N. gonorrhoeae is not dependent on any
gonococcal genome sequence has facilitated the soluble competence factor (Biswas et al., 1977). Like
H. influenzae, N. gonorrhoeae only takes up DNA that
contains a genus-specific uptake sequence. In N. gonor-
Accepted 17 October, 2005. *For correspondence. E-mail rhoeae, the DNA uptake sequence (DUS) is a 10-base
jpdillard@wisc.edu; Tel. (+1) 608 265 2837; Fax (+1) 608 262 8418. sequence (GCCGTCTGAA) that appears very frequently

© 2005 The Authors

Journal compilation © 2005 Blackwell Publishing Ltd

in its own chromosome and that of other Neisseria spp.
The 2.15 Mb genome sequence of strain FA1090 contains
1965 copies of the DUS (Davidsen et al., 2004), thus
averaging one DUS for every 1096 bp. The DUS some-
times occurs in inverted repeats between genes forming
possible transcriptional terminators (Goodman and
Scocca, 1988). Transformation of N. gonorrhoeae with
DUS-bearing supercoiled or linear DNA is equally effi-
cient. It was reported that the transformation efficiency of
single-stranded DNA (ssDNA) is comparable to that of
double-stranded DNA (dsDNA) (Stein, 1991). However,
this idea is still controversial (see Remaining questions
below).
This review will focus on the studies that have eluci-
dated the four major steps in gonococcal transformation:
donation of DNA, binding and uptake of transforming
DNA, processing, and integration into the gonococcal
chromosome. It should be noted that many of the
mechanisms involved in N. gonorrhoeae transformation
are the same in Neisseria meningitidis transformation.
Therefore, where appropriate, insights derived from stud-
ies of N. meningitidis will also be described and significant
differences will be mentioned.
Step 1. DNA donation
Most descriptions of natural transformation begin with
DNA uptake. The DNA is simply presumed to be in the
environment, and the donor is thought to have no role
other than having made the supreme sacrifice. However,
the donor may be an important player. What set of events
has resulted in the presence of naked DNA in proximity
to the competent recipient? In the case of N. gonorrhoeae,
the bacteria only live inside human beings and only take
up DNA containing the genus-specific DUS; therefore, the
donated DNA must be from siblings of the infecting strain
or from a co-infecting gonococcal strain. In rare instances
the DNA might come from another species, including
other Neisseria or H. influenzae (Kroll et al., 1998). Trans-
fer from other Neisseria might be expected, as all Neis-
seria species use the same DUS. In fact, transfer from
commensal Neisseria spp. appears to be responsible for
the creation of penA alleles that confer penicillin resis-
tance in N. gonorrhoeae (Spratt et al., 1992). Similarly,
argF of N. meningitidis has regions that appear to have
been acquired from commensal Neisseria (Zhou and
Spratt, 1992). Transfer from H. influenzae would not be as
likely as Haemophilus spp. use a different DUS. However,
the H. influenzae chromosome was found to contain four
copies of the Neisseria DUS, and one of these is thought
to have facilitated the transfer of the bio locus from
Haemophilus to N. meningitidis and N. gonorrhoeae (Kroll
et al., 1998). Most species are thought to donate DNA
© 2005 The Authors
Journal compilation © 2005 Blackwell Publishing Ltd, Molecular Microbiology, 59, 376–385
Natural transformation of Neisseria gonorrhoeae 377
for subsequent transformation by autolysis. However,
N. gonorrhoeae can also donate DNA for transformation
by type IV secretion.
Type IV secretion
Type IV secretion systems (T4SS) are one of more
than five multicomponent secretion systems utilized by
Gram-negative bacteria to secrete macromolecular sub-
strates across the cell envelope. T4SS are noted for
their ability to secrete both protein and DNA substrates,
although not all systems secrete both. The presence of
a T4SS encoded on the gonococcal genetic island
(GGI) has recently been described. The GGI is present
in approximately 80% of N. gonorrhoeae strains and
appears to have been horizontally acquired (Dillard and
Seifert, 2001). Many of the T4SS homologues encoded
in the GGI have been shown to be necessary for the
secretion of gonococcal chromosomal DNA into the
extracellular milieu during growth (Hamilton et al., 2001;
2005). The DNA secreted by donors is taken up by
recipient bacteria when donors and recipients are
cultured together (Dillard and Seifert, 2001; Hamilton
et al., 2001).
The DNA has been detected in culture medium by use
of a fluorescent DNA-binding dye, as well as by co-culture
transformation. The addition of DNase I to the culture
supernatants eliminates the increased fluorescence seen
in the wild-type culture medium as compared with that
from the T4SS mutants (Hamilton et al., 2001; 2005), con-
firming that DNA is the material being detected. Also
DNase I reduces transfer of chromosomal markers 250-
fold, indicating that DNA is present in the medium and is
not transferred by conjugation (Dillard and Seifert, 2001).
Theoretically, the DNA could be released by autolysis of
a portion of the population. However, all measures of
autolysis have shown no differences between the wild-
type strain and the T4SS mutants. Neither the release of
a cytoplasmic protein (Dillard and Seifert, 2001) nor the
release of radiolabelled RNA (our unpublished observa-
tion) was different between wild type and mutants during
growth. Furthermore, cell death was not reduced in the
T4SS mutants as compared with wild type when analysed
by live/dead staining and fluorescent microscopy or
fluorometry (Hamilton et al., 2005). Thus, the T4SS
releases DNA into the medium by a method different from
autolysis.
Little is known about the secreted DNA; how it is pro-
cessed, if secretion is linked to chromosome replication,
or if secreted DNA is single- or double-stranded. However,
it is easy to imagine type IV secretion as a mechanism of
DNA donation for lateral gene transfer within a population
of gonococci without sacrificing vertical transmission or
risking lysis and death of the entire population.
378 H. L. Hamilton and J. P. Dillard
Autolysis
Autolysis is a curious phenomenon. As many microbiolo-
gists study survival strategies, the idea that a bacterium
would act to promote its own death and lysis seems inher-
ently wrong. In spite of our feelings on the subject,
N. gonorrhoeae autolyses readily; in vitro, autolysis nor-
mally occurs during stationary phase or during non-growth
conditions, such as exhaustion of nutrients, non-optimal
temperature, pH or osmolarity (Hebeler and Young, 1975).
The major autolysin in vitro is thought to be an N-
acetylmuramyl-L-alanine amidase (Hebeler and Young,
1976). An enzyme with this specificity is also the major
autolysin in S. pneumoniae, and recent evidence sug-
gests that the pneumococcal enzyme acts in the lysis of
a portion of the population to provide donor DNA for
transformation (Steinmoen et al., 2002). Two amidase
homologues are encoded in the gonococcal chromosome
sequence (GenBank Accession No. AE004969). One of
these is similar to Escherichia coli AmpD and is predicted
to act in the breakdown of peptidoglycan monomers in the
cytoplasm, but not in breakdown of the cell wall. The
second is similar to E. coli AmiC and would be predicted
to act on macromolecular peptidoglycan. The AmiC homo-
logue is likely responsible for the peptidoglycan hydrolysis
activity identified by Hebeler and Young (1976) and might
play an important part in gonococcal autolysis. Membrane
instability also contributes to autolysis (Cacciapuoti et al.,
1978) and a recent report describes mutations in a
phospholipase that result in decreased autolysis in
N. meningitidis and N. gonorrhoeae (Bos et al., 2005).
DNA released through gonococcal autolysis can con-
tribute to DNA donated for natural transformation (Nor-
lander et al., 1979). For the 20% of gonococcal strains
that lack the T4SS and for all N. meningitidis strains, autol-
ysis is likely the only mechanism of DNA donation. A
number of questions arise when considering donation of
DNA by cell death. Is cell lysis mediated by host defences,
or do the bacteria commit suicide by autolysis? How can
autolysis be controlled such that one sibling dies while
another does not? Is autolysis part of a programmed cell
death event that processes DNA to make it better for
transport to, or incorporation into, another cell? Which is
better for transformation, DNA released by autolysis or by
secretion? Both from an immunological standpoint and for
understanding DNA donation, the process of autolysis
warrants further investigation.
One possible mechanism for initiating a programmed
cell death event was described for N. gonorrhoeae.
Gonococci were found to carry a phase-variable type III
restriction-modification enzyme, designated NgoX. The
coding sequence of the methylase subunit carried 8–12
pentanucleotide repeats near the 5′ end of the gene,
allowing the sequence to slip in and out of frame by
Journal compilation © 2005 Blackwell Publishing Ltd, Molecular Microbiology, 59, 376–385
slipped-strand mispairing, similar to the well-characterized
mechanism of opacity protein variation in gonococci
(Stern et al., 1986; Belland et al., 1996; Snyder et al.,
2001a). The type III restriction-modification system meth-
ylase is required for sequence recognition. Therefore, the
switching on of the methylase with the restriction subunit
already produced could result in destruction of the chro-
mosome (Bourniquel and Bickle, 2002). A second type III
methylase (NG0641) carries 37 copies of a 4 bp repeat
and would similarly be expected to phase-vary (Snyder
et al., 2001a). Phase-variable expression of these lethal
activities could allow for random death of a portion of the
population.
Most gonococci also contain a toxin-antitoxin gene pair
on the chromosome (J.P. Dillard and H.S. Seifert, unpub-
lished). Although the gonococcal toxin–antitoxin (TA) sys-
tem has not been characterized, related protein-based
killer systems cause cell death or cell cycle arrest. Inter-
estingly, the TA systems appear to respond to the growth
state of the cell (Gerdes et al., 2005). As gonococci are
known to autolyse under non-growth conditions, the TA
system would be one explanation for the linkage of autol-
ysis to growth conditions and could explain the massive
death and lysis occurring in cultures following the exhaus-
tion of nutrients (Morse and Bartenstein, 1974).
Step 2. DNA binding and uptake
The process of DNA binding and uptake by the ever-
competent gonococcus has been intensely studied and
recently reviewed (Chen and Dubnau, 2004). A major
remaining question is which, if any, protein acts as a
receptor for transforming DNA. Because of its strong
preference for DUS-containing transforming DNA,
N. gonorrhoeae must distinguish between self DNA
(carrying the DUS) and non-self DNA (lacking the DUS).
Thus, DNA binding events are classified into two types:
non-specific and specific. Once specific binding of DUS-
containing DNA has occurred, uptake across the outer
membrane, the murein layer and the inner membrane
follow.
Nearly all naturally transformable bacteria, both Gram-
positive and Gram-negative, utilize type IV pili (Tfp) or a
type IV pilus-like apparatus for the process of DNA uptake.
Although the Gram-positive species do not have type IV
pili, the DNA uptake proteins show significant similarity to
proteins required for assembly of Tfp or the related type
II secretion systems (Chen and Dubnau, 2004). Very early
in the studies of gonococcal competence, a correlation
was established between the expression of type IV pili and
competence. Piliated gonococci were found to be more
transformable than non-piliated bacteria (Sparling, 1966).
At high DNA concentrations, the transformation frequency
can exceed 20% (Gunn and Stein, 1996), whereas non-
© 2005 The Authors

piliated variants yield transformation frequencies near
1 × 10−7 (Sparling, 1966). Both non-specific and specific
DNA binding has been measured (Dougherty et al.,
1979). Only strains expressing retractable Tfp bound DNA
non-specifically, suggesting the hypothesis that electro-
static interactions between DNA and positively charged
patches on Tfp might lead to non-specific binding (Aas
et al., 2002a). It is unclear whether the pilus or a different
pilus-like apparatus mediates uptake of DUS-carrying
DNA. Mutations in genes involved in pilus biogenesis [pilE
(pilin subunit), pilF (putative NTP-binding protein), pilD
(prepilin peptidase), pilG (putative inner membrane
protein) and pilQ (outer membrane secretin protein)]
reduce natural transformation to undetectable or nearly
undetectable levels, and reduce binding of DUS-
containing DNA to the level of non-specific DNA (Drake
and Koomey, 1995; Freitag et al., 1995; Boyle-Vavra and
Seifert, 1996; Aas et al., 2002a). PilC, which is localized
to the pilus tip or to the outer membrane, has also been
implicated in competence (Jonsson et al., 1991; Rudel
et al., 1995).
Even small amounts or altered forms of pilin are suffi-
cient for gonococcal transformation. S- and L-pilin forms
result from recombination events in the pilE expression
locus; S-pilin variants display intermediate piliation and
secrete a soluble form of the pilin monomer, while L-pilin
variants are the result of a tandem duplication in pilE that
generates an overlong pilin that is not assembled into pili.
Gibbs et al. (1989) demonstrated that S-pilin variants
display wild-type transformation frequencies, while the
competence of gonococcal L-pilin variants is reduced
approximately 35-fold. S-pilin variants make some nor-
mally processed pilin (Long et al., 1998), and it is likely
that L-pilin variants make a small amount as well, thereby
fulfilling the need for pilin, if not pili, for transformation. The
fact that PilE and the pilus assembly proteins are neces-
sary for natural transformation suggested that Tfp mediate
DNA transformation directly; however, this hypothesis has
not been proven, leaving open the possibility that a pilus-
like apparatus might be enough to mediate natural trans-
formation. In fact, Long et al. (2003) demonstrated that
gonococcal pilin variants that elaborate virtually no
observable pili but that still express pilin are competent for
natural transformation and that this transformation is
dependent on the pilus-related proteins PilQ and PilT. This
finding generates support for the hypothesis that full,
extended pili are not necessary for natural transformation
and that a pilus-like apparatus is sufficient.
Two proteins known to be involved in gonococcal com-
petence do not have a demonstrated role in Tfp biogene-
sis: PilT and ComE (Biswas et al., 1989; Chen and
Gotschlich, 2001; Aas et al., 2002a,b). The dud-1 mutant
was one of the first reported transformation-deficient
mutants (Biswas et al., 1989) and the mutation was later
© 2005 The Authors
Journal compilation © 2005 Blackwell Publishing Ltd, Molecular Microbiology, 59, 376–385
Natural transformation of Neisseria gonorrhoeae 379
mapped to pilT (Wolfgang et al., 1998). PilT is necessary
for twitching motility, which is believed to occur via retrac-
tion of gonococcal Tfp. PilT, whose homologue has been
shown to be an NTPase (Herdendorf et al., 2002; Aukema
et al., 2005), is not necessary for specific binding of DNA
but is rather involved in uptake of DNA (Aas et al., 2002a;
Herdendorf et al., 2002). Uptake of DNA across the outer
membrane, i.e. the transfer to a DNase-resistant state,
has been demonstrated to be distinct from DNA binding
(Aas et al., 2002a), and components involved in compe-
tence have been linked to each step. ComE was identified
as the gonococcal orthologue of the B. subtilis compe-
tence protein, ComEA. There are four copies of comE in
the gonococcal chromosome and sequential mutation of
comE1–4 additively diminished transformation efficiency
(Chen and Gotschlich, 2001). ComE binds DNA non-spe-
cifically and its presence is necessary for efficient DNA
uptake (Chen and Gotschlich, 2001). As ComE is likely
not localized to the gonococcal surface, it has been pro-
posed that ComE binds DNA within the periplasm after
specific DNA uptake has occurred (Chen and Gotschlich,
2001; Aas et al., 2002a).
Two pilin-like proteins, ComP and PilV, have been
shown to affect the level of specific DNA binding (Aas
et al., 2002b). Mutants lacking one or the other of these
proteins still produce pili. However, mutants defective in
comP and pilV are greatly reduced in piliation. Increased
levels of ComP result in high transformation frequencies,
whereas increased PilV decreases transformation.
Although no regulation of competence has been observed
in wild-type Neisseria, control of the relative levels of
these two proteins could be used to modulate compe-
tence (Aas et al., 2002b). Mutants defective in pilC1 and
pilC2, or mutants lacking one of five newly discovered
pilin-like proteins PilH–L were not transformable or were
significantly reduced in transformation competence and
were found not to localize ComP to the pili (Winther-
Larsen et al., 2005).
Mutations in the additional genes comA, comL, dca and
tpc also diminish natural transformation. ComL, compe-
tence lipoprotein, is associated with the murein layer and
is thought to be involved either directly or indirectly in
puncturing the peptidoglycan layer for incoming DNA
entry into the gonococcus (Facius et al., 1996; Fusseneg-
ger et al., 1996a). In addition to competence defects,
mutants in tpc display a reduction in total murein hydro-
lase activity and have cell separation defects; they grow
in tetrapacs rather than as diplococci (Fussenegger et al.,
1996b). The involvement of ComL and Tpc might indicate
the need for localized openings in the gonococcal pepti-
doglycan for DNA uptake to occur, or may be due to
pleiotropic effects on cell wall metabolism. Dca is encoded
in the division and cell wall gene cluster, but was not found
to have any effect on peptidoglycan metabolism. It is
380 H. L. Hamilton and J. P. Dillard

Donation Binding and Processing and Homologous

Uptake Recombination

ComP OM
Autolysis
V
C OM
Q
ComE
ComL

Tpc
PG N? PG

D
IM
T F G
ATP

ADP ComA IM

Type IV secretion

A
C A
A
RM A X
B D

RM

Fig. 1. Gonococcal transformation occurs in four steps: DNA donation, binding and uptake of DNA, processing and homologous recombination.
Type IV secretion of DNA and autolysis of gonococci serve as two mechanisms for the donation of DNA for natural transformation. Binding and
uptake requires many pilus-related proteins (single letters indicate Pil gene products) as well as ComP, ComL, ComE, ComA and Tpc. During
uptake, plasmid DNA is processed into linear double-stranded DNA (dsDNA), and at least some of this incoming dsDNA is converted to single-
stranded DNA (ssDNA). Once in the cytoplasm, dsDNA might be processed by restriction-modification enzymes (RM) as well as by RecBCD
nuclease (B, C, D). ssDNA is bound by cytoplasmic RecA (A), which mediates homologous recombination into the gonococcal chromosome.
Small black boxes indicate DNA uptake sequences, which are necessary for efficient uptake. OM, outer membrane; PG, peptidoglycan layer; IM,
inner membrane.

predicted to be an inner membrane protein, and it is
required for competence in N. gonorrhoeae but not
N. meningitidis (Snyder et al., 2001b). ComA is predicted
to be an inner membrane protein involved in transport of
DNA into the cytosol (Facius and Meyer, 1993). This idea
is supported by studies of the related protein ComEC in
B. subtilis. ComEC is a polytopic membrane protein
required for DNA uptake and is thought to form a channel
for DNA transport (Draskovic and Dubnau, 2005).
Taken together, these data suggest a model for gono-
coccal competence (Fig. 1). Many components involved
in Tfp biogenesis are necessary for competence, and
DNA is bound by an as yet unidentified component that
is likely associated with the pilus or a pilus-like apparatus.
ComP and PilV influence specific binding of DUS-contain-
ing DNA. Once across the outer membrane, ComE binds
Journal compilation © 2005 Blackwell Publishing Ltd, Molecular Microbiology, 59, 376–385
DNA and aids in transformation, while Tpc and ComL may
allow the DNA to cross the peptidoglycan layer. PilT is
necessary for DNA uptake, theoretically by retracting the
Tfp or the subunits of a pilus-like apparatus. ComA might
help the DNA cross the inner membrane into the cytosol.
Once within the cytoplasm, this DNA might be processed
by gonococcal enzymes.
Step 3. DNA processing
The fate of transforming DNA in N. gonorrhoeae is depen-
dent on a number of factors, including its size and nature.
Transformation of N. gonorrhoeae with plasmid DNA is
1000-fold less efficient than with chromosomally derived
loci (Eisenstein et al., 1977). Sox et al. (1979) demon-
strated that transformation of N. gonorrhoeae with a
© 2005 The Authors

native 7.1 kb, penicillinase-producing plasmid often gen-
erated larger or smaller versions of the plasmid in gono-
coccal transformants. E. coli-propagated versions of the
same gonococcal plasmids did not transform recipient
gonococci. These experiments provided evidence that
plasmid DNA is subject to restriction by the gonococcus
during transformation (Sox et al., 1979). N. gonorrhoeae
encodes on the order of 16 methyltransferases, many of
which have corresponding endonucleases that, together,
generate an impressive restriction barrier for transforming
plasmids (Stein et al., 1995). The capacity for plasmid
transformation can be improved by inactivating several of
these restriction-modification systems (Gunn and Stein,
1996). Large plasmids are processed into linear dsDNA
pieces during transformation, before any restriction of
DNA that occurs inside the gonococcus (Biswas et al.,
1986).
The restriction barrier to plasmid DNA transformation in
N. gonorrhoeae is not evident during transformation with
chromosomal loci. Although initial studies of gonococcal
transformation suggested that transforming DNA enters
the cell as double-stranded molecules (Biswas and Spar-
ling, 1981), more recent studies suggest that at least
some ssDNA is formed during transformation and that this
conversion occurs primarily in the periplasm (Chaussee
and Hill, 1998). While the enzyme(s) responsible for this
activity have not been identified, a similar mechanism of
the degradation of one DNA strand during transformation
exists in Gram-positive bacteria (reviewed in Dubnau,
1999). The lack of an observable restriction barrier during
transformation of chromosomal loci argues that this DNA,
at least once it is in the cytoplasm, is single-stranded and
therefore resistant to endonucleases. Incoming ssDNA
would not be restricted, and the integrated heteroduplex
would be methylated on the resident strand and therefore
not restricted. As plasmid transformation results from the
re-assembly of an intact plasmid by the annealing of
overlapping imported strands, both strands would be
unmodified and therefore susceptible to restriction. A
similar restriction barrier is observed for plasmids intro-
duced into N. gonorrhoeae by conjugation (Butler and
Gotschlich, 1991).
The size of the imported DNA has not been extensively
studied. Experiments with plasmids (mentioned above)
show that it is easy to recreate a 7 kb plasmid following
transformation, but impossible to recreate a 42 kb plas-
mid. This result might indicate that the 42 kb plasmid was
cut at multiple sites before import, and suggests that
imported fragments are less than 42 kb. Linkage studies
performed using chromosomal markers conferring anti-
biotic resistance can now be re-analysed for size using
the chromosome sequence data. Sarubbi et al. (1974)
showed linkage of antibiotic resistance markers rif-1 and
str-7, mutations known to occur in rpoB and rpsL respec-
© 2005 The Authors
Journal compilation © 2005 Blackwell Publishing Ltd, Molecular Microbiology, 59, 376–385
Natural transformation of Neisseria gonorrhoeae 381
tively. These genes are separated by 11 kb in strain
FA1090. Similarly, linkage was found with str-7 (rpsL) and
spc-3 (rpsE), which are 13 kb apart. So, 11 kb and 13 kb
fragments can be imported. However, no linkage was
found between rif-1 and spc-3, indicating that the ∼25 kb
fragment carrying both of these markers is rarely or never
imported.
Step 4. Homologous recombination into the
gonococcal chromosome
Homologues of enzymes involved in homologous recom-
bination in E. coli have been identified in N. gonorrhoeae,
and a number of these proteins are required for gonococ-
cal transformation. Gonococcal genes encoding RecA, B,
C, D, O, Q and X are present in the gonococcal chromo-
some, suggesting the existence of both RecBCD and
RecF pathways for homologous recombination, both of
which are dependent on RecA (Koomey and Falkow,
1987; Mehr and Seifert, 1998; Stohl and Seifert, 2001).
RecA is necessary to maintain newly acquired chromo-
somal markers via homologous recombination into the
genome, and RecA-deficient gonococci are not transform-
able (Koomey and Falkow, 1987).
Mehr and Seifert demonstrated that the RecBCD path-
way of homologous recombination, but not the RecF path-
way, is necessary for efficient gonococcal transformation.
However, gonococci are only 40-fold reduced in transfor-
mation efficiency in a RecBCD-pathway mutant as well as
a RecBCD-/RecF-pathway mutant, suggesting the possi-
bility of a third, as yet undescribed pathway for homolo-
gous recombination during natural transformation (Mehr
and Seifert, 1998). Additionally, the homologous recombi-
nation activity that remains in these mutants might reflect
a difference in the nature of the transforming DNA; single-
stranded transforming DNA might bypass RecBCD and
bind directly to RecA to mediate recombination (Mehr and
Seifert, 1998). Finally, gonococci lacking RecX, which is
presumed to regulate RecA activity, are fivefold reduced
in DNA transformation (Stohl and Seifert, 2001; Stohl
et al., 2003).
Advantages and implications
It is notable that many naturally transformable species are
human pathogens. This might be an artefact of our
human-centric view of the world or the current state of
funding. However, it is clear that transformation impacts
N. gonorrhoeae pathogenesis and consequently, its
evolutionary survival. The genome sequence of
N. gonorrhoeae strain FA1090 reveals the presence of at
least six genetic islands, presumably acquired via trans-
formation from different bacterial species, which may
encode any number of fitness advantages to the organism
382 H. L. Hamilton and J. P. Dillard
(GenBank Accession No. AE004969). The GGI, which is Are there effects on recipient gonococci apart from
present in 80% of gonococcal strains (but not FA1090), acquisition of genes? N. meningitidis shows increased
has the characteristics of a horizontally acquired mobile phase variation in response to transforming DNA from
genetic element that might be transferred between gono- heterologous Neisseria species or N. meningitidis strains,
coccal strains by natural transformation (Hamilton et al., due to titrating DNA repair proteins (Alexander et al.,
2005). In addition to the acquisition of large genetic 2004). A similar phenomenon may also occur in
islands, gonococcal transformation is known to generate N. gonorrhoeae. In addition, it might be advantageous for
hybrid porin alleles and is likely to also generate other gonococci to use the recognition of incoming DNA as the
advantageous alleles which affect gonococcal pathogen- indication of the presence of significant numbers of other
esis (Hobbs et al., 1994; Cooke et al., 1998). gonococci, i.e. DNA could act as a quorum-sensing mol-
ecule. Presumably genes might be turned on in response
Remaining questions to transforming DNA, much like genes are regulated in the
response to detection of other well-known quorum-
Although the gonococcus is an excellent model for the
sensing molecules.
study of natural transformation, many questions still
Could the methylation state of incoming DNA aid in
remain. How is DNA donation controlled? Is cell lysis a
distinguishing self gonococcal DNA from non-self gono-
stochastic or a regulated process? What are the identities
coccal DNA? The many restriction-modification systems
of the autolysins? Both a phospholipase and a peptidogly-
of N. gonorrhoeae might differentially methylate genomic
can hydrolase appear to be involved in this process, and
DNA. Taking up DNA of non-self rather than self DNA (i.e.
mechanisms of autolysis were characterized decades ago
DNA from a different gonococcal strain) could be more
(Hebeler and Young, 1976; Cacciapuoti et al., 1978).
advantageous for the purposes of allelic diversification
However, no one has yet been able to identify a nonauto-
and fitness.
lytic mutant. Does type IV secretion of DNA occur in all
Does N. gonorrhoeae possess a competence nuclease
cells or only a portion of the population, and how is it
to degrade one incoming strand during transformation
regulated?
with dsDNA? It is interesting that most models of trans-
What binds the DUS on DNA during transformation? As
formation in both Gram-positive and Gram-negative bac-
mentioned previously, the receptor for species-specific
teria depict a competence nuclease that degrades one
DNA for natural transformation has not yet been identified.
strand of the double-stranded transforming DNA, but only
The prime candidates for this function would appear to be
one competence nuclease has yet been identified, EndA
PilE, ComP, or another protein that is controlled by the
of S. pneumoniae (reviewed in Dubnau, 1999). DNA deg-
presence of ComP. However, no DNA binding to PilE or
radation studies like those performed by Provvedi et al.
ComP could be demonstrated (Mathis and Scocca, 1984;
(2001) in B. subtilis might aid in elucidating the fate of
Aas et al., 2002a), and a ComP-dependent DNA-binding
incoming DNA in N. gonorrhoeae. Along a similar line
protein has not yet been found.
what is the identity of the specific enzymes that linearize
Are gonococci as efficiently transformed with ssDNA as
circular DNA molecules during transformation with plas-
with dsDNA? Stein has demonstrated that ssDNA gener-
mid DNA?
ated by M13 phage transformed gonococci at similar
Natural transformation appears to be an ever-present
levels as double-stranded donor DNA (Stein, 1991); how-
and essential mechanism for the acquisition and
ever, some researchers adhere to the belief that only
exchange of genetic material in the gonococcus. This
dsDNA transforms. ssDNA was also reported to transform
process has undoubtedly contributed to the success of
H. influenzae with an efficiency on the order of ∼50% that
N. gonorrhoeae as a human pathogen. N. gonorrhoeae
of dsDNA (Postel and Goodgal, 1966), although this result
remains an ideal organism for the study of natural trans-
has also been challenged (Mulder and Doty, 1968).
formation and will be a crucial tool for understanding bac-
Recently a third Gram-negative bacterial species,
terial competence for many years to come.
Pseudomonas stutzeri, was shown to be transformed by
ssDNA (Meier et al., 2002). Additionally, N. meningitidis
Acknowledgements
PilQ, which forms the outer membrane ‘pore’ of the pilus
apparatus and is required for natural transformation, has We acknowledge the Gonococcal Genome Sequencing
been recently found to bind ssDNA better than dsDNA Project supported by USPHS/NIH Grant #AI38399, and
(Frye et al., 2004). These facts argue that ssDNA should B.A. Roe, L. Song, S.P. Lin, X. Yuan, S. Clifton, T. Ducey,
be a good substrate for transformation of N. gonorrhoeae. L. Lewis and D.W. Dyer at the University of Oklahoma. We
Additionally, DNA secreted by the gonococcal T4SS, would like to acknowledge the support from our laboratory
which is known to transform recipient bacteria, is presum- through NIH Grant AI47958 to J.P.D. and traineeship on
ably single-stranded (Hamilton et al., 2001). the NIH T32 Grant AI055397 to H.L.H. We apologize to

© 2005 The Authors

Journal compilation © 2005 Blackwell Publishing Ltd, Molecular Microbiology, 59, 376–385

those whose work contributed to the fields discussed in
this review but were not specifically mentioned.
References
Aas, F.E., Wolfgang, M., Frye, S., Dunham, S., Lovold, C.,
and Koomey, M. (2002a) Competence for natural transfor-
mation in Neisseria gonorrhoeae: components of DNA
binding and uptake linked to type IV pilus expression. Mol
Microbiol 46: 749–760.
Aas, F.E., Lovold, C., and Koomey, J.M. (2002b) An inhibitor
of DNA binding and uptake events dictates the proficiency
of genetic transformation in Neisseria gonorrhoeae: mech-
anism of action and links to type IV pilus expression. Mol
Microbiol 46: 1441–1450.
Alexander, H.L., Richardson, A.R., and Stojiljkovic, I. (2004)
Natural transformation and phase variation modulation in
Neisseria meningitidis. Mol Microbiol 52: 771–783.
Aukema, K.G., Kron, E.M., Herdendorf, T.J., and Forest, K.T.
(2005) Functional dissection of a conserved motif within
the pilus protein PilT. J Bacteriol 187: 611–618.
Belland, R.J., Morrison, S.G., and Hogan, D. (1996) A phase-
variable type III restriction-modification system in Neisseria
gonorrhoeae. In Abstracts of the Tenth International Patho-
genic Neisseria Conference. Zollinger, W.D., Frasch, C.E.,
and Deal, C.D. (eds). http://neisseria.org/ipnc.
Biswas, G.D., and Sparling, P.F. (1981) Entry of double-
stranded deoxyribonucleic acid during transformation of
Neisseria gonorrhoeae. J Bacteriol 145: 638–640.
Biswas, G.D., Sox, T., Blackman, E., and Sparling, P.F.
(1977) Factors affecting genetic transformation of Neis-
seria gonorrhoeae. J Bacteriol 129: 983–992.
Biswas, G.D., Burnstein, K.L., and Sparling, P.F. (1986) Lin-
earization of donor DNA during plasmid transformation in
Neisseria gonorrhoeae. J Bacteriol 168: 756–761.
Biswas, G.D., Lacks, S.A., and Sparling, P.F. (1989) Trans-
formation-deficient mutants of piliated Neisseria gonor-
rhoeae. J Bacteriol 171: 657–664.
Bos, M.P., Tefsen, B., Voet, P., Weynants, V., Putten,
J.P.M., and Tommassen, J. (2005) Function of neisserial
outer membrane phospholipase A in autolysis and
assessment of its vaccine potential. Infect Immun 73:
2222–2231.
Bourniquel, A.A., and Bickle, T.A. (2002) Complex restriction
enzymes: NTP-driven molecular motors. Biochimie 84:
1047–1059.
Boyle-Vavra, S., and Seifert, H.S. (1996) Uptake-sequence-
independent DNA transformation exists in Neisseria gon-
orrhoeae. Microbiology 142: 2839–2845.
Butler, C.A., and Gotschlich, E.C. (1991) High-frequency
mobilization of broad-host-range plasmids into Neisseria
gonorrhoeae requires methylation in the donor. J Bacteriol
173: 5793–5799.
Cacciapuoti, A.F., Wegener, W.S., and Morse, S.A. (1978)
Cell envelope of Neisseria gonorrhoeae: phospholipase
activity and its relationship to autolysis. Infect Immun 20:
418–420.
Chaussee, M.S., and Hill, S.A. (1998) Formation of single-
stranded DNA during transformation of Neisseria gonor-
rhoeae. J Bacteriol 180: 5117–5122.
© 2005 The Authors
Journal compilation © 2005 Blackwell Publishing Ltd, Molecular Microbiology, 59, 376–385
Natural transformation of Neisseria gonorrhoeae 383
Chen, I., and Dubnau, D. (2004) DNA uptake during bacterial
transformation. Nat Rev Microbiol 2: 241–249.
Chen, I., and Gotschlich, E.C. (2001) ComE, a competence
protein from Neisseria gonorrhoeae with DNA-binding
activity. J Bacteriol 183: 3160–3168.
Cooke, S.J., Jolley, K., Ison, C.A., Young, H., and Heckels,
J.E. (1998) Naturally occurring isolates of Neisseria gon-
orrhoeae, which display anomalous serovar properties,
express PIA/PIB hybrid porins, deletions in PIB or novel
PIA molecules. FEMS Microbiol Lett 162: 75–82.
Davidsen, T., Rodland, E.A., Lagesen, K., Seeberg, E., Tor-
bjorn, R., and Tonjum, T. (2004) Biased distribution of DNA
uptake sequences towards genome maintenance genes.
Nucleic Acids Res 32: 1050–1058.
Dillard, J.P., and Seifert, H.S. (2001) A variable genetic island
specific for Neisseria gonorrhoeae is involved in providing
DNA for natural transformation and is found more often in
disseminated infection isolates. Mol Microbiol 41: 263–
278.
Dougherty, T.J., Asmus, A., and Tomasz, A. (1979) Specific-
ity of DNA uptake in genetic transformation of gonococci.
Biochem Biophys Res Commun 86: 97–104.
Drake, S.L., and Koomey, M. (1995) The product of the pilQ
gene is essential for the biogenesis of type IV pili in Neis-
seria gonorrhoeae. Mol Microbiol 18: 975–986.
Draskovic, I., and Dubnau, D. (2005) Biogenesis of a putative
channel protein, ComEC, required for DNA uptake: mem-
brane topology, oligomerization and formation of disulphide
bonds. Mol Microbiol 55: 881–896.
Dubnau, D. (1999) DNA uptake in bacteria. Annu Rev Micro-
biol 53: 217–244.
Eisenstein, B.I., Sox, T., Biswas, G., Blackman, E., and Spar-
ling, P.F. (1977) Conjugal transfer of the gonococcal pen-
icillinase plasmid. Science 195: 998–1000.
Facius, D., and Meyer, T.F. (1993) A novel determinant
(comA) essential for natural transformation competence in
Neisseria gonorrhoeae and the effect of a comA mutant on
pilin variation. Mol Microbiol 10: 699–712.
Facius, D., Fussenegger, M., and Meyer, T.F. (1996)
Sequential action of factors involved in natural competence
for transformation of Neisseria gonorrhoeae. FEMS Micro-
biol Lett 137: 159–164.
Finkel, S.E., and Kolter, R. (2001) DNA as a nutrient: novel
role for bacterial competence gene homologs. J Bacteriol
183: 6288–6293.
Freitag, N.E., Seifert, H.S., and Koomey, M. (1995) Charac-
terization of the pilF-pilD pilus-assembly locus of Neisseria
gonorrhoeae. Mol Microbiol 16: 575–586.
Frye, S., Assalkhou, R., Balasingham, S., Tuven, H.K.,
Benam, A.V., Homberset, H., and Tonjum, T. (2004) The
meningococcal secretin PilQ: actions and interactions. In
14th International Pathogenic Neisseria Conference.
Apicella, M.A., and Seifert, H.S. (eds). http://neisseria.org/
ipnc.
Fudyk, T.C., Maclean, I.W., Simonsen, J.N., Njagi, E.N.,
Kimani, J., Brunham, R.C., and Plummer, F.A. (1999)
Genetic diversity and mosaicism at the por locus of Neis-
seria gonorrhoeae. J Bacteriol 181: 5591–5599.
Fussenegger, M., Facius, D., Meier, J., and Meyer, T.F.
(1996a) A novel peptidoglycan-linked lipoprotein
(ComL) that functions in natural transformation com-
384 H. L. Hamilton and J. P. Dillard
petence of Neisseria gonorrhoeae. Mol Microbiol 19:
1095–1105.
Fussenegger, M., Kahrs, A.F., Facius, D., and Meyer, T.F.
(1996b) Tetrapac (tpc), a novel genotype of Neisseria gon-
orrhoeae affecting epithelial cell invasion, natural transfor-
mation competence and cell separation. Mol Microbiol 19:
1357–1372.
Gerdes, K., Christensen, S.K., and Lonbner-Olsen, A. (2005)
Prokaryotic toxin–antitoxin stress response loci. Nat Rev
Microbiol 3: 371–382.
Gibbs, C.P., Reimann, B.Y., Schultz, E., Kaufmann, A., Haas,
R., and Meyer, T.F. (1989) Reassortment of pilin genes in
Neisseria gonorrhoeae occurs by two distinct mechanisms.
Nature 338: 651–652.
Goldberg, I.D., Steinberg, V.I., Siddiqui, A., Hart, E.J., and
Schaper, D. (1978) Attempts to isolate bacteriophage
specific for Neisseria gonorrhoeae. In Immunobiology of
Neisseria gonorrhoeae. Brooks, G.F., Gotschlich, E.C.,
Holmes, K.K., Sawyer, W.D. and Young, F.E. (eds). Wash-
ington, DC: American Society for Microbiology, pp. 55–59.
Goodgal, S.H. (1982) DNA uptake in Haemophilus transfor-
mation. Annu Rev Genet 16: 169–192.
Goodman, S.D., and Scocca, J.J. (1988) Identification and
arrangement of the DNA sequence recognized in specific
transformation of Neisseria gonorrhoeae. Proc Natl Acad
Sci USA 85: 6982–6986.
Gunn, J.S., and Stein, D.C. (1996) Use of a nonselective
transformation technique to construct a multiply restriction/
modification-deficient mutant of Neisseria gonorrhoeae.
Mol Gen Genet 251: 509–517.
Hamilton, H.L., Schwartz, K.J., and Dillard, J.P. (2001) Inser-
tion-duplication mutagenesis of Neisseria: use in charac-
terization of DNA transfer genes in the gonococcal genetic
island. J Bacteriol 183: 4718–4726.
Hamilton, H.L., Dominguez, N.M., Schwartz, K.J., Hackett,
K.T., and Dillard, J.P. (2005) Neisseria gonorrhoeae
secretes chromosomal DNA via a novel type IV secretion
system. Mol Microbiol 55: 1704–1721.
Hebeler, B.H., and Young, F.E. (1975) Autolysis of Neisseria
gonorrhoeae. J Bacteriol 122: 385–392.
Hebeler, B.H., and Young, F.E. (1976) Mechanism of autol-
ysis of Neisseria gonorrhoeae. J Bacteriol 126: 1186–
1193.
Herdendorf, T.J., McCaslin, D.R., and Forest, K.T. (2002)
Aquifex aeolicus PilT, homologue of a surface motility pro-
tein, is a thermostable oligomeric NTPase. J Bacteriol 126:
6465–6471.
Hobbs, M.M., Seiler, A., Achtman, M., and Cannon, J.G.
(1994) Microevolution within a clonal population of patho-
genic bacteria: recombination, gene duplication and hori-
zontal genetic exchange in the opa gene family of
Neisseria meningitidis. Mol Microbiol 12: 171–180.
Jonsson, A.B., Nyberg, G., and Normark, S. (1991) Phase
variation of gonococcal pili by frameshift mutation in pilC,
a novel gene for pilus assembly. EMBO J 10: 477–488.
Koomey, M. (1998) Competence for natural transformation in
Neisseria gonorrhoeae: a model system for studies of hor-
izontal gene transfer. APMIS Suppl 84: 56–61.
Koomey, J.M., and Falkow, S. (1987) Cloning of the recA
gene of Neisseria gonorrhoeae and construction of gono-
coccal recA mutants. J Bacteriol 169: 790–795.
Journal compilation © 2005 Blackwell Publishing Ltd, Molecular Microbiology, 59, 376–385
Kroll, J.S., Wilks, K.E., Farrant, J.L., and Langford, P.R.
(1998) Natural genetic exchange between Haemophilus
and Neisseria: intergeneric transfer of chromosomal genes
between major human pathogens. Proc Natl Acad Sci USA
95: 12381–12385.
Long, C.D., Madraswala, R.N., and Seifert, H.S. (1998) Com-
parisons between colony phase variation of Neisseria gon-
orrhoeae FA1090 and pilus, pilin, and S-pilin expression.
Infect Immun 66: 1918–1927.
Long, C.D., Tobiason, D.M., Lazio, M.P., Kline, K.A., and
Seifert, H.S. (2003) Low-level pilin expression allows for
substantial DNA transformation competence in Neisseria
gonorrhoeae. Infect Immun 71: 6279–6291.
Lorenz, M.G., and Wackernagel, W. (1994) Bacterial gene
transfer by natural genetic transformation in the environ-
ment. Microbiol Rev 58: 563–602.
Mathis, L.S., and Scocca, J.J. (1984) On the role of pili in
transformation of Neisseria gonorrhoeae. J Gen Microbiol
130: 3165–3173.
Mehr, I.J., and Seifert, H.S. (1998) Differential roles of homol-
ogous recombination pathways in Neisseria gonorrhoeae
pilin antigenic variation, DNA transformation and DNA
repair. Mol Microbiol 30: 697–710.
Meier, P., Berndt, C., Weger, N., and Wackernagel, W.
(2002) Natural transformation of Pseudomonas stutzeri by
single-stranded DNA requires type IV pili, competence
state, and comA. FEMS Microbiol Lett 207: 75–80.
Morse, S.A., and Bartenstein, L. (1974) Factors affecting
autolysis of Neisseria gonorrhoeae. Proc Soc Exp Biol Med
145: 1418–1421.
Mulder, C., and Doty, P. (1968) Residual activity of denatured
transforming DNA of Haemophilus influenzae: a naturally
occurring cross-linked DNA. J Mol Biol 32: 423–435.
Norlander, L., Davies, J., and Normark, S. (1979) Genetic
exchange mechanisms in Neisseria gonorrhoeae. J Bac-
teriol 138: 756–761.
Postel, E.H., and Goodgal, S.H. (1966) Uptake of ‘single-
stranded’ DNA in Haemophilus influenzae and its ability to
transform. J Mol Biol 16: 317–327.
Provvedi, R., Chen, I., and Dubnau, D. (2001) NucA is
required for DNA cleavage during transformation of Bacil-
lus subtilis. Mol Microbiol 40: 634–644.
Rudel, T., Facius, D., Barten, R., Scheuerpflug, I., Nonnen-
macher, E., and Meyer, T.F. (1995) Role of pili and the
phase-variable PilC protein in natural competence for
transformation of Neisseria gonorrhoeae. Proc Natl Acad
Sci USA 92: 7986–7990.
Sarubbi, F.A., Jr, Blackman, E., and Sparling, P.F. (1974)
Genetic mapping of linked antibiotic resistance loci in Neis-
seria gonorrhoeae. J Bacteriol 120: 1284–1292.
Snyder, L.A.S., Butcher, S.A., and Saunders, N.J. (2001a)
Comparative whole-genome analyses reveal over 100
putative phase-variable genes in the pathogenic Neisseria
spp. Microbiology 147: 2321–2332.
Snyder, L.A.S., Saunders, N.J., and Shafer, W.M. (2001b) A
putatively phase variable gene (dca) required for natural
transformation in Neisseria gonorrhoeae but not Neisseria
meningitidis is located within the division cell wall (dcw)
gene cluster. J Bacteriol 183: 1233–1241.
Snyder, L.A.S., Davies, J.K., and Saunders, N.J. (2004)
Microarray genomotyping of key experimental strains of
© 2005 The Authors

Neisseria gonorrhoeae reveals gene complement diversity
and five new neisserial genes associated with Minimal
Mobile Elements. BMC Genomics 5: 23.
Solomon, J.M., and Grossman, A.D. (1996) Who’s competent
and when: regulation of natural genetic competence in
bacteria. Trends Genet 12: 150–155.
Sox, T.E., Mohammed, W., Blackman, E., Biswas, G., and
Sparling, P.F. (1978) Conjugative plasmids in Neisseria
gonorrhoeae. J Bacteriol 134: 278–286.
Sox, T.E., Mohammed, W., and Sparling, P.F. (1979) Trans-
formation-derived Neisseria gonorrhoeae plasmids with
altered structure and function. J Bacteriol 138: 510–518.
Sparling, P.F. (1966) Genetic transformation of Neisseria
gonorrhoeae to streptomycin resistance. J Bacteriol 92:
1364–1371.
Spratt, B.G., Bowler, L.D., Zhang, Q.Y., Zhou, J., and Smith,
J.M. (1992) Role of interspecies transfer of chromosomal
genes in the evolution of penicillin resistance in pathogenic
and commensal Neisseria species. J Mol Evol 34: 115–
125.
Stein, D.C. (1991) Transformation of Neisseria gonorrhoeae:
physical requirements of the transforming DNA. Can J
Microbiol 37: 345–349.
Stein, D.C., Gunn, J.S., Radlinska, M., and Piekarowicz, A.
(1995) Restriction and modification systems of Neisseria
gonorrhoeae. Gene 157: 19–22.
Steinmoen, H., Knutsen, E., and Havarstein, L.S. (2002)
Induction of natural competence in Streptococcus pneumo-
© 2005 The Authors
Journal compilation © 2005 Blackwell Publishing Ltd, Molecular Microbiology, 59, 376–385
Natural transformation of Neisseria gonorrhoeae 385
niae triggers lysis and DNA release from a subfraction of
the cell population. Proc Natl Acad Sci USA 99: 7681–
7686.
Stern, A., Brown, M., Nickel, P., and Meyer, T.F. (1986)
Opacity genes in Neisseria gonorrhoeae: control of phase
and antigenic variation. Cell 47: 61–71.
Stohl, E.A., and Seifert, H.S. (2001) The recX gene potenti-
ates homologous recombination in Neisseria gonorrhoeae.
Mol Microbiol 40: 1301–1310.
Stohl, E.A., Brockman, J.P., Burke, K.L., Morimatsu, K.,
Kawalczykowski, S.C., and Seifert, H.S. (2003) Escheri-
chia coli RecX inhibits RecA recombinase and coprotease
activities in vitro and in vivo. J Biol Chem 278: 2278–
2285.
Winther-Larsen, H.C., Wolfgang, M., Dunham, S., van
Putten, J.P.M., Dorward, D., Lovold, C., et al. (2005) A
conserved set of pilin-like molecules controls type IV pilus
dynamics and organelle-associated functions in Neisseria
gonorrhoeae. Mol Microbiol 56: 903–917.
Wolfgang, M., Lauer, P., Park, H.S., Brossay, L., Hebert, J.,
and Koomey, M. (1998) PilT mutations lead to simulta-
neous defects in competence for natural transformation
and twitching motility in piliated Neisseria gonorrhoeae.
Mol Microbiol 29: 321–330.
Zhou, J., and Spratt, B.G. (1992) Sequence diversity within
argF, fbp and recA genes of natural isolates of Neisseria
meningitidis: interspecies recombination within the argF
gene. Mol Microbiol 6: 2135–2146.

States following Cretaceous thickening of the
lithosphere (26). These stable isotope data alone
do not preclude Pliocene surface uplift for the
northern Sierra Nevada due to mantle de-
lamination as has been proposed for the south-
ern Sierra Nevada (6–8). In fact, our estimates
of post-Eocene uplift are within the range of
elevation change predictable for mantle delam-
ination and may actually constrain the magni-
tude of mantle control on Pliocene surface uplift
in the northern Sierra Nevada (Fig. 4). However,
they are incompatible with a model of an
Eocene low-elevation landscape that experi-
enced substantial (1.5 to 2.0 km) Pliocene
surface uplift. Calling upon a Pliocene age for
major surface uplift in the northern Sierra
Nevada would either require post-Eocene level-
ing of the mountains with subsequent renewed
surface uplift or a tectonic and mechanical
decoupling of the northern and southern Sierra
Nevada, both of which are unrecognized in the
geologic record.
Antibiotic Stress Induces Genetic
Transformability in the Human
Pathogen Streptococcus pneumoniae
Marc Prudhomme,* Laetitia Attaiech,* Guillaume Sanchez,
Bernard Martin, Jean-Pierre Claverys†
Natural transformation is a widespread mechanism for genetic exchange in bacteria.
Aminoglycoside and fluoroquinolone antibiotics, as well as mitomycin C, a DNA-damaging
agent, induced transformation in Streptococcus pneumoniae. This induction required an intact
competence regulatory cascade. Furthermore, mitomycin C induction of recA was strictly dependent
on the development of competence. In response to antibiotic stress, S. pneumoniae, which lacks
an SOS-like system, exhibited genetic transformation. The design of antibiotherapy should take into
consideration this potential of a major human pathogen to increase its rate of genetic exchange
in response to antibiotics.
acterial transformation, originally discov-
B ered in the human pathogen S. pneumoniae
(1), relies on a process that is inherent
to the species, is independent of extrachromo-
somal elements, and can be considered the only
programmed mechanism for generalized genetic
exchange in bacteria. It allows the uptake and
integration of exogenous DNA in the recipient
genome and is considered to be a form of para-
sexuality (2). Transformation is believed to con-
tribute to the genetic plasticity of S. pneumoniae
and to play a central role in the adaptation of this
pathogen to host defenses (3). We sought to
establish whether transformation is induced in
response to antibiotic stress.
Laboratoire de Microbiologie et Génétique Moléculaires,
UMR 5100 CNRS-Université Paul Sabatier, 118 route de
Narbonne, 31062 Toulouse Cedex 9, France.
*These authors contributed equally to this work.
†To whom correspondence should be addressed. E-mail:
claverys@ibcg.biotoul.fr
www.sciencemag.org
References and Notes
1. P. Molnar, P. England, Nature 346, 29 (1990).
2. C. Beaumont, R. A. Jamieson, M. H. Nguyen, B. Lee,
Nature 414, 738 (2001).
3. J. R. Unruh, Geol. Soc. Am. Bull. 103, 1395 (1991).
4. J. Wakabayashi, T. L. Sawyer, J. Geol. 109, 539 (2001).
5. C. H. Jones, G. L. Farmer, J. Unruh, Geol. Soc. Am. Bull.
116, 1408 (2004).
6. G. Zandt et al., Nature 431, 41 (2004).
7. M. N. Ducea, J. Saleeby, Contrib. Mineral. Petrol. 133,
169 (1998).
8. J. Saleeby, M. N. Ducea, D. Clemens-Knott, Tectonics 22,
doi:10.1029/2002TC001374 (2003).
9. W. R. Dickinson, R. V. Ingersoll, S. A. Graham, Geol. Soc.
Am. Bull. 90, 1458 (1979).
10. B. Wernicke et al., Science 271, 190 (1996).
11. M. A. House, B. P. Wernicke, K. A. Farley, Nature 396, 66
(1998).
12. M. A. Poage, C. P. Chamberlain, Tectonics 21, 1601 (2002).
13. M. K. Clark, G. Maheo, J. Saleeby, K. A. Farley, GSA Today
15, 4 (2005).
14. G. M. Stock, R. S. Anderson, R. C. Finkel, Earth Surf.
Process. Landforms 30, 985 (2005).
15. E. E. Small, R. S. Anderson, Science 270, 277 (1995).
16. W. Lindgren, U.S. Geol. Surv. Prof. Pap. 73 (1911).
17. W. H. Yeend, U.S. Geol. Surv. Prof. Pap. 722 (1974).
18. Available as supporting material on Science Online.
In S. pneumoniae, competence for genetic
transformation is a transient physiological state
allowing efficient DNA uptake (4) and a pre-
viously unrecognized capacity to kill non-
competent cells (5, 6), a phenomenon referred
to as pneumococcal fratricide (6) or sobrinicide
(7). The development of competence requires
transcriptional activation of the com regulon,
which comprises 105 to 124 genes (8, 9), in-
cluding recA (10). The RecA protein plays a
key role in transformation by catalyzing homol-
ogous recombination between the internalized
DNA and the recipient genome. The com regulon
is induced when a competence-stimulating
peptide (CSP), encoded by comC and exported
through a dedicated secretion apparatus ComAB,
accumulates in the medium and stimulates its
receptor, the membrane-bound histidine kinase
ComD (4). It is assumed that ComD then auto-
phosphorylates and transphosphorylates its
cognate response regulator ComE (11), which
SCIENCE VOL 313 7 JULY 2006
REPORTS
19. C. N. Garzione, J. Quade, P. G. DeCelles, N. B. English,
Earth Planet. Sci. Lett. 183, 215 (2000).
20. D. B. Rowley, R. T. Pierrehumbert, B. S. Currie,
Earth Planet. Sci. Lett. 188, 253 (2001).
21. A. Mulch, C. Teyssier, M. A. Cosca, O. Vanderhaeghe,
T. W. Vennemann, Geology 32, 525 (2004).
22. M. A. Poage, C. P. Chamberlain, Am. J. Sci. 301, 1 (2001).
23. N. L. Ingraham, B. E. Taylor, Water Resources Res. 27, 77
(1991).
24. IAEA, http://isohis.iaea.org/userupdate/GNIP2001yearly.xls
(2001).
25. C. D. Henry, M. E. Perkins, Geology 29, 719 (2001).
26. P. G. DeCelles, Am. J. Sci. 304, 105 (2004).
27. Spatial Climate Analysis Service, Oregon State University,
www.ocs.oregonstate.edu/prism, created April 2005.
28. We acknowledge C. H. Jones and two anonymous
reviewers for thorough and constructive reviews. This work
was supported by NSF grant EAR-0309011 to C.P.C.
Supporting Online Material
www.sciencemag.org/cgi/content/full/313/5783/87/DC1
Materials and Methods
Figs. S1 and S2
Table S1
References and Notes
8 February 2006; accepted 22 May 2006
10.1126/science.1125986
in turn activates the expression of the so-called
early com genes (4), including comAB,
comCDE, and comX. The latter encodes an
alternative sigma factor sX (12), which most
probably recognizes a sequence (TACGAATA,
hereafter called Pcin) conserved in the putative
promoter regions of the late com genes (4, 9).
The early control of competence induction is
not yet fully understood. It was first suggested
that competence induction relies simply on passive
CSP accumulation, but we favor an alternative
model in which CSP production could be
temporarily increased in response to changes
in environmental conditions (4, 13). We further
propose that competence in S. pneumoniae is
a general stress response, playing a role simi-
lar to that of the SOS response in Escherichia
coli (4).
We tested this hypothesis by investigating
the effect of mitomycin C, a DNA-damaging
agent known to induce the SOS response, on
the com regulon. To monitor competence in-
duction (14), we used transcriptional fusions
with the luc gene from the firefly (Photinus
pyralis) encoding luciferase, the activity of
which can be monitored directly in living S.
pneumoniae cells (14). We first used a fusion of
luc with ssbB, a representative of the late com
genes (8, 9), which encodes a single-stranded
DNA–binding protein (15). Luciferase activity
was monitored during growth with various con-
centrations of mitomycin C. Expression of the
reporter was stimulated by exposure to 25 to 60
ng mlj1 of mitomycin C (Fig. 1A and fig.
S1A), indicating that ssbB was induced when
cells were grown in the presence of the DNA-
damaging agent. No induction of the same
fusion was detected in a comA mutant
background Ethat is, in a strain unable to export
CSP and therefore to develop spontaneous
competence (Fig. 1B)^, demonstrating that an
89
REPORTS

A wild type B comA - D E

0.12 12 0.9
15 1.2

RLU/OD (x10 )
-3
RLU/OD (x10 )

0.09 9
-1

CSP MC 0.6

OD492

OD492
10 MC MC 0.8

OD492
0.06 6
0.3
5 0.4 0.03 3

0 0 0
0 100 200 300 0 100 200 300 0 50 100 150 200 0 50 100 150 200
min at 37°C min at 37°C
C
Pcin Pa Fig. 1. Mitomycin C (MC) induction of ssbB and recA. (A and B) Luciferase
cinA recA luc cat 'recA dinF
activity expressed in relative luminescence units (RLU)/optical density (OD)
(triangles) and OD492 (squares) of cultures in CþY medium of (A) strain
R895 (ssbB::luc) and (B) strain R1313 (ssbB::luc, comA–) without and with
mitomycin C (40 ng mlj1). Curves of luciferase activity [with (red triangles) and without (gray triangles) mitomycin C] and OD [with (black squares)
and without (gray squares) mitomycin C] represent compilations of data from 19 and 32 replicate cultures, respectively in (A) and (B). Standard
deviations are indicated for luciferase activities only. (C) Structure of the recA operon with the recA::luc fusion generated by the integration of plasmid
pR432. Plasmid sequences are not drawn to scale. Pcin and Pa are indicated by the small branched arrows. (D and E) Luciferase activity (triangles) and
OD492 (squares) of cultures in CþY medium of strain R1624 (recA::luc, DcomC). In (D), solid blue triangles and solid black squares indicate the
presence of CSP (100 ng mlj1), whereas open gray symbols indicate the absence of CSP. In (E), symbols indicate the following concentrations of
mitomycin C: solid gray symbols (0 ng mlj1), red triangles and open gray squares (25 ng mlj1), open black symbols (100 ng mlj1), and solid black
symbols (400 ng mlj1). Arrows indicate the addition of CSP or mitomycin C after 70 min of incubation.

intact competence regulatory cascade was Fig. 2. Antibiotics induce wild type comA -
required for induction by mitomycin C. These the competence regulon. (A
data established that a DNA-damaging agent to F) Luciferase activity (tri- A D 1.0
induced the com regulon of S. pneumoniae. angles) and OD492 (squares) 30
Mitomycin C is known to trigger the SOS of cultures of strain R895
(ssbB::luc, wild type) [(A) to 20
response in E. coli. DNA damage caused by Nf Nf 0.5
mitomycin C blocks the replication fork, gen- (C)], strain R1313 (ssbB::luc,
10
erating a single-stranded DNA region to which comA–) [(D) and (F)], and
RecA binds to form a nucleoprotein filament strain R1047 (E) (ssbB::luc, 0 0
comA–) [(D) to (F)] in CþY
(RecA*) (16). The coprotease activity of B E
medium with (red and black 15 1.0
RecA* then catalyzes the self-cleavage of the
RLU/OD (x10 )

symbols) and without (gray

-1

SOS repressor LexA (17), which leads to in- symbols) antibiotics. Antibi- Kn Kn
10

OD492
duction of the SOS genes, including recA. In S. otics were added after 70
pneumoniae, the recA gene has been shown to 0.5
min of incubation (arrows) in
be expressed from two promoters: Pcin, which the following concentrations: 5
generates a 5.7-kb-long transcript in competent 11 mg mlj1 of norfloxacin
cells, and Pa, a sA promoter that directs the (Nf) [(A) and (D)], 31.25 mg 0 0
synthesis of a 4.3-kb-long transcript (18) (Fig. mlj1 of kanamycin (Kn) [(B)
1C). The competence-specific induction of and (E)], and 12.5 mg mlj1 10 C F 1.0
Sm Sm
recA (that is, expression from Pcin) accounts of streptomycin (Sm) [(C) and
for 95% of the transformation (10). To establish (F)]. Curves of luciferase ac-
whether mitomycin C could induce recA ex- tivity (with standard devia- 0.5
5
pression, a recA::luc fusion was constructed tions) and OD represent
(Fig. 1C) and validated by measuring its in- compilations of data from 8
duction with CSP (Fig. 1D). Mitomycin C was cultures [(A) and (D)], 15 cul-
found to activate this fusion with kinetics sim- tures [(C) and (F)], or 16 cul- 0 0
ilar to that of ssbB (fig. S2). The recA::luc tures [(B) and (E)] of strains 0 100 200 300 0 100 200 300 400
construct was introduced in a strain unable to with the respective antibi-
otics. See Fig. 1 legend for min at 37°C
develop spontaneous competence because of
details.
the deletion of the entire comC coding region.
The DNA-damaging agent did not induce luc
expression in this genetic background (Fig. 1E), mechanism (7), instead uses the competence mechanisms probably reflects the need for a
demonstrating that the induction of recA by regulatory cascade to coordinate a response to slow accumulation of inducing lesions in both
mitomycin C occurs only from Pcin and is mitomycin C. Induction of the com regulon in species.
therefore strictly dependent on the ability of S. pneumoniae and of the SOS response in E. Several antibiotics are known to induce
this compound to induce competence. This coli occurs only after prolonged incubation, at the SOS response in SOS-proficient bacteria
observation supports the hypothesis that S. È2.5 hours (Fig. 1A) and 2 hours, respectively (20, 21). To investigate the parallels between
pneumoniae, which lacks an SOS-like induction (19). The similar delay of unrelated regulatory competence induction and the SOS response,

90 7 JULY 2006 VOL 313 SCIENCE www.sciencemag.org

 REPORTS

Fig. 3. Induction of genetic transformation by streptomycin. (A and B)
Luciferase activity (A) and growth [(B), OD492 and colony-forming units
(CFUs) per milliliter] of cultures of strain R895 with 0 (open squares),
6.25 (open triangles), 12.5 (solid gray triangles), or 25 (solid black
squares) mg mlj1 of streptomycin (Sm). Streptomycin was added after 70
min of incubation [arrow in (A)]. See Fig. 1 legend for details. Cell survival
was monitored by plating aliquots [dotted lines in (B)]. (C) Luciferase activity
of strain R895 grown without (squares) and with (triangles) 6.25 mg mlj1 of
streptomycin. CSP (100 ng mlj1) was added (right arrow) after 182 min
(solid black triangles). (D) Genetic transformation in aliquots from cultures in
(C) taken after 195 min of incubation and mixed with chromosomal DNA
carrying a marker conferring resistance to streptomycin (14). The monitoring
of transformation with a marker conferring resistance to streptomycin is
unrelated to the use of streptomycin to induce competence. The number of
streptomycin-resistant chromosomal transformants obtained corresponds to
transformation frequencies of 0.65% (Sm) and 0.57% (SmþCSP).

we measured luciferase synthesis from the
ssbB::luc fusion during growth using a wide
range of concentrations of various antibiotics
(table S2). Among the protein synthesis inhib-
itors that were tested, kanamycin and strepto-
mycin triggered competence (Fig. 2, B and C,
and fig. S1A), but erythromycin and tetracy-
cline did not (fig. S3, A and B). The fluo-
roquinolones (norfloxacin, levofloxacin, and
moxifloxacin, the latter two of which are used
for the treatment of respiratory tract infections),
which target type II topoisomerases, DNA gyrase,
and topoisomerase IV (22), were found to
induce the ssbB::luc fusion (Fig. 2A and fig.
S1, B and C). No induction was detected with
the DNA gyrase inhibitor novobiocin (fig.
S3B), the RNA polymerase inhibitor rifampi-
cin (fig. S3B), the glycopeptide antibiotic
vancomycin (fig. S3B), or with the b-lactams
ampicillin (fig. S3A) and the third-generation
cephalosporin, cefotaxime (fig. S3A). No cor-
relation could be made between the intensity of
growth inhibition and the induction of compe-
tence (table S2). Similarly to mitomycin C, the
induction of ssbB::luc by aminoglycosides and
norfloxacin required an intact competence
regulatory cascade, because no induction could
be detected in a comA mutant (Fig. 2, D to F).
Induction of the com regulon normally al-
lows competent cells to take up and integrate
exogenous DNA. To test whether antibiotic-
induced competence resulted in a bona fide
transformation, transforming DNA was added
to a culture treated with streptomycin. For the
transformation assay, we selected an intermedi-
ate concentration of streptomycin (625 ng mlj1),
which did not cause severe killing (Fig. 3B).
Chromosomal transformants were readily
obtained in the streptomycin-treated culture,
whereas no transformants were present in the
control culture (Fig. 3, C and D). The addition
of CSP (100 ng mlj1) to the streptomycin-
treated culture did not further increase ssbB::luc
expression (Fig. 3C) or the yield of transfor-
mants (Fig. 3D), demonstrating full competence
induction by the antibiotic. Chromosomal trans-
formants were also obtained in parallel cultures
treated with mitomycin C (60 ng mlj1) and
norfloxacin (10 mg mlj1) (fig. S4).
The induction by mitomycin C and fluo-
roquinolones of the SOS response in E. coli and
of competence in S. pneumoniae suggests that
SOS and competence play similar roles in both
species. However, the parallel is only partial,
because competence was not induced by anti-
biotics that disrupt cell wall integrity (such as
b-lactams), which is contrary to the SOS sys-
tem in E. coli (21). Aminoglycosides that in-
duce the com regulon (such as kanamycin and
streptomycin) did not trigger an SOS response
in E. coli but instead induced heat-shock pro-
tein expression (23), suggesting that similar
stress signals are processed differently in the
two species. Mitomycin C and fluoroquinolones
may generate a common signal—chromosome
replication arrest—owing to the formation of
interstrand cross-links (mitomycin C) or the pres-
ence of covalently bound topoisomerases (fluo-
roquinolones) (24). The situation with ribosome
inhibitors is more complex, because some such
as kanamycin and streptomycin act as com-
petence inducers (Fig. 2, B and C), whereas
others such as erythromycin and tetracycline do
not (fig. S3, A and B). A parallel can be made
with the situation in E. coli: Kanamycin and
streptomycin, which leave the ribosomal A site
empty, induce a heat-shock–like response,
whereas erythromycin and tetracycline, which
either fill the A site with aminoacyl–transfer
RNA (tRNA) or block it, trigger a cold-shock–
like response (23). The level of ppGpp, which
is produced by ribosomes that are stalled by a
lack of charged tRNAs, decreases upon the
addition of erythromycin or tetracycline (or a
reduction in growth temperature) and increases
as the growth temperature is increased (23, 25),
suggesting that this nucleotide could act as a
signal (4).
Whatever the underlying mechanism(s) may
be, the induction of the com regulon by various
antibiotics supports the hypothesis that com-
petence is a general stress response of S.
pneumoniae (4, 7) and is consistent with the
proposal that CSP is not an effector of quorum
sensing (7) but is an alarmone that conveys a
stress signal (4). As a coordinator of compe-
tence, CSP enhances the efficiency of transfor-
mation as a rescue process in two ways: by
increasing the number of potential transfor-
mants (3) and by triggering fratricide (5) through
allolysis (defined as lysis in trans) and the
release of DNA from CSP-nonresponsive pneu-
mococcal cells. Whenever these cells differ
from the competent population (for example,
during cocolonization), allolysis provides a
source of genetically diverse DNA. Through
the coupling of fratricide and transformation,
CSP could thus play a crucial role in generating
genetic diversity under stress conditions for a
species that seems unable to rely on inducible
mutagenic repair (such as the SOS response)
(7). Consequently, the high incidence of asymp-
tomatic carriage of this pathogen is a major
concern, because inappropriate antibiotic treat-
ments could accelerate the occurrence of addi-
tional resistant clones and promote the evolution
of virulence.
References and Notes
1. F. Griffith, J. Hyg. (London) 27, 113 (1928).
2. J. Maynard Smith, C. G. Dowson, B. G. Spratt, Nature
349, 29 (1991).
3. J. P. Claverys, M. Prudhomme, I. Mortier-Barrière,
B. Martin, Mol. Microbiol. 35, 251 (2000).
4. J. P. Claverys, L. S. Håvarstein, Front. Biosci. 7, 1798
(2002).
5. S. Guiral, T. J. Mitchell, B. Martin, J. P. Claverys, Proc.
Natl. Acad. Sci. U.S.A. 102, 8710 (2005).
6. L. S. Håvarstein, B. Martin, O. Johnsborg, C. Granadel,
J. P. Claverys, Mol. Microbiol. 59, 1297 (2006).
7. J. P. Claverys, M. Prudhomme, B. Martin, Annu. Rev.
Microbiol. (2006).

www.sciencemag.org SCIENCE VOL 313 7 JULY 2006 91

REPORTS
8. A. Dagkessamanskaia et al., Mol. Microbiol. 51, 1071
(2004).
9. S. Peterson et al., Mol. Microbiol. 51, 1051 (2004).
10. I. Mortier-Barrière, A. de Saizieu, J. P. Claverys, B. Martin,
Mol. Microbiol. 27, 159 (1998).
11. E. V. Pestova, L. S. Håvarstein, D. A. Morrison, Mol.
Microbiol. 21, 853 (1996).
12. P. Luo, D. A. Morrison, J. Bacteriol. 185, 349 (2003).
13. B. Martin, M. Prudhomme, G. Alloing, C. Granadel,
J. P. Claverys, Mol. Microbiol. 38, 867 (2000).
14. To monitor antibiotic-induced competence, cells
were incubated in CþY medium with an initial pH
value adjusted so that spontaneous competence
induction remained a rare event (see supporting
online material).
15. D. E. Grove, S. Willcox, J. D. Griffith, F. R. Bryant, J. Biol.
Chem. 280, 11067 (2005).
16. S. C. Kowalczykowski, Annu. Rev. Biophys. Biophys. Chem. 25. M. Cashel, D. R. Gentry, V. J. Hernandez, D. Vinella, in
20, 539 (1991). Escherichia coli and Salmonella typhimurium: Cellular
17. J. W. Little, Proc. Natl. Acad. Sci. U.S.A. 81, 1375 and Molecular Biology, F. C. Neidhardt, Ed. (American
(1984). Society for Microbiology Press, Washington, DC, 1996),
18. B. Martin, P. Garcı́a, M. P. Castanié, J. P. Claverys, Mol. pp. 1458–1496.
Microbiol. 15, 367 (1995). 26. We thank D. Lane for critical reading of the manuscript.
19. J. W. Beaber, B. Hochhut, M. K. Waldor, Nature 427, 72 This work was supported in part by European Union grant
(2004). QLK2-CT-2000-00543.
20. I. Phillips, E. Culebras, F. Moreno, F. Baquero,
Supporting Online Material
J. Antimicrob. Chemother. 20, 631 (1987).
www.sciencemag.org/cgi/content/full/313/5783/89/DC1
21. C. Miller et al., Science 305, 1629 (2004).
Materials and Methods
22. C. Levine, H. Hiasa, K. J. Marians, Biochim. Biophys. Acta
Figs. S1 to S4
1400, 29 (1998).
Tables S1 and S2
23. R. A. VanBogelen, F. C. Neidhardt, Proc. Natl. Acad. Sci.
References
U.S.A. 87, 5589 (1990).
24. K. Drlica, X. Zhao, Microbiol. Mol. Biol. Rev. 61, 377 27 March 2006; accepted 1 June 2006
(1997). 10.1126/science.1127912

Presymptomatic Detection In order to evaluate the application of PMCA

for the detection of prions in blood during the
presymptomatic phase, 46 hamsters were inocu-
of Prions in Blood lated intraperitoneally with 10% brain homoge-
nate of the 263K scrapie strain, and 38 control
Paula Saá,1,2 Joaquı́n Castilla,1 Claudio Soto1* animals were injected with phosphate-buffered
saline (PBS). At different times during the
Prions are thought to be the proteinaceous infectious agents responsible for transmissible incubation period, groups of animals were killed,
spongiform encephalopathies (TSEs). PrPSc, the main component of the infectious agent, is also blood was collected, and the buffy coat fraction
the only validated surrogate marker for the disease, and its sensitive detection is critical for was separated (13). Samples of the buffy coat
minimizing the spread of the disease. We detected PrPSc biochemically in the blood of hamsters were resuspended directly on healthy hamster
infected with scrapie during most of the presymptomatic phase of the disease. At early stages of brain homogenate and subjected to 144 PMCA
the incubation period, PrPSc detected in blood was likely to be from the peripheral replication of cycles. Three different aliquots were tested from
prions, whereas at the symptomatic phase, PrPSc in blood was more likely to have leaked from the each sample. To refresh the substrate, after a
brain. The ability to detect prions biochemically in the blood of infected but not clinically sick round of PMCA cycling, samples were diluted
animals offers a great promise for the noninvasive early diagnosis of TSEs. 10-fold into normal brain homogenate, followed
by another round of 144 PMCA cycles. This
rion diseases, also called transmissible from the onset of infection to the appearance of procedure was repeated seven times, because

P spongiform encephalopathies (TSEs), are

a group of fatal and infectious neurode-
generative diseases, including Creutzfeldt-Jakob
disease (CJD) in humans and bovine spongiform
encephalopathy (BSE), scrapie, and chronic
wasting disease (CWD) in animals. Prions are
composed mainly or exclusively of the misfolded
prion protein (PrPSc) (1), which replicates in the
body, transforming the normal prion protein
(PrPC) into more of the misfolded isoform.
Although prion diseases are rare in humans,
the established link between a new variant form
of CJD (vCJD) and BSE (2–4) has raised con-
cern about a potential epidemic in the human
population. Over the past few years, BSE has
become a substantial health problem affecting
many countries (5), and it seems now apparent
that vCJD can be iatrogenically transmitted from
human to human by blood transfusion (6, 7).
Exacerbating this state of affairs is the lack of a
reliable test to identify individuals incubating
the disease during the long and silent period
1
George and Cynthia Mitchell Center for Alzheimer’s
Disease Research, Departments of Neurology, Neuroscience
and Cell Biology, and Biochemistry and Molecular Biology,
University of Texas Medical Branch, 301 University
Boulevard, Galveston, TX 77555–0646, USA. 2Centro de
Biologı́a Molecular, Universidad Autónoma de Madrid,
Madrid, Spain.
*To whom correspondence should be addressed. E-mail:
clsoto@utmb.edu
clinical symptoms (8–10).
PrPSc is not only the main component of the
infectious agent and the most likely cause of
the disease, but it is also the only validated
surrogate marker for TSEs (9). However, PrPSc
concentration is high enough for routine bio-
chemical detection only in the brain and some
lymphoid tissues at a time close to the symp-
tomatic stage of the disease (9). The develop-
ment of highly sensitive presymptomatic assays
for the biochemical detection of PrPSc is critical
for minimizing the spread of the disease (9).
One important aim in prion diagnosis is the
noninvasive and presymptomatic biochemical
detection of PrPSc in biological fluids, particu-
larly using blood, a fluid known to contain in-
fectivity even before the onset of clinical signs
(6, 11, 12).
PrPSc has been detected in the blood of sick
animals by means of the protein misfolding cy-
clic amplification (PMCA) technology (13).
PMCA produces accelerated prion replication,
which dramatically amplifies the quantity of
PrPSc present in a sample (14, 15). In a cyclical
process, large quantities of PrPC are converted
into the misfolded form triggered by the pres-
ence of minute and otherwise undetectable
amounts of PrPSc. The method is highly specific
for the detection of PrPSc and leads to a several-
million-fold increase in sensitivity as compared
to that of standard Western blot assays (13).
according to our results, this enables the detection
of 20 to 50 molecules of monomeric hamster PrP,
which seems to correspond to a single unit of
infectious oligomeric PrPSc (16).
The first group of hamsters was killed 2
weeks after intraperitoneal inoculation. None of
the five infected or control animals showed any
detectable quantity of PrPSc in their blood (Fig.
1 and Table 1). Thus, the PrPSc present in the
inoculum disappeared to undetectable levels
during the first few days after inoculation. PrPSc
was, however, readily detectable in blood 1
week later (20 days after inoculation) in 50% of
the animals infected but in none of the controls
(Fig. 1 and Table 1). The highest percentage of
positive animals during the presymptomatic
phase was observed 40 days after intraperi-
toneal inoculation, in which the sensitivity of
PrPSc detection was 60%. After 60 days, the
detection of PrPSc in blood became harder. In-
deed, only one out of five animals scored pos-
itive at 70 days, whereas none of the five
infected hamsters had detectable PrPSc in their
blood 80 days after inoculation (Table 1). At
the symptomatic stage, which in this experi-
ment was at 114.2 T 5.6 days, 80% of animals
had PrPSc in their blood (Fig. 1). We never
detected a false positive result in any of the 38
control samples analyzed (Table 1).
The distribution of PrPSc detection at dif-
ferent times of the incubation period showed

92 7 JULY 2006 VOL 313 SCIENCE www.sciencemag.org

 Downloaded from mbio.asm.org on May 23, 2016 - Published by mbio.asm.org
RESEARCH ARTICLE

Epidemic Klebsiella pneumoniae ST258 Is a Hybrid Strain

Liang Chen,a Barun Mathema,a,b Johann D. D. Pitout,c,d,e Frank R. DeLeo,f Barry N. Kreiswirtha
Public Health Research Institute Center, New Jersey Medical School, Rutgers University, Newark, New Jersey, USAa; Department of Epidemiology, Mailman School of
Public Health, Columbia University, New York, New York, USAb; Division of Microbiology, Calgary Laboratory Services, Calgary, Alberta, Canadac; Departments of Pathology
and Laboratory Medicine,d Department of Microbiology, Immunology, and Infectious Diseases,e University of Calgary, Calgary, Alberta, Canada; Laboratory of Human
Bacterial Pathogenesis, Rocky Mountain Laboratories, National Institute of Allergy and Infectious Diseases, National Institutes of Health, Hamilton, Montana, USAf

ABSTRACT Carbapenem-resistant Enterobacteriaceae (CRE), especially Klebsiella pneumoniae carbapenemase (KPC)-producing

K. pneumoniae, pose an urgent threat in health facilities in the United States and worldwide. K. pneumoniae isolates classified as
sequence type 258 (ST258) by multilocus sequence typing are largely responsible for the global spread of KPC. A recent compara-
tive genome study revealed that ST258 K. pneumoniae strains are two distinct genetic clades; however, the molecular origin of
ST258 largely remains unknown, and our understanding of the evolution of the two genetic clades is incomplete. Here we com-
pared the genetic structures and single-nucleotide polymorphism (SNP) distributions in the core genomes of strains from two
ST258 clades and other STs (ST11, ST442, and ST42). We identified an ~1.1-Mbp region on ST258 genomes that is homogeneous
to that of ST442, while the rest of the ST258 genome resembles that of ST11. Our results suggest ST258 is a hybrid clone— 80% of
the genome originated from ST11-like strains and 20% from ST442-like strains. Meanwhile, we sequenced an ST42 strain that
carries the same K-antigen-encoding capsule polysaccharide biosynthesis gene (cps) region as ST258 clade I strains. Comparison
of the cps-harboring regions between the ST42 and ST258 strains (clades I and II) suggests the ST258 clade I strains evolved from
a clade II strain as a result of cps region replacement. Our findings unravel the molecular evolution history of ST258 strains, an
important first step toward the development of diagnostic, therapeutic, and vaccine strategies to combat infections caused by
multidrug-resistant K. pneumoniae.
IMPORTANCE Recombination events and replacement of chromosomal regions have been documented in various bacteria, and
these events have given rise to successful pathogenic clones. Here we used comparative genomic analyses to discover that the
ST258 K. pneumoniae genome is a hybrid— 80% of the chromosome is homologous to ST11 strains, while the remaining 20% is
homologous to that of ST442. Meanwhile, a recent study indicated that ST258 strains can be segregated into two ST258 clades,
with distinct capsule polysaccharide gene (cps) regions. Our analysis suggests ST258 clade I strains evolved from clade II through
homologous recombination of cps region. Horizontal transfer of the cps region appears to be a key element driving the molecular
diversification in K. pneumoniae strains. These findings not only extend our understanding of the molecular evolution of ST258
but are an important step toward the development of effective control and treatment strategies for multidrug-resistant K. pneu-
moniae.

Received 19 May 2014 Accepted 22 May 2014 Published 24 June 2014

Citation Chen L, Mathema B, Pitout JDD, DeLeo FR, Kreiswirth BN. 2014. Epidemic Klebsiella pneumoniae ST258 is a hybrid strain. mBio 5(3):e01355-14. doi:10.1128/mBio.01355-
14.
Editor George Jacoby, Lahey Hospital and Medical Center
Copyright © 2014 Chen et al. This is an open-access article distributed under the terms of the Creative Commons Attribution-Noncommercial-ShareAlike 3.0 Unported license,
which permits unrestricted noncommercial use, distribution, and reproduction in any medium, provided the original author and source are credited.
Address correspondence to Barry N. Kreiswirth, kreiswba@njms.rutgers.edu.
This article is a direct contribution from a member of the American Academy of Microbiology.

K lebsiella pneumoniae carbapenemase (KPC) has emerged as a se-

rious clinical challenge in health care facilities in the United States
and worldwide (1). The blaKPC-harboring plasmid encoding the car-
bapenemase has been found in numerous K. pneumoniae sequence
types (STs)/clones as well as in other Gram-negative species; how-
ever, the vast majority of the global KPC-producing K. pneumoniae
isolates are associated with a single multilocus sequence type—
ST258 (2–4). K. pneumoniae ST258 emerged as a notable clinical
problem in the middle 2000s in the United States and remains the
main ST in the United States and elsewhere (3–6).
Recently, two KPC-harboring K. pneumoniae ST258 clinical
isolates were sequenced to closure (7). These genomes were used
as references for a comparative genome analysis of 83 ST258 clin-
ical isolates collected between 2002 and 2012 from geographically
diverse sources. Phylogenetic analysis of the core genome of these
isolates revealed that ST258 K. pneumoniae strains are comprised
of two distinct genetic clades (ST258 clades I and II), largely due to
an ~215-kb region of divergence (RD) that includes genes in-
volved in capsular polysaccharide (CPS) biosynthesis (7). Further
genotyping analysis with 2 cps-associated genes, wzi and wzy, in
ST258 and other unrelated K. pneumoniae strains identified the
ST258 clade I genotype in genetically distinct ST42 strains (7).
Interestingly, a GenBank BLAST search using nucleotides encom-
passing the ST258 clade II cps region indicated this region is highly
similar to that of a Brazilian ST442 strain, Kp13, which harbors
cpsKp13 (8, 9).

May/June 2014 Volume 5 Issue 3 e01355-14 ®

mbio.asm.org 1
 Downloaded from mbio.asm.org on May 23, 2016 - Published by mbio.asm.org
Chen et al.

TABLE 1 Features of completely sequenced ST258, ST11, and ST442 genomes

Result for strain:
Parameter a NJST258_1 NJST258_2 HS11286 JM45 ATCC BAA-2146 Kp13
ST no. 258 258 11 11 11 442
Yr 2010 2010 2011 2010 2010 2009
Country United States United States China China United States Brazil
bla genes blaKPC-3, blaOXA-9, blaKPC-3, blaSHV-11
blaKPC-2, blaKPC-2, blaNDM-1, blaCTX-M-15, blaKPC-2, blaCTX-M-2,
blaTEM-1 blaCTX-M-14, blaCTX-M-24, blaCMY-6, blaOXA-1, blaOXA-9, blaTEM-1,
blaTEM-1 blaVEB-3 blaTEM-1, blaSHV-11 blaSHV-12, blaSHV-110
Size (bp) 5,263,329 5,293,301 5,333,942 5,273,813 5,435,369 5,307,003
G⫹C content (%) 57.4 57.5 57.5 57.5 57.3 57.5
No. of CDS 5,475 5,434 5,316 4,872 5,315 5,189
rRNA (n) 25 25 25 25 25 24
16S 8 8 8 8 8 8
23S 8 8 8 8 8 8
5S 9 9 9 9 9 8
tRNA (n) 77 86 87 83 85 86
Plasmids (n) 5 3 6 2 4 6
Prophages (n) 8 7 7 7 8 2
ICEs (n) 2 2 2 1 3 1
IS elements (n) 22 19 23 14 23 31
IS family (n) IS1294 (3), IS1400 (1), IS1294 (2), IS1400 (1), IS1 (1), IS5 (1), IS1400 (1), IS5 (1), IS1 (1), IS1222 (1), IS1 (7), IS10 (1),
IS5 (7), IS5075 (1), IS5 (7), IS903 (2), IS903 (7), ISEc22 ISEcp1 (1), ISKpn1 IS1400 (1), IS2 (2), IS1222 (1), IS4321
IS903 (2), ISKpn1 ISKpn1 (5), (1), ISEcp1 (6), (6), ISKpn18 (6) IS26 (2), IS6100 (1), (2), IS3 (1), IS5 (3),
(5), ISKpn18 (2), ISKpn18 (2) ISKpn1 (6), IS66 (3), ISEc36 (2), IS903 (7), ISEc11
ISSm1 (1) ISKpn18 (1) ISEcp1 (2), ISKpn1 (2), ISEcp1 (1),
(5), ISKpn18 (2), ISKpn1 (5),
ISKpn21 (1) ISKpn21 (1)
a ICE, integrated conjugative element; IS, insertion sequence; CDS, coding sequences.

The emergence and global spread of ST258 and recent reports
that this clone has diversified as a result of recombination and
replacement of the cps region raise the question of its evolutionary
history. One speculation is that ST11 (allelic profile 3-3-1-1-1-1-
4), a highly predominant multidrug-resistant clone in Asia and
South America (10–12), and a single-locus variant of ST258 (al-
lelic profile 3-3-1-1-1-1-79) gave rise to the ST258 clone through
the acquisition of the tonB79 allele (13).
To better understand the phylogeny of the ST11 and ST258
lineages, we compared the genomes of three ST11 strains
(HS11286, JM45, and ATCC BAA-2146), three ST258 strains
(NJST258_1, NJST258_2, and Kp1787 [a representative ST258
clade I strain present in our collection]), and an ST42 strain,
Kp1832. The comparative analysis of these genomes indicates that
large and repeated chromosomal exchanges in K. pneumoniae
have occurred between ST11 and ST258, with a significant role for
ST442 in the recent molecular evolution of epidemic ST258
strains.
RESULTS
Large ~1.1-Mbp recombination region in ST258. To elucidate
the phylogenetic relationship among ST258, ST11, and ST442
strains, we first compared the genome sequences of six closed
Klebsiella pneumoniae strains (Table 1). The size of the chromo-
somes was on average ~5.3 Mbp, but the number of mobile ge-
netic elements (MGEs), including plasmids, prophages, inte-
grated conjugative elements (ICEs), and insertion sequences (IS),
varied (Table 1; see Fig. S1 in the supplemental material). Consis-
tent with multilocus sequence typing (MLST) (Fig. 1A), which
indicates ST11 and ST258 differ by a single locus (the tonB allele
distinguishes the two sequence types), the 3 ST11 and 2 ST258
genomes have 7 of 8 prophages in common, and all harbor
ICEKp258.1 (see Fig. S1). Sequence comparison among the tonB
alleles shows tonB79 (in ST258) differs from tonB4 (in ST11) by
four single-nucleotide polymorphisms (SNPs) and differs from
tonB14 (in ST442) by a single SNP. Of note, the three ST11 strains
(HS11286, JM45, and ATCC BAA-2146) harbor three different cps
operons, a finding similar to the distinguishing cps genotypes in
ST258 clade I and II strains and which supports the observation
that cps switching provides K. pneumoniae the plasticity to change
its antigenic nature (Fig. 1B).
Comparative genome and SNP distribution analyses of the
core chromosome region, as depicted in Fig. 1B, uncovered a
number of surprising findings given the MLST results for ST11
and ST258. Except for differences in the tonB allele and the region
encoding the capsular polysaccharide biosynthetic machinery, the
6 genomes of ST11 and ST258 have a high degree of identity (re-
gions of the same color in Fig. 1B). However, further analysis of
the RD and flanking nucleotides revealed that the differences be-
tween ST11 and ST258 were expansive, covering an ~1.1-Mbp
contiguous region corresponding to nucleotide positions
1,660,631 to 2,723,681 in strain NJST258_1 (Fig. 1). Significantly,
the ~1.1-Mbp region identified in ST258 clade I and II strains has
identical chromosomal nucleotide boundaries (Fig. 2).
Analysis of SNPs in the genomes of ST258 strains (NJST258_1,
NJST258_2, and Kp1787) and ST11 strains (HS11286, JM45, and
ATCC BAA-2146) indicated that these strains differ by an average
of 9,647 SNPs, and 98.1% (9,460 SNPs) of these polymorphisms
are concentrated in the contiguous ~1.1-Mbp region (identified
above), which represents 20% of the genome (Fig. 3). By compar-
ison, the genomes of ST258 strains and ST442 strain Kp13 differed
by 21,095 SNPs, consistent with their genetically distinct MLST

2 ®

mbio.asm.org May/June 2014 Volume 5 Issue 3 e01355-14

 Downloaded from mbio.asm.org on May 23, 2016 - Published by mbio.asm.org
Klebsiella pneumoniae ST258 Is a Hybrid Strain

A
~1.1 MB

NJST258_1
MLST mdh infB tonB gapA phoE pgi rpoB

ST11 1 3 4 3 1 1 3
ST258 1 3 79 3 1 1 3
ST442 2 20 14 10 9 1 11
ST42 1 6 15 2 8 3 1
B
100
Kp13
(ST442) 0
cps 100
HS11286
(ST11) 0
cps 100

JM45
(ST11) 0
cps 100

ATCC BAA-2146
0
(ST11) cps
100

NJST258_1/_2
(ST258 clade II) 0
cps ICE Kp258.2
100
Kp1787
(ST258 clade I) 0
cps
100
Kp1832
(ST42) 0
cps
FIG 1 (A) MLST allele locations on NJST258_1 genome. The light green arrow denotes the genome of NJST258_1, and the light blue region shows the ~1.1-Mbp
putative recombination region between the ST11 and ST442 genomes. The chromosomal positions of the seven MLST housekeeping genes (gapA, infB, mdh, pgi,
phoE, rpoB, and tonB) are illustrated beneath the genome arrow of NJST258_1, and the corresponding allele numbers for ST11, ST258, ST442, and ST42 are listed
below the gene names. (B) Core genome SNP distributions in the ST11, ST258 (clades I and II), ST442, and ST42 strains. The number of SNPs (y axis) per 1,000 nt
is plotted according to the position on the NJST258_1 genome (x axis). Different homogeneous regions (⬎98% identity based on SNP comparisons) are color
coded. Specifically, the ~52-kb cps-containing regions in Kp1787 (ST258 clade I) and Kp1832 (ST42), which are nearly identical in these strains, are shaded in
green. ICEKp258.2 and cps are illustrated by small vertical bars, and the same cps regions are shown in the same color.

profiles (Fig. 1B and SNP matrix in Fig. 3). Most significantly, the
SNP mapping revealed contiguous ~1.1-Mbp regions that were
nearly identical in ST258 clade II (NJST258_1 and NJST258_2)
and ST442 (Kp13) strains, differing by only 206 SNPs (1.0%). As
depicted in Fig. 1B, the comparative genomic organization and
SNP results provide additional support to the idea that the ST258
clade II strain is a hybrid strain containing 80% (~4.2 Mbp) of the
chromosome from ST11 and 20% (~1.1 Mbp) from ST442
(Fig. 1B).
The ~1.1-Mbp chromosomal region in ST258 clade I and II
strains contains the ~215-kb RD and the cps gene cluster (Fig. 2).
ICEKp258.2 is common to the prototype ST258 strains shown in
Fig. 1B but is absent from ST442 strain Kp13. To determine the
level of conservation of ICEKp258.2 among ST258 clinical iso-
lates, we analyzed the DNA contigs of 83 additional ST258 ge-
nomes sequenced in our previous study (7). We found that
ICEKp258.2 is conserved in all of the queried ST258 genomes, and
the insertion of this element in ST258 clade I and II genomes is at
the same tRNA-Asn site (data not shown).
cps replacement in ST258 strains. In our previous study, we
identified seven ST42 strains that harbored cps genetic markers
(wzy and wzi) that are identical to those in ST258 clade I strains
(7), and we hypothesized that this unrelated sequence type (ST42)
was the donor for the cps region in ST258 clade I strains. As a first
step toward testing this hypothesis, we used Illumina Miseq to
sequence the DNA in the cps region of ST42 and ST258 clade I
strains and that in strain Kp1832, a representative ST42 isolate in
our strain collection. The gross organization between ST42 and
ST258 strains indicates their distal genetic relatedness (Fig. 1B), a
finding consistent with MLST data. An SNP analysis confirmed
that there was significant genome divergence between ST258 and
ST42 strains. A total of 31,157 SNPs distinguished the three ST258
strains from the ST42 strain, Kp1832, and 27% of these SNPs
(8,444 SNPs) were located in the ~1.1-Mbp recombination region
(Fig. 3).
Since the two sequenced reference strains (NJST258_1 and
NJST258_2) were genotyped as ST258 clade II strains, we created
a de novo genome sequence of the prototypic ST258 clade I strain

May/June 2014 Volume 5 Issue 3 e01355-14 ®

mbio.asm.org 3

Chen et al.
cps
cps ICEKp258.2
FIG 2 Upstream and downstream junction SNPs for the ~1.1-Mbp recombination fragment and cps region in ST258, ST442, and ST42 strains. The start site of
the replacement of the ~52-kb cps-harboring region is the same as that of the ~215-kb RD in ST258 II clades (7). Sequences were obtained from Kp1878 (an ST258
clade I strain), NJST258_1 (an ST258 clade II strain), JM45 (an ST11 strain), Kp1832 (an ST42 strain), and Kp13 (an ST442 strain).
(Kp1787) to use as a clade I reference genome. An ~450-kb region
of the Kp1787 genome, which contains the entire ~215-kb RD,
was used as a reference to examine recombination events between
Kp1832 (ST42) and Kp1787 (an ST258 clade I strain). The com-
parison revealed a nearly identical region (2 SNPs) spanning
~52 kb that contains the cps region; the alignment of the two
regions maps to the start of this DNA replacement at the same
location in the RD region (Fig. 2). In addition, the ICEKp258.2
element is absent from strain Kp1832, and there is nucleotide
divergence outside the aforementioned ~52-kb region (Fig. 1 and
2).
Comparative sequence analysis further showed that the neigh-
boring sequences upstream and downstream from this ~52-kb
region are identical in ST258 clade I and II strains (Fig. 2). In
addition, the SNP distribution among the 85 ST258 genomes re-
ported in our previous study revealed that 592 (89%) of the 664
SNPs in the RD are located within the ~52-kb cps-harboring re-
gion (7). Together, the genomic findings are consistent with the
hypothesis that clade I evolved rapidly through the acquisition of
the cps region from an ST42 strain. The evidence provided above
strongly suggests that replacement of the original (presumably
ST258 clade II) cps region in clade I contributes largely to the
noted phylogenetic difference between the two ST258 clades.
blaKPC-harboring genetic element. We and others have re-
ported that the blaKPC gene in ST258 strains is carried exclusively
by a Tn3-based transposon, Tn4401 (5, 7, 14). Tn4401 is 10 kb in
length, delimited by two 39-bp imperfect inverted repeat (IR) se-
quences, and harbors the blaKPC gene, a Tn3 transposase gene
(tnpA), a Tn3 resolvase gene (tnpR), and two insertion sequences,
ISKpn6 and ISKpn7 (15) (see Fig. S2 in the supplemental mate-
rial). In contrast, ST11 and ST442 strains harbor blaKPC-
containing elements that are distinct from those in ST258 strains
(see Fig. S2) and share only ~2 kb of sequence with Tn4401. Col-
4 ®
mbio.asm.org
Downloaded from mbio.asm.org on May 23, 2016 - Published by mbio.asm.org
ICEKp258.2
tonB
lectively, these findings suggest ST258 strains are hybrid strains
that arose from an ancestral ST11 strain that acquired an ~1.1-
Mbp contiguous chromosomal segment from an ST442-like
strain by DNA recombination/replacement. The identification of
distinct blaKPC-harboring elements in ST258, ST11, and ST442
strains indicates blaKPC was acquired by ST258 strains via horizon-
tal gene transfer (rather than by vertical gene transmission from
ST11 or ST442 parental strains) after the recombination events.
DISCUSSION
The current rise of KPC-producing K. pneumoniae infections in
U.S. health care facilities has been overwhelmingly associated with
strains typed as ST258. To better understand the evolutionary
history of this epidemic clone, we compared the genome se-
quences of ST258 strains, single-locus variant ST11 strains, and
other selected K. pneumoniae strain types. Notably, we discovered
that ST258 strains are hybrid strains comprised of genomic DNA
from ST11 (~80%) and ST442 (~20%)-like strains—presumably
the product of a large chromosomal replacement event.
Recombination events and replacement of large chromosomal
regions have been documented in various bacteria, and there are
reported examples where the hybrid strains are associated with
epidemiological success. Robinson and Enright were the first to
report a naturally occurring bacterial hybrid—in this case, in a
Staphylococcus aureus strain known as ST239 (16). This hybrid
strain is a pandemic methicillin-resistant S. aureus (MRSA) strain
responsible for ~90% of the nosocomial infections throughout
mainland Asia and much of South America (17). ST239 is com-
prised of large chromosomal regions from two distantly related
lineages, ST8 and ST30. Approximately 20% of the ST8 genome
was replaced with an ~550-kb contiguous chromosomal fragment
from an ST30 donor strain, thereby creating ST239. This appar-
ently rare molecular event has not been explained or reproduced
May/June 2014 Volume 5 Issue 3 e01355-14
 Downloaded from mbio.asm.org on May 23, 2016 - Published by mbio.asm.org
Klebsiella pneumoniae ST258 Is a Hybrid Strain

46
21
_2

A-
BA
58

86
87

32
T2

12

C
45
17

13

18
S1
JS

C
JM
Kp

Kp

Kp
AT
N

H
NJST258_1 175/43 508/360 8180/8027 8056/7742 12829/12661 21035/217 31304/8566
NJST258_2 445/337 8117/8004 7993/7719 12766/12638 20966/194 31240/8543
Kp1787 8087/7968 8056/7771 12743/12609 21284/503 30927/8222
JM45 3621/3370 6784/6681 28841/8086 30395/7718
HS11286 8183/7915 28522/7761 30080/7414
ATCC BAA-2146 33493/12721 35032/12338
Kp13 31848/8590

(ST42)

(ST258 clade I)

(ST442) (ST11)

(ST258 clade II)

FIG 3 (A) SNP matrix for different K. pneumonaie strains. The matrix is illustrated as (total no. of SNPs/no. of SNPs in the ~1.1-Mbp recombination region).
Green shading indicates the number of SNPs in ST258 strains compared to that in ST11 strains. Orange shading indicates the number of SNPs in ST258 strains
compared to that in ST442 strains. (B) Phylogenetic analysis of the eight isolates based upon 52,135 concatenated SNPs in the core genome.

in the laboratory. In group B Streptococcus (GBS), large single
chromosomal replacement events and multiple localized recom-
bination events occur naturally and can be reproduced in the lab-
oratory (18). For GBS, conjugation is the molecular pathway for
genomic movement (19). It is worth noting that genetic replace-
ment of the GBS cps region between unrelated sequence types is
the common mechanism by which this species alters its surface
antigen composition (18). Similarly, cps region replacement-
associated capsular switching has also been suggested as being an
intrinsic feature throughout the evolutionary history of Strepto-
coccus pneumoniae (20).
Here we discovered that ST258 clade II strains are hybrid strains in
which 20% of the K. pneumoniae ST11 genome was replaced with a
homologous ~1.1-Mbp contiguous region from a strain in the ST442
lineage. This region, which includes the previously described region
of difference (RD) and capsular polysaccharide biosynthetic genes,
has molecular scars of multiple localized recombination events, sim-
ilar to the phenomenon in Streptococcus agalactiae (18, 19). The find-
ing that ST442 and ST258 have a contiguous chromosomal region in
common and that the nucleotide boundaries between the ST442 and
ST258 clade I and II genomes are indistinguishable (Fig. 2) is evidence
that the recombination event creating an ST11 and ST442 hybrid
strain likely occurred once, thereby creating the ST258 clade II lineage
(Fig. 4).
Based on recent genome-scale studies, there have been numer-
ous putative chromosomal recombination events involving the

May/June 2014 Volume 5 Issue 3 e01355-14 ®

mbio.asm.org 5

Chen et al.
ST11-like
ICEKp258.2
ST258
clade II
cps 258-2
ST442-like
cps Kp13
(cps 258-2)
FIG 4 Hypothesized evolutionary history in K. pneumoniae ST258 strains.
region encoding the CPS biosynthetic machinery (7). ST258
clades I and II have distinct cps regions, and this is also true of the
three ST11 strains analyzed in this study (Fig. 1B), further sup-
porting the notion that DNA exchange in and around the cps
regions may be a general mechanism used by K. pneumoniae to
rapidly diversify and that novel clades arise through cps switching
between ST258 (clade I or II) and unrelated K. pneumoniae se-
quence types.
The presence, absence, and diversity of genetic landmarks
within the acquired ~1.1-Mbp region provide clues into ST258’s
recent evolutionary origin and the extent of its genomic plasticity.
ICEKp258.2, which is absent in ST442, is present in both ST258
clades, where its chromosomal insertion site is conserved, suggest-
ing that this ICE was acquired after the major genome recombi-
nation event that gave rise to ST258 (Fig. 4). In support of the
notion that ICEKp258.2 is a relatively recent acquisition, the G⫹C
content of ICEKp258.2 is 37.1%, significantly less than the
~57.5% G⫹C content of the entire K. pneumoniae chromosome.
Taken together, these observations provide evidence that
ICEKp258.2 is exogenous and was likely acquired once by ST258,
before the recombination events involving the cps regions in
clades I and II (Fig. 4). Moreover, the replacement of the ST258
clade II cps region with that from ST42 (thus creating clade I)
occurred after the acquisition of ICEKp258.2 (Fig. 4).
Adler and colleagues investigated the association of the
ICEKp258.2 with ST258 by testing160 K. pneumoniae strains with
diverse sequence types for the presence of pilV, a gene carried on
ICEKp258.2 (21). They found that pilV was present only in ST258
and genetically related strains. Based on sequence analysis,
ICEKp258.2 harbors a type IV pilus gene cluster and a type III
restriction-modification system. A type IV pilus could increase the
uptake and exchange of DNA, such as plasmids, as well as facilitate
adherence to living and nonliving surfaces— e.g., the human gut
or the environment (22)—which may in part explain the high
transmissibility of ST258 strains and the movement of KPC genes.
Additionally, a type III restriction-modification system could
serve in “host specificity” regarding the exchange of certain com-
patible plasmids and other mobile elements (23). Restriction of
plasmids and specific mobile elements may explain the differences
observed between ST11 (which lacks ICEKp258.2) and ST258, as
6 ®
mbio.asm.org
Downloaded from mbio.asm.org on May 23, 2016 - Published by mbio.asm.org
ST42
cps
(cps 258-1)
ST258
clade I
cps 258-1
the former is associated with a broad range of plasmids and car-
bapenemases (KPC, VIM, IMP, NDM, and OXA-48) (10–12, 24–
28), whereas ST258 strains predominantly harbor KPC. Taken
together, the association of ICEKp258.2 with ST258 K. pneu-
moniae strains raises the possibility that this element may contrib-
ute to epidemiological success of this sequence type. To investigate
whether ICEKp258.2 could potentially be an “epidemic clone-
specific” target, we are currently investigating the impact of alter-
ing the type IV pilus gene cluster and the type III restriction-
modification system in this element.
Taken together, our findings underscore the role of recombi-
nation in the rapid evolution of clinical strains of K. pneumoniae
in both creating hybrid clones and in more localized chromo-
somal replacements that alter antigenic presentation and ulti-
mately divert the host response.
MATERIALS AND METHODS
Sequence information. Data used in comparative analysis were down-
loaded from the NCBI database (http://www.ncbi.nlm.nih.gov/genome/
genomes/815), including complete genome sequences and annotation of
K. pneumoniae isolates HS11286 (CP003200) (29), JM45 (CP006656),
ATCC BAA-2146 (CP006659) (30), Kp13 (CP003999) (8),
NJST258_1(CP006923) (7), and NJST258_2 (CP006918) (7). Additional
sequence data were retrieved from our recent study on K. pneumoniae
ST258 (7).
Genome sequencing and assembly. Strain Kp1832 was selected from
one of the seven ST42 K. pneumoniae isolates that carry the same wzy and
wzi genes as ST258 cps-1 strains (7). Genomic DNA isolation and library
preparation were performed as described previously (7). The genome was
sequenced using an Illumina MiSeq platform, which generated 250-bp
paired-end reads. De novo assembly for Kp1832 and Kp1787 (a represen-
tative ST258 clade I strain, selected from our previous study [7]) was
accomplished by using a combination of CLC genomic workbench (v
7.0.3; CLC Bio, Aarhus, Denmark), Mira (31), and Velvet (32). The best
assemblies from each method were combined using Geneious Pro soft-
ware in order to generate the supercontig for the cps-harboring element.
Comparative genomics analysis. Visualization of circular genome
comparisons was performed using the BLAST ring image generator
(BRIG) (33). Prophages were identified by PHAST (34). Insertion se-
quences were identified using the IS Finder database (http://www
-is.biotoul.fr). De novo assembled contigs from Kp1832 and Kp1787 were
ordered and oriented relative to the NJST258_1 genome and then com-
May/June 2014 Volume 5 Issue 3 e01355-14

bined together as a pseudochromosome using the Mauve contig mover
(35). Multiple genome sequence alignments and comparison analysis
were then performed with Mauve (35). For core genomic analysis, SNPs
located on the MGEs, including prophases, ICEs, and insertion elements,
as well as those on rRNAs and tRNAs, were excluded. The concatenated
SNPs were used to generate a consensus phylogenetic tree by the maxi-
mum likelihood method based on the Tamura-Nei model with the MEGA
5 software (36). The SNP distribution among different genome sequences
was inferred from genome alignment using Mauve (35), and SNPs were
counted on a 1,000-nucleotide (nt) window based on the nucleotide po-
sition on NJST258_1.
Nucleotide sequence accession numbers. Illumina short read data for
ST258 have been deposited in the Sequence Read Archive (SRA) database
under accession no. SRP036874 (7). The Illumina short read data for
Kp1832 have been deposited in the SRA database under accession no.
SRX512850.
SUPPLEMENTAL MATERIAL
Supplemental material for this article may be found at http://mbio.asm.org/
lookup/suppl/doi:10.1128/mBio.01355-14/-/DCSupplemental.
Figure S1, EPS file, 25.9 MB.
Figure S2, EPS file, 1.1 MB.
ACKNOWLEDGMENTS
This work was supported in part by a grant (to B.N.K.) from the National
Institutes of Health (1R01AI090155) and the Intramural Research Pro-
gram of the National Institute of Allergy and Infectious Diseases, National
Institutes of Health.
B.N.K. discloses that he holds two patents that focus on using DNA
sequencing to identify bacterial pathogens.
REFERENCES
1. CDC. 2013. Antibiotic resistance threats in the United States, 2013. CDC,
Atlanta, GA.
2. Patel G, Bonomo RA. 2013. “Stormy waters ahead”: global emergence of
carbapenemases. Front. Microbiol. 4:48. http://dx.doi.org/10.3389/
fmicb.2013.00048.
3. Munoz-Price LS, Poirel L, Bonomo RA, Schwaber MJ, Daikos GL,
Cormican M, Cornaglia G, Garau J, Gniadkowski M, Hayden MK,
Kumarasamy K, Livermore DM, Maya JJ, Nordmann P, Patel JB,
Paterson DL, Pitout J, Villegas MV, Wang H, Woodford N, Quinn JP.
2013. Clinical epidemiology of the global expansion of Klebsiella pneu-
moniae carbapenemases. Lancet Infect. Dis. 13:785–796. http://
dx.doi.org/10.1016/S1473-3099(13)70190-7.
4. Tzouvelekis LS, Markogiannakis A, Psichogiou M, Tassios PT, Daikos
GL. 2012. Carbapenemases in Klebsiella pneumoniae and other
Enterobacteriaceae: an evolving crisis of global dimensions. Clin. Micro-
biol. Rev. 25:682–707. http://dx.doi.org/10.1128/CMR.05035-11.
5. Kitchel B, Rasheed JK, Patel JB, Srinivasan A, Navon-Venezia S, Car-
meli Y, Brolund A, Giske CG. 2009. Molecular epidemiology of KPC-
producing Klebsiella pneumoniae isolates in the United States: clonal ex-
pansion of multilocus sequence type 258. Antimicrob. Agents Chemother.
53:3365–3370. http://dx.doi.org/10.1128/AAC.00126-09.
6. Schwaber MJ, Lev B, Israeli A, Solter E, Smollan G, Rubinovitch B,
Shalit I, Carmeli Y, Israel Carbapenem-Resistant Enterobacteriaceae
Working Group. 2011. Containment of a country-wide outbreak of
carbapenem-resistant Klebsiella pneumoniae in Israeli hospitals via a na-
tionally implemented intervention. Clin. Infect. Dis. 52:848 – 855. http://
dx.doi.org/10.1093/cid/cir025.
7. DeLeo FR, Chen L, Porcella SF, Martens CA, Kobayashi SD, Porter AR,
Chavda KD, Jacobs MR, Mathema B, Olsen RJ, Bonomo RA, Musser
JM, Kreiswirth BN. 2014. Molecular dissection of the evolution of
carbapenem-resistant multilocus sequence type 258 Klebsiella pneu-
moniae. Proc. Natl. Acad. Sci. U. S. A. 111:4988 – 4993. http://dx.doi.org/
10.1073/pnas.1321364111.
8. Ramos PI, Picão RC, Almeida LG, Lima NC, Girardello R, Vivan AC,
Xavier DE, Barcellos FG, Pelisson M, Vespero EC, Médigue C, Vascon-
celos AT, Gales AC, Nicolás MF. 2014. Comparative analysis of the
complete genome of KPC-2-producing Klebsiella pneumoniae Kp13 re-
May/June 2014 Volume 5 Issue 3 e01355-14
Downloaded from mbio.asm.org on May 23, 2016 - Published by mbio.asm.org
Klebsiella pneumoniae ST258 Is a Hybrid Strain
veals remarkable genome plasticity and a wide repertoire of virulence and
resistance mechanisms. BMC Genomics 15:54. http://dx.doi.org/10.1186/
1471-2164-15-S2-P54.
9. Chen L, Chavda KD, Findlay J, Peirano G, Hopkins K, Pitout JD,
Bonomo RA, Woodford N, Deleo FR, Kreiswirth BN. 14 April 2014.
Multiplex PCR for identification of two capsular types in epidemic KPC-
producing Klebsiella pneumoniae ST258 strains. Antimicrob. Agents Che-
mother. http://dx.doi.org/10.1128/AAC.02673-14.
10. Qi Y, Wei Z, Ji S, Du X, Shen P, Yu Y. 2011. ST11, the dominant clone
of KPC-producing Klebsiella pneumoniae in China. J. Antimicrob. Che-
mother. 66:307–312. http://dx.doi.org/10.1093/jac/dkq431.
11. Pereira PS, de Araujo CF, Seki LM, Zahner V, Carvalho-Assef AP,
Asensi MD. 2013. Update of the molecular epidemiology of KPC-2-
producing Klebsiella pneumoniae in Brazil: spread of clonal complex 11
(ST11, ST437 and ST340). J. Antimicrob. Chemother. 68:312–316. http://
dx.doi.org/10.1093/jac/dks396.
12. Chiu SK, Wu TL, Chuang YC, Lin JC, Fung CP, Lu PL, Wang JT, Wang
LS, Siu LK, Yeh KM. 2013. National surveillance study on carbapenem
non-susceptible Klebsiella pneumoniae in Taiwan: the emergence and
rapid dissemination of KPC-2 carbapenemase. PLoS One 8:e69428. http://
dx.doi.org/10.1371/journal.pone.0069428.
13. Breurec S, Guessennd N, Timinouni M, Le TA, Cao V, Ngandjio A,
Randrianirina F, Thiberge JM, Kinana A, Dufougeray A, Perrier-Gros-
Claude JD, Boisier P, Garin B, Brisse S. 2013. Klebsiella pneumoniae
resistant to third-generation cephalosporins in five African and two Viet-
namese major towns: multiclonal population structure with two major
international clonal groups, CG15 and CG258. Clin. Microbiol. Infect.
19:349 –355. http://dx.doi.org/10.1111/j.1469-0691.2012.03805.x.
14. Cuzon G, Naas T, Truong H, Villegas MV, Wisell KT, Carmeli Y, Gales
AC, Venezia SN, Quinn JP, Nordmann P. 2010. Worldwide diversity of
Klebsiella pneumoniae that produce ␤-lactamase blaKPC-2 gene. Emerg.
Infect. Dis. 16:1349 –1356. http://dx.doi.org/10.3201/eid1609.091389.
15. Naas T, Cuzon G, Villegas MV, Lartigue MF, Quinn JP, Nordmann P.
2008. Genetic structures at the origin of acquisition of the beta-lactamase
blaKPC gene. Antimicrob. Agents Chemother. 52:1257–1263. http://
dx.doi.org/10.1128/AAC.01451-07.
16. Robinson DA, Monk AB, Cooper JE, Feil EJ, Enright MC. 2005. Evo-
lutionary genetics of the accessory gene regulator (agr) locus in Staphylo-
coccus aureus. J. Bacteriol. 187:8312– 8321. http://dx.doi.org/10.1128/
JB.187.24.8312-8321.2005.
17. Harris SR, Feil EJ, Holden MT, Quail MA, Nickerson EK, Chantratita
N, Gardete S, Tavares A, Day N, Lindsay JA, Edgeworth JD, de Len-
castre H, Parkhill J, Peacock SJ, Bentley SD. 2010. Evolution of MRSA
during hospital transmission and intercontinental spread. Science 327:
469 – 474. http://dx.doi.org/10.1126/science.1182395.
18. Bellais S, Six A, Fouet A, Longo M, Dmytruk N, Glaser P, Trieu-Cuot
P, Poyart C. 2012. Capsular switching in group B Streptococcus CC17
hypervirulent clone: a future challenge for polysaccharide vaccine devel-
opment. J. Infect. Dis. 206:1745–1752. http://dx.doi.org/10.1093/infdis/
jis605.
19. Brochet M, Rusniok C, Couvé E, Dramsi S, Poyart C, Trieu-Cuot P,
Kunst F, Glaser P. 2008. Shaping a bacterial genome by large chromo-
somal replacements, the evolutionary history of Streptococcus agalactiae.
Proc. Natl. Acad. Sci. U. S. A. 105:15961–15966. http://dx.doi.org/
10.1073/pnas.0803654105.
20. Wyres KL, Lambertsen LM, Croucher NJ, McGee L, von Gottberg A,
Liñares J, Jacobs MR, Kristinsson KG, Beall BW, Klugman KP, Parkhill
J, Hakenbeck R, Bentley SD, Brueggemann AB. 2013. Pneumococcal
capsular switching: a historical perspective. J. Infect. Dis. 207:439 – 449.
http://dx.doi.org/10.1093/infdis/jis703.
21. Adler A, Khabra E, Chmelnitsky I, Giakkoupi P, Vatopoulos A, Mathers
AJ, Yeh AJ, Sifri CD, De Angelis G, Tacconelli E, Villegas MV, Quinn
J, Carmeli Y. 2014. Development and validation of a multiplex PCR assay
for identification of the epidemic ST-258/512 KPC-producing Klebsiella
pneumoniae clone. Diagn. Microbiol. Infect. Dis. 78:12–15. http://
dx.doi.org/10.1016/j.diagmicrobio.2013.10.003.
22. Giltner CL, Nguyen Y, Burrows LL. 2012. Type IV pilin proteins: versa-
tile molecular modules. Microbiol. Mol. Biol. Rev. 76:740 –772. http://
dx.doi.org/10.1128/MMBR.00035-12.
23. Rao DN, Dryden DT, Bheemanaik S. 2014. Type III restriction-
modification enzymes: a historical perspective. Nucleic Acids Res. 42:
45–55. http://dx.doi.org/10.1093/nar/gkt1373.
24. Samuelsen Ø, Thilesen CM, Heggelund L, Vada AN, Kümmel A,
®
mbio.asm.org 7

Chen et al.
Sundsfjord A. 2011. Identification of NDM-1-producing Enterobacteri-
aceae in Norway. J. Antimicrob. Chemother. 66:670 – 672. http://
dx.doi.org/10.1093/jac/dkq483.
25. Lascols C, Peirano G, Hackel M, Laupland KB, Pitout JD. 2013. Sur-
veillance and molecular epidemiology of Klebsiella pneumoniae isolates
that produce carbapenemases: first report of OXA-48-like enzymes in
North America. Antimicrob. Agents Chemother. 57:130 –136. http://
dx.doi.org/10.1128/AAC.01686-12.
26. Kristóf K, Tóth A, Damjanova I, Jánvári L, Konkoly-Thege M, Kocsis B,
Koncan R, Cornaglia G, Szego E, Nagy K, Szabó D. 2010. Identification
of a blaVIM-4 gene in the internationally successful Klebsiella pneu-
moniae ST11 clone and in a Klebsiella oxytoca strain in Hungary. J. An-
timicrob. Chemother. 65:1303–1305. http://dx.doi.org/10.1093/jac/
dkq133.
27. Ma L, Lu PL, Siu LK, Hsieh MH. 2013. Molecular typing and resistance
mechanisms of imipenem-non-susceptible Klebsiella pneumoniae in
Taiwan: results from the Taiwan Surveillance of Antibiotic Resistance
(TSAR) study, 2002-2009. J. Med. Microbiol. 62:101–107. http://
dx.doi.org/10.1099/jmm.0.050492-0.
28. Voulgari E, Zarkotou O, Ranellou K, Karageorgopoulos DE, Vrioni G,
Mamali V, Themeli-Digalaki K, Tsakris A. 2013. Outbreak of OXA-48
carbapenemase-producing Klebsiella pneumoniae in Greece involving an
ST11 clone. J. Antimicrob. Chemother. 68:84 – 88. http://dx.doi.org/
10.1093/jac/dks356.
29. Liu P, Li P, Jiang X, Bi D, Xie Y, Tai C, Deng Z, Rajakumar K, Ou HY.
2012. Complete genome sequence of Klebsiella pneumoniae subsp. pneu-
8 ®
mbio.asm.org
Downloaded from mbio.asm.org on May 23, 2016 - Published by mbio.asm.org
moniae HS11286, a multidrug-resistant strain isolated from human spu-
tum. J. Bacteriol. 194:1841–1842. http://dx.doi.org/10.1128/JB.00043-12.
30. Leski T, Vora GJ, Taitt CR. 2012. Multidrug resistance determinants from
NDM-1-producing Klebsiella pneumoniae in the USA. Int. J. Antimicrob.
Agents 40:282–284. http://dx.doi.org/10.1016/j.ijantimicag.2012.05.019.
31. Chevreux B, Wetter T, Suhai S. 1999. Genome sequence assembly using
trace signals and additional sequence information, p 45–56. In Computer
Science and Biology: Proceedings of the German Conference on Bioinfor-
matics (GCB). GCB, Hannover, Germany.
32. Zerbino DR, Birney E. 2008. Velvet: algorithms for de novo short read
assembly using de Bruijn graphs. Genome Res. 18:821– 829. http://
dx.doi.org/10.1101/gr.074492.107.
33. Alikhan NF, Petty NK, Ben Zakour NL, Beatson SA. 2011. BLAST Ring
Image Generator (Brig): simple prokaryote genome comparisons. BMC
Genomics 12:402. http://dx.doi.org/10.1186/1471-2164-12-402.
34. Zhou Y, Liang Y, Lynch KH, Dennis JJ, Wishart DS. 2011. PHAST: a fast
phage search tool. Nucleic Acids Res. 39:W347–W352. http://dx.doi.org/
10.1093/nar/gkq1255.
35. Darling AE, Mau B, Perna NT. 2010. progressiveMauve: multiple ge-
nome alignment with gene gain, loss and rearrangement. PLoS One
5:e11147. http://dx.doi.org/10.1371/journal.pone.0011147.
36. Tamura K, Peterson D, Peterson N, Stecher G, Nei M, Kumar S. 2011.
MEGA5: molecular evolutionary genetics analysis using maximum
likelihood, evolutionary distance, and maximum parsimony methods.
Mol. Biol. Evol. 28:2731–2739. http://dx.doi.org/10.1093/molbev/
msr121.
May/June 2014 Volume 5 Issue 3 e01355-14
Abundance of Antibiotic Resistance Genes in Bacteriophage following
Soil Fertilization with Dairy Manure or Municipal Biosolids, and
Evidence for Potential Transduction
Joseph Ross,a Edward Toppa,b
Agriculture and Agri-Food Canada, London, ON, Canadaa; Department of Biology, University of Western Ontario, London, ON, Canadab

Animal manures and municipal biosolids recycled onto crop production land carry antibiotic-resistant bacteria that can influ-
ence the antibiotic resistome of agricultural soils, but little is known about the contribution of bacteriophage to the dissemina-
tion of antibiotic resistance genes (ARGs) in this context. In this work, we quantified a set of ARGs in the bacterial and bacterio-
phage fractions of agricultural soil by quantitative PCR. All tested ARGs were present in both the bacterial and phage fractions.
We demonstrate that fertilization of soil with dairy manure or human biosolids increases ARG abundance in the bacterial frac-
tion but not the bacteriophage fraction and further show that pretreatment of dairy manure can impact ARG abundance in the
bacterial fraction. Finally, we show that purified bacteriophage can confer increased antibiotic resistance to soil bacteria when
combined with selective pressure. The results indicate that soilborne bacteriophage represents a substantial reservoir of antibi-
otic resistance and that bacteriophage could play a significant role in the horizontal transfer of resistance genes in the context of
an agricultural soil microbiome. Overall, our work reinforces the advisability of composting or digesting fecal material prior to
field application and suggests that application of some antibiotics at subclinical concentrations can promote bacteriophage-me-
diated horizontal transfer of ARGs in agricultural soil microbiomes.

O veruse of antibiotics has contributed to the spread of antibi-

otic resistance due to the release of antibiotics, antibiotic-
resistant bacteria, and antibiotic resistance genes (ARGs) into the
environment (1). This phenomenon is mediated by horizontal
gene transfer (HGT) of mobile genetic elements—such as plasmids,
transposons, and integrons— between bacterial cells through conju-
gation and viral transduction (2–4).
Transduction has been shown to occur in environmental matri-
ces, including freshwater and wastewater; moreover, bacterial 16S
ribosomal RNA sequences have been observed in the viral fraction of
wastewater, confirming the ability of bacteriophage to carry bacterial
genes (5, 6). In fact, while much attention has been paid to conjuga-
tion, more recent work has additionally implicated bacteriophage as a
major vehicle for horizontal gene transfer and recombination (7–10).
Using high-throughput sequencing of murine fecal phage popu-
lations, Modi et al. (11) demonstrated that antibiotic treatment leads
to the enrichment of genes conferring resistance to the administered
drug as well as to unrelated antibiotics in the phage genome. Further-
more, bacteriophage from drug-treated mice provided cultured naive
microbiota with increased levels of resistance to the corresponding
drug. Overall, this work concluded that antibiotic residues potentiate
the transduction-mediated dissemination of ARGs.
Antibiotic resistance genes have been found in the bacterio-
phage DNA fraction of various environmental matrices, such as
activated sludge (12), urban sewage and river water (13), and
wastewater effluents from hospitals and wastewater treatment
plants (14). The aforementioned studies indicate that bacterio-
phage represents a reservoir of ARGs across a broad selection of
environments. However, it remains to be determined if bacterio-
phage has such a role in agricultural soils.
In the present study, a set of ARGs previously detected in ma-
nured soils was quantified in bacteria and in bacteriophage recov-
ered from agricultural soil using quantitative PCR (qPCR) (15–17).
The impact of soil amendment either with dairy manure or with
biosolids on the abundance and distribution of ARGs in bacteria
and bacteriophage was determined. In the case of dairy manure,
the effect of various preapplication treatments on the composition
of ARGs in the bacterial and bacteriophage fractions was deter-
mined. Finally, we tested whether bacteriophage enriched from
biosolids conferred increased antibiotic resistance to soilborne
bacteria when combined with selective pressure.
MATERIALS AND METHODS
Field operations and soil sampling. The soil samples used in the experi-
ments described here were obtained from field experiments undertaken dur-
ing the 2014 growing season on the Agriculture and Agri-Food Canada re-
search farm in London, Ontario, Canada (42.984°N, 81.248°W). The field
installations and methods were described in detail by Marti et al. (17). Briefly,
the soil is a silt loam (gray-brown luvisol) with the following properties: a pH
of 7.5; a cation exchange capacity of 13.2; a sand, silt, and clay composition of
18%, 67%, and 15%, respectively; and an organic matter content of 3.4%.
Manure for application in the spring of 2014 was obtained from dairy farms in
the London area, and biosolids that were aerobically digested and dewatered
were obtained from the municipality of Tilsonburg, Ontario, Canada. The
dairy manures were variously untreated (raw), anaerobically digested, me-
chanically dewatered, or composted. The dewatered and composted manures
Received 22 July 2015 Accepted 1 September 2015
Accepted manuscript posted online 4 September 2015
Citation Ross J, Topp E. 2015. Abundance of antibiotic resistance genes in
bacteriophage following soil fertilization with dairy manure or municipal
biosolids, and evidence for potential transduction. Appl Environ Microbiol
81:7905–7913. doi:10.1128/AEM.02363-15.
Editor: D. W. Schaffner
Address correspondence to Edward Topp, ed.topp@agr.gc.ca.
Supplemental material for this article may be found at http://dx.doi.org/10.1128
/AEM.02363-15.
Copyright © 2015, American Society for Microbiology. All Rights Reserved.

November 2015 Volume 81 Number 22 Applied and Environmental Microbiology aem.asm.org 7905
Ross and Topp
were applied at a rate of 5 dry metric tons/ha, and the raw and the digested
manures were applied at a rate of 80,000 liters/ha.
Soil cores (2 cm wide, 15 cm deep) were taken immediately after ma-
nure or biosolid application, as well as at 7 and 30 days postapplication
(days 0, 7, and 30, respectively), in order to assess any temporal impact on
ARG composition. Six cores were sampled from each of three replicated
plots using a T sampler rinsed with 70% ethanol between samplings.
Cores were bulked into a labeled Ziploc bag, mixed by hand until they
were homogeneous, and transported to the laboratory in a cooler with
cool packs. Thus, three independent soil samples (each representing the
average for 6 cores from a given plot) from control, dairy manure, and
biosolids treatments were analyzed at each sampling time.
Extraction of bacterial and phage DNA fractions from soil. Twenty-
five-gram portions of soil were mixed with 25 ml of sterile SM buffer (100
mM NaCl, 8 mM MgSO4 · 7H2O, 50 mM Tris-HCl, pH 7.5) in a 50-ml
Falcon tube and thoroughly mixed by vortexing for 1 h. The resulting
slurry was centrifuged at 2,500 ⫻ g for 4 min to pellet the soil, and the
supernatant was passed through a 0.22-␮m-pore-size Durapore mem-
brane filter (Millipore) under vacuum. The filter was rinsed with 10 ml of
SM buffer. Bacteria were retained on the filter, whereas bacteriophage
passed through and was in the filtrate.
The membrane was transferred to a 15-ml Falcon tube and incubated
with 1 ml GITC (5 M guanidine isothiocyanate, 100 mM EDTA, 0.5%
N-lauroylsarcosine) buffer for 1 h at 75°C. DNA was extracted using a
modification of the Qiagen DNeasy blood and tissue kit procedure.
Briefly, 0.5 ml of buffer AL was added to each tube before vortexing for 1
min, adding 0.5 ml 95% ethanol, vortexing again, and loading the con-
tents of the tubes onto the provided spin columns in 700-␮l increments.
The columns were then washed once with 500 ␮l of buffer AW1 and twice
with 700 ␮l of buffer AW2. The columns were centrifuged for 2 min at
maximum speed to remove residual wash buffer, before they were eluted
twice with 50 ␮l buffer AE. The resulting 100 ␮l of DNA, now enriched for
bacterial DNA, was further purified and concentrated by ethanol precip-
itation, using a 1:10 volume of 3 M sodium acetate, 3 volumes of 95%
ethanol, and RNA-grade glycogen (Life Technologies) as a coprecipitate.
The pellet was resuspended in 12 ␮l of 1⫻ TE (10 mM Tris-HCl, pH 7.5,
1 mM EDTA), quantified with a NanoDrop ND1000 microspectropho-
tometer (NanoDrop Technologies, Wilmington, DE), and diluted to a
final concentration of 1 ng/␮l using deionized water. The DNA extract
was aliquoted and stored at ⫺20°C.
The soil filtrate was filtered once more using a 0.22-␮m-pore-size
Sterivex syringe filter. The phage in the filtrate was concentrated using
Amicon Ultra-15 filtration devices (30-kDa-molecular-mass cutoff) per
the manufacturer’s instructions. The resulting 400 to 600 ␮l was further
enriched for bacteriophage by ultracentrifugation on CsCl density gradi-
ents of 1.7, 1.5, and 1.35 g/ml as described in reference 18. The resulting 1
ml of fluid containing enriched phage was desalted using Amicon Ultra-
0.5 filtration devices (30-kDa-molecular-mass cutoff) per the manufac-
turer’s instructions. The 30 ␮l of concentrate was adjusted to 1 ml with
sterile SM buffer and stored at 4°C. To ensure the removal of nonphage
DNA prior to lysis of the phage particles, 0.5 ml of enriched bacteriophage
was treated with 10 U of Turbo DNase (Life Technologies) for 1 h at 37°C
and then for 5 min at 65°C. To isolate bacteriophage DNA, 0.4 mg of
proteinase K (final concentration, 0.7 mg/ml) was incubated with the
DNase-treated bacteriophage preparation for 1 h at 37°C to digest the
capsid proteins. The resulting lysate was extracted twice with phenol-
chloroform-isoamyl alcohol, and the DNA was further purified and con-
centrated by ethanol precipitation as described above.
Each time that the above-described procedures were carried out, a
series of conventional PCRs was performed in order to assess the quality of
bacteriophage isolation. The presence of bacterial rrnS DNA (encoding
the 16S rRNA) was assessed using the primers described in Table 1. PCRs
were carried out with the following cycle conditions: 1 cycle at 95°C for 10
min followed by 30 cycles of 95°C for 15 s, 55°C for 15 s, and 72°C for 1
min. GoTaq Flexi reagents (Promega) used in the reaction mixtures, and
7906 aem.asm.org Applied and Environmental Microbiology
the reaction mixtures were composed of 1⫻ Green GoTaq Flexi buffer; 1.5
mM MgCl2; 0.2 mM (each) dATP, dGTP, dCTP, and dTTP; 0.4 ␮M each
primer; 1 U of GoTaq DNA polymerase; and deionized water to 25 ␮l. The
reactions were carried out at the following stages of the extraction proce-
dure: for bacterial DNA, after the full procedure, including ethanol pre-
cipitation; for bacteriophage particles, after Turbo DNase treatment but
before proteinase K lysis; and for bacteriophage DNA, after the full pro-
cedure, including ethanol precipitation. The ideal result was a positive
signal from the bacterial DNA preparations and a negative result from
bacteriophage particle preparations, indicating the absence of bacterial
DNA outside the bacteriophage particles. Thus, any bacterial DNA de-
tected in the enriched bacteriophage DNA fraction must be due to a legit-
imate uptake of that DNA into the phage particle during the viral replica-
tion cycle. PCR products were analyzed by agarose gel electrophoresis; a
representative gel is shown in Fig. S1 in the supplemental material. In all
cases, no bacterial DNA (above the levels seen for a no-template negative
control) was detected for either intact phage particles or purified phage
DNA. To ensure that the lack of amplification from bacteriophage parti-
cles or bacteriophage DNA was not simply due to the presence of PCR
inhibitors, these reactions were also performed in the presence of the same
DNA used for the positive control, with positive results being obtained
(see Fig. S1 in the supplemental material).
Quantification of gene target copies. The primers and probes (Sigma-
Aldrich, Toronto, ON, Canada) used in the present study are summarized
in Table 1. The resistance genes quantified were strA, strB, sul1, aadA, and
blaOXA-20, which encode resistance to streptomycin (strA and strB), sulfa-
methazine (sul1), aminoglycosides (aadA), and ␤-lactams (blaOXA-20).
Note that the amount of blaOXA-20 in dairy manure-amended soils was
consistently below the limit of detection, and therefore, blaOXA-20 is not
discussed. As a control for total bacterial DNA abundance among the
various samples, rrnS (encoding the 16S rRNA) was also quantified. PCR
amplification was performed using a Bio-Rad CFX96 real-time PCR in-
strument with Bio-Rad CFX Manager software, version 3.0. The reactions
were performed with the Brilliant II quantitative PCR (qPCR) master mix
(Agilent, Toronto, ON, Canada) for the TaqMan PCR and the Brilliant II
SYBR green qPCR master mix (Agilent) for the SYBR green PCR. Two
microliters of template DNA (corresponding to 2 ng of either bacterio-
phage or bacterial DNA) was added, and deionized water was used to
reach a final volume of 25 ␮l. Negative controls without template DNA
were run in triplicate. Each reaction was run in triplicate with the follow-
ing cycle conditions: 1 cycle at 95°C for 10 min followed by 40 cycles of
95°C for 15 s, the annealing temperature indicated in Table 1 for 35 s, and
72°C for 1 min. For the SYBR green assay, a melting curve step was added
in order to check the purity of the PCR product. This step consisted of a ramp
of the temperature from 65 to 95°C at an increment of 0.5°C and a hold for 5
s for each step. The plasmids used to generate standard curves were created as
described by Marti et al. (16). Standard curves were prepared with a 10-fold
serial dilution of the known concentration of the plasmid solution for each
marker, in order to have final concentrations of the plasmid ranging from 107
to 10° copies per microliter. The identities of the quantified gene targets were
ensured on the basis of hybridization when using TaqMan chemistry or melt-
ing behavior when using SYBR green. Gene abundance data are presented as
the number of gene copies (GC) per nanogram of bacterial or phage DNA
included in the reaction mixture.
Transduction of soil bacteria by biosolids-derived phage. Total bac-
teriophage was extracted and enriched from biosolids (three independent
replicates) essentially as described above for soil, except that the initial
centrifugation was increased to 15 min at 10,000 ⫻ g and the filtration
steps were repeated two to three times to ensure the adequate removal of
bacteria. Phage was stored at 4°C in SM buffer. In order to ensure that the
aforementioned phage purification procedure, while stringent, yielded an
adequate number of infection-competent phage particles, the titer of co-
liphage within the resulting bacteriophage enrichments was estimated
using the single-agar-layer method essentially as described in reference 23.
Escherichia coli K-12 (grown in tryptone-yeast extract-glucose broth) was
November 2015 Volume 81 Number 22
 Antibiotic Resistance in Soil Phage

TABLE 1 Primers and probes used for conventional or quantitative PCR in this study
Primer specificity and Product Annealing Final primer
primer namea Sequence (5= to 3=)b size (bp) temp (°C) concn (nM) Target Reference
Universal bacteria
(conventional
PCR)
GM5-F CCTACGGGAGGCAGCAG 550 55 400 rrnS gene 19
907-R CCGTCAATTCCTTTGAGTTT 400

Universal bacteria
BACT1369F CGGTGAATACGTTCYCGG 123 59 300 rrnS gene 20
PROK1492R GGWTACCTTGTTACGACTT 300
TM1389F HEX-CTTGTACACACCGCCCGTC-BHQI 300

strA
strA-F TATGGTTGTTTGCCATGGTG 149 62 400 Streptomycin 15
strA-R TTCTCTTCGGCGTTAGCAAT 400 Phosphotransferase A

strB
strB-F ATCGCTTTGCAGCTTTGTTT 143 61 300 Streptomycin 21
strB-R ATGATGCAGATCGCCATGTA 300 Phosphotransferase B
strB-P HEX-ATGCCTCGGAACTGCGT-BHQI 200

sul1
sul1-F GACTGCAGGCTGGTGGTTAT 105 64 200 Sulfamethazine 16
resistance gene 1
sul1-R GAAGAACCGCACAATCTCGT 200

aadA
aadA-F CAGCGCAATGACATTCTTGC 293 63 200 Aminoglycoside 21
adenyltransferase A
aadA-R GTCGGCAGCGACA(C/T)CCTTCG 200
aadA-P HEX-TGGTAGGTCCAGCGGCGGAG-BHQI 300

blaOXA-20
blaOXA-20-F TGATGATTGTCGAAGCCAAA 100 60 400 Beta-lactamase, class D 22
(oxacillinase)
blaOXA-20-R GCCTGTAGGCCACTCTACCC 400
a
F, forward primer; R, reverse primer; P, probe.
b
HEX, 2=,4=,5=,7=-tetrachloro-6-carboxy-4,7-dichlorofluorescein succinimidyl ester; BHQI, black hole quencher 1.

used as the host, and tryptone-yeast extract-glucose agar was the medium.
Control plates that received the enriched bacteriophage preparation but
no strain K-12 bacteria had no visible colonies, corroborating the conclu-
sion from the rrnS-specific PCR that the phage preparations were free of
bacterial contamination. The average titer of coliphage was determined to
be about 300 PFU/ml in our enrichments. For comparison, Colomer-
Lluch et al. (13) reported coliphage titers of about 104 PFU/ml and 102
PFU/ml in urban sewage and river water, respectively; Ashelford et al. (24)
reported about 107 phage per gram of soil associated with plant roots in
sugar beet fields. While our coliphage titer was low by comparison to the
titers in the aforementioned studies, it should be noted that these were
assessments of the bacteriophage titers in environmental matrices that
had not undergone phage purification procedures and that the titer of
K-12-infecting coliphage reported here is likely a vast underestimate of
the total number of infection-competent bacteriophage recovered.
We chose to assess the impact of biosolids-derived phage on soilborne
microbiota without the complication of soil particles, which might adsorb
to antibiotics or bacteriophage and thus interfere with the effective free
concentration of drugs or phage particles. Accordingly, total bacteria were
extracted from soil in an aqueous context as follows: 20 g of soil was mixed
with 20 ml of SM buffer in a 50-ml Falcon tube by vortexing for 1 h. The
resulting slurry was centrifuged at 2,000 ⫻ g for 2 min to remove most soil
and large particulates, and the supernatant was removed and placed into
two 15-ml Falcon tubes (5 ml per tube). One tube served as a no-phage
control and received 500 ␮l of SM buffer. The other tube was amended
with 500 ␮l of biosolids-derived phage; this was performed for three in-
dependent biological isolates of phage and soil-derived supernatant. After
1 h of incubation at 30°C, ampicillin, cefoxitin, or sulfamethazine was
added to various final concentrations (see Table 2). The resulting trans-
duction reaction mixtures were incubated for 3 days at 30°C, serially di-
luted, and plated on either selective or control Chromocult agar. Thus, we
chose to focus on resistant versus total coliform bacteria as a subset of the
total population of potential bacterial transductants generated in this as-
say. Control plates lacked antibiotics (in order to count the total number
of coliforms), while selective plates contained ampicillin (32 ␮g/ml), cefoxitin
(32 ␮g/ml), or sulfamethazine (512 ␮g/ml) in order to count the number of
resistant coliform bacteria. Drug concentrations were based on breakpoints
(i.e., MICs) reported by the Canadian Integrated Program for Antimicrobial
Resistance Surveillance (25). The frequency of resistant coliforms was calcu-
lated by dividing the resistant coliform count (in numbers of CFU per milli-
liter) by the total coliform count (also in numbers of CFU per milliliter)
counted in the absence of antibiotics. Phage-only controls were also included
and yielded no bacterial growth. In all cases, the Chromocult plates were
incubated at 37°C overnight. Note that these 3 antibiotics were chosen on the
basis of an initial qualitative screen of 10 antibiotics (see Fig. S2 in the supple-
mental material). These 10 antibiotics were chosen to represent several major

November 2015 Volume 81 Number 22 Applied and Environmental Microbiology aem.asm.org 7907
Ross and Topp

FIG 1 Abundance of gene targets in the bacterial and bacteriophage DNA fractions of soil amended with dairy manure. Bar graphs represent the number of gene
copies per nanogram of template DNA, as measured by qPCR. The gene targets quantified are indicated above the corresponding graphs. As indicated at the
bottom, soil either did not receive manure (control [Con]) or was amended with raw (Raw) dairy manure or manure that had been digested (Dig), dewatered
(DW), or composted (Comp) prior to application. Soil from these plots was sampled immediately after application, as well as 7 and 30 days later (days 0, 7, and
30, respectively). A no-template control (⫺) was also included in the qPCRs. Data are presented as the mean and standard deviation from three biological
replicates. *, a statistically significant difference between the bacterial and corresponding phage DNA fractions. A statistically significant difference between a
given manure treatment and the corresponding control not receiving manure is indicated by the letter a or b for the bacterial or phage fraction, respectively.

classes (e.g., penicillins, aminoglycosides, macrolides, cephalosporins, car-
bapenems, quinolones, sulfonamides).
Calculations and statistics. Each condition was analyzed in triplicate.
In the qPCR experiments, three technical replicates were performed for
each of three independent biological isolates for each sample. In the trans-
duction experiment, we report the averages from three independent bio-
logical replicates. Statistically significant treatment effects were deter-
mined using an unpaired t test without assuming equal variance. Data
were analyzed using SigmaPlot software, version 12.5 (Systat Software
Inc.). The significance level was set at a P value of 0.05.
RESULTS
Impact of manure application on abundance and distribution
of antibiotic resistance genes in bacteriophage and bacterial
fractions. The abundance of rrnS in the bacterial fraction did not
vary with manure application (Fig. 1, top). Soil carried about 105
gene copies (GC) per ng of template DNA at day 0 and about 106
GC/ng DNA 7 and 30 days following manure application. The
increase in rrnS copy numbers at days 7 and 30 relative to the
numbers at day 0 was statistically significant for most of the cor-
responding treatments, namely, the control, raw, digested, dewa-
tered, and composted treatments for day 0 versus day 7 and the
control, digested, and composted treatments for day 0 versus day
30. For simplicity, the results of these statistical analyses are not
indicated in Fig. 1. Overall, there was a time-dependent increase in
bacterial abundance following manure application. The absence
of any effect of manure preapplication treatment on rrnS abun-
dance indicates that the overall abundance of bacteria was unaf-
fected by these treatments. Thus, any changes in ARG abundance

7908 aem.asm.org Applied and Environmental Microbiology November 2015 Volume 81 Number 22

seen between manure treatments would not simply reflect a gross
variation in the overall bacterial abundance.
As expected, the rrnS gene target was far less abundant in the
bacteriophage DNA fraction for all treatments, where it was pres-
ent at a level of about 102 GC/ng DNA (Fig. 1, top). Importantly,
this level was not significantly different from the background level
of rrnS measured in a no-template control. This means that any
ARGs observed in the phage fraction do not simply reflect the
rampant acquisition of bacterial genes overall, since the rrnS gene
is very rarely acquired by bacteriophage in our soil samples. As
with the bacterial fraction, the rrnS copy number in the bacterio-
phage DNA fraction was consistent among all treatments, indicat-
ing little or no impact of manure application or pretreatment of
manure on the overall bacterial DNA abundance in bacterio-
phage.
A no-template control was also evaluated for each of the ARG
primer/probe sets, and no gene targets were detected, indicating
low background levels for each of the ARGs. Within the bacterial
fraction, each of the detected ARGs was significantly more abun-
dant in soil receiving raw manure than in control soil not receiving
manure (Fig. 1). For instance, the level of sul1 increased from
about 101 GC/ng DNA in control soil to about 105 GC/ng DNA in
soil amended with raw manure. Furthermore, ARG abundance
was significantly higher for soils amended with digested and de-
watered manure than for control soil, indicating that these treat-
ments are ineffective at mediating the dissemination of ARGs by
manure application. In contrast, composting of manure prior to
field application did not increase the abundance of ARG targets
above that in soils not receiving manure, indicating the efficacy of
this practice. The ARG abundance in the bacterial fraction tran-
siently increased on day 7 relative to that on day 0 and then de-
creased by day 30. This suggests that there may have been an initial
proliferation of ARG-harboring bacteria in the days following ma-
nure application, followed by their decline by day 30.
In soils not receiving manure, each ARG quantified in the bac-
terial fraction was also detected in the phage fraction (Fig. 1).
Moreover, ARG abundance was not significantly lower in the
phage fraction than in the corresponding bacterial fraction in soils
not receiving manure, with the exception of the abundances of
strA (not detected on day 7) and strB (not detected on day 7 or 30).
Strikingly, aadA was significantly more abundant in the phage
fraction than in the bacterial fraction on days 7 and 30 (Fig. 1,
bottom). In stark contrast to the findings for the bacterial fraction,
ARG abundance in the phage fraction did not generally respond
significantly to manure treatment or time (Fig. 1). Instead, the
phage fraction appeared to harbor a constant reservoir of ARGs at
a background abundance roughly matching that seen in the bac-
terial fraction in the absence of manure. This abundance did not
respond to amendment with manure, regardless of preapplication
treatment.
Impact of biosolids application on abundance and distribu-
tion of antibiotic resistance genes in bacteriophage and bacte-
rial fractions. The abundance of rrnS was between 104 and 105
GC/ng DNA in untreated soil and did not increase significantly
following application of aerobically digested biosolids (Fig. 2). In
fact, there was a significant decrease on days 0 and 7 relative to that
in untreated soil. Thus, any increase in ARG abundance in treated
versus control soil would not simply reflect an increase in bacterial
DNA abundance.
Following biosolids application, the gene targets strA, strB, and
November 2015 Volume 81 Number 22 Applied and Environmental Microbiology
Antibiotic Resistance in Soil Phage
sul1 were more abundant in the bacterial fraction of treated soil
than it that of control soil (Fig. 2). The abundance of these gene
targets declined thereafter through day 30. In contrast, aadA and
blaOXA-20 were not enriched by biosolids application. Further-
more, other than aadA on day 30, both aadA and blaOXA-20 other-
wise remained at comparable levels in the treated and control soils
through the period of investigation.
Every gene target was detected in the bacteriophage fraction in
control soils, with the exception of strA, which was undetected in
both the bacterial and bacteriophage fractions (Fig. 2). Only sul1
was significantly less abundant in the bacteriophage fraction than
in the bacterial fraction for untreated soil. ARG abundance in the
bacteriophage fraction did not respond significantly to biosolids
treatment or time for most samples, with the exception of sul1
abundance. On day 0, the abundance of this gene target increased
slightly but significantly in treated soil relative to that in control
soil (Fig. 2). As was observed with dairy manure, the phage frac-
tion appears to harbor a constant reservoir of ARGs at the back-
ground abundance, and the abundance in the phage fraction had
little or no response to amendment with biosolids.
Taken together, these results suggest that soilborne bacteriophage
represents a reservoir of antibiotic resistance genes. Amendment of
soil with dairy manure or biosolids had no significant effect on the
abundance of gene targets in the bacteriophage fraction but in-
creased the abundance in the bacterial fraction.
Potential for transduction. Aqueous suspensions of soil bac-
teria were incubated for 3 days with or without bacteriophage
enriched from biosolids and with or without supplementation
with antibiotics (Table 2). Note that we chose to quantify coliform
bacteria as a representative subset of bacteria within the total soil-
borne microbiome. In the absence of added antibiotics, there was
no effect of bacteriophage supplementation on the abundance of
viable antibiotic-resistant coliform bacteria (Table 2). However,
after 3 days in the presence of 1/10 the breakpoint concentration
of cefoxitin, the abundance of cefoxitin-resistant coliforms was
3.7-fold higher in the presence than in the absence of bacterio-
phage (Table 2). Likewise, phage conferred a 6.3-fold increase in
the abundance of sulfamethazine-resistant coliforms when incu-
bated in the presence of 1/100 of the sulfamethazine breakpoint
concentration and a 7.1-fold increase in the presence of 1/10 of its
breakpoint concentration (Table 2). Phage appeared to confer a
slight decrease (less than 2-fold) in the abundance of cefoxitin-
resistant coliforms when incubated in the presence of 1/100 of the
cefoxitin breakpoint concentration. On the other hand, the abun-
dance of ampicillin-resistant coliforms did not vary with any
treatment. Taken together, these results suggest that bacterio-
phage from biosolids increased the abundance of coliform bacte-
ria resistant to sulfamethazine or cefoxitin in the presence but not
the absence of each antibiotic.
DISCUSSION
Distribution of antibiotic resistance genes in bacterial and bac-
teriophage fractions of agricultural soils. Several studies have
detected ARGs in the bacteriophage metagenome (or phageome)
of a wide variety of environmental matrices, including activated
sludge (9, 12), urban sewage and river water (13), and various
wastewater effluents from hospitals and wastewater treatment
plants (14, 26). When specific ARGs are detected by real-time PCR
in the bacteriophage populations of the above-described environ-
ments, their levels are only about 10-fold lower than those in the
aem.asm.org 7909
Ross and Topp

FIG 2 Abundance of gene targets in the bacterial and bacteriophage DNA fractions of soil amended with aerobically digested biosolids. Bar graphs represent the
number of gene copies per nanogram of template DNA, as measured by qPCR. The gene targets quantified are indicated above the corresponding graphs. As
indicated at the bottom of each graph, soil was untreated (control [Con]; sampled immediately before application) or received biosolids and was sampled
immediately after application (day 0) or at 7 and 30 days postapplication. A no-template control (⫺) was also included in the qPCRs. Data are presented as the
mean and standard deviation from three replicates. *, a statistically significant difference between the corresponding bacterial and phage DNA fractions. A
statistically significant difference between a given biosolids-treated time point and the corresponding untreated control is indicated by the letter a or b for the
bacterial or phage fraction, respectively.

corresponding bacterial fractions, on average (reviewed in refer-
ence 2). The present study further demonstrates that the phageome of
agricultural soil harbors ARGs. The ARGs tend to be present in
similar numbers in the corresponding bacterial and phage DNA
fractions from agricultural soils, at least in the absence of manure
or biosolids application. ARG copy abundance in the bacterial
fraction rose sharply in response to dairy manure application and,
in the case of strA, strB, and sul1, also rose in response to biosolids
application. In the case of manure, some preapplication treatment
options, specifically, composting, attenuated the enrichment ef-
fect. As measured by rrnS gene target copy abundance, there was
no stimulation of total bacterial populations with the addition of
manure, regardless of manure treatment. In contrast, the abun-
dance of ARG target copies did respond to manure application. It
is likely that the ARGs originated from bacteria that were intro-
duced by application of dairy manure and biosolids, especially
given that composting of the manure attenuated this enrichment
effect.
In contrast to the responsiveness of the ARG target abun-
dance in the bacterial fraction, ARG levels were virtually iden-
tical in the phageome, regardless of treatment or the time of
sampling. It is unclear why ARG levels in the phageome did not
respond to manure application or pretreatment in parallel to
the bacteriome.
Here, we used rrnS (the 16S rRNA gene) as an indicator of
overall bacterial abundance; rrnS is also useful as a representative
gene for measurement of overall bacterial gene acquisition by bac-
teriophage (6). Since rrnS is ubiquitous in bacterial species, it
should represent a baseline level of bacterial gene acquisition by
phage. Notably, rrnS levels in the phage fraction of all soils sam-
pled in this study were lower than those in the corresponding
bacterial fraction by several orders of magnitude (indeed, in most

7910 aem.asm.org Applied and Environmental Microbiology November 2015 Volume 81 Number 22
 Antibiotic Resistance in Soil Phage

TABLE 2 Abundance of ampicillin-, cefoxitin-, and sulfamethazine-resistant coliforms recovered from soil suspensions following 3 days of
incubation in presence or absence of bacteriophage enriched from biosolidsa
Drug-resistant coliform count/total coliform count
Drug concnb during incubation Bacteriophage supplementation Ampicillin Cefoxitin Sulfamethazine
0 ⫺ 0.000337 ⫾ 0.000052 0.00724 ⫾ 0.00917 0.245 ⫾ 0.145
⫹ 0.00101 ⫾ 0.00099 0.00335 ⫾ 0.00273 0.129 ⫾ 0.037

1:100 ⫺ 0.00143 ⫾ 0.00142 0.00315 ⫾ 0.00044 0.00326 ⫾ 0.00291

⫹ 0.00252 ⫾ 0.01480 0.00168 ⫾ 0.00067* 0.0207 ⫾ 0.0132*

1:10 ⫺ 0.411 ⫾ 0.323 0.0512 ⫾ 0.0239 0.0342 ⫾ 0.0459

⫹ 0.116 ⫾ 0.031 0.192 ⫾ 0.060* 0.243 ⫾ 0.098*
a
Suspensions were incubated without the addition of antibiotics or with one of the antibiotics at either a 1:100 or a 1:10 clinical breakpoint concentration. Data represent the
mean ⫾ standard deviation (n ⫽ 3) of the count of resistant coliforms (in numbers of CFU per milliliter) normalized to the total coliform count (in numbers of CFU per milliliter).
*, statistically significant differences between phage treatment and no phage treatment.
b
1:100 and 1:10, final concentrations 100-fold and 10-fold, respectively, below the breakpoint concentrations of the indicated antibiotics (ampicillin, 32 ␮g/ml; cefoxitin, 32 ␮g/ml;
sulfamethazine, 512 ␮g/ml).

cases, rrnS levels in the phage fraction were not even above the
background rrnS levels, defined in control qPCRs lacking a tem-
plate). This suggests that bacterial gene uptake by phage is a rela-
tively rare event, in agreement with previous estimates of about
one transduction event per every 108 phage infections (27).
Potential transduction of soil coliform bacteria by biosolids-
derived phage. We chose to focus on coliform bacteria as a subset
of the total population of potential bacterial transductants gener-
ated in our experimental setup. This had the advantage of allowing
us to quantify a relatively homogeneous set of bacterial species; the
counting of total viable bacteria from a complex matrix like soil is
cumbersome, and we would have no idea if increases in the pop-
ulation of viable bacteria were due to the transduction of the var-
ious species or due to some shift in the complex dynamics of
interspecies competition due to the perturbations of both antibi-
otic selection and increased phage infective burden. By focusing
on coliforms, we thus took a more reductionist approach. A major
drawback is that we likely grossly underestimated the true fre-
quencies of transduction in agricultural soil, as well as the true
extent of the influence of subclinical selective pressure. Neverthe-
less, our results serve as a proof of concept that subclinical con-
centrations of some antibiotics can potentiate the phage-mediated
horizontal transfer of resistance genes into potential human
pathogens in the context of an agricultural soil microbiome.
We assume that the enrichment of antibiotic-resistant coli-
forms in the presence of bacteriophage is due to transduction,
although we use this term broadly, as the mechanism of gene
transfer and whether it is generalized or specialized (4) have not
been proven. A few lines of evidence support the hypothesis that
transduction was at play. First, the effect occurred specifically
when enriched bacteriophage was incubated with soil-derived
bacteria. Second, PCR analyses of the purified bacteriophage
preparations yielded no evidence of bacterial DNA, and phage-
only plating controls gave no indication of bacterial growth, indi-
cating that the bacteriophage enrichments were not contaminated
with bacterial DNA or viable coliform bacteria. The increased re-
sistance in the presence of enriched bacteriophage must therefore
be specific to the phage; the resistance genes may have originated
from bacteria in the biosolids from which the bacteriophage was
enriched, or the bacteriophage may be promoting the transfer of
resistance genes within the soil bacterial community.
Transduction is known to be feasible in soil (28) and is un-
doubtedly a widespread phenomenon in many environments. For
instance, retail chicken meat carries a number of phage capable of
transferring antimicrobial resistance; of 243 phage randomly iso-
lated from chicken meat, about 25% was able to transduce into E.
coli resistance to 1 or more of the 35 antimicrobials tested (29).
Based on factors like phage versus bacterial abundance, the num-
ber of phage in a transduction-competent state, and the physical
conditions of various environments, Muniesa et al. (8) concluded
that phage-mediated horizontal transfer between intestinal bacte-
ria or between intestinal and indigenous bacteria in extraintestinal
environments was probable. Moreover, bacteriophage isolated
from antibiotic-treated mouse feces can confer antibiotic resis-
tance to cultured murine intestinal microbiota (11). In-feed anti-
biotics induce prophage in swine fecal microbiomes (30), and
horizontal gene transfer (including transduction) is subject to se-
lective pressure (31). For instance, a study examining the inci-
dence of HGT of ARGs in thousands of microbial genomes found
that the genes involved in HGT events are 25-fold more likely to
become fixed in human-associated bacteria than in diverse envi-
ronmental isolates, due to selective pressure (32). Combined with
our observation of probable transduction only in the presence of
selective pressure, it is tempting to speculate that transduction is
generally potentiated by antibiotic selection.
It is unclear why ampicillin selection did not confer the trans-
duction of ampicillin resistance and why cefoxitin selection (at
1/100 of the breakpoint concentration) caused a slight but signif-
icant decrease in cefoxitin resistance in the presence versus ab-
sence of phage. In the case of the latter, the decrease was quite
small (less than 2-fold), and a larger increase in resistance (nearly
4-fold) was conferred by phage when cefoxitin selection was ap-
plied at 1/10 of the cefoxitin breakpoint concentration, perhaps
indicating that a lack of adequate selection by the lower cefoxitin
concentration simply failed to promote transduction; the 2-fold
decrease might thus represent an anomaly. It would be interesting
to test whether the transduction frequency also correlates with the
incubation time, as precedent exists for transduction efficiency
peaking after a defined time period: in early transduction experi-
ments with soil, Zeph et al. (28) noted that optimal incubation
times existed for P1 transduction of chloramphenicol resistance
into E. coli recipients. It is possible that different antibiotics are
selecting for transduction by different sets of phage, each of which
yields peak transduction after specific periods of time. Shousha et

November 2015 Volume 81 Number 22 Applied and Environmental Microbiology aem.asm.org 7911
Ross and Topp
al. (29) observed that of 243 bacteriophage randomly isolated
from chicken meat, about a quarter was able to transduce resis-
tance to one or more of the five antibiotics tested into E. coli.
Resistance to kanamycin was transduced the most often, followed
by that to chloramphenicol, while only a few phage transduced
tetracycline or ampicillin resistance. It thus seems likely that some
ARGs are subject to transduction less frequently than others. In
our experiment, it is curious that ampicillin selection yielded no
transduction, while cefoxitin selection did, since beta-lactamases
or ampC promoter or attenuator mutations (33) might reasonably
yield resistance to both antibiotics. One possibility is that selection
with ampicillin did not yield an increase in transduction as robust
as that seen for selection with cefoxitin. Thus, different types of
selection might display different potentiating effects on the trans-
duction frequency. It would be interesting to test whether selec-
tion with various antibiotics during transduction yields increased
resistance to unrelated antibiotics in our experimental setup.
Overall, our results suggest that the application of subclinical
concentrations of antibiotics can promote bacteriophage-medi-
ated horizontal transfer of antibiotic resistance genes in agricul-
tural soil microbiomes. Further work is required to determine if
antibiotics entrained into soil through application of animal or
human wastes might increase the risk of spreading antibiotic re-
sistance to potential human pathogens.
ACKNOWLEDGMENTS
This research was supported by competitive funding through the AAFC
Growing Forward 2 program. J.R. was supported by the NSERC Visiting
Fellowship in Government Laboratories program.
We thank S. Gordon for valued technical assistance.
REFERENCES
1. Martinez JL. 2009. Environmental pollution by antibiotics and by antibi-
otic resistance determinants. Environ Pollut 157:2893–2902. http://dx.doi
.org/10.1016/j.envpol.2009.05.051.
2. Muniesa M, Colomer-Lluch M, Jofre J. 2013. Potential impact of envi-
ronmental bacteriophages in spreading antibiotic resistance genes. Future
Microbiol 8:739 –751. http://dx.doi.org/10.2217/fmb.13.32.
3. Marti E, Variatza E, Balcazar J. 2014. The role of aquatic ecosystems as
reservoirs of antibiotic resistance. Trends Microbiol 22:36 – 41. http://dx
.doi.org/10.1016/j.tim.2013.11.001.
4. Frost L, Leplae R, Summers A, Toussaint A. 2005. Mobile genetic
elements: the agents of open source evolution. Nat Rev Microbiol 3:722–
732. http://dx.doi.org/10.1038/nrmicro1235.
5. Lang AS, Zhaxybayeva O, Beatty JT. 2012. Gene transfer agents: phage-
like elements of genetic exchange. Nat Rev Microbiol 10:472– 482. http:
//dx.doi.org/10.1038/nrmicro2802.
6. Sander M, Schmieger H. 2001. Method for host-independent detection
of generalized transducing bacteriophages in natural habitats. Appl Envi-
ron Microbiol 67:1490 –1493. http://dx.doi.org/10.1128/AEM.67.4.1490
-1493.2001.
7. Penadés J, Chen J, Quiles-Puchalt N, Carpena N, Novick R. 2015.
Bacteriophage-mediated spread of bacterial virulence genes. Curr Opin
Microbiol 23:171–178. http://dx.doi.org/10.1016/j.mib.2014.11.019.
8. Muniesa M, Imamovic L, Jofre J. 2011. Bacteriophages and genetic
mobilization in sewage and faecally polluted environments. Microb Bio-
technol 4:725–734. http://dx.doi.org/10.1111/j.1751-7915.2011.00264.x.
9. Parsley LC, Consuegra EJ, Kakirde KS, Land AM, Harper WF, Liles
MR. 2010. Identification of diverse antimicrobial resistance determinants
carried on bacterial, plasmid, or viral metagenomes from an activated
sludge microbial assemblage. Appl Environ Microbiol 76:3753–3757.
http://dx.doi.org/10.1128/AEM.03080-09.
10. Kenzaka T, Tani K, Nasu M. 2010. High-frequency phage-mediated gene
transfer in freshwater environments determined at single-cell level. ISME
J 4:648 – 659. http://dx.doi.org/10.1038/ismej.2009.145.
11. Modi S, Lee H, Spina C, Collins J. 2013. Antibiotic treatment expands
7912 aem.asm.org Applied and Environmental Microbiology
the resistance reservoir and ecological network of the phage metagenome.
Nature 499:219 –222. http://dx.doi.org/10.1038/nature12212.
12. Calero-Cáceres W, Melgarejo A, Colomer-Lluch M, Stoll C, Lucena
F, Jofre J, Muniesa M. 2014. Sludge as a potential important source of
antibiotic resistance genes in both the bacterial and bacteriophage frac-
tions. Environ Sci Technol 48:7602–7611. http://dx.doi.org/10.1021
/es501851s.
13. Colomer-Lluch M, Jofre J, Muniesa M. 2011. Antibiotic resistance genes
in the bacteriophage DNA fraction of environmental samples. PLoS One
6:e17549. http://dx.doi.org/10.1371/journal.pone.0017549.
14. Marti E, Variatza E, Balcazar J. 2014. Bacteriophages as a reservoir of
extended spectrum ␤-lactamase and fluoroquinolone resistance genes in
the environment. Clin Microbiol Infect 20:O456 –O459. http://dx.doi.org
/10.1111/1469-0691.12446.
15. Rahube T, Marti R, Scott A, Tien YC, Murray R, Sabourin L, Zhang Y,
Duenk P, Lapen D, Topp E. 2014. Impact of fertilizing with raw or
anaerobically digested sewage sludge on the abundance of antibiotic-
resistant coliforms, antibiotic resistance genes, and pathogenic bacteria in
soil and on vegetables at harvest. Appl Environ Microbiol 80:6898 – 6907.
http://dx.doi.org/10.1128/AEM.02389-14.
16. Marti R, Tien YC, Murray R, Scott A, Sabourin L, Topp E. 2014. Safely
coupling livestock and crop production systems: how rapidly do antibiotic
resistance genes dissipate in soil following a commercial application of
swine or dairy manure? Appl Environ Microbiol 80:3258 –3265. http://dx
.doi.org/10.1128/AEM.00231-14.
17. Marti R, Scott A, Tien Y-C, Murray R, Sabourin L, Zhang Y, Topp E.
2013. The impact of manure fertilization on the abundance of antibiotic
resistant bacteria and frequency of detection of antibiotic resistance genes
in soil, and on vegetables at harvest. Appl Environ Microbiol 79:5701–
5709. http://dx.doi.org/10.1128/AEM.01682-13.
18. Thurber R, Haynes M, Breitbart M, Wegley L, Rohwer F. 2009. Labo-
ratory procedures to generate viral metagenomes. Nat Protoc 4:470 – 483.
http://dx.doi.org/10.1038/nprot.2009.10.
19. Muyzer G, Dewaal E, Uitierlinden A. 1993. Profiling of complex micro-
bial populations by denaturing gradient gel electrophoresis analysis of
polymerase chain reaction-amplified genes coding for 16S rRNA. Appl
Environ Microbiol 59:695–700.
20. Suzuki M, Taylor L, DeLong E. 2000. Quantitative analysis of small-
subunit rRNA genes in mixed microbial populations via 5=-nuclease as-
says. Appl Environ Microbiol 66:4605– 4614. http://dx.doi.org/10.1128
/AEM.66.11.4605-4614.2000.
21. Walsh F, Ingenfeld A, Zampicolli M, Hilber-Bodmer M, Frey J, Duffy B.
2011. Real-time PCR methods for quantitative monitoring of streptomycin
and tetracycline resistance genes in agricultural ecosystems. J Microbiol
Methods 86:150 –155. http://dx.doi.org/10.1016/j.mimet.2011.04.011.
22. Bert F, Branger C, Lambert-Zechovsky N. 2002. Identification of PSE
and OXA beta-lactamase genes in Pseudomonas aeruginosa using PCR-
restriction fragment length polymorphism. J Antimicrob Chemother 50:
11–18. http://dx.doi.org/10.1093/jac/dkf069.
23. United States Environmental Protection Agency. 2001. Method 1602:
male-specific (F⫹) and somatic coliphage in water by single agar layer
(SAL) procedure. United States Environmental Protection Agency, Wash-
ington, DC.
24. Ashelford K, Day M, Fry J. 2003. Elevated abundance of bacteriophage
infecting bacteria in soil. Appl Environ Microbiol 69:285–289. http://dx
.doi.org/10.1128/AEM.69.1.285-289.2003.
25. Government of Canada. 2010. Canadian Integrated Program for Antimi-
crobial Resistance Surveillance (CIPARS) 2007. Public Health Agency of
Canada, Guelph, ON, Canada.
26. Colomer-Lluch M, Jofre J, Muniesa M. 2014. Quinolone resistance
genes (qnrA and qnrS) in bacteriophage particles from wastewater
samples and the effect of inducing agents on packaged antibiotic resis-
tance genes. J Antimicrob Chemother 69:1265–1274. http://dx.doi.org
/10.1093/jac/dkt528.
27. Chibani-Chennoufi S, Bruttin A, Dillmann M-L, Brüssow H. 2004.
Phage-host interaction: an ecological perspective. J Bacteriol 186:3677–
3686. http://dx.doi.org/10.1128/JB.186.12.3677-3686.2004.
28. Zeph L, Onaga M, Stotzky G. 1988. Transduction of Escherichia coli by
bacteriophage P1 in soil. Appl Environ Microbiol 54:1731–1737.
29. Shousha A, Awaiwanont N, Sofka D, Smulders FJ, Paulsen P, Szos-
tak MP, Humphrey T, Hilbert F. 2015. Bacteriophages isolated from
November 2015 Volume 81 Number 22

chicken meat and the horizontal transfer of antimicrobial resistance
genes. Appl Environ Microbiol 81:4600 – 4606. http://dx.doi.org/10
.1128/AEM.00872-15.
30. Allen H, Looft T, Bayles D, Humphrey S, Levine U, Alt D, Stanton T.
2011. Antibiotics in feed induce prophages in swine fecal microbiomes.
mBio 2:e00260-11. http://dx.doi.org/10.1128/mBio.00260-11.
31. Perry J, Wright G. 2014. Forces shaping the antibiotic resistome. Bioes-
says 36:1179 –1184. http://dx.doi.org/10.1002/bies.201400128.
November 2015 Volume 81 Number 22 Applied and Environmental Microbiology
Antibiotic Resistance in Soil Phage
32. Smillie C, Smith M, Friedman J, Cordero O, David L, Alm E. 2011.
Ecology drives a global network of gene exchange connecting the
human microbiome. Nature 480:241–244. http://dx.doi.org/10.1038
/nature10571.
33. Mulvey M, Bryce E, Boyd D, Ofner-Agostini M, Land A, Simor A,
Paton S. 2005. Molecular characterization of cefoxitin-resistant Esche-
richia coli from Canadian hospitals. Antimicrob Agents Chemother 49:
358 –365. http://dx.doi.org/10.1128/AAC.49.1.358-365.2005.
aem.asm.org 7913
REVIEW doi:10.1038/nature10886

RNA-guided genetic silencing

systems in bacteria and archaea
Blake Wiedenheft1,2†, Samuel H. Sternberg3 & Jennifer A. Doudna1–4

Clustered regularly interspaced short palindromic repeat (CRISPR) are essential components of nucleic-acid-based
adaptive immune systems that are widespread in bacteria and archaea. Similar to RNA interference (RNAi) pathways
in eukaryotes, CRISPR-mediated immune systems rely on small RNAs for sequence-specific detection and silencing
of foreign nucleic acids, including viruses and plasmids. However, the mechanism of RNA-based bacterial immunity is
distinct from RNAi. Understanding how small RNAs are used to find and destroy foreign nucleic acids will provide new
insights into the diverse mechanisms of RNA-controlled genetic silencing systems.

B
acteria and archaea are the most diverse and abundant organisms
on the planet, thriving in habitats that range from hot springs to
humans. However, viruses outnumber their microbial hosts in
every ecological setting, and the selective pressures imposed by these
rapidly evolving parasites has driven the diversification of microbial
defence systems1–3. Historically, our understanding of antiviral immu-
nity in bacteria has focused on restriction-modification systems,
abortive-phage phenotypes, toxin–antitoxins and other innate defence
systems4,5. More recently, bioinformatic, genetic and biochemical stud-
ies have revealed that many prokaryotes use an RNA-based adaptive
immune system to target and destroy genetic parasites (reviewed in refs
6–12). Such adaptive immunity, previously thought to occur only in
eukaryotes, provides an example of RNA-guided destruction of foreign
genetic material by a process that is distinct from RNA interference
(RNAi) (Fig. 1).
In response to viral and plasmid challenges, bacteria and archaea
integrate short fragments of foreign nucleic acid into the host chromo-
some at one end of a repetitive element known as CRISPR (clustered
regularly interspaced short palindromic repeat)13–15. These repetitive
loci serve as molecular ‘vaccination cards’ by maintaining a genetic
record of prior encounters with foreign transgressors. CRISPR loci
are transcribed, and the long primary transcript is processed into a
library of short CRISPR-derived RNAs (crRNAs)16–21 that each con-
tain a sequence complementary to a previously encountered invading
nucleic acid. Each crRNAs is packaged into a large surveillance complex
that patrols the intracellular environment and mediates the detection
and destruction of foreign nucleic acid targets15,22–27.
CRISPRs were originally identified in the Escherichia coli genome
in 1987, when they were described as an unusual sequence element
consisting of a series of 29-nucleotide repeats separated by unique
32-nucleotide ‘spacer’ sequences28. Repetitive sequences with a similar
repeat–spacer–repeat pattern were later identified in phylogenetically
diverse bacterial and archaeal genomes, but the function of these repeats
remained obscure until many spacer sequences were recoginized as
being identical to viral and plasmid sequences29–31. This observation
led to the hypothesis that CRISPRs provide a genetic memory of infec-
tion29, and the detection of short CRISPR-derived RNA transcripts
suggested that there may be functional similarities between CRISPR-
based immunity and RNAi30,32. In this Insight, we review three stages of
CRISPR-based adaptive immunity and compare mechanistic aspects of
these immune systems to other RNA-guided genetic silencing pathways.
1
Howard Hughes Medical Institute, 4000 Jones Bridge Road, Chevy Chase, Maryland 20815-6789, USA. 2Department of Molecular and Cell Biology, University of California, Berkeley, California 94720,
USA. 3Department of Chemistry, University of California, Berkeley, California 94720, USA. 4Physical Biosciences Division, Lawrence Berkeley National Laboratory, Berkeley, California 94720, USA.
†Present address: Department of Immunology and Infectious Diseases, Montana State University, Bozeman, Montana, USA
Architecture and composition of CRISPR loci
The defining feature of CRISPR loci is a series of direct repeats
(approximately 20–50 base pairs) separated by unique spacer
sequences of a similar length 11,33,34 (Fig. 2). The repeat sequences
within a CRISPR locus are conserved, but repeats in different CRISPR
loci can vary in both sequence and length. In addition, the number
of repeat–spacer units in a CRISPR locus varies widely within and
among organisms35.
The sequence diversity of these repetitive loci initially limited their
detection and obscured their relationship, but computational methods
have been developed for detecting repeat patterns rather than related
sequences33,34,36–38. One of the first-generation pattern-recognition algo-
rithms identified the repeat–spacer–repeat architecture in phylogeneti-
cally diverse bacterial and archaeal genomes, but related structures were
not identified in eukaryotic chromosomes39. Comparative analyses of
the sequences adjacent to the CRISPR loci have revealed an (A+T)-rich
‘leader’ sequence that has been shown to serve as a promoter element
for CRISPR transcription39–42. In addition to the leader sequence, Jansen
et al.39 identified a set of four CRISPR-associated (cas) genes known as
cas1–4 that are found exclusively in genomes containing CRISPRs. Based
on sequence similarity to proteins of known function, Cas3 was predicted
to be a helicase and Cas4 a RecB-like exonuclease39.
Subsequent bioinformatic analyses have shown that CRISPR loci are
flanked by a large number of extremely diverse cas genes32,43. The cas1
gene is a common component of all CRISPR systems, and phylogenetic
analyses of Cas1 sequences indicate there are several versions of the
CRISPR system. Providing additional evidence for the classification
of distinct CRISPR types, neighbourhood analysis has identified con-
served arrangements of between four and ten cas genes that are found
in association with CRISPR loci harbouring specific repeat sequences35.
These distinct immune systems have been divided into three major
CRISPR types on the basis of gene conservation and locus organiza-
tion10. More than one CRISPR type is often found in a single organism,
indicating that these systems are probably mutually compatible and could
share functional components10. Despite the variation in number and
diversity of cas genes, the distinguishing feature of all type I systems is that
they encode a cas3 gene. The Cas3 protein contains an N-terminal HD
phosphohydrolase domain and a C-terminal helicase domain32,39,43,44. In
some type I systems, the Cas3 nuclease and helicase domains are encoded
by separate genes (cas3ʹʹ and cas3ʹ, respectively), but in each case they are
thought to participate in degrading foreign nucleic acids22,44–46 (Fig. 2).

1 6 F E B R UA RY 2 0 1 2 | VO L 4 8 2 | N AT U R E | 3 3 1
© 2012 Macmillan Publishers Limited. All rights reserved
 INSIGHT REVIEW
Type II CRISPR systems consist of just four cas genes, one of which
is always cas9 (formerly referred to as csn1). Cas9 is a large protein
that includes both a RuvC-like nuclease domain and an HNH nucle-
ase domain. Studies in Streptococcus pyogenes and Streptococcus ther-
mophilus have indicated that Cas9 may participate in both CRISPR RNA
processing and target destruction14,15,17. Two variations of the type III
system have been identified (known as III-A and III-B). This division
is supported by the functional differences reported in Staphylococcus
epidermidis and Pyrococcus furiosus47,48. The immune system in S. epi-
dermidis (type III-A) targets plasmid DNA in vivo, whereas the purified
components of the type III-B system in P. furiosus have been found to
cleave only single-stranded RNA substrates in vitro. The functional dis-
tinction between these two closely related systems suggests there could
be other mechanistic differences between the distinct CRISPR subtypes.
Integration of new information into CRISPR loci
Acquisition of foreign DNA is the first step of CRISPR-mediated immu-
nity (Fig. 2 and 3). During this stage, a short segment of DNA from an
invading virus or plasmid (known as the protospacer) is integrated pref-
erentially at the leader end of the CRISPR locus14,15. Although metagen-
omic studies performed on environmental samples indicate that
CRISPRs evolve rapidly in dynamic equilibrium with resident phage
populations13,49,50, the type II system in S. thermophilus is currently the
only CRISPR system that has been shown to robustly acquire new phage
or plasmid sequences in a pure culture. Phage-challenge experiments in
S. thermophilus have indicated that a small proportion of the cells in a
population will typically incorporate a single virus-derived sequence at
the leader end of a CRISPR locus14,15,51,52. The CRISPR-repeat sequence
is duplicated for each new spacer seqenced added, thus maintaining the
repeat–spacer–repeat architecture. Although the mechanism of spacer
integration and replication of the repeat sequence is still unknown,
studies in S. thermophilus and E. coli have indicated that several Cas
proteins are involved in the process14,15,22,53. Mutational analysis of the
CRISPR-mediated interference
Foreign DNA
CRISPR locus
Repeat Repeat
CRISPR
transcription
Cas or RNase III
crRNA
5′ 3′
crRNA-guided surveillance complex
Cas protein(s)
Seed
5′ 3′
Target interference
Figure 1 | Parallels and distinctions between CRISPR RNA-guided
silencing systems and RNAi. CRISPR systems and RNAi recognize
long RNA precursors that are processed into small RNAs, which act as
sequence-specific guides for targeting complementary nucleic acids. In
CRISPR systems, foreign DNA is integrated into the CRISPR locus, and long
transcripts from these loci are processed by a CRISPR-associated (Cas) or
RNase III family nuclease16–21,64. The short CRISPR-derived RNAs (crRNAs)
assemble with Cas proteins into large surveillance complexes that target
destruction of invading genetic material15,22,24–27,48. In some eukaryotes, long
double-stranded RNAs are recognized as foreign, and a specialized RNase III
family endoribonuclease (Dicer) cleaves these RNAs into short-interfering
RNAs (siRNAs) that guide the immune system to invading RNA viruses76.
PIWI-interacting RNAs (piRNAs) are transcribed from repetitive clusters in
the genome that often contain many copies of retrotransposons and primarily
3 3 2 | N AT U R E | VO L 4 8 2 | 1 6 F E B R UA RY 2 0 1 2
© 2012 Macmillan Publishers Limited. All rights reserved
cas genes in S. thermophilus demonstrated that csn2 (previously known
as cas7) is required for new spacer sequence acquisition14. This gene is
not conserved in other CRISPR types, which suggests that either the
mechanism of adaptation in S. thermophilus is distinct from the other
types or that there are functional orthologues of Csn2 in other systems.
Furthermore, gene deletion experiments in both S. thermophilus and
E. coli have shown that neither cas1 nor cas2 genes are required for
CRISPR RNA processing or targeted interference22,53,54. These genetic
studies suggest a role for Cas1 and Cas2 in the integration of foreign
DNA into the CRISPR.
The role of Cas1 in CRISPR-mediated immunity is still uncertain;
however, biochemical and structural data indicate a function for Cas1 in
new–spacer–sequence acquisition54–56. Cas1 proteins from Pseudomo-
nas aeruginosa56, E. coli54 and Sulfolobus solfataricus55 have been purified
and studied biochemically. The Cas1 protein from S. solfataricus has
been shown to bind nucleic acids with high affinity (Kd ranging from 20
to 50 nM), but without sequence preference55. The Cas1 protein from
E. coli also binds to DNA with a preference for mismatched or abasic
substrates57. This observation is consistent with a recent study show-
ing a physical and genetic interaction between E. coli Cas1 and several
proteins associated with DNA replication and repair54.
Activity assays with Cas1 from P. aeruginosa and E. coli indicate that
Cas1 is a metal-dependent nuclease. The Cas1 protein from P. aeruginosa
is a DNA-specific nuclease, whereas the Cas1 protein from E. coli had a
nuclease activity on a wider range of nucleic acid substrates54,56. These
in vitro assays suggest that Cas1 proteins interact with nucleic acids in a
non-sequence-specific manner.
Crystal structures for five different Cas1 proteins are currently avail-
able (Protein Data Bank (PDB) identifiers: 3GOD, 3NKD, 3LFX, 3PV9
and 2YZS)54,56. Although the amino acid sequences for these proteins are
extremely diverse (less than 15% sequence identity), their tertiary and
quaternary structures are similar. All Cas1 proteins seem to share a two-
domain architecture consisting of an N-terminal β-strand domain and a
Eukaryotic RNA-interference
Foreign RNA
Nucleus
Source piRNA locus miRNA locus
Repeat
Drosha
? miRNA siRNA
RNA biogenesis piRNA 3′ Dicer
RNA-induced silencing complex
AGO/PIWI
Seed
RNA-guided
interference 5′
Target interference
act by restricting transposon mobility76–78. The biogenesis of piRNAs is
not yet fully understood. MicroRNAs (miRNAs) are also encoded on the
chromosome, and primary miRNA transcripts form stable hairpin structures
that are sequentially processed (shown by red triangles) by two RNase III
family endoribonucleases (Drosha and Dicer)79. miRNAs do not participate
in genome defence but are major regulators of endogenous gene expression80.
Like crRNAs, eukaryotic piRNAs, siRNAs and miRNAs associate with
proteins that facilitate complementary interactions with invading nucleic
acid targets27,60,69,79. In eukaryotes, the Argonaute proteins pre-order the 5ʹ
region of the guide RNA into a helical configuration, reducing the entropy
penalty of interactions with target RNAs69. This high-affinity binding site,
called the ‘seed’ sequence, is essential for target sequence interactions. Recent
studies indicate that the CRISPR system may use a similar seed-binding
mechanism for enhancing target sequence interactions26,27,53,60.
 REVIEW INSIGHT

C-terminal α-helical domain (Fig. 3). The C-terminal domain contains a
conserved divalent metal-ion binding site, and alanine substitutions of the
metal-coordinating residues inhibit Cas1-catalysed DNA degradation54,56.
The metal ion is surrounded by a cluster of basic residues that form a
strip of positive charge across the surface of the C-terminal domain. This
positively charged surface may serve as an electrostatic snare to position
nucleic-acid substrates near the catalytic metal ions56 (Fig. 3). The Cas1
protein forms a stable homodimer that is formed through interactions
between the two β-strand domains, which are related by a pseudo-two-
fold axis of symmetry54,56. This organization creates a saddle-like structure
that can be modelled onto double-stranded DNA without steric clashing.
β-hairpins, one from each of the two symmetrically related molecules,
hang on opposite faces of the double-stranded DNA (like stirrups on a
saddle). Although this feature of the Cas1 structure did not initially stand
out as a potential DNA-binding site, comparative analysis of the avail-
able Cas1 structures reveals a conserved set of positively charged residues
along each of the β-hairpins that could contact the phosphate backbone.
The two β-hairpins, which are symmetrically related, might participate
in sequence-specific interactions with the CRISPR repeat, whereas the
large positively charged surface on the C-terminal α-helical domain could
account for the high-affinity, non-sequence-specific interactions that have
been observed in vitro.
In spite of these structural studies and biochemical results, it is still only
possible to speculate on the role of Cas1 in the integration of new spacer
sequences, and many steps associated with the integration process still
need to be explained. For example, new spacer sequences are inserted

Invading virus

cas gene casette CRISPR

r Spacer Spacer
ce cas cas
spa Leader/promoter Repeat Repeat Repeat
o to
M Pr
PA
**

New spacer
Stage 1: Acquisition sequence
CRISPR trancription
Spacer Spacer
Leader/promoter Repeat Repeat Repeat Repeat
Repeat duplication

Stage 2: CRISPR RNA biogenesis

Type I Type II Type III

CRISPR transcription CRISPR transcription CRISPR transcription

RNase III
Leader CRISPR-
Leader specific
Leader endoribonuclease
tracrRNA tracrRNA tracrRNA ?
CRISPR- 5′ 3′
specific endoribonuclease

5′

3′3′ 5′ 3′ 5′ crRNA
5′ 3′
∼30-nt spacer trimming
No crRNA 3′ crRNA
trimming trimming
3′-Seed?
Cas3
5′ 5′-Seed?
Seed 5′ 3′
3′ ** 3′ PAM
PAM **
3′ 5′ 3′
3′
Target DNA
Ca
s3

Target either RNA or DNA

Target DNA

Stage 3: Interference

Figure 2 | Diversity of CRISPR-mediated adaptive immune systems
in bacteria and archaea. A diverse set of CRISPR-associated (cas)
genes (grey arrows) encode proteins required for new spacer sequence
acquisition (Stage 1), CRISPR RNA biogenesis (Stage 2) and target
interference (Stage 3). Each CRISPR locus consists of a series of direct
repeats separated by unique spacer sequences acquired from invading
genetic elements (protospacers). Protospacers are flanked by a short
motif called the protospacer adjacent motif (PAM, **) that is located on
the 5ʹ (type I) or 3ʹ (type II) side in foreign DNA10,51,52,59,67. Long CRISPR
transcripts are processed into short crRNAs by distinct mechanisms. In
type I and III systems, a CRISPR-specific endoribonuclease (yellow ovals
and green circles, respectively) cleaves 8 nucleotides upstream of each
spacer sequence16,18–21,64. In type III systems, the repeat sequence on the 3ʹ
end of the crRNA is trimmed by an unknown mechanism (green pacman,
right). In type II systems, a trans-acting antisense RNA (tracrRNA) with
complementarity to the CRISPR RNA repeat sequence forms an RNA
duplex that is recognized and cleaved by cellular RNase III (brown ovals)17.
This cleavage intermediate is further processed at the 5ʹ end resulting in
a mature, approximately 40-nucleotide crRNA with an approximately
20-nucleotide 3ʹ-handle. In each system, the mature crRNA associates with
one or more Cas proteins to form a surveillance complex (green rectangles).
Type I systems encode a Cas3 nuclease (blue pacman), which may be
recruited to the surveillance complex following target binding24,27,44. A short
high-affinity binding site called a seed-sequence has been identified in some
type I systems27,60, and genetic experiments suggest that type II systems have
a seed sequence located at the 3ʹ end of the crRNA spacer sequence53.

1 6 F E B R UA RY 2 0 1 2 | VO L 4 8 2 | N AT U R E | 3 3 3
© 2012 Macmillan Publishers Limited. All rights reserved
 INSIGHT REVIEW
preferentially at the leader end of the CRISPR, but the mechanism of
leader end recognition is unknown. One simple model suggests that the
leader sequence contains a recognition element that recruits the integra-
tion machinery. It is equally possible that integration relies on single-
stranded regions of the CRISPR DNA that are made available during
transcription. Transcription-associated recombination is involved in
genome stability58, and a mechanism that couples integration together
with transcription would link the process of adaptation to CRISPR RNA
expression, ensuring that spacers from the most recent virus or plasmid
are transcribed first.
The integration machinery must be able to distinguish foreign DNA
from that of the host genome. The molecular cues that are involved in the
distinction of ‘self ’ from ‘non-self ’ are still unknown, but sequencing of
CRISPR loci following phage challenge suggests that spacer sequences
are not selected at random15,29,51,52,59,60. Mapping spacer sequences onto
viral genomes reveals a short sequence motif proximal to the protospacer,
which is referred to as the protospacer adjacent motif (PAM). PAM
sequences are only a few nucleotides long, and the precise sequence var-
ies depending on the CRISPR system type59. This variation suggests that
one or more of the Cas proteins associated with each immune system is
involved in PAM recognition, but the mechanism governing this specific-
ity is unknown.
CRISPR RNA biogenesis
Spacer acquisition is the first step of immunization, but successful protec-
tion from bacteriophage or plasmid challenge requires the CRISPR to be
transcribed and processed into short CRISPR-derived RNAs (crRNAs).
crRNAs were first detected by small RNA profiling in Archaeoglobus
fulgidus61 and S. solfataricus62. Northern-blot analysis using probes
against the repeat sequence of the CRISPR revealed a ‘ladder-like’ pat-
tern of RNA consistent with a long precursor CRISPR RNA transcript
(pre-crRNA) that was processed at approximately 60-nucleotide inter-
vals. In fact, the 3ʹ ends of cloned crRNAs were mapped to the middle
of the CRISPR repeat61, which suggested that the repeat sequence was
recognized and cleaved.
The need for crRNAs in CRISPR-mediated defence was demon-
strated initially by investigation of a CRISPR-specific endoribonucle-
ase in E. coli called Cas6e (formerly known as Cse3 or CasE)22. Cas6e
specifically binds and cleaves within each repeat sequence of the long
pre-crRNA, resulting in a library of crRNAs that each contain a unique
spacer sequence flanked by fragments of the adjacent repeats. Mutation
of a conserved histidine blocks crRNA biogenesis and leaves the cell
susceptible to phage infection22.
The Cas6e protein consists of a double ferredoxin-like fold that selec-
tively associates with specific RNA repeats and does not associate with
DNA or CRISPR RNAs containing a non-cognate repeat sequence
18,20,22,63
(Fig. 4). Crystal structures of Cas6e bound to a CRISPR RNA
repeat reveal a combination of sequence- and structure-specific interac-
tions that explain the molecular mechanism of substrate recognition18,20.
The repeat sequence of the E. coli CRISPR is partially palindromic, and
the RNA forms a stable (approximately 20-nucleotide) stem loop22,35. A
positively charged β-hairpin in Cas6e interacts with the major groove of
the RNA duplex, which positions the 3ʹ strand of the crRNA stem along a
conserved, positively charged cleft on one face of the protein18,20 (Fig. 4).
RNA binding induces a conformational change that disrupts the bot-
tom base pair of the stem and positions the scissile phosphate within
the enzyme active site for site-specific cleavage20. CRISPR RNA cleav-
age occurs 8 nucleotides upstream of the spacer sequence, which results
in 61-nucleotide mature crRNAs consisting of a 32-nucleotide spacer
flanked by 8 nucleotides of the repeat sequence on the 5ʹ end (known
as the 5ʹ-handle) and 21 nucleotides of the remaining repeat sequence
on the 3ʹ end (Fig. 4). Cas6e remains tightly bound to the 3ʹ stem-loop20
and may serve as a nucleation point for assembly of a large effector com-
plex, Cascade (CRISPR-associated complex for antiviral defence), that is
required for phage silencing in the next stage of the immune system22,24,26
(discussed later).
3 3 4 | N AT U R E | VO L 4 8 2 | 1 6 F E B R UA RY 2 0 1 2
© 2012 Macmillan Publishers Limited. All rights reserved
Crystal structures of CRISPR-specific endoribonucleases from two
other immune systems offer additional insights into the co-evolution-
ary relationship between these specialized enzymes and their cognate
RNAs16,19,21 (Fig. 4). In P. aeruginosa, Cas6f (previously known as Csy4)
specifically binds and cleaves the CRISPR-RNA-repeat 8 nucleotides
upstream of the spacer sequence, which leaves a similar 8-nucleotide
5ʹ-handle on mature crRNAs19. The co-crystal structure of Cas6f bound
to its cognate RNA reveals interesting parallels between the method of
RNA binding used by Cas6f and Cas6e18–20. Like Cas6e, the P. aeruginosa
Cas6f protein recognizes the sequence and shape of a stable stem-loop
in the crRNA repeat sequence by interacting extensively with the major
groove of the double-stranded RNA. However, the structural elements
responsible for this interaction are distinct between the two proteins18–20
(Fig. 4). The Cas6f protein has a two-domain architecture, which con-
sists of an N-terminal ferredoxin-like fold similar to that in Cas6e, but
its C-terminal domain is structurally distinct. An arginine-rich helix
in the C-terminal domain of Cas6f inserts into the major groove of the
crRNA duplex, and the bottom of the crRNA is positioned for sequence-
specific hydrogen-bonding contacts in the RNA major groove. These
contacts position the scissile phosphate of the crRNA in the enzyme
active site so that cleavage occurs 8 nucleotides upstream of the spacer
sequence18–20 (Fig. 4).
Although Cas6f and Cas6e recognize the sequence and shape of the
crRNA hairpin in their respective systems, CRISPR RNA repeats in
other CRISPR systems are thought to be unstructured35. For example,
the Cas6 protein from P. furiosus associates with CRISPR transcripts that
are expected to contain unstructured repeats64. The specific recognition
of an unstructured RNA repeat requires a distinct mechanistic solution
for RNA substrate discrimination. Remarkably, crystallographic studies
of the Cas6 protein from P. furiosus have revealed the same duplicated
ferredoxin-like fold observed in the Cas6e protein, but with a different
mode of RNA recognition involving the opposite face of the protein
(Fig. 4). In Cas6, the two ferredoxin-like folds clamp the 5ʹ end of the
single-stranded RNA repeat sequence in place21. Although the RNA in
this structure is disordered in the enzyme active site, biochemical studies
have shown that cleavage occurs 8 nucleotides upstream of the spacer
sequence16,64. While the nucleotide sequences at the cleavage site vary for
each of the different Cas6 proteins, all Cas6 family endoribonucleases
cleave their cognate RNA 8 nucleotides upstream of the spacer sequence
using a metal-ion-independent mechanism.
Despite advances in our understanding of crRNA biogenesis, the
diversity of cas genes has obscured identification of the protein fac-
tors responsible for CRISPR RNA processing in some systems. Type II
immune systems consist of four cas genes, none of which have a detect-
able sequence similarity to known CRISPR-specific endoribonucleases.
Recently, a different CRISPR RNA processing mechanism has been
reported that involves RNase-III-mediated cleavage of double-stranded
regions of the CRISPR RNA repeats17. The first indication of this mecha-
nism came from deep sequencing of RNA from S. pyogenes. An abundant
transcript containing a 25-nucleotide sequence that was complemen-
tary to the CRISPR repeat was identified. This RNA, termed tracrRNA
(trans-activating CRISPR RNA), is coded on the opposite strand and just
upstream of the CRISPR locus. Genetic and biochemical experiments
demonstrated that tracrRNA and pre-crRNA are co-processed by RNase
III, which produces cleavage products with a 2 nucleotide 3ʹ overhang17.
In vivo processing of CRISPR RNAs required Cas9 (previously known as
Csn1), although a precise role for this enzyme in RNA processing has not
yet been defined. The essential role of cellular proteins that are not solely
involved in CRISPR-mediated defence, such as RNase III, indicates that
different host factors may be involved as ancillary components of these
immune systems.
crRNA-guided interference
The third stage of CRISPR-mediated immunity is target interference
(Fig. 2). Here crRNAs associate with Cas proteins to form large CRISPR-
associated ribonucleoprotein complexes that can recognize invading

a b
Leader recognition
β-hairpin
~20 Å
DNA binding?
Recombination/repair
Figure 3 | Steps leading to new spacer integration. a, The Cas1 protein
forms a stable homodimer where the two molecules (green and grey) are
related by a pseudo-two-fold axis of symmetry (PBD ID: 3GOD)54,56. This
organization creates a saddle-like structure in the N-terminal domain,
in which β-hairpins (blue) from each symmetrically related molecule
hang (like stirrups) that are separated by approximately 20 Å, and may
interact with the phosphodiester backbone of double-stranded DNA. An
electrostatic surface representation (bottom) reveals a cluster of basic
residues (blue) that form a positively charged strip across the metal-
binding surface of the C-terminal domain. This strip may serve as an
electrostatic trap that positions DNA substrates proximally to catalytic
nucleic acids. Foreign nucleic acids are identified by base-pairing interac-
tions between the crRNA spacer sequence and a complementary sequence
from the intruder. Phage- and plasmid-challenge experiments performed
in several model systems have demonstrated that crRNAs complementary
to either the coding or the non-coding strand of the invading DNA can
provide immunity14,22,47,60,65,66. This is indicative of an RNA-guided DNA-
targeting system, and indeed a pathway for DNA silencing has recently
been demonstrated in S. thermophilus15. DNA sequencing and Southern
blots indicated that both strands of the target DNA are cleaved within
the region that is complementary to the crRNA spacer sequence15. This
mechanism efficiently eliminates foreign DNA sequences, which have
been specified by the spacer region of the crRNA, but avoids targeting
the complementary DNA sequences in the CRISPR region of the host
chromosome. The mechanism for distinguishing self from non-self is
built into the crRNA. The spacer sequence of each crRNA is flanked by
a portion of the adjacent CRISPR repeat sequence, and any complemen-
tarity beyond the spacer into the adjacent repeat region signals self and
prevents the destruction of the host chromosome67.
However, not all CRISPR systems target DNA. In vitro experiments
using enzymes from the type III-B CRISPR system of P. furiosus have
shown that this system cleaves target RNA rather than DNA48. All DNA
targeting systems encode a complementary DNA sequence for each
crRNA in the CRISPR locus and therefore require a mechanism for distin-
guishing self (CRISPR locus) from non-self (invading DNA). In contrast,
systems that target RNA may not be required to make this distinction
because most CRISPR loci are transcribed only in one direction and thus
do not generate complementary RNA targets. CRISPR systems that tar-
get RNA may be uniquely capable of defending against viruses that have
RNA-based genomes. However, adaptation of the CRISPR in response
to a challenge by an RNA-based virus will probably require the invading
RNA to be reverse-transcribed into DNA before it can be integrated into
the CRISPR locus.
© 2012 Macmillan Publishers Limited. All rights reserved
REVIEW INSIGHT
Leader
CRISPR DNA Repeat Repeat
Protospacer selection
PAM
Viral DNA Protospacer
0 10 20 30
Base pairs
New ‘spacer’ acquisition
Leader
CRISPR DNA Repeat Repeat Repeat
metal ions (green sphere). b, CRISPR adaptation occurs by integrating
fragments of foreign nucleic acid preferentially at the leader end of the
CRISPR, forming new repeat-spacer units in the process. Protospacers
are chosen non-randomly and may be selected from regions flanking the
protospacer adjacent motif (PAM). Coordinated cleavage of the foreign
DNA and integration of the protospacer into the leader-end of the CRISPR
occurs through a mechanism that duplicates the repeat sequence and
thus preserves the repeat-spacer-repeat architecture of the CRISPR locus.
Although the protein components required for this process have not been
conclusively identified, Cas1 and other general recombination or repair
factors have been implicated (blue ovals)32,54,56.
Cas proteins directly participate in target binding. Recent bio-
chemical studies have shown that CRISPR-associated complexes
facilitate target recognition by enhancing sequence-specific
hybridization between the CRISPR RNA and complementary target
sequences27. A short high-affinity binding site located at one end of
the crRNA spacer sequence governs the efficiency of target binding,
and viruses that acquired a single mismatch in this region were able
to escape detection by the immune system60. This high-affinity bind-
ing site is functionally analogous to the ‘seed’ sequence (Fig. 1) that
has been identified in eukaryotic microRNAs (miRNAs)68. Struc-
tural and biochemical studies have shown that Argonaute proteins
facilitate target recognition by pre-ordering the nucleotides at the
5ʹ end of the miRNA in a helical configuration69. This pre-ordering
reduces the entropic penalty that is associated with helix forma-
tion and provides a thermodynamic advantage for target binding
within this region. A similar mechanism may occur during crRNA
target binding, providing an interesting example of convergent evo-
lution between CRISPR-based immunity in prokaryotes and RNAi
in eukaryotes (Fig. 1).
Structural and biochemical studies have been performed on
CRISPR-associated complexes isolated from three different type I
CRISPR systems24–27,48. These complexes������������������������
seem to share some gen-
eral morphological features, but the precise special arrangement of
the Cas proteins and their interactions with the crRNA have been
unclear. Sub-nanometre-resolution structures of the CRISPR-asso-
ciated complex from E. coli (Cascade) have recently been determined
using cryo-electron microscopy26. This complex is comprised of an
unequal stoichiometry of 5 functionally essential Cas proteins and
a 61-nucleotide crRNA22,24,26. The structure reveals a sea-horse-
shaped architecture in which the crRNA is displayed along a helical
arrangement of protein subunits that protect the crRNA from deg-
radation26. The 5ʹ and 3ʹ ends of the crRNA form unique structures
1 6 F E B R UA RY 2 0 1 2 | VO L 4 8 2 | N AT U R E | 3 3 5
 INSIGHT REVIEW
Type I
Cleavage Cleavage
Cleavage
Cas6e Cas6f tracrRNA
pre-crRNA:
Cas6e/Cas6f
5′ trimming
crRNAs:
8-nt 5′-handle
Figure 4 | Diverse mechanisms of CRISPR RNA biogenesis. CRISPR
RNA repeats are specifically recognized and cleaved by diverse
mechanisms. In type I CRISPR systems, Cas6e (PDB ID: 2Y8W) and Cas6f
(PDB ID: 2XLK) recognize the major groove of the crRNA stem-loop
primarily through electrostatic interactions using a β-hairpin and α-helix,
respectively18,19,20. Cleavage occurs at the double-stranded–single-stranded
junction (black arrows), leaving an 8-nt 5ʹ-handle on mature crRNAs. In
type II CRISPR systems, tracrRNA hybridizes to the pre-crRNA repeat
to form duplex RNAs that are substrates for endonucleolytic cleavage by
host RNase III (PDB ID: 2EZ6), an activity that may also require Cas9
(ref. 17). Subsequent trimming (red arrows) by an unidentified nuclease
that are anchored at opposite ends of the Cascade complex, dis-
playing the 32-nucleotide spacer sequence for base-pairing with
complementary targets.
The structure of Cascade bound to a 32-nucleotide target sequence26
reveals a concerted conformational change that could be a signal for
recruiting Cas3. Cas3 — the trans-acting nuclease of type I CRISPR sys-
tems — may function as a target ‘slicer’ in a similar way to Argonaute in
RNAi pathways22,44,46,70. Although Cas3 was implicated previously in the
process of self versus non-self discrimination, recent studies have dem-
onstrated that Cascade recognizes the PAM directly and that mutations
in the PAM decrease Cascade’s affinity for the target60. The importance
of the PAM is highlighted by the recovery of phage and plasmid escape
mutants, which frequently contain a single mutation in the PAM15,51–53,60.
The structure of Cascade indicates that the PAM is positioned near the
‘tail’ of the sea-horse-shaped complex. High-resolution structures and
mutational analysis of the nucleic acid and protein components in this
and related systems are needed to determine the mechanisms of target
authentication and degradation.
Applications of CRISPR structure and function
The sequence diversity of CRISPR loci, even within closely related
strains, has been used for high-resolution genotyping and forensic medi-
cine. This technique, known as spoligotyping (spacer oligotyping), has
been applied successfully to the analysis of human pathogens, including
Mycobacterium tuberculosis71, Corynebacterium diphtheriae72 and Salmo-
nella enterica73. Spoligotyping was developed long before the function
of CRISPRs was understood, but now that studies have begun to reveal
the biological function and mechanism of CRISPR-mediated genetic
silencing, new opportunities for creative applications have emerged.
3 3 6 | N AT U R E | VO L 4 8 2 | 1 6 F E B R UA RY 2 0 1 2
© 2012 Macmillan Publishers Limited. All rights reserved
Type II Type III
Cleavage
Cleavage
RNase III Cas6
Cas9 RNase III Cas6
3′ trimming ?
?
20-nt 3′-handle 8-nt 5′-handle
removes leftover repeat sequences from the 5ʹ end. Cas6 (PDB ID: 3PKM)
in type III-B CRISPR systems specifically recognizes single-stranded
RNA, upstream of the scissile phosphate, on a face of the protein opposite
that of the previously identified active site residues16,21,64. The remainder of
the repeat substrate probably wraps around the protein (red dashed line)
to allow cleavage 8 nucleotides upstream of the repeat-spacer junction.
Subsequent 3ʹ trimming (red arrows) generates mature crRNAs of two
discrete lengths. The N-terminal domain of all Cas 6 family proteins
adopts a ferredoxin-like fold (light blue).The C-terminal domain of Cas6
and Cas6e also adopts a ferredoxin-like fold but the C-terminal domain of
Cas6f is structurally distinct (dark blue).
Laboratory strains of bacteria are grown in high-density bioreactors for
many different applications in the food industry, and they are becoming
increasingly important in the production of biofuels. CRISPR systems
offer a natural mechanism for adapting economically important bacteria
for resistance against multiple phages.
The biochemical activities of various Cas proteins may have use-
ful applications in molecular biology in much the same way that DNA
restriction enzymes have revolutionized cloning and DNA manipulation.
A wide range of CRISPR-specific endoribonucleases that recognize small
RNA motifs with high affinity expand the number of tools available for
manipulating nucleic acids. In addition, a crRNA-guided ribonucleopro-
tein complex in P. furiosus was shown to cleave target RNAs48. Site-specific
cleavage of target RNA molecules could have a range of uses, from gen-
erating homogeneous termini after in vitro transcription to targeting a
specific intracellular messenger RNA for inactivation in a similar way to
RNAi. CRISPRs also provide a new mechanism for limiting the spread of
antibiotic resistance or the transfer of virulence factors by blocking hori-
zontal gene transfer15,47. In addition, CRISPRs participate in a regulatory
mechanism that alters biofilm formation in P. aeruginosa74,75. Although
the clinical relevance of CRISPRs remains to be demonstrated, the oppor-
tunities for creative implementation of this new gene-regulation system
are perceivably vast.
Future directions of CRISPR biology
The discovery of some of the fundamental mechanisms of CRISPR-
based adaptive immunity has raised new questions and highlighted the
areas with the greatest potential for future research. Although CRISPR
RNA processing and targeting steps are now understood in some detail,
how and when target sequences are identified during a phage infection

or plasmid transformation are still unclear. Furthermore, why DNA or
RNA target sequences are chosen, and their fate once they are bound
to a crRNA-targeting complex is not well understood. In addition, the
mechanisms by which foreign sequences are selected and integrated
into CRISPR loci are almost entirely unknown. Some CRISPR loci seem
to be considerably more active than others, at least under laboratory
conditions, so selection of the model organisms will be important. The
diversity and prevalence of CRISPR systems throughout microbial com-
munities ensures that new findings and applications in this field will be
forthcoming in the years ahead. ■
1. Hoskisson, P. A. & Smith, M. C. Hypervariation and phase variation in the
bacteriophage ‘resistome’. Curr. Opin. Microbiol. 10, 396–400 (2007).
2. Rodriguez-Valera, F. et al. Explaining microbial population genomics through
phage predation. Nature Rev. Microbiol. 7, 828–836 (2009).
3. Weinbauer, M. G. Ecology of prokaryotic viruses. FEMS Microbiol. Rev. 28,
127–181 (2004).
4. Labrie, S. J., Samson, J. E. & Moineau, S. Bacteriophage resistance mechanisms.
Nature Rev. Microbiol. 8, 317–327 (2010).
5. Stern, A. & Sorek, R. The phage-host arms race: shaping the evolution of
microbes. Bioessays 33, 43–51 (2010).
6. Al-Attar, S., Westra, E. R., van der Oost, J. & Brouns, S. J. Clustered regularly
interspaced short palindromic repeats (CRISPRs): the hallmark of an ingenious
antiviral defense mechanism in prokaryotes. Biol. Chem. 392, 277–289 (2011).
7. Deveau, H., Garneau, J. E. & Moineau, S. CRISPR/Cas system and its role in
phage-bacteria interactions. Annu. Rev. Microbiol. 64, 475–493 (2010).
8. Horvath, P. & Barrangou, R. CRISPR/Cas, the immune system of bacteria and
archaea. Science 327, 167–170 (2010).
9. Karginov, F. V. & Hannon, G. J. The CRISPR system: small RNA-guided defense
in bacteria and archaea. Mol. Cell 37, 7–19 (2010).
10. Makarova, K. S. et al. Evolution and classification of the CRISPR-Cas systems.
Nature Rev. Microbiol. 9, 467–477 (2011).
11. Marraffini, L. A. & Sontheimer, E. J. CRISPR interference: RNA-directed adaptive
immunity in bacteria and archaea. Nature Rev. Genet. 11, 181–190 (2010).
12. Sorek, R., Kunin, V. & Hugenholtz, P. CRISPR-a widespread system that provides
acquired resistance against phages in bacteria and archaea. Nature Rev.
Microbiol. 6, 181–186 (2008).
13. Andersson, A. F. & Banfield, J. F. Virus population dynamics and acquired virus
resistance in natural microbial communities. Science 320, 1047–1050 (2008).
14. Barrangou, R. et al. CRISPR provides acquired resistance against viruses in
prokaryotes. Science 315, 1709–1712 (2007).
This study provided the first direct evidence for CRISPR immune system
function, demonstrating that new spacer acquisition provides resistance
against future phage infection in a cas gene-dependent manner.
15. Garneau, J. E. et al. The CRISPR/Cas bacterial immune system cleaves
bacteriophage and plasmid DNA. Nature 468, 67–71 (2010).
This study showed that CRISPRs can acquire new spacers upon plasmid
challenge and provided the first example of crRNA-guided cleavage of
double-stranded DNA in vivo.
16. Carte, J., Wang, R., Li, H., Terns, R. M. & Terns, M. P. Cas6 is an endoribonuclease
that generates guide RNAs for invader defense in prokaryotes. Genes Dev. 22,
3489–3496 (2008).
17. Deltcheva, E. et al. CRISPR RNA maturation by trans-encoded small RNA and
host factor RNase III. Nature 471, 602–607 (2011).
This study identified a new CRISPR RNA processing pathway that involves
a trans-acting RNA complementary to the CRISPR repeat sequence and
demonstrated that processing of this duplex is mediated by cellular RNase III.
18. Gesner, E. M., Schellenberg, M. J., Garside, E. L., George, M. M. & MacMillan, A. M.
Recognition and maturation of effector RNAs in a CRISPR interference pathway
Nature Struct. Mol. Biol. 18, 688–692 (2011).
19. Haurwitz, R. E., Jinek, M., Wiedenheft, B., Zhou, K. & Doudna, J. A. Sequence-
and structure-specific RNA processing by a CRISPR endonuclease. Science
329, 1355–1358 (2010).
20. Sashital, D. G., Jinek, M. & Doudna, J. A. An RNA induced conformational
change required for CRISPR RNA cleavage by the endonuclease Cse3. Nature
Struct. Mol. Biol. 18, 680–687 (2011).
21. Wang, R., Preamplume, G., Terns, M. P., Terns, R. M. & Li, H. Interaction of the
Cas6 riboendonuclease with CRISPR RNAs: recognition and cleavage. Structure
19, 257–264 (2011).
22. Brouns, S. J. et al. Small CRISPR RNAs guide antiviral defense in prokaryotes.
Science 321, 960–964 (2008).
This paper revealed that long CRISPR RNA transcripts are processed
into small RNAs (crRNAs) by a dedicated endoribonuclease and that the
processed RNAs are bound by a Cas protein complex (Cascade), which
together with Cas3 are required for phage protection.
23. Hale, C., Kleppe, K., Terns, R. M. & Terns, M. P. Prokaryotic silencing (psi)RNAs in
Pyrococcus furiosus. RNA 14, 2572–2579 (2008).
24. Jore, M .M. et al. Structural basis for CRISPR RNA-guided DNA recognition by
Cascade. Nature Struct. Mol. Biol. 18, 529–536 (2011).
25. Lintner, N. G. et al. Structural and functional characterization of an archaeal
clustered regularly interspaced short palindromic repeat (crispr)-associated
complex for antiviral defense (CASCADE). J. Biol. Chem. 286, 21643–21656
(2011).
© 2012 Macmillan Publishers Limited. All rights reserved
REVIEW INSIGHT
26. Wiedenheft, B. et al. Structures of the RNA-guided surveillance complex from a
bacterial immune system. Nature 477, 486–489 (2011).
This paper presented the first sub-nanometer resolution structures
of Cascade, showing how the crRNA is accommodated within a large
ribonucleoprotein complex that is involved in foreign nucleic acid surveillance.
27. Wiedenheft, B. et al. RNA-guided complex from a bacterial immune system
enhances target recognition through seed sequence interactions. Proc. Natl
Acad. Sci. USA 108, 10092–10097 (2011).
28. Ishino, Y., Shinagawa, H., Makino, K., Amemura, M. & Nakata, A. Nucleotide
sequence of the iap gene, responsible for alkaline phosphatase isozyme
conversion in Escherichia coli, and identification of the gene product. J. Bacteriol.
169, 5429–5433 (1987).
29. Bolotin, A., Ouinquis, B., Sorokin, A. & Ehrlich, S. D. Clustered regularly
interspaced short palindrome repeats (CRISPRs) have spacers of
extrachromosomal origin. Microbiology 151, 2551–2561 (2005).
30. Mojica, F. J., Diez-Villasenor, C., Garcia-Martinez, J. & Soria, E. Intervening
sequences of regularly spaced prokaryotic repeats derive from foreign genetic
elements. J. Mol. Evol. 60, 174–182 (2005).
31. Pourcel, C., Salvignol, G. & Vergnaud, G. CRISPR elements in Yersinia pestis
acquire new repeats by preferential uptake of bacteriophage DNA, and provide
additional tools for evolutionary studies. Microbiology 151, 653–663 (2005).
32. Makarova, K. S., Grishin, N. V., Shabalina, S. A., Wolf, Y. I. & Koonin, E. V.
A putative RNA-interference-based immune system in prokaryotes:
computational analysis of the predicted enzymatic machinery, functional
analogies with eukaryotic RNAi, and hypothetical mechanisms of action. Biol.
Direct 1, 7 (2006).
33. Grissa, I., Vergnaud, G. & Pourcel, C. CRISPRFinder: a web tool to identify
clustered regularly interspaced short palindromic repeats. Nucleic Acids Res.
35, W52–W57 (2007).
34. Rousseau, C., Gonnet, M., Le Romancer, M. & Nicolas, J. CRISPI: a CRISPR
interactive database. Bioinformatics 25, 3317–3318 (2009).
35. Kunin, V., Sorek, R. & Hugenholtz, P. Evolutionary conservation of sequence and
secondary structures in CRISPR repeats. Genome Biol. 8, R61 (2007).
36. Bland, C. et al. CRISPR recognition tool (CRT): a tool for automatic detection of
clustered regularly interspaced palindromic repeats. BMC Bioinformatics 8, 209
(2007).
37. Dsouza, M., Larsen, N. & Overbeek, R. Searching for patterns in genomic data.
Trends Genet. 13, 497–498 (1997).
38. Edgar, R. C. PILER-CR: fast and accurate identification of CRISPR repeats. BMC
Bioinformatics 8, 18 (2007).
39. Jansen, R., Embden, J. D., Gaastra, W. & Schouls, L. M. Identification of genes
that are associated with DNA repeats in prokaryotes. Mol. Microbiol. 43,
1565–1575 (2002).
40. Pougach, K. et al. Transcription, processing and function of CRISPR cassettes in
Escherichia coli. Mol. Microbiol. 77, 1367–1379 (2010).
41. Pul, U. et al. Identification and characterization of E. coli CRISPR-cas promoters
and their silencing by H-NS. Mol. Microbiol. 75, 1495–1512 (2010).
42. Westra, E. R. et al. H-NS-mediated repression of CRISPR-based immunity in
Escherichia coli K12 can be relieved by the transcription activator LeuO. Mol.
Microbiol. 77, 1380–1393 (2010).
43. Haft, D. H., Selengut, J., Mongodin, E. F. & Nelson, K. E. A guild of 45 CRISPR-
associated (Cas) protein families and multiple CRISPR/Cas subtypes exist in
prokaryotic genomes. PLoS Comput. Biol. 1, e60 (2005).
44. Sinkunas, T. et al. Cas3 is a single-stranded DNA nuclease and ATP-dependent
helicase in the CRISPR/Cas immune system. EMBO J. 30, 1335–1342 (2011).
45. Han, D. & Krauss, G. Characterization of the endonuclease SSO2001 from
Sulfolobus solfataricus P2. FEBS Lett. 583, 771–776 (2009).
46. Mulepati, S. & Bailey, S. Structural and biochemical analysis of the nuclease
domain of the clustered regularly interspaced short palindromic repeat
(CRISPR) associated protein 3(CAS3). J. Biol. Chem. 286, 31896–31903
(2011).
47. Marraffini, L. A. & Sontheimer, E. J. CRISPR interference limits horizontal gene
transfer in staphylococci by targeting DNA. Science 322, 1843–1845 (2008).
48. Hale, C. R. et al. RNA-guided RNA cleavage by a CRISPR RNA-Cas protein
complex. Cell 139, 945–956 (2009).
Most CRISPR systems appear to target DNA, but this article reported the
isolation of a multisubunit ribonucleoprotein complex (Cmr-complex) from P.
furriosus that cleaves target RNAs (not DNA) at a fixed distance from the 3΄
end of the crRNA-guide.
49. Tyson, G. W. & Banfield, J. F. Rapidly evolving CRISPRs implicated in acquired
resistance of microorganisms to viruses. Environ. Microbiol. 10, 200–207
(2008).
50. Snyder, J. C., Bateson, M. M., Lavin, M. & Young, M. J. Use of cellular CRISPR
(clusters of regularly interspaced short palindromic repeats) spacer-based
microarrays for detection of viruses in environmental samples. Appl Environ
Microbiol 76, 7251–7258 (2010).
51. Deveau, H. et al. Phage response to CRISPR-encoded resistance in
Streptococcus thermophilus. J. Bacteriol. 190, 1390–1400 (2008).
52. Horvath, P. et al. Diversity, activity, and evolution of CRISPR loci in Streptococcus
thermophilus. J. Bacteriol. 190, 1401–1412 (2008).
53. Sapranauskas, R. et al. The Streptococcus thermophilus CRISPR/Cas system
provides immunity in Escherichia coli. Nucleic Acids Res. 39, 9275–9282
(2011).
54. Babu, M. et al. A dual function of the CRISPR-Cas system in bacterial antivirus
immunity and DNA repair. Mol. Microbiol. 79, 484–502 (2011).
55. Han, D., Lehmann, K. & Krauss, G. SSO1450--a CAS1 protein from Sulfolobus
1 6 F E B R UA RY 2 0 1 2 | VO L 4 8 2 | N AT U R E | 3 3 7
 INSIGHT REVIEW
solfataricus P2 with high affinity for RNA and DNA. FEBS Lett. 583, 1928–
1932 (2009).
56. Wiedenheft, B. et al. Structural basis for DNase activity of a conserved protein
implicated in CRISPR-mediated antiviral defense. Structure 17, 904–912
(2009).
57. Chen, C. S. et al. A proteome chip approach reveals new DNA damage
recognition activities in Escherichia coli. Nature Methods 5, 69–74 (2008).
58. Aguilera, A. The connection between transcription and genomic instability.
EMBO J. 21, 195–201 (2002).
59. Mojica, F. J., Diez-Villasenor, C., Garcia-Martinez, J. & Almendros, C. Short motif
sequences determine the targets of the prokaryotic CRISPR defence system.
Microbiology 155, 733–740 (2009).
60. Semenova, E. et al. Interference by clustered regularly interspaced short
palindromic repeat (CRISPR) RNA is governed by a seed sequence. Proc. Natl
Acad. Sci. USA 108, 10098–10103 (2011).
61. Tang, T. H. et al. Identification of 86 candidates for small non-messenger
RNAs from the archaeon Archaeoglobus fulgidus. Proc. Natl Acad. Sci. USA 99,
7536–7541 (2002).
62. Tang, T. H. et al. Identification of novel non-coding RNAs as potential antisense
regulators in the archaeon Sulfolobus solfataricus. Mol. Microbiol. 55, 469–481
(2005).
63. Ebihara, A. et al. Crystal structure of hypothetical protein TTHB192 from
Thermus thermophilus HB8 reveals a new protein family with an RNA
recognition motif-like domain. Protein Sci. 15, 1494–1499 (2006).
64. Carte, J., Pfister, N. T., Compton, M. M., Terns, R. M. & Terns, M. P. Binding and
cleavage of CRISPR RNA by Cas6. RNA 16, 2181–2188 (2010).
65. Gudbergsdottir, S. et al. Dynamic properties of the Sulfolobus CRISPR/Cas and
CRISPR/Cmr systems when challenged with vector-borne viral and plasmid
genes and protospacers. Mol. Microbiol. 79, 35–49 (2011)
66. Manica, A., Zebec, Z., Teichmann, D. & Schleper, C. In vivo activity of CRISPR-
mediated virus defence in a hyperthermophilic archaeon. Mol. Microbiol. 80,
481–491 (2011).
67. Marraffini, L. A. & Sontheimer, E. J. Self versus non-self discrimination during
CRISPR RNA-directed immunity. Nature 463, 568–571 (2010).
This study identified a mechanism for distingishing self from non-self,
which relies on base-pairing potential in regions outside the crRNA
spacer sequence.
68. Bartel, D. P. MicroRNAs: genomics, biogenesis, mechanism, and function. Cell
116, 281–297 (2004).
69. Parker, J. S., Parizotto, E. A., Wang, M., Roe, S. M. & Barford, D. Enhancement
of the seed-target recognition step in RNA silencing by a PIWI/MID domain
protein. Mol. Cell 33, 204–214 (2009).
70. Beloglazova, N. et al. Structure and activity of the Cas3 HD nuclease MJ0384, an
3 3 8 | N AT U R E | VO L 4 8 2 | 1 6 F E B R UA RY 2 0 1 2
© 2012 Macmillan Publishers Limited. All rights reserved
effector enzyme of the CRISPR interference. EMBO J. 30, 4616–4627 (2011).
71. Groenen, P. M., Bunschoten, A. E., van Soolingen, D. & van Embden, J. D.
Nature of DNA polymorphism in the direct repeat cluster of Mycobacterium
tuberculosis; application for strain differentiation by a novel typing method. Mol.
Microbiol. 10, 1057–1065 (1993).
72. Mokrousov, I., Limeschenko, E., Vyazovaya, A. & Narvskaya, O. Corynebacterium
diphtheriae spoligotyping based on combined use of two CRISPR loci.
Biotechnol. J. 2, 901–906 (2007).
73. Liu, F. et al. Novel virulence gene and clustered regularly interspaced short
palindromic repeat (CRISPR) multilocus sequence typing scheme for subtyping
of the major serovars of Salmonella enterica subsp. enterica. Appl. Environ.
Microbiol. 77, 1946–1956 (2011).
74. Cady, K. C. & O’Toole, G. A. Non-identity targeting of yersinia-subtype CRISPR-
prophage interaction requires the Csy and Cas3 proteins. J. Bacteriol. 193,
3433–3445 (2011).
75. Zegans, M. E. et al. Interaction between bacteriophage DMS3 and host CRISPR
region inhibits group behaviors of Pseudomonas aeruginosa. J. Bacteriol. 191,
210–219 (2009).
76. Obbard, D. J., Gordon, K. H. J., Buck, A. H. & Jiggins, F. M. The evolution of RNAi
as a defence against viruses and transposable elements. Philos. Trans. R. Soc. B
Biol. Sci. 364, 99–115 (2009).
77. Aravin, A. A., Hannon, G. J. & Brennecke, J. The Piwi-piRNA pathway provides an
adaptive defense in the transposon arms race. Science 318, 761–764 (2007).
78. Aravin, A. A. et al. Double-stranded RNA-mediated silencing of genomic tandem
repeats and transposable elements in the D-melanogaster germline. Curr Biol
11, 1017–1027 (2001).
79. Bartel, D. P. MicroRNAs: target recognition and regulatory functions. Cell 136,
215–233 (2009).
80. Guo, H., Ingolia, N. T., Weissman, J. S. & Bartel, D. P. Mammalian microRNAs
predominantly act to decrease target mRNA levels. Nature 466, 835–840
(2010).
Acknowledgements B.W. is a Howard Hughes Medical Institute Fellow of the Life
Sciences Research Foundation. S.H.S. acknowledges support from the National
Science Foundation and National Defense Science & Engineering Graduate
Research Fellowship programs. J.A.D. is an Investigator of the Howard Hughes
Medical Institute.
Author Information Reprints and permissions information is available at
www.nature.com/reprints. The authors declare no competing financial
interests. Readers are welcome to comment on the online version of this
article at www.nature.com/nature. Correspondence should be addressed to
J.A.D. (doudna@berkeley.edu).
 Prospects & Overviews
The phage-host arms race:
Shaping the evolution of microbes
Adi Stern and Rotem Sorek
Bacteria, the most abundant organisms on the planet,
are outnumbered by a factor of 10 to 1 by phages that
infect them. Faced with the rapid evolution and turnover
of phage particles, bacteria have evolved various mech-
anisms to evade phage infection and killing, leading to
an evolutionary arms race. The extensive co-evolution of
both phage and host has resulted in considerable diver-
sity on the part of both bacterial and phage defensive
and offensive strategies. Here, we discuss the unique
and common features of phage resistance mechanisms
and their role in global biodiversity. The commonalities
between defense mechanisms suggest avenues for the
discovery of novel forms of these mechanisms based on
their evolutionary traits.
.
Keywords:
arms race; bacteria; evolution; phage; resistance
DOI 10.1002/bies.201000071
Department of Molecular Genetics, Weizmann Institute of Science,
Rehovot, Israel
*Corresponding author:
Rotem Sorek
E-mail: rotem.sorek@weizmann.ac.il
Abbreviations:
Abi, abortive infection; Cas, CRISPR-associated sequence; CRISPR,
clustered regularly interspersed palindromic repeats; crRNA, CRISPR RNA;
LGT, lateral gene transfer; MTase, methyltransferase; REase, restriction
endonuclease; RM, restriction modification; TA, toxin-antitoxin.
Bioessays 33: 43–51,ß 2010 WILEY Periodicals, Inc.
Review essays
Introduction
Phage-host relationships have been studied intensively since
the early days of molecular biology. In the late 1970s, while
viruses were found to be ubiquitous, it was assumed that they
were present in relatively low numbers and that their effect on
microbial communities was low [1]. With the increasing avail-
ability of new molecular techniques that allow studies of
microbial communities without the need to culture them
[2], it is now realized that viruses greatly outnumber bacteria
in the ocean and other environments, with viral numbers
(107–108/mL) often 10-fold larger than bacterial cell counts
(106/mL) [3–5]. Thus, bacteria are confronted with a constant
threat of phage predation.
The Red Queen hypothesis (Box 1) posits that competitive
environmental interactions, such as those displayed by hosts
and parasites, will lead to continuous variation and selection
towards adaptation of the host, and counter-adaptations on
the side of the parasite. Arguably, nowhere is this evolutionary
trend so pronounced as in phage-microbe interactions. This is
due to the extremely rapid evolution and turnover of phage
particles [6], causing acute pressure on microbial communities
to evade infection and killing by phages. In fact, the arms race
between phages and bacteria is predicted to have had an
impact on global nutrient cycling [7], on global climate
[6, 7], on the evolution of the biosphere [8], and also on the
evolution of virulence in human pathogens [9].
This review focuses on the evolution of three of the most
well-studied microbial defense mechanisms against phages:
the restriction-modification (RM) system, the recently discov-
ered ‘‘clustered regularly interspersed palindromic repeats’’
(CRISPR) loci together with CRISPR-associated (cas) genes,
and the abortive infection (Abi) system (summarized in
Table 1). We first describe these defense systems, and the
counter-adaptations that evolved in the phage to allow escape
from bacterial defense. Next, we discuss features that are
common to many microbial defense systems, such as rapid
evolution, tendency for lateral gene transfer (LGT), and the
selfish nature of these systems. We elaborate on defense
systems that have gained new functions in the host genome.
Finally, the exciting hypothesis that many other prokaryotic
defense systems are still to be discovered is discussed. Our
review mainly focuses on the evolutionary angle of the phage-
www.bioessays-journal.com 43
 A. Stern and R. Sorek Prospects & Overviews ....
material and another that protects host genetic material from
Box 1 restriction (Fig. 1A). Both activities are mediated by recog-
nition of a specific DNA sequence, on average 4–8 bp long.
The Red Queen hypothesis
Review essays

Protection is normally conferred by modification (usually

The Red Queen hypothesis was originally proposed by
Van Valen [10], and is also termed the evolutionary arms
race hypothesis. As the Red Queen tells Alice in Lewis
Carroll’s Through the Looking-Glass: ‘‘Now, here, you
see, it takes all the running you can do, to keep in the
same place. If you want to get somewhere else, you must
run at least twice as fast as that!’’ The original theory
proposed that in tight co-evolutionary interactions, such
as those in a prey-predator relationship, changes (e.g.,
running faster) on the one side may lead to near extinc-
tion of the other side. The only way the second side can
maintain its fitness is by counter-adaptation (running
even faster). This will lead to an uneasy balance between
prey and predator, where species have to constantly
evolve to stay at the same fitness level. The metaphor
of an evolutionary arms race has been found to be
relevant for many biological processes, but nowhere is
this metaphor as apt as in host-parasite relationships.
host arms race; for deeper mechanistic descriptions of phage
resistance systems we refer the readers to an excellent recent
review [11]. In addition, the discussion here is restricted to
active defense mechanisms; passive host adaptations, such as
mutations at the phage receptors, are not discussed.
Restriction-modification systems:
Degradation of foreign DNA
Arguably, the most well-studied phage defense mechanism is
the RM system [12], which is present in over 90% of sequenced
bacterial and archaeal genomes [13]. This system is composed
of two activities: one that restricts incoming foreign genetic
methylation) of specific bases in this recognition sequence
in the host genome. Accordingly, all non-methylated DNA
recognition sequences are recognized as foreign and are
cleaved. Genetically, the minimal composition of RM systems
consists of a methyltransferase (MTase) gene that performs the
defense activity and a restriction endonuclease (REase) gene
that performs the foreign restriction activity. Since the REases
undergo rapid evolution, it is often the presence of an MTase
that serves as the basis for identification of RM systems in
newly sequenced genomes [14].
Phages have evolved to evade the ubiquitous RM systems
in a variety of ways (Fig. 1B). Some phages have acquired an
MTase, or stimulate the host MTase so that it confers protec-
tion to the phage genome [15]. Other phages code for proteins
that target and shut down the REase. One interesting example
is the Ocr protein of phage T7, which blocks the active site of
some REases by mimicking 24 bp of bent B-form DNA [16].
Alternatively, some phages incorporate unusual bases in their
genomes, thus throwing off the REase. For example, some
Bacillus subtilis phages replace thymine with 5-hydroxyme-
thyluracil [15]. These phages code a protein that then further
inhibits the host protein uracil-DNA glycosylase from cleaving
uracil bases from the phage DNA [17, 18]. Interestingly, this
inhibitory protein is also a DNA mimic (reviewed in ref. 19).
The T-even phages T2, T4, and T6 also contain unusual bases
in their genomes and may further post-synthetically glycosy-
late their DNA to avoid REase restriction [15]. Yet another
evasion mechanism that has been reported is alteration of
the restriction recognition sites. For example, some phages
employ ‘‘palindrome avoidance’’: since Type II REases often
recognize symmetrical (palindromic) sequences, some phages
tend to avoid containing such sites in their genome [20].
The well-studied model organisms Escherichia coli and T4
phage display a fascinating example of a co-evolutionary arms
race [11]. The battle purportedly begins with the T4 phage genome,
which contains the modified base hydroxymethylcytosine instead

Table 1. A summary of the major defense mechanisms described in this article

Name Mechanism Phylogenetic breadth

Restriction-modification The restriction enzyme cleaves specific patterns
in the incoming foreign DNA, while the modification
enzyme protects host DNA from cleavage by
unique biochemical modification
CRISPR/Cas Fragments of phage DNA are integrated into
CRISPR loci, which are then transcribed and
processed into short non-coding RNAs. These
RNAs, along with the associated Cas proteins,
guide the way to interfere with the phage nucleic
acids, in an as-yet-unknown mechanism
Abortive infection Premature cellular death occurs upon phage entry,
blocking the expansion of the phage to neighboring
cells. Notably, Abi systems include a large collection
of mechanisms with little or no known evolutionary
relationship, apart from a very similar phenotype
Appear in 90% of all sequenced
prokaryotic genomes [13]
Appear in 40% and 90% of all
sequenced bacterial and archaeal
genomes, respectively [27]. The
distribution of CRISPR/Cas-bearing
species across the phylogenetic tree
of life is highly patchy [38, 59]
Currently known Abi systems display
a sporadic phylogenetic distribution in
Gamma-proteobacteria and in Firmicutes [44],
yet recent reports suggest that some
systems may have an even broader
phylogenetic range [54, 111]

44 Bioessays 33: 43–51,ß 2010 WILEY Periodicals, Inc.

.... Prospects & Overviews
Figure 1. Restriction-modification (RM) defense system. A: A general
illustration of function, exemplified by type II RM enzymes. B:
Examples of strategies employed by phages to evade restriction:
(1) incorporation of unusual bases protects from restriction [15];
(2) masking of the restriction sites by phage proteins [112]; (3) stimu-
lation of MTase activity causes the phage DNA to be protected;
(4) neutralization of REase by phage proteins that mimic DNA [113].
of cytosine [15]. To counter-attack the phage, E. coli K-12
possesses a unique form of REase, the McrBC enzyme, which
cleaves only modified DNA substrates (such as that of T4) [21].
The T4 phage rises to the challenge by glucosylating its
genome, and is thus impervious to McrBC [15]. However,
E. coli cT596 encodes the GmrS-GmrD system that can also
restrict glucosylated DNA [22, 23]. In continuation of the
battle, some T4 phages encode the IPI protein, which in its
processed form (IPI) disables the GmrS-GmrD system [24].
Evidently, the battle is far from an end, since bacterial strains
have been found to overcome the IPI protein of T4 [24]. In
fact, it is likely that many bacterial defense systems,
coupled with their cognate phage evasion strategies, have
undergone similar attack and counter-attack cycles, where
a change on one side selects for changes that can overcome
the opponent.
Bioessays 33: 43–51,ß 2010 WILEY Periodicals, Inc.
A. Stern and R. Sorek
CRISPR/Cas: Acquired adaptive immunity
The existence of a unique nucleotide arrangement in E. coli,
Review essays
comprised of a cluster of direct repeats interspersed with
variable sequences (termed spacers; both around 30 bp long),
was first described in 1987 [25]. These clusters, together with
associated cas genes (Fig. 2A), were later found to exist in
40% of sequenced bacterial genomes and 90% of archaeal
genomes [26–28]. The first inkling that this system comprises a
phage-defense mechanism arose when spacer sequences were
found to be highly similar to DNA from foreign origin, i.e. from
phage, plasmid or transposon DNA [29–31]. In 2007, exper-
imental evidence was presented that this system indeed con-
fers resistance to phage infection [32]. This led to the striking
insight that similar to vertebrates, prokaryotes possess
acquired (and inherited) immunity [27, 33–35]. Following
phage infection, it has been found that a small portion of
bacterial cells integrated new spacers identical to the phage
genomic sequence (termed proto-spacer), resulting in CRISPR-
mediated phage resistance [32, 36, 37]. Further experiments
have shown that the CRISPR locus is transcribed into a single
RNA transcript, which is then further cleaved by the Cas
proteins to generate smaller CRISPR RNA (crRNA) units, each
including one targeting spacer [37, 38]. These units then
interfere with the incoming foreign genetic material by comp-
lementary base-pairing with either DNA [36, 37] or RNA [39]
from the foreign element (Fig. 2A).
While the study of CRISPR/Cas is still in its infancy,
examples of phages that are resistant to CRISPR/Cas interfer-
ence have nevertheless been noted (Fig. 2B). Following the
first round of infection and acquisition of novel spacers by the
bacterial CRISPR, phages that have mutated, recombined or
lost their proto-spacer target sequence in the second round of
infection are then resistant to CRISPR [32, 40–42]. It also seems
likely that phages have evolved mechanisms that directly
target the CRISPR/Cas machinery. To date, only vague hints
of such mechanisms exist. For example, it has been shown
that one of the proteins encoded by the T7 phage phosphor-
ylates the CasB protein [43]. It remains to be shown whether
this feat of the phage affects CRISPR/Cas functioning.
Abortive infection: Cellular suicide
If a phage has successfully entered the host cell and avoided
restriction by the host RM systems and by CRISPR, it proceeds to
develop, replicate, and release its progeny. Abi is a collective
term describing host mechanisms that interrupt with phage
development at different stages of phage transcription, genome
replication, and phage packaging [44]. Abi-mediated resistance
leads to death (suicide) of the cell, and is thought to occur since
corruption of host functions has already been initiated by the
phage. However, this death confers an advantage to surround-
ing bacterial cells since it confines the infection to the sacrificed
cell and prevents the spread of infectious particles.
Although several Abi systems have been discovered, with
the majority of them encoded on plasmids, the mechanism by
which they operate remains largely unknown. Frequently, Abi
is mediated by a single gene encoded on a plasmid or on a
prophage, which displays little or no homology to any known
45
 A. Stern and R. Sorek
A) Repeats
cas genes
Review essays
Spacers
transcription
Cas processing
proteins
B)
?
P
Figure 2. The CRISPR/Cas system. A: Mechanism of action: tran-
scription from the repeat-spacer CRISPR locus generates a long
non-coding RNA, with repeats that may sometimes assume a sec-
ondary structure. Cleavage of the repeat sequences by the Cas
proteins generates crRNAs that target the phage DNA or RNA, and
interfere with phage infection. B: Phages can evade CRISPR interfer-
ence by mutation or recombination of the targeted proto-spacer
sequence. Another putative evasion mechanism is phosphorylation
of the Cas proteins. However, this remains to be verified.
proteins. Some Abi genes have been shown to target phage
genes involved in DNA replication [45, 46], and others have
been shown to target the host translation apparatus [47, 48].
For instance, in E. coli K-12, the Lit protease encoded by the
defective e14 prophage is only activated in the presence of a
short polypeptide called Gol, which is produced by the T4
phage. Once active, this protease cleaves the translation
elongation factor Ef-Tu, thus leading to translational arrest
and cell death (Fig. 3A). Similarly, the Prr protein (also part of
a defective prophage in E. coli cT196) cleaves tRNALys in
response to the presence of the T4 peptide Stp [49].
Mutations in these phage peptides suppress activation of
Lit or Prr and rescue the infecting phage [47, 49] (Fig. 3B).
46
Prospects & Overviews ....
CRISPR
locus
CRISPR
pre-crRNA
Processed,
mature crRNAs
Since Abi genes have a toxic effect on their host, they are
under tight regulation [50]. In fact, many similarities can be
drawn between Abi systems (and also RM systems) and toxin-
antitoxin (TA) systems. TA systems are composed of a stable
toxin and an unstable antitoxin [51]. Normally, the antitoxin
binds and inhibits the toxin. However, a decrease in the levels
of the unstable antitoxin activates the toxin and leads to
growth arrest or cell death. It has been shown that this situ-
ation occurs in response to phage infection in two known TA
modules, mazEF and hok-sok, which can thus cause Abi [52,
53]. Moreover, the AbiQ system was recently found to function
as a protein-RNA TA pair [54], albeit exactly how this system
interferes with phage replication remains unknown. Thus, it
appears that some TA systems may be a subtype of Abi
systems. Interestingly, RM systems may also be viewed as
TA systems, since loss of the MTase (which like the antitoxin
is often unstable) results in a toxic effect of the REase [55, 56].
Commonalities among phage defense
systems
One of the most striking features common to all phage defense
systems is their high genetic variability, which occurs as a
Bioessays 33: 43–51,ß 2010 WILEY Periodicals, Inc.
.... Prospects & Overviews A. Stern and R. Sorek

to counteract the invading phage, thus contributing to the

extensive co-evolution with the prokaryotic parasites.
However, there is an interesting alternative view, not necess-

Review essays
arily mutually exclusive, which may explain the high varia-
bility and mobility of phage defense systems. Accordingly,
these systems are selfish elements, as elaborated in the next
section.

Defense systems: A burden or a benefit?

In this section, the advantages and disadvantages of microbial
immune systems are reviewed. Such a discussion is incom-
plete without examining the pros and cons of the phage-host
relationship. Evidently, for a single cell, phage predation, and
killing of the host are a strong disadvantage. However, phages
may also contribute directly to host fitness by supplying a pool
of new and possibly beneficial genes [73]. As such, phages
serve as vessels of LGT, which is one of the most important
forces in microbial evolution (Box 2).
What is the associated cost of encoding phage defense
mechanisms? An obvious primary cost is the energy cost
linked to carrying additional genetic cargo. The fact that,

Figure 3. The Abi system. A: Illustration of the mechanism of the

E. coli K-12 Abi Lit system. When the T4 phage peptide Gol is syn-
thesized, it binds and activates the bacterial (prophage-encoded) Lit
protein, which then cleaves the elongation factor EF-Tu. This leads
to the arrest of protein synthesis and to bacterial cell death, with the
phage trapped inside. B: A mutation at the Gol polypeptide reduces
the activation of Lit, and rescues the phage.
consequence of the co-evolutionary arms race with phages
[57]. This is manifested in the enormous numbers and types of
RM systems [12, 58], of CRISPR subtypes [59, 60] and in the
variety of Abi systems [44]. Moreover, the gene sequences of
these systems often display a high evolutionary rate ([61–63];
A. Stern and R. Sorek, unpublished data). This variability is
most easily exemplified in the diversity of RM systems: four
inherently different groups of RM systems exist, classified as
Type I–IV based on their subunit composition, mode of action,
and cofactor requirement (reviewed in ref. 12). Each type of RM
system includes an assortment of restriction enzymes that
recognize different recognition sequences. For instance,
nearly 4000 Type II enzymes are known of today, which
are further divided into 11 overlapping subclasses [13, 64].
Interestingly, Type II REase sequences have frequently been
found to be ORFan sequences, i.e. have no significant sim-
ilarity to any other protein [65, 66]. Initially, this was seen as
evidence for convergent evolution [67]. However, crystallo-
graphic studies have shown structural similarity among
Type II REases, most likely representing a rapid evolutionary
rate since their divergence from a common ancestor [68].
Another trait common to all defense systems presented
here is their propensity to undergo LGT, sometimes between
distantly related prokaryotes [69–72]. Often these systems
reside on plasmids, on prophages, or on genomic islands of
foreign origin, or are linked to transposase genes. This mobi-
lity allows rapid acquisition and dissemination of new systems
Box 2
Role of defense systems in limiting LGT
LGT is the process whereby genetic material is incorp-
orated from a non-parental organism. There are three
major mechanisms of LGT: (a) transduction by phages;
(b) transformation of naked DNA; and (c) bacterial con-
jugation, which involves transfer of DNA via direct con-
nections generated between a pair of cells. All three
mechanisms involve the entry of foreign DNA into the
host cell. LGT is recognized as one of the most important
mechanisms of genetic innovation in both bacteria and
archaea [82–84]. In fact, it is now realized that much of
the genomic diversity in prokaryotes is a consequence of
LGT, rather than allelic differences at the same loci [85,
86]. However, when viewed from the point of view of a
single cell, LGT has a negligible chance of contributing to
the fitness of the organism. Evidently, transduction by
lytic phages may severely compromise the survival of a
bacterial colony. Plasmids and transposons may also
decrease the fitness of an organism by integrating into
crucial regions of the genomes, by adding an energetic
cost involved with replication of excess DNA, and by
expressing harmful genes (e.g. refs. 87–89). Finally, even
if a new beneficial gene sequence is obtained, there is
only a small probability that it will integrate into the
current cellular network of an organism without causing
deleterious effects [90, 91]. Thus, defense mechanisms
such as those described here are crucial for maintaining
the genetic identity of organisms, and protecting it
against the constant bombardment of potentially detri-
mental foreign DNA.

Bioessays 33: 43–51,ß 2010 WILEY Periodicals, Inc. 47

 A. Stern and R. Sorek
often, only a small portion of a given bacterial population
bears a plasmid or a prophage with a given defense mechan-
ism may be seen as evidence for this cost. A second intriguing
Review essays
cost of encoding a phage defense mechanism is the risk of
autoimmunity. Autoimmunity, classically defined in mamma-
lian immune systems, is the failure of the immune system to
recognize what is self and what is foreign, resulting in an
immune response against self. In RM systems, this is a clear
danger since the restriction enzyme is often more stable than
the protecting methylase [74]. Abi mechanisms are equally
lethal, since errant function of the Abi gene will lead to cell
death. Finally, the CRISPR system is also not ‘‘immune’’ to
errors, and the frequent acquisition of self genetic material
also leads to spacers that target the self-genome, potentially
incurring autoimmunity ([75], see also ref. 76).
Paradoxically, phages themselves often bear anti-phage
defense systems, enabling their acquisition by host cells. To
cite a few examples, the HindIII RM gene complex was found
on a cryptic prophage in the Haemophilus influenzae genome
[77], the abiN Abi gene is encoded on a prophage in
Lactococcus lactis subsp. cremoris S114, and even a CRISPR
array was found within a Clostridium difficile prophage [78].
These results initially seem counterintuitive, since what
benefit is there for the phage to carry such systems? One
possible explanation is that this allows superinfection exclu-
sion, thus preventing other phages from infecting an already
infected cell [11, 78, 79]. However, this phenomenon has also
been viewed as evidence for the selfish nature of phage
defense mechanisms.
The behavior of defense mechanisms as selfish mobile
elements has been extensively discussed for the case of RM
systems [55, 80], but is also applicable for many Abi systems
that operate as TA systems, which also have ‘‘selfish’’ proper-
ties [51, 81]. The main lines of evidence in favor of this view is
that (a) RM systems destroy any other invading RM system,
(b) any attempt to lose the RM system will result in the death of
the host, and (c) RM systems are prone to extensive mobility,
and are often associated with plasmids, phages, transposons,
and integrons [55]. These characteristics of the RM systems
lead to an increase of their relative frequency in the bacterial
population. Thus, according to this hypothesis, the defense
incurred by RM systems on host cells is a mere by-product of
the fact that RM systems defend themselves.
Exaptations: Alternative functions of
defense systems
Intriguingly, a number of anti-phage defense systems have
evolved to gain a distinct function in cellular regulation that is
independent of phage restriction. Such evolutionary events
were coined ‘‘exaptations’’ [92], a term used to describe the
use of a biological structure or function for a purpose other
than that for which it initially evolved. For instance, several
RM systems have lost their REase activity, leaving an orphan
MTase that can now take part in epigenetic modifications. The
Dam methylase in E. coli and the CcrM methylase in
Caulobacter crescentus, both of which originated from RM
systems [93], are two such examples. Methylation by Dam
has been linked to several important regulatory processes,
48
Prospects & Overviews ....
such as mismatch repair by the MutHLS complex, binding
of the replication-initiation complex to methylated OriC, and
regulation of bacterial pathogenicity [93]. On the other hand,
the CcrM methylase has been shown to affect the cell cycle in
alpha proteobacteria that encode this gene [93].
Another interesting example of exaptation of RM systems
is evident in phase-variable Type III RM systems [94], whose
genes can be reversibly inactivated due to tandem repeat tracts
in their sequences. These repeats initiate a mechanism called
slipped-strand mispairing, leading to a change in the number
of repeats after DNA replication and possible frame-shift
mutations [94]. Several regulatory roles have been suggested
for phase-variable RM systems: (a) to allow regulated removal
of the barrier against foreign DNA, thus allowing potentially
beneficial uptake of DNA [95, 96], (b) autolytic self-DNA
degradation or ‘‘bacterial suicide’’ [97–99] (discussed further
below), and (c) epigenetic gene regulation via differential meth-
ylation of the genome [100]. The latter phenomenon, which
allows switching different genes on and off, has been linked
to pathogenicity of bacterial species by allowing colonization,
immune evasion, and adaptation to novel environments [94].
TA modules, which include some Abi systems, are also
known to participate in a variety of other cellular processes.
For example, one of the most studied TA loci, mazEF, aborts
translation by cleaving mRNA molecules in response to differ-
ent stress signals [101, 102], one of which is phage infection
[52]. An ongoing debate exists whether this action is reversible
or not: reversible effects have been attributed to bacteriostatic
effects, which allow reduced growth rate of each cell during
nutritional stress [101, 102]. On the other hand, irreversible
effects of mazEF have been attributed to programmed cell
death that occurs in a subpopulation of cells, permitting
the survival of the population of a whole [103].
Interestingly, in Myxococcus xanthus it was shown that the
toxin MazF exists without the antitoxin MazE, and has
adopted a key transcriptional regulator as an alternative anti-
toxin [104]. MazF mediates programmed cell death during
multicellular development of this organism. To summarize,
these accounts exemplify the broad evolutionary diversifica-
tion of different microbial defense mechanisms, and their
potential to cross boundaries from phage-encoded mechan-
isms, to anti-phage systems, to regulatory host mechanisms.
Conclusions
Despite our growing understanding of microbial immunity,
much still remains obscure. Do defense systems work separ-
ately or in unison? What is the cost of each system? Which
phages are targeted by which systems? For instance, while
most characterized defense systems work against double-
stranded DNA phages, RNA viruses might also be abundant
[105]. Defense systems that target such viruses are yet to be
discovered.
The recently discovered CRISPR system epitomizes our
incomplete understanding of the complexity of bacterial
defense systems. The discovery that almost half of all prokar-
yotes possess acquired inherited immunity came as a surprise
to the scientific community, given our initial tendency to view
prokaryotes as less complex organisms. However, since it is
Bioessays 33: 43–51,ß 2010 WILEY Periodicals, Inc.
.... Prospects & Overviews
now realized that prokaryotes are faced with a constant threat
of predation, it is becoming clearer that the microbial immune
system must be highly complex. Nevertheless, the CRISPR
system and the Abi systems have a highly sporadic distri-
bution across the prokaryotic phylogeny (Table 1).
Furthermore, while RM systems are present in 90% of pro-
karyotic species [13], this system is far from infallible, since it
has been shown that phages have a non-negligible probability
of escaping the REase and being methylated by the MTase.
[106]. Thus, all three mechanisms reviewed here are either
somewhat leaky, or span only part of the phylogenetic breadth
of bacteria and archaea. Given the intensive arms race
between bacteria and phages, it seems probable that there
are other yet unknown defense mechanisms waiting to be
revealed.
So how does one set about the search for novel prokaryotic
defense systems, and how does one learn more about existing
ones? It is likely that many solutions to these questions will be
made possible, thanks to advances in sequencing technologies
such as whole-transcriptome studies and metagenomics. For
instance, a seminal work by Andersson and Banfield [42]
reconstructed viral and host population genomes from com-
munity genomic data, and showed how CRISPR spacers cor-
relate with the location of coexisting viruses and hosts. In
addition, controlled experimental evolution studies, such as
those performed in Pseudomonas fluorescens and its phage,
enable a direct survey of the changes that occur in the phage
and host population with and without infection [107–109].
These studies are also likely to shed light on current defense
systems and possibly discover novel ones.
More specifically, the search for novel defense mechan-
isms may be based on the well-established common properties
of both eukaryotic and prokaryotic immune systems: their
high rate of evolution, and their tendency to undergo LGT.
Thus, it may be that a reservoir of novel defense mechanisms
lies in the most variable regions of bacterial and archaeal
genomes, known as genomic islands [110]. A strategy that
focuses on such islands to search for novel phage resistance
mechanisms might lead, in the future, to surprising
discoveries.
Acknowledgments
R.S. is an EMBO Young Investigator. He was supported, in
part, by the ISF-FIRST program (grant 1615/09), NIH
R01AI082376-01, ERC-StG, the Wolfson Family Trust miRNA
research program, the Minerva foundation, and the Yeda-Sela
Center for basic research. A.S. was supported by the Clore
post-doctoral fellowship.
References
1. Torrella F, Morita RY. 1979. Evidence by electron micrographs for a
high incidence of bacteriophage particles in the waters of Yaquina Bay,
Oregon: ecological and taxonomical implications. Appl Environ
Microbiol 37: 774–8.
2. Riesenfeld CS, Schloss PD, Handelsman J. 2004. Metagenomics:
genomic analysis of microbial communities. Annu Rev Genet 38: 525–
52.
3. Bergh O, Borsheim KY, Bratbak G, Heldal M. 1989. High abundance of
viruses found in aquatic environments. Nature 340: 467–8.
Bioessays 33: 43–51,ß 2010 WILEY Periodicals, Inc.
A. Stern and R. Sorek
4. Chibani-Chennoufi S, Bruttin A, Dillmann ML, Brussow H. 2004.
Phage-host interaction: an ecological perspective. J Bacteriol 186:
3677.
5. Wommack KE, Colwell RR. 2000. Virioplankton: viruses in aquatic
Review essays
ecosystems. Microbiol Mol Biol Rev 64: 69.
6. Fuhrman JA. 1999. Marine viruses and their biogeochemical and eco-
logical effects. Nature 399: 541–8.
7. Suttle CA. 2007. Marine viruses – major players in the global ecosystem.
Nat Rev Microbiol 5: 801–12.
8. Comeau AM, Krisch HM. 2005. War is peace – dispatches from the
bacterial and phage killing fields. Curr Opin Microbiol 8: 488–94.
9. Brussow H, Canchaya C, Hardt WD. 2004. Phages and the evolution of
bacterial pathogens: from genomic rearrangements to lysogenic con-
version. Microbiol Mol Biol Rev 68: 560–602.
10. Van Valen L. 1973. A new evolutionary law. Evol Theor 1: 1–30.
11. Labrie SJ, Samson JE, Moineau S. 2010. Bacteriophage resistance
mechanisms. Nat Rev Microbiol 8: 317–27.
12. Tock MR, Dryden DT. 2005. The biology of restriction and anti-restric-
tion. Curr Opin Microbiol 8: 466–72.
13. Roberts RJ, Vincze T, Posfai J, Macelis D. 2010. REBASE – a data-
base for DNA restriction and modification: enzymes, genes and
genomes. Nucleic Acids Res 38: D234–6.
14. Roberts RJ, Vincze T, Posfai J, Macelis D. 2005. REBASE – restriction
enzymes and DNA methyltransferases. Nucleic Acids Res 33:
D230.
15. Kruger DH, Bickle TA. 1983. Bacteriophage survival: multiple mech-
anisms for avoiding the deoxyribonucleic acid restriction systems of their
hosts. Microbiol Rev 47: 345–60.
16. Bandyopadhyay PK, Studier FW, Hamilton DL, Yuan R. 1985.
Inhibition of the type I restriction-modification enzymes EcoB and
EcoK by the gene 0.3 protein of bacteriophage T7. J Mol Biol 182:
567–78.
17. Wang Z, Mosbaugh DW. 1989. Uracil-DNA glycosylase inhibitor gene
of bacteriophage PBS2 encodes a binding protein specific for uracil-
DNA glycosylase. J Biol Chem 264: 1163–71.
18. Wang Z, Mosbaugh DW. 1988. Uracil-DNA glycosylase inhibitor of
bacteriophage PBS2: cloning and effects of expression of the inhibitor
gene in Escherichia coli. J Bacteriol 170: 1082–91.
19. Putnam CD, Tainer JA. 2005. Protein mimicry of DNA and pathway
regulation. DNA Repair (Amst) 4: 1410–20.
20. Rocha EP, Danchin A, Viari A. 2001. Evolutionary role of restriction/
modification systems as revealed by comparative genome analysis.
Genome Res 11: 946–58.
21. Raleigh EA, Wilson G. 1986. Escherichia coli K-12 restricts DNA con-
taining 5-methylcytosine. Proc Natl Acad Sci USA 83: 9070–4.
22. Bair CL, Rifat D, Black LW. 2007. Exclusion of glucosyl-hydroxyme-
thylcytosine DNA containing bacteriophages is overcome by the injected
protein inhibitor IPI. J Mol Biol 366: 779–89.
23. Bair CL, Black LW. 2007. A type IV modification dependent restriction
nuclease that targets glucosylated hydroxymethyl cytosine modified
DNAs. J Mol Biol 366: 768–78.
24. Rifat D, Wright NT, Varney KM, Weber DJ, et al. 2008. Restriction
endonuclease inhibitor IPI of bacteriophage T4: a novel structure for a
dedicated target. J Mol Biol 375: 720–34.
25. Ishino Y, Shinagawa H, Makino K, Amemura M, et al. 1987. Nucleotide
sequence of the iap gene, responsible for alkaline phosphatase isozyme
conversion in Escherichia coli, and identification of the gene product.
J Bacteriol 169: 5429.
26. Jansen R, van Embden JDA, Gaastra W, Schouls LM. 2002.
Identification of genes that are associated with DNA repeats in prokar-
yotes. Mol Microbiol 43: 1565–75.
27. Sorek R, Kunin V, Hugenholtz P. 2008. CRISPR – a widespread system
that provides acquired resistance against phages in bacteria and arch-
aea. Nat Rev Microbiol 6: 181–6.
28. Grissa I, Vergnaud G, Pourcel C. 2007. The CRISPRdb database and
tools to display CRISPRs and to generate dictionaries of spacers and
repeats. BMC Bioinformatics 8: 172.
29. Bolotin A, Quinquis B, Sorokin A, Ehrlich SD. 2005. Clustered regularly
interspaced short palindrome repeats (CRISPRs) have spacers of
extrachromosomal origin. Microbiology 151: 2551–61.
30. Mojica FJ, Diez-Villasenor C, Garcia-Martinez J, Soria E. 2005.
Intervening sequences of regularly spaced prokaryotic repeats derive
from foreign genetic elements. J Mol Evol 60: 174–82.
31. Pourcel C, Salvignol G, Vergnaud G. 2005. CRISPR elements in
Yersinia pestis acquire new repeats by preferential uptake of bacterio-
phage DNA, and provide additional tools for evolutionary studies.
Microbiology 151: 653–63.
49
 A. Stern and R. Sorek
32. Barrangou R, Fremaux C, Deveau H, Richards M, et al. 2007. CRISPR
provides acquired resistance against viruses in prokaryotes. Science
315: 1709–12.
33. Horvath P, Barrangou R. 2010. CRISPR/Cas, the immune system of
Review essays
bacteria and archaea. Science 327: 167–70.
34. Marraffini LA, Sontheimer EJ. 2010. CRISPR interference: RNA-
directed adaptive immunity in bacteria and archaea. Nat Rev Genet
11: 181–90.
35. van der Oost J, Jore MM, Westra ER, Lundgren M, et al. 2009.
CRISPR-based adaptive and heritable immunity in prokaryotes.
Trends Biochem Sci 34: 401–7.
36. Marraffini LA, Sontheimer EJ. 2008. CRISPR interference limits hori-
zontal gene transfer in staphylococci by targeting DNA. Science 322:
1843.
37. Brouns SJ, Jore MM, Lundgren M, Westra ER, et al. 2008. Small
CRISPR RNAs guide antiviral defense in prokaryotes. Science 321: 960–
4.
38. Makarova KS, Grishin NV, Shabalina SA, Wolf YI, et al. 2006.
A putative RNA-interference-based immune system in prokaryotes:
computational analysis of the predicted enzymatic machinery, functional
analogies with eukaryotic RNAi, and hypothetical mechanisms of action.
Biol Direct 1: 7.
39. Hale CR, Zhao P, Olson S, Duff MO, et al. 2009. RNA-guided
RNA cleavage by a CRISPR RNA-Cas protein complex. Cell 139:
945–56.
40. Deveau H, Barrangou R, Garneau JE, Labonte J, et al. 2008. Phage
response to CRISPR-encoded resistance in Streptococcus thermophi-
lus. J Bacteriol 190: 1390–400.
41. Heidelberg JF, Nelson WC, Schoenfeld T, Bhaya D. 2009. Germ
warfare in a microbial mat community: CRISPRs provide insights into
the co-evolution of host and viral genomes. PLoS One 4: e4169.
42. Andersson AF, Banfield JF. 2008. Virus population dynamics and
acquired virus resistance in natural microbial communities. Science
320: 1047–50.
43. Qimron U, Tabor S, Richardson CC. 2010. New details about bac-
teriophage T7-host interactions. Microbe 5: 117–20.
44. Chopin MC, Chopin A, Bidnenko E. 2005. Phage abortive infection in
lactococci: variations on a theme. Curr Opin Microbiol 8: 473–9.
45. Durmaz E, Klaenhammer TR. 2007. Abortive phage resistance mech-
anism AbiZ speeds the lysis clock to cause premature lysis of phage-
infected Lactococcus lactis. J Bacteriol 189: 1417.
46. Bidnenko E, Ehrlich D, Chopin MC. 1995. Phage operon involved in
sensitivity to the Lactococcus lactis abortive infection mechanism
AbiD1. J Bacteriol 177: 3824.
47. Georgiou T, Yu YTN, Ekunwe S, Buttner MJ, et al. Specific peptide-
activated proteolytic cleavage of Escherichia coli elongation factor Tu.
1998. Proc Natl Acad Sci USA 95: 2891.
48. Morad I, Chapman-Shimshoni D, Amitsur M, Kaufmann G. 1993.
Functional expression and properties of the tRNA (Lys)-specific core
anticodon nuclease encoded by Escherichia coli prrC. J Biol Chem 268:
26842.
49. Snyder L. 1995. Phage-exclusion enzymes: a bonanza of biochemical
and cell biology reagents? Mol Microbiol 15: 415–20.
50. Blower TR, Fineran PC, Johnson MJ, Toth IK, et al. 2009. Mutagenesis
and functional characterization of the RNA and protein components of
the toxIN abortive infection and toxin-antitoxin locus of Erwinia.
J Bacteriol 191, 6029–39.
51. Gerdes K, Christensen SK, Lobner-Olesen A. 2005. Prokaryotic toxin-
antitoxin stress response loci. Nat Rev Microbiol 3: 371–82.
52. Hazan R, Engelberg-Kulka H. 2004. Escherichia coli mazEF-mediated
cell death as a defense mechanism that inhibits the spread of phage P1.
Mol Genet Genomics 272: 227–34.
53. Pecota DC, Wood TK. 1996. Exclusion of T4 phage by the hok/sok killer
locus from plasmid R1. J Bacteriol 178: 2044.
54. Fineran PC, Blower TR, Foulds IJ, Humphreys DP, et al. 2009. The
phage abortive infection system, ToxIN, functions as a protein-RNA
toxin-antitoxin pair. Proc Natl Acad Sci USA 106: 894–9.
55. Kobayashi I. 2001. Behavior of restriction-modification systems as
selfish mobile elements and their impact on genome evolution.
Nucleic Acids Res 29: 3742–56.
56. Hayes F. 2003. Toxins-antitoxins: plasmid maintenance, programmed
cell death, and cell cycle arrest. Science 301: 1496–9.
57. Hoskisson PA, Smith MC. 2007. Hypervariation and phase variation in
the bacteriophage ‘resistome’. Curr Opin Microbiol 10: 396–400.
58. Orlowski J, Bujnicki JM. 2008. Structural and evolutionary classifi-
cation of Type II restriction enzymes based on theoretical and exper-
imental analyses. Nucleic Acids Res 36: 3552.
50
Prospects & Overviews ....
59. Haft DH, Selengut J, Mongodin EF, Nelson KE. 2005. A guild of 45
CRISPR-associated (Cas) protein families and multiple CRISPR/Cas
subtypes exist in prokaryotic genomes. PLoS Comput Biol 1: e60.
60. Kunin V, Sorek R, Hugenholtz P. 2007. Evolutionary conservation of
sequence and secondary structures in CRISPR repeats. Genome Biol 8:
R61.
61. Sharp PM, Kelleher JE, Daniel AS, Cowan GM, et al. 1992. Roles of
selection and recombination in the evolution of type I restriction-modi-
fication systems in enterobacteria. Proc Natl Acad Sci USA 89: 9836–40.
62. Murray NE, Daniel AS, Cowan GM, Sharp PM. 1993. Conservation of
motifs within the unusually variable polypeptide sequences of type I
restriction and modification enzymes. Mol Microbiol 9: 133–43.
63. Zheng Y, Roberts RJ, Kasif S. 2004. Identification of genes with fast-
evolving regions in microbial genomes. Nucleic Acids Res 32: 6347–57.
64. Roberts RJ, Belfort M, Bestor T, Bhagwat AS, et al. 2003.
A nomenclature for restriction enzymes, DNA methyltransferases, hom-
ing endonucleases and their genes. Nucleic Acids Res 31: 1805–12.
65. Kroger M, Hobom G, Schutte H, Mayer H. 1984. Eight new restriction
endonucleases from Herpetosiphon giganteus-divergent evolution in a
family of enzymes. Nucleic Acids Res 12: 3127.
66. Mullings R, Bennett SP, Brown NL. 1988. Investigation of sequence
homology in a group of type-II restriction/modification isoschizomers.
Gene 74: 245–51.
67. Wilson G, Murray NE. 1991. Restriction and modification systems.
Annu Rev Genet 25: 585–627.
68. Venclovas C, Timinskas A, Siksnys V. 1994. Five-stranded beta-sheet
sandwiched with two alpha-helices: a structural link between restriction
endonucleases EcoRI and EcoRV. Proteins 20: 279–82.
69. Nelson KE, Clayton RA, Gill SR, Gwinn ML, et al. 1999. Evidence for
lateral gene transfer between Archaea and bacteria from genome
sequence of Thermotoga maritima. Nature 399: 323–9.
70. Jeltsch A, Pingoud A. 1996. Horizontal gene transfer contributes to the
wide distribution and evolution of type II restriction-modification sys-
tems. J Mol Evol 42: 91–6.
71. Haaber J, Moineau S, Hammer K. 2009. Activation and transfer of the
chromosomal phage resistance mechanism AbiV in Lactococcus lactis.
Appl Environ Microbiol 75: 3358–61.
72. Godde JS, Bickerton A. 2006. The repetitive DNA elements called
CRISPRs and their associated genes: evidence of horizontal transfer
among prokaryotes. J Mol Evol 62: 718–29.
73. Gogarten JP, Townsend JP. 2005. Horizontal gene transfer, genome
innovation and evolution. Nat Rev Microbiol 3: 679–87.
74. Naito T, Kusano K, Kobayashi I. 1995. Selfish behavior of restriction-
modification systems. Science 267: 897–9.
75. Stern A, Keren L, Wurtzel O, Amitai G, et al. 2010. Self-targeting by
CRISPR: gene regulation or autoimmunity? Trends Genet 26: 335–40.
76. Marraffini LA, Sontheimer EJ. 2010. Self versus non-self discrimination
during CRISPR RNA-directed immunity. Nature 463: 568–71.
77. Hendrix RW, Smith MC, Burns RN, Ford ME, et al. 1999. Evolutionary
relationships among diverse bacteriophages and prophages: all the
world’s a phage. Proc Natl Acad Sci USA 96: 2192–7.
78. Sebaihia M, Wren BW, Mullany P, Fairweather NF, et al. 2006. The
multidrug-resistant human pathogen Clostridium difficile has a highly
mobile, mosaic genome. Nat Genet 38: 779–86.
79. Lu MJ, Henning U. 1994. Superinfection exclusion by T-even-type
coliphages. Trends Microbiol 2: 137–9.
80. Nakayama Y, Kobayashi I. 1998. Restriction-modification gene com-
plexes as selfish gene entities: roles of a regulatory system in their
establishment, maintenance, and apoptotic mutual exclusion. Proc
Natl Acad Sci USA 95: 6442–7.
81. Makarova KS, Wolf YI, Koonin EV. 2009. Comprehensive comparative-
genomic analysis of type 2 toxin-antitoxin systems and related mobile
stress response systems in prokaryotes. Biol Direct 4: 19.
82. Gogarten JP, Doolittle WF, Lawrence JG. 2002. Prokaryotic evolution
in light of gene transfer. Mol Biol Evol 19: 2226–38.
83. Koonin EV, Makarova KS, Aravind L. 2001. Horizontal gene transfer in
prokaryotes: quantification and classification. Annu Rev Microbiol 55:
709–42.
84. Dagan T, Artzy-Randrup Y, Martin W. 2008. Modular networks and
cumulative impact of lateral transfer in prokaryote genome evolution.
Proc Natl Acad Sci USA 105: 10039–44.
85. Whitaker RJ, Banfield JF. 2006. Population genomics in natural
microbial communities. Trends Ecol Evol 21: 508–16.
86. Medini D, Serruto D, Parkhill J, Relman DA, et al. 2008. Microbiology in
the post-genomic era. Nat Rev Microbiol 6: 419–30.
87. Modi RI, Adams J. 1991. Coevolution in bacterial-plasmid populations.
Evolution 45: 656–67.
Bioessays 33: 43–51,ß 2010 WILEY Periodicals, Inc.
.... Prospects & Overviews
88. Zund P, Lebek G. 1980. Generation time-prolonging R plasmids: cor-
relation between increases in the generation time of Escherichia coli
caused by R plasmids and their molecular size. Plasmid 3: 65–9.
89. Bouma JE, Lenski RE. 1988. Evolution of a bacteria/plasmid associ-
ation. Nature 335: 351–2.
90. Kedzierska B, Glinkowska M, Iwanicki A, Obuchowski M, et al. 2003.
Toxicity of the bacteriophage lambda cII gene product to Escherichia coli
arises from inhibition of host cell DNA replication. Virology 313: 622–8.
91. Lercher MJ, Pal C. 2008. Integration of horizontally transferred genes
into regulatory interaction networks takes many million years. Mol Biol
Evol 25: 559–67.
92. Brosius J, Gould SJ. 1992. On ‘‘genomenclature’’: a comprehensive
(and respectful) taxonomy for pseudogenes and other ‘‘junk DNA’’. Proc
Natl Acad Sci USA 89: 10706–10.
93. Marinus MG, Casadesus J. 2009. Roles of DNA adenine methylation in
host-pathogen interactions: mismatch repair, transcriptional regulation,
and more. FEMS Microbiol Rev 33: 488–503.
94. Srikhanta YN, Maguire TL, Stacey KJ, Grimmond SM, et al. 2005. The
phasevarion: a genetic system controlling coordinated, random switch-
ing of expression of multiple genes. Proc Natl Acad Sci USA 102: 5547–
51.
95. Ando T, Xu Q, Torres M, Kusugami K, et al. 2000. Restriction-modi-
fication system differences in Helicobacter pylori are a barrier to inter-
strain plasmid transfer. Mol Microbiol 37: 1052–65.
96. Donahue JP, Israel DA, Peek RM, Blaser MJ, et al. 2000. Overcoming
the restriction barrier to plasmid transformation of Helicobacter pylori.
Mol Microbiol 37: 1066–74.
97. Dybvig K, Sitaraman R, French CT. 1998. A family of phase-variable
restriction enzymes with differing specificities generated by high-fre-
quency gene rearrangements. Proc Natl Acad Sci USA 95: 13923–8.
98. Saunders NJ, Peden JF, Hood DW, Moxon ER. 1998. Simple sequence
repeats in the Helicobacter pylori genome. Mol Microbiol 27: 1091–8.
99. Hamilton HL, Dillard JP. 2006. Natural transformation of Neisseria
gonorrhoeae: from DNA donation to homologous recombination. Mol
Microbiol 59: 376–85.
100. Seib KL, Peak IR, Jennings MP. 2002. Phase variable restriction-
modification systems in Moraxella catarrhalis. FEMS Immunol Med
Microbiol 32: 159–65.
Bioessays 33: 43–51,ß 2010 WILEY Periodicals, Inc.
A. Stern and R. Sorek
101. Pedersen K, Christensen SK, Gerdes K. 2002. Rapid induction and
reversal of a bacteriostatic condition by controlled expression of toxins
and antitoxins. Mol Microbiol 45: 501–10.
102. Christensen SK, Pedersen K, Hansen FG, Gerdes K. 2003. Toxin-
Review essays
antitoxin loci as stress-response-elements: ChpAK/MazF and ChpBK
cleave translated RNAs and are counteracted by tmRNA. J Mol Biol 332:
809–19.
103. Engelberg-Kulka H, Amitai S, Kolodkin-Gal I, Hazan R. 2006.
Bacterial programmed cell death and multicellular behavior in bacteria.
PLoS Genet 2: e135.
104. Nariya H, Inouye M. 2008. MazF, an mRNA interferase, mediates pro-
grammed cell death during multicellular Myxococcus development. Cell
132: 55–66.
105. Lang AS, Rise ML, Culley AI, Steward GF. 2009. RNA viruses in the
sea. FEMS Microbiol Rev 33: 295–323.
106. Korona R, Levin BR. 1993. Phage-mediated selection and the
evolution and maintenance of restriction-modification. Evolution 47:
556–75.
107. Paterson S, Vogwill T, Buckling A, Benmayor R, et al. 2010.
Antagonistic coevolution accelerates molecular evolution. Nature 464:
275–8.
108. Buckling A, Wei Y, Massey RC, Brockhurst MA, et al. Antagonistic
coevolution with parasites increases the cost of host deleterious
mutations. Proc Biol Sci 2006. 273: 45–9.
109. Wichman HA, Badgett MR, Scott LA, Boulianne CM, et al. 1999.
Different trajectories of parallel evolution during viral adaptation.
Science 285: 422–4.
110. Juhas M, van der Meer JR, Gaillard M, Harding RM, et al. 2009.
Genomic islands: tools of bacterial horizontal gene transfer and evol-
ution. FEMS Microbiol Rev 33: 376–93.
111. Pandey DP, Gerdes K. 2005. Toxin-antitoxin loci are highly abundant in
free-living but lost from host-associated prokaryotes. Nucleic Acids Res
33: 966–76.
112. Iida S, Streiff MB, Bickle TA, Arber W. 1987. Two DNA antirestriction
systems of bacteriophage P1, darA, and darB: characterization of darA-
phages. Virology 157: 156–66.
113. Dryden DT, Tock MR. 2006. DNA mimicry by proteins. Biochem Soc
Trans 34: 317–9.
51
Gene, 27 (1984) 131-149 131
Elsevier

GENE 919

The use of transposon Tn5 mutagenesis in the rapid generation of correlated physical and genetic
maps of DNA segments cloned into multicopy plasmids - a review?

(Escherichia coli; rhizobia; agrobacteria; recombinant DNA; nitrogen fixation; transposition frequency;
insertional specificity; polarity; bacteriophage L and Mu; eukaryotic genes)

F.J. de Bruijn * and J.R. Lupski **

* Department of Cellular and Developmental Biology, Harvard University, Cambridge. MA 02138. Department of Molecular
Biology. Massachusetts General Hospital,Jackson 11, Boston, MA 02114, and ** Biochemistry Department, New York Universit?
Medical Center, 550 First Avenue, New York, NY 10016 (U.S.A.) Tel. (212) 340-5137

(Received August 15th, 1983)

(Accepted September 6th, 1983)

SUMMARY

The properties of transposon Tn5 that render it useful for in vivo mutagenesis of cloned DNA sequences
are reviewed. Transposition frequency, insertional specificity, polarity and stability of Tn5 insertion mutations
are among the topics discussed. Examples are cited from the published literature which illustrate the applications
of Tn5 mutagenesis to the analysis of cloned prokaryotic and eukaryotic genes. A methods section is included
which outlines precisely how to carry out transposon Tn5 mutagenesis analysis of cloned DNA segments.

INTRODUCTION

Transposable genetic elements carrying antibiotic

resistance genes (Tn elements or transposons) have
+ Dedicated to the memory of Ahmad I. Bukhari, our friend and
mentor, who inspired writing this review.
proven to be extremely useful tools in bacterial
* To whom correspondence and reprint requests should be sent genetics (see Kleckner et al., 1977):
at his present address: Max-Planck-Institut fur Ziichtungsfor- (1) The transposition of a Tn element into a gene
schung, Abt. Schell, D-5000 Koln, Vogelsang 30 (F.R.G.) Tel. usually leads to inactivation of that gene, with such
(49)-221-5062315 or 5062341.
insertion mutations being generally nonleaky and
Abbreviations: Ap’, ampicillin resistance; LB, Luria broth, SM
stable.
buffer, YM broth, see MATERIALS AND METHODS, section (2) Such insertion mutations in multi-gene ope-
a; Cm’, chloramphenicol resistance; Em’, erythromycin resis- rons usually exert polar effects on genes located
tance; IR, inverted repeat; Km’, kanamycin resistance; kb, kilo- downstream from the insertion site. This is presum-
base pairs; m.o.i., multiplicity of infection; M-STET, see MATE- ably due to the presence of transcriptional as well as
RIALS AND METHODS, section el; Nm’, neomycin resis-
translational termination signal sequences carried by
tance; NPT II, neomycin phosphotransferase type II; p.f.u.,
plaque-forming units; RIA, radioimmune assay; SDS, sodium
the Tn element.
dodecyl sulfate; Sm’, streptomycin resistance; Tc’, tetracycline (3) A Tn insertion mutation represents a new
resistance; ::, novel joint. genetic (e.g., antibiotic resistance) and physical

0378-I 119/84/$03.00 0 1984 Elsevier Science Publishers

132

marker at the site of insertion. The latter can be Tn5 mutagenesis has proven extremely useful in
physically identified, e.g., by DNA-DNA hetero- the molecular genetic analysis of the Gram-negative
duplexing or by restriction analysis. Rhizobiaeceae family. Rhizobia induce the formation
In this review we shall focus on the properties of of specialized structures (nodules) on the roots or
a 5700-bp transposon Tn5 (Berg et al., 1975; Berg, stems of their host plant, in which the bacteria carry
1977), encoding the NPTII, which confers resistance out the nitrogen-fixation process (see Sprent, 1979).
to a number of aminoglycoside antibiotics, such as Agrobacteria induce the formation of crown-gall
neomycin and kanamycin (and streptomycin upon tumors on their host plant (see Kahl and Schell,
certain bacterial species, such as Rhizobium meliloti; 1982). Since the phenotype of any bacterial mutation
Putnoky et al., 1983) upon its host (Berg et al., 1978). in a gene involved in the symbiotic associations is
Transposon Tn.5 is composed of two inverted only observable in plantum, conventional muta-
repeats of 1535 bp (IS.50; Berg, 1983) containing genesis and screening techniques are of limited use.
genes responsible for transposition and its regu- A plasmid vector for the introduction of transposon
lation, flanking an approximately 2700-bp region Tn5 and the finding that Tn5 is capable of trans-
carrying the structural gene for NPTII (see Rezni- position and inducing insertion mutations (see
koff, 1982; for an exhaustive review of the biology of below; Beringer et al., 1978; Van Vliet et al., 1978),
Tn5, see Berg and Berg, 1983). made possible the pre-selection of random muta-
Tn5 mutagenesis has been successfully employed tions, thereby effectively reducing the number of
in the genetic analysis of a number of bacterial infected plants to be screened. This has resulted in
species. TnS-induced insertion mutations have been the identification of a number of Rhizobium loci in-
generated in chromosomal genes and operons of volved in nodulation (nod) and nitrogen fixation (nif)
Escherichia coli and Salmonella typhimurium, such as (Beringer et al., 1978; Rolfe et al., 1980; 1981; Dun-
his (Kleckner et al., 1977; 1979; Biek and Roth, can, 1981; Meade et al., 1982; Scott et al., 1982;
1980), lac (Berg et al., 1980; Miller et al., 1980), ilv Buchanan-Wollaston et al., 1980; Lamb et al., 1982;
(Berg et al., 1979), trg (Harayama et al., 1979), ksg Ma et al., 1982; Forrai et al., 1983) and Agrobac-
(Fouts and Barbour, 1981), uvrD, mutB and mutD terium loci involved in tumor formation and opine
(Shanabruch et al., 1981) and pnp (Portier, 1980). catabolism (Garlinkel and Nester, 1980; and for
Tn5 mutagenesis has been used to generate muta- reviews see Bevan and Chilton, 1982 and Kahl and
tions and then to facilitate cloning of genes of bac- Schell, 1982).
terial species such as Caulobacter crescentus Ruvkun and Ausubel (1981) extended the Tn5
(Purucker et al., 1982), Myxococcus xanthus (Kuner mutagenesis methodology for the Rhizobiaceae by
and Kaiser, 1980; Shimkets et al., 1983) and Erwinia developing a general method for site-directed Tn5
herbicola (Gantotti et al., 1981). Tn5 mutagenesis mutagenesis. It involves the cloning of Rhizobium (or
has also been used to study genes coding for bio- Agrobacterium) nif, nod or tumor formation genes
luminescence of Vibrio fischeri (Engebrecht et al., into multicopy plasmids, followed by Tn5 muta-
1983) and Vibrio harvqyi (Belas et al., 1982) IgAl genesis in E. coli, reintroduction of the TnS-mutated
proteases of Haemophilus infuenzae (Bricker et al., sequences into Rhizobium (or Agrobacterium) and
1983) heme biosynthesis in R. meliloti (Leong et al., forced gene-replacement of the corresponding Rhizo-
1982), phase variation of Salmonella (Silverman bium (or Agrobacterium) wild-type sequences with
and Simon, 1980), naphthalene degradation genes the TnS-mutated counterparts. Such analysis has
(Schell, 1983), type I fimbriae of Klebsiella pneu- made possible the combined physical and genetic
moniae (Purcell and Clegg, 1983). and acetohy- characterization of additional nod and nifgenes in a
droxy acid synthetase in S. typhimurium (Shaw et al., variety of Rhizobium species (Ruvkun et al., 1982;
1980). The structure and organization of bacterial Ausubel, 1982; Corbin et al., 1981; 1983; Fuhrmann
episomes and plasmids such as mini-F (Phua et al., and Hennecke, 1982; Scott et al., 1982) and the
1982) and CloDF13 (Hakkaart et al., 1981) and construction of correlated physical and genetic maps
bacteriophages such as Pl (Heilmann et al., 1980) of the regions of the tumor inducing (Ti) plasmid of
and Mu (Giphart-Gassler et al., 1982) have also A. tumefaciens (Garlinkel et al., 1981; Klee et al.,
been facilitated by the use of Tn5 mutagenesis. 1983; for reviews see Bevan and Chilton, 1982; Kahl
and Schell, 1982).

TnS-induced mutations have been used to clone
genes that lack a readily identifiable phenotype, to
direct suppressive integration of ColEl derivatives
into the chromosome of dnaA strains of E. coli
(Sasakawa and Yoshikawa, 1980) and to construct
transposable genetic elements carrying genes of
interest by flanking such genes with two complete
copies of Tn.5 (Guarente et al., 1980), ,or by cloning
the gene(s) of interest directly within the transposon
Tn5 sequences (Meyer et al., 1979). The selectable
genetic marker introduced by Tn5 into cryptic plas-
mids or those without a readily selectable phenotype
have been useful for the genetic transfer and identifi-
cation of such TnS-tagged plasmids (Johnston et al.,
1978; Lamb et al., 1982; Hooykaas et al., 1981).
Finally, a simple deletion derivative of transposon
Tn5 has been proposed to be useful in the creation
of a “mobile recombinational switch”, since it is
capable of turning on the expression of genes down-
stream from its insertion site in one orientation, but
not the other (Berg, 1980).
Tn5 mutagenesis for construction of correlated
physical and genetic maps was first successfully
applied to the characterization of the chromosomal
nitrogen fixation (nif) gene cluster of K. pneumoniue
by Riedel et al. (1979). These authors physically
mapped a large number of transposon insertions in
the nifgene cluster (isolated and genetically mapped
by Merrick et al., 1978; 1980), by cleavage of total
chromsomal DNA of strains carrying nif: :Tn.5 inser-
tions with appropriate restriction endonucleases (see
Fig. 2), followed by Southern transfer and probing of
the resulting filters with cloned subfragments of the
nifregion. Altered patterns of hybridization, due to
the presence of transposon insertions allowed the
physical mapping of the transposon. This combined
with the genetic mapping studies resulted in the
construction of a correlated physical and genetic
map of the nzf region by unambiguously ordering
Tn-induced insertion mutations, providing an esti-
mation of the minimum sizes of individual genes and
an assignment of specific genes to restriction frag-
ments (Riedel et al., 1979).
In this review we will focus on an extension of the
transposon Tn5 mutagenesis methodology described
by Riedel et al. (1979), namely the rapid construction
of correlated physical and genetic maps of pro-
karyotic (and eukaryotic) DNA segments cloned
into multicopy plasmids. We will discuss a number
133
of parameters of transposon Tn5 mutagenesis such
as vectors for the introduction of Tn5, transposition
frequency and insertional specificity of this trans-
poson, stability and polarity of TnS-induced inser-
tion mutations and convenient restriction sites within
Tn5 for physical mapping, based on our observations
as well as the observations from the literature cited.
In the interest of clarity, we will cite a number of the
best studied cases to illustrate our points, but will not
attempt to provide an exhaustive reference list for
each point discussed. In addition, we include a
“cook-book” style Materials and Methods section,
describing the experimental techniques involved in
the Tn5 mutagenesis protocol we employ.
MATERIALS AND METHODS
(a) Media and solutions (see Miller, 1972)
A broth: 10 g Bactotryptone, 2.5 g NaCl per liter.
Plates are solidified with 1.1% agar.
A top agar: 1 broth plus 0.6% agar.
YM broth: 2 broth plus 0.2% maltose plus 0.01%
yeast extract.
LB (Luria broth): 10 g Bactotryptone, 5 g yeast
extract, 5 g NaCl per liter with the pH adjusted to 7.5
with NaOH.
SM buffer: 0.1 M NaCl, 0.02 M Tris pH 7.5,
0.01 M MgSO,, 0.01% gelatin.
(b) Bacteriophage and bacterial strains
The A::Tn5 phage was originally constructed by
Dr. Douglas Berg (Berg et al., 1975). The 1 bacterio-
phage vector we used as a source of Tn5 was 1467,
which was obtained from Dr. Nancy Kleckner.
Phage 1467 has the following genotype: lb221
rex: :Tn5 ~1857, Oam29, Pam80. The b22 1 mutation
is a deletion in the 1 genome that removes att, the
phage attachment site. This prevents the. bacterio-
phage from undergoing lysogeny. Gene rex is non-
essential for Agrowth and in this case contains a Tn5
insertion. The ~1857 mutation results in a tempera-
ture-sensitive repressor. Both the 0 and P genes are
involved in phage i DNA replication, so amber
mutations in these genes will prevent phage replica-
tion in a Su” background.
134

Bacterial strain Ql (fhr-1 Zeu-6 thi-1 IucYl, supE44 physically verified using one of the rapid plasmid
ton,421 T5’ $80’) or strain LE392 (F ~, hsdR5 14 (rk , preparation protocols described below. One must be
mk ) supE44 supF58 lacy1 galK2 gal722 metB1 certain that the plasmid is a monomer. Plasmid-
trpR.5.5 A- ) is used as host for propagating A467. In bearing cells are then grown to saturation in 5 ml YM
our experience LE392 appears to contain a better broth (supplemented with antibiotic) at 30°C. The
suppressor and thus gives higher titre phage stock. culture is diluted lOO-fold and grown at 30’ C in 5 ml
Any bacterial strain that is neither Su’ nor 3, YM broth plus antibiotics to an A,,, of 0.8 (approx.
resistant can be used as the host for plasmids to be 10” cells/ml). Then 1 ml of cells is mixed with 1,: :Tn5
mutagenized with A: :TnS. phage at an m.o.i. of 1- 10 and placed at 30°C for
2 h. Infected cells are then plated in 0.2-ml aliquots
(c) Preparation of 12467 (L::TnS) phage stocks onto LB agar plates containing 20 pg/ml kanamycin
and the antibiotic used to select for the plasmid. The
Cells of LE392 are infected with 2: :Tn5 phage and plates are incubated at 30°C for 48 h. Km’ colonies
incubated overnight at 37 o C to obtain fresh plaques. are washed off the plates by adding 5 ml 209, sucrose
Simultaneously, a fresh culture of LE392 is grown in solution (20 O0 sucrose, 10 mM Tris pH 7.8, 1 mM
5 ml YM broth to saturation. The next day the EDTA) to each plate using a sterile glass spreading
LE392 culture is diluted in LB and regrown to an rod. The cell suspension is decanted into a sterile
AhOO(Bausch and Lomb Spectronic 20) of approx. 30-ml Corex centrifuge tube and the cells are collect-
0.8. One fresh plaque from the above plate is picked ed by centrifugation at 7000 rev./min for 10 min in a
with a sterile Pasteur pipette and placed into 1 ml of Beckman JA-17 rotor. The cells are resuspended in
SM buffer (see Maniatis et al., 1982). Then 0.25 ml 5 ml sucrose solution or LB broth and plasmid DNA
of SM buffer containing phage is added to 0.5 ml of is isolated by the cleared lysate procedure of Clewell
the regrown LE392 cells. After mixing, the cells are and Helinski (1969) or using a scaled-up (5 x )
incubated at 37°C for 20 min to allow phage version ofthe alkaline lysis protocol described below.
absorption. To the cell and phage mixture 7.5 ml of Plasmid DNA thus prepared is then used to trans-
melted ,? top agar equilibrated at 45 “C is added. form competent cells of the desired bacterial strain,
Aliquots of 2.5 ml are then spread per A plate. An in order to screen for the phenotype of the Tn5
equivalent amount of LE392 cells with no phage mutated plasmid. When characterizing Tn5 inser-
added is plated as a control. The plates are incubated tions into a cloned gene with a readily observable
at 37°C for 6-8 h and inspected for lysis by com- phenotype, the bacterial strain of choice will carry a
paring with the LE392 control. When confluent lysis mutation in such gene, and the TnS-mutated plas-
is observed on the plates containing the A::TnS-in- mids are directly screened for their ability to comple-
fected cells, and the control plates show a visible ment the mutation. For example, in the case of the
bacterial lawn, the phage is harvested by scraping the Tn5 mutagenesis analysis of the cloned glnALG
top agar into a sterile screw-top Corex centrifuge (glnAR) operon of K. pneutnoniae, plasmid DNA
tube and adding 2.5 ml SM buffer plus 0.5 ml chlo- from I::TnS-infected cells was used to transform
roform. The mixture is then vortexed thoroughly, competent cells of a GlnA R- K. pneumoniue
incubated on ice for 5 min, and centrifuged in a strain, and the transformants were screened for their
Beckman JA-17 rotor at 10000 rev./min, 4°C for GlnA and GlnR phenotype (De Bruijn and Ausubel.
5 min. After decanting the supernatant into a sterile 1981; see Fig. 4). Thus ghzA : :Tn5 and
screw-top tube, a few drops of chloroform are added glnR(LG): :Tn5 insertions were genetically identified.
and the lysate is stored at 4°C. The phage are then When characterizing Tn5 insertions into a cloned
titered on LE392 (see Miller, 1972). gene without a genetically identifiable phenotype,
such as a eukaryotic gene, DNA from A: :Tn5 infect-
(d) Mutagenesis of plasmid DNA with transposon ed cells is used to transform any desired strain of
Tn5 E. coli, followed by screening of the transformants by
other means such as antibody binding assays (see
The plasmid to be mutagenized is transformed EXPERIMENTAL, SeCtiOn j).
into any Su’ /1” bacterial strain and its presence

(e) Rapid plasmid minipreps
(1) Boiling method [modified from methods by De
Bruijn and Ausubel (1981), Holmes and Quigley
(1982) and Lupski et al. (1982)]
Cells from a 5-ml saturated culture, grown in LB
plus the appropriate antibiotic, are placed in a 15-ml
Corex tube and pelleted by centrifuging at
7000 rev./min for 10 min in a Beckman JA-17 rotor.
The cell pellet is resuspended in 50 ~1of 25 % sucrose
by vigorous vortexing. To this suspension 300 ~1 of
M-STET (5% Triton X-100,50 mM EDTA, 50 mM
Tris pH 8, 5 % sucrose) is added. After mixing, 50 ~1
of 20 mg/ml lysozyme in H,O is added. This mixture
is gently vortexed, transferred to a 1.5-ml eppendorf
tube using a Pasteur pipette, heated in a boiling water
bath for 2 min, and placed on ice. The tubes are spun
for 15 min in an eppendorf table-top centrifuge
placed in the cold room. The viscous pellet is gently
removed with a pipette, and to the supematant 50 ~1
of 3 M K. acetate (pH 4.8) and 500 ~1of isopropanol
are added. The DNA is allowed to precipitate at
- 20’ C for l-2 h, then pelleted by centrifugation for
15 min in the eppendorf centrifuge kept at 4’ C. The
white pellet is dried and resuspended in 120 ~1H,O.
The chromosomal DNA remains as a pellet, while
the plasmid DNA readily goes into solution. A quick
spin is performed to again pellet this chromosomal
DNA, and the resuspended plasmid DNA is passed
over a 400 ~1 Sephadex G-50 (medium) column
equilibrated with 10 mM Tris pH 7.5, 1 mM EDTA.
The Sephadex minicolumns are made by puncturing
the bottom of a 400 ~1 eppendorf tube, using a 20
gauge needle heated over a bunsen burner, covering
this hole with a small glass wool plug, followed by
sterilization and placing the tube inside a larger
1.5-ml eppendorf tube which acts as a support. The
400~~1eppendorf is then tilled to overflowing with
equilibrated G-50 Sephadex and centrifuged at top
speed for 75 s in an IEC table-top centrifuge to pack
the column. The 120~1 of resuspended plasmid
DNA is then loaded on top of the packed column
and the tube spun again in a 1.5-ml eppendorf tube
at top speed for 45 s. In this manner 24 minipreps
can be done simultaneously. Enough purified plas-
mid DNA is usually recovered from 5 ml of liquid
culture to perform 10-12 restriction enzyme digests.
135
(2) Alkali lysis method [modified from method by
Bimboim and Doly (1979)]
A plasmid-bearing strain is inoculated from a
single colony into 3 ml of LB plus appropriate
antibiotic. The cells are grown to A,,, (Spectronic
20) of approximately 0.4-0.5, and 300 pg chloram-
phenicol is added to amplify the plasmid. After
allowing approx. 5 h for amplification, the cells are
collected in a 1.5-ml eppendorf tube by performing
2 x 30 s centrifugations in an eppendorf table-top
centrifuge. The pellet is resuspended in 200 ~1resus-
pension buffer (25 mM Tris pH 8.0,50 mM glucose,
10 mM EDTA, 2 mg/ml lysozyme, 50 pg/ml RNase)
and incubated on ice for 5 min or until the viscosity
changes. It is then mixed with 200 ~1 of freshly pre-
pared alkali lysing buffer (0.5% SDS, 0.2 N NaOH)
until the solution becomes clear; then 200 ~1 3 M
K * acetate pH 4.8 is added, and a white precipitate
appears. This is mixed well, but gently, and incubated
for 5 min at 0°C. The solution is centrifuged for
5 min at 4°C and the supematant is transferred to
another eppendorf tube. The DNA is precipitated by
adding 500 ~1isopropanol and placing it at - 20’ C.
After centrifuging for 5 min in the cold room, the
pellet is resuspended in 50 ~1 H,O and passed over
a G-50 Sephadex microcolumn as described above.
EXPERIMENTAL
(a) Vectors used for transposon Tn5 mutagenesis
A number of different vectors have been employed
for the generation of TnS-induced insertion muta-
tions. For the mutagenesis of bacterial genera such
as Rhizobium, Agrobacterium, Caulobacterium and
Erwinia, a promiscuous Pl-type R-factor (RP4),
carrying a copy of bacteriophage Mu and of Tn5
(pJB4JI; Beringer et al., 1978) has been utilized.
This plasmid is extremely unstable in the above-
mentioned bacterial species, presumably due to Mu-
mediated illegitimate recombination events upon
introduction, and is therefore called a “suicide plas-
mid”.
Selection for Km’ or Nm’ afer mating the plasmid
across from E. coli results in the transposition of the
Tn5 from the “suicide plasmid” into the genome of
the recipient cell, followed by loss of the plasmid
136

vector DNA sequences (Beringer et al., 1978; Van mutagenesis has been made possible even in bac-
Vliet et al., 1978; Ely and Croft, 1982; Gantotti et al., terial species such as S. typhimurium, which are
1981; Duncan, 1981; Walton and Moseley, 1981). normally not sensitive to A infection. Palva and
However, Tn5 mutagenesis using pJB4JI and deri- Lindstrom (198 1) have shown that introduction of
vative vectors has generated ambiguous results in the E. coli malB region, encoding the 3, receptor
some bacterial species such as Rhizobium meliloti protein (LamB), into S. typhimurium allows II phage
(Meade et al., 1982). These authors reported that a adsorption and DNA injection, but not replication,
large percentage of the induced insertion mutations in this bacterium.
carried phage Mu DNA sequences in addition to Other vectors useful for the introduction of Tn5
Tn5 sequences at the site of insertion. Therefore, include plasmid ColEl derivatives carrying a tem-
when using this Tn5 mutagenesis protocol, caution perature sensitive replication mutation in addition to
should be exercised in analyzing presumptive Tn5- the Tn5 transposon, allowing the selection for Tn5
induced insertion mutations; the resulting strains transposition at elevated temperatures (Laird and
should be checked for the presence of both Tn5 and Young, 1980; Meade, H., Brown, S. and Ausubel,
Mu DNA sequences by Southern blotting (Meade F., unpublished). This allows Tn5 mutagenesis in
et al., 1982; Forrai et al., 1983). any bacterial species which can be transformed with
An alternative vector has been constructed con- ColEl type plasmid DNA. In addition, strains
sisting of the pBR325 replicon (Bolivar et al., 1977) containing a single copy of Tn5 integrated into the
into which the mobilization (mob) site of the promi- chromosome, either as a simple Tn5 insertion or as
scuous P-type plasmid RP4 has been inserted and part of an integrated /i : : Tn5 prophage, have been
carrying a copy of Tn5 (pSUP2021; Simon et al., used to generate Tn5 insertions (for examples see
1983). Such a plasmid can be mobilized at a very Fuhrmann and Hennecke, 1982; Pannekoek et al.,
high frequency to bacterial species such as Rhizobium 1980; Hakkaart et al., 1981).
or Agrobacterium, when provided in tram wit transfer
functions. Once transferred conjugally, it is unable to
replicate in the recipient Rhizobium or Agrobacterium (b) Transposon mutagenesis protocol
cells, and addition of Km or Nm to the growth
medium selects for the transposition of the Tn5 from For transposon Tn5 mutagenesis of DNA seg-
the plasmid replicon into the genome of the cell. This ments cloned into (multicopy) plasmids in E. coli, the
vector may circumvent some of the problems asso- ;i : : Tn5 (1467) vector has proven to be the most
ciated with the pJB4JI plasmid described above. useful one, due to the ease with which high titer
For Tn5 mutagenesis in E. coli, S. typhimurium, stocks of this conditionally defective phage can be
and Myxococcus xanthus, bacteriophage vectors prepared (see MATERIALS AND METHODS,
have been used. In the case of S. t_vphimurium a section c), the high frequency with which Tn5 trans-
derivative of P22, incapable of lytic growth or stable poses from the A : : Tn5 vector, and the efficiency
lysogenization and carrying Tn5 (DB 2062; D. Bot- with which the vector is lost (segregated away) after
stein) has been employed (see for example Shana- Tn5 transposition. The protocol routinely used is
bruch et al., 1981). In Myxococcus bacteriophage outlined in Fig. 1 and can be broken down into the
Pl : : Tn5, capable of adsorption and injection of following steps:
phage DNA, but incapable of self-propagation in (1) Infection of E. coli cells harboring the (multi-
this bacterium, has been used (Kuner and Kaiser, copy) plasmid carrying the cloned DNA segment of
1981). In bacterial species that are sensitive to interest with the A: : Tn5 (see MATERIALS AND
infection by phage A, a ,? derivative deleted for the att METHODS, section d).
site and carrying amber mutations in genes 0 and P (2) Growth of the infected cells on plates contain-
(rendering it incapable of integration or replication in ing kanamycin (or neomycin) and the antibiotic used
Su” strains) with Tn5 inserted in a nonessential gene to select for the plasmid, to select for Tn5 trans-
(2467; Lb221 rex : : Tn5 ~1857 Oam29 Pam80; Berg position from 2 : : Tn5 into the chromosome of the
et al., 1975; N. Kleckner, unpublished observations) cell or into the resident plasmids.
has proven to be most effective. Recently i : : Tn5 (3) Preparation of plasmid DNA from the cells
 137

Tn5 MUTAGENESIS PROTOCOL

® INFECT PLATE LB+Cm(or····)

+ Km

((¡ E.coli ). 5
>.: :Tn5 +
( >,: replicot1on plosmid (Cm r
and intec;¡rotion or Te r etc) r
Cm r Km colonies
defic1ent)

Wild type
or EXTRACT
mutant . - - - TRANSFORM . . - - PLASMID
strain DNA

r r SCREEN EXTRACT RESTRICTION

Select Cm Km PHENOTYPE PLASMID - + MAP
DNA
Fig. l. Tn5 mutagenesis protocol. The general transposon Tn5 mutagenesis protocol employed in our studies includes the following
steps: (1) lnfection of cells harboring the recombinant plasmid to be mutagenized with .t:Tn5 (see MATERIALS AND METHODS,
section d). (2) Plating of the infected cells on LB plates containing Km (to select for Tn5) and the antibiotic corresponding to the
resistance gene carried by the plasmid vector (such as Cm, Te, or Ap). (3) Incubation ofthe plates at 30 º C for 48 h. (4) Extraction
of plasmid DNA from the Km' (and Cm', Te' or Ap') cells thus obtained (see MATERIALS AND METHODS, section d).
(5) Transformation of cells from mutant strains of E. co/i, in case the genetic phenotype ofthe Tn5-mutated plasmids can be determined
by genetic complementation, and wild-type cells if another method ofphenotypic identification is employed (see EXPERIMENTAL,
sectionj). (6) Screening ofthe phenotype ofthe resulting transformants (see MATERIALS AND METHODS, section d). (7) Extraction
ofplasmid DNA from the transformants ofinterest (see MATERIALS AND METHODS, section e). (8) Restriction mapping ofthe
Tn5-containing plasmids to determine the physical location of the Tn5 within the plasmid (see EXPERIMENTAL, section k).

TnS
IR IR
Km' obtained in the previous step (see MATERIALS AND
METHODS, section d).
cr,-c, :,:
-=-
"'"'X D
t

-
"' ;¡;
ce..
-C,)<
(4) Transformation o f competent cells o f E. coli
00

- =
) < ...,
::, u, "'
.-3, = o
Ql:::r I»
g, = 3
:,: = -
�
1kb
Fig. 2. Partial restriction map of Tn5. In this figure the most
useful restriction sites within transposon Tn5 are indicated. Tn5
does not contain any sites for the restriction endonucleases
EcoRI, Ball, Kpnl, Pvul, Clal, or Sst 1, and therefore these endo-
nucleases are useful to determine the approximate physical
location of a Tn5 with respect to target plasmid DNA fragments
generated by these restriction endonucleases. In addition, Tn5-
containing DNA segments can be easily re-cloned using these
endonucleases. Hindlll, Xhol, Pst I, Bg/II and Hpal cleave within
the in verted repeats (IRs) and are useful to determine the exact
physical location ofthe Tn5 insertion, relative to (a) given site(s)
within the plasmid target sequences and their distance away from
the ends ofTn5 are 1195 bp, 485 bp, 680 bp, 1515 bp and 185 bp,
respectively. Hpal is also extremely useful for the DNA sequence
;;·
:. 5 .,
c.
carrying (a) mutation(s) in the gene(s) ofinterest, to
carry out complementation studies using the wild-
type and Tn5-mutagenized D N A segments, or a
wild-type strain i f a screening method different from
genetic complementation is employed.
analysis of Tn5-target junction sequences (see EXPERIMEN-
TAL, section 1). Sa/1 and Smal can be employed to determine
the relative orientation of the Tn5, since these restriction endo-
nucleases cleave Tn5 asymmetrically. BamHI is convenient for
mapping the Tn5 insertion site, since it cleaves Tn5 approxi-
mately in the middle. The cleavage sites shown are derived from
Jorgensen et al. (1979) and Auerswald et al. (1981). The nucleo-
tide sequence for the Km' gene encoding NPTII has also been
determined by Beck et al. (1982).
138

(5) Isolation of plasmid DNA from single colonies the useful cleavage sites within Tn5 shown in Fig. 2
of the transformed cells using a small-scale plasmid (see EXPERIMENTAL, section k; for an example see
DNA preparation protocol (see MATERIALS AND Fig. 3).
METHODS, section e), followed by restriction endo- (6) Construction of a correlated physical and
nuclease mapping of the Tn.5 insertions employing genetic map of the cloned DNA segments by com-
bining the genetic complementation data obtained in
step 4 with the physical mapping data obtained in
step 5 (for an example see Fig. 4).
This protocol has been most effective in the
analysis of DNA segments carried by multicopy
plasmids, due to the relative ease with which large
numbers of independent Tn.5 insertions can be
generated and with which plasmid DNA can be pre-
pared from l-5-ml cultures for restriction endo-
nuclease analysis (see MATERIALS AND METHODS,
sections d and e). However, it can also be employed
in the analysis of DNA segments carried by low-
copy-number or single copy plasmids. In these cases,
the number of transpositions generated in a standard
H P H H P H experiment is reduced and the amount of cells used
for small scale plasmid DNA preparation has to be
W-1 _J
C B A
increased, but all other criteria remain the same.

Fig. 3. Example of an experiment to determine the location of

(c) Transposition frequency of Tn5
plasmid-borne Tn5 insertions. Hind111 cleavage pattern was
obtained for a series of plasmid::Tn5 isolates, run on a 0.75”,,
agarose gel and stained with ethidium bromide. The plasmid The transposition frequency of Tn.5 from the
used, pEGSI contains a single Hind111 site. TnS is cleaved twice /I : : Tn5 vector into the chromosome of its recipient
by HitzdIII, within the inverted repeats (IRS; see Fig. 2). There- cell is approximately 5 x 10 5 Km’ colonies per
fore cleavage of pEG8l::TnS plasmid with Hind111 generates
p.f.u. of the 2 : : Tn5 phage (Johnson et al., 1982)
three fragments, one fragment is a constant. internal 3400-bp
and from a site in the chromosomal fat operon (or
Tn5 fragment (B) and the other two reprcscnt Tn-target
function fragments (A and C). By measuring the size of the A an F’luc) into resident iL phage approx.
junction fragments and subtracting I I95 bp (distance between 5 x 10 ’ Km’ transductants per /1 phage produced
the Hind111 site in the inverted repeat and the end of Tn5), the (Egner and Berg, 1980).
exact position ofthe Tn5 insertion, relative to the unique Hind111 In our experiments we routinely infect 1 x 10”
site in pEG81, can be determined. The left-hand lane S shows the
cells with l-10 x 10” p.f.u. of iL: : Tn5 and obtain
DNA M, standards used, consisting of phage 1 DNA cleaved
with Hind111 + KpnI. and their sizes arc indicated in kb.
l-2 x 10” Km’ colonies, representing an approxi-
Additional M, markers used are shown in the right-hand lane S mate transposition frequency of 2 x 10 ’ to
and consist of M l3mp8 cleaved with HnelII; the size of these 1 x 10Ph Km’ colonies per p.f.u. of L : : Tn5 (De
fragments is 2527, 1623 and 849 bp, respectively. Lanes l-10 Bruijn et al., 1983). The transposition frequency of
show the Hind111 digestion pattern of ten pEG8 1::Tn5 plasmids
Tn5 into a supercoiled, multicopy plasmid may be
(Lupskiet al., 1983b)carryingTn5insertionsprogressivelyfarther
higher than the transposition frequency into a chro-
away from the closest unique Hind111 site of pEG8l. At the
bottom, the hatched boxes represent the pEG81 vector DNA mosomal DNA sequence, due to the possible prefer-
sequences, the open box represents TnS sequences (the wide ence of Tn elements to insert into actively replicating
open boxes represent the inverted repeats while the narrow open or supercoiled DNA molecules (Fitts and Taylor,
box represents the internal portion of Tn5 including the Km’ 1980). This would explain the relative ease with
gene), and the line represents the remainder of the cloned DNA
which 100-200 independent Tn5 insertion mutations
fragment carried by pEG81. The position of HirzdIII and Psi1
restriction sites is indicated by H and P. respectively. Restriction
into a plasmid can be generated from the infection of
fragment lengths A, B, and C correspond to those shown in the a 1 ml, mid-log culture of plasmid containing cells
above gel. (De Bruijn and Ausubel, 198 1; Ruvkun and Ausubel,
 139

glnRlLGl glnA

pFB514

,
1
lkb ,
VGlnA-
VGlnA+R- Tn5 insertions
VGlnA+R+

Fig. 4. Example of a physical and genetic map of a cloned DNA segment, as generated by transposon Tn5 mutagenesis. The triangles
indicate the physical map position of independent Tn5 insertions within the cloned DNA fragment as determined by first analyzing
pFBS 14::Tn5 plasmid DNA with EcoRI, which does not cleave Tn5 (see Fig. 2), in order to determine into which pFBS 14 EcoRI fragment
the transposon had inserted. The precise physical location of each Tn5 insertion was determined relative to the Hitid sites flanking
the cloned DNA segment, by cleavage of the pFB514::TnS plasmids with Hind111 and determining the size of the TnS-target DNA
junction fragments (De Bruijn and Ausubel, 1981; see also Fig. 3). The phenotype of the pFB514::TnS plasmids was determined by
examining their ability to complement GlnA and/or GlnR - mutations of Klebsiellu pneumoniae. The solid, open and dotted triangles
represent GlnA _, GlnA + R- and GlnA + R + pFB514::TnS plasmids respectively. The boxed-in segments represent plasmid vector
DNA sequences and the horizontal line represents the cloned K. pnezmzoniue DNA fragment. The combination ofthe physical and genetic
data resulted in the construction of the correlated physical and genetic map of the cloned glnAR(glnALG) region of K. pneumoniue, as
shown (De Bruijn and Ausubel, 1981). This analysis enabled these authors to determine the location and physical boundaries of the
structural gene for glutamine synthetase (ghA) and the identification of a previously unidentified locus glnR(ghLG), immediately
adjacent to gInA, involved in positive activation of the nitrogen fixation (nif) genes in K. pneumoniae.

1982; Lupski et al., 1982; 1983b,c; De Bruijn et al.,
1983).
The transposition frequency of a given transposon
may vary greatly depending on its host organism.
For example, the transposition frequency of Tn5 in
K. pneumoniae (from which this transposon was
originally isolated) appears to be even higher than in
E. coli, and therefore difficulties have sometimes
been encountered generating and stably maintaining
a single Tn5 insertion in K. pneumoniae because of
secondary Tn5 transpositions (see EXPERIMENTAL,
section f).
(d) Insertional specificity of Tn5
A variety of Tn elements, such as Tn3, Tn5, Tn9,
TnlO and phage Mu, have been examined for their
insertional specificity or insertion site preference (for
a review see Kleckner, 1981). Tn5 has been ranked
among the Tn elements with a relatively low inser-
tional specificity based on a number of different
experiments. Shaw and Berg (1979) examined the
insertional specificity of Tn5 on a “macro-level” by
determining the distribution of TnS-induced auxo-
trophic mutations in E. coli. These authors con-
cluded that even though a slight preference for
insertion into certain genes appeared to exist, never-
theless a large variety of TnS-induced auxotrophic
mutations could be readily identified. These results
were extended by Berg et al. (1980), who examined
the distribution of Tn5 insertions into the E. coli lac
operon and concluded that less than 57; of the
lac: : Tn5 insertions could not be separated by
genetic recombination, further suggesting a low
insertional specificity of Tn5. Independent results
from Miller et al. (1980) support this conclusion.
In our experiments we examined this question by
determining the distribution of 1500 independent
Tn5 insertions into a multicopy plasmid carrying
three genes whose expression could be monitored
easily (pGR100; hisD, Tc’, Ap’; Riedel, 1980). By
determining the number of Tn5 insertions leading to
insertional inactivation of each of the three genes and
comparing these numbers with the percentage of
plasmid DNA each individual gene occupies, an
essentially random distribution pattern was observed
(De Bruijn et al., 1983). In addition, by determining
the physical location of 92 Tn5 insertions in the
140

cloned K. pneumoniae glnALG(glnAR) operon (De inverted repeat (Bossi and Ciampi, 1981; Berg,
Bruijn and Ausubel, 1981), 88 Tn5 insertions in the 1983). Thus, there may be some preferential sites for
cloned K. pneumoniae hisDG0 region (De Bruijn insertion of Tn5 at the nucleotide sequence level.
et al., 1983), 28 Tn5 insertions in a plasmid carrying This does not appear to affect the usefulness of Tn5
the K. pneumoniae glnF(ntrA) gene (De Bruijn and in the generation of correlated physical and genetic
Ausubel, 1983), 61 Tn5 insertions in the cloned maps.
R. meliloti glnA (GSI) region (De Bruijn, F.J. and
Ausubel, F.M., in preparation), 88 Tn.5 insertions in (e) Polarity of TnS-induced insertion mutations
the cloned E. coli dnaG region (Lupski et al., 1982)
and 57 insertions in a plasmid carrying a cDNA Tn-induced insertion mutations generally exert a
fragment of Plasmodium knowlesi (see EXPERIMEN- strong polar effect on genes located distally (down-
TAL, section j; Lupski et al., 1983b), we have further stream) to their insertion site within the same operon
analyzed insertional specificity of Tn5. We found an due to transcriptional termination signals carried by
essentially random distribution pattern of Tn5 inser- the Tn element (see Kleckner, 1981). Transposon
tions into the variety of cloned DNA sequences listed Tn5 has been no exception to this rule (Berg et al.,
above and no evidence for the existence of prominent 1980). Therefore, TnS-induced insertion mutations
insertional hotspots, and encountered no difficulty have been employed extensively to determine the
generating Tn5 insertions every 50-100 bp in any of genetic organization of a variety of operons such as
the cloned DNA sequences mutagenized to date. the S. typhimurium his operon (Kleckner et al., 1975;
Our results confirm earlier experiments regarding the 1977; 1979), the E. coli ilv operon, the E. coli lac
insertional specificity of Tn5 made in a large number operon (Berg et al., 1980) and the complex
of other laboratories (see Garfinkel et al., 1981; K. pneumoniae nifgene cluster (Merrick et al., 1978;
Corbin et al., 1982; 1983; Ruvkun et al., 1982; Laird 1980; Riedel et al., 1979).
and Young, 1980; Shanabruch and Walker, 1980; However, in some instances incomplete polarity
Hakkaart et al., 1981). has been observed; for example, when analyzing
Even though Tn5 appears to insert randomly, or certain chromosomal TnS-induced insertion muta-
at least randomly enough to make it useful for in- tions in the E. coli lac operon (Berg et al., 1980).
sertion mutagenesis of cloned DNA sequences, at These authors observed that approximately one third
the nucleotide sequence level Tn5 may have some of the Tn5 insertions in the 1acZ gene resulted in
increased frequency of insertion into DNA se- constitutive, low-level expression of the downstream
quences that have homology to the Tn5 ends. We located lacy gene, while the remaining two thirds of
previously isolated Tn5 insertions into a recom- the 1acZ : : Tn5 insertions exerted a strictly polar
binant plasmid containing a cDNA insert coding for effect of lacy. The observed non-polar or partially
a Plasmodium knowlesi surface antigen (Lupski et al., polar phenotype of these 1acZ : : Tn5 insertion muta-
1983b). This cDNA insert has subsequently been tions was ascribed to the presence of promoter-like
shown to consist of 36 bp repeated in tandem eight DNA sequences at the ends of Tn5, initiating low-
times (Godson et al., 1983). By sequencing the level transcription of downstream genes, regardless
cDNA .. : Tn5 junctions of six Tn5 insertions we of the relative orientation of the Tn5 (Berg et al.,
have determined the exact location of the Tn5 inserts 1980).
into the individual repeats. All of the Tn5 inserts This effect is observed even more readily when
have been located within 10 nucleotides of the indi- analyzing TnS-induced insertion mutations in ope-
vidual repeats. whose DNA sequence appeared to rons (or segments thereof) that are carried by multi-
have some homology to the ends of Tn5. Three Tn5 copy plasmids. Strong, polar effects of such TnS-
inserts have been located between the same nucleo- induced insertion have been observed in the case of
tides in different repeats and thus they occupied the the cloned ent gene cluster of E. coli (Laird and
same relative position (Lupski, et al., 1983a). It has Young, 1980), the uvrB locus of E. coli (Pannekoek
been shown previously that Tn5 appears to integrate et al., 1980), the cloned glnALG(glnAR) operon of
with increased frequency into sequences which show K. pneumoniae (De Bruijn and Ausubel, 1981; De
some degree of homology to the end of the Tn5 Bruijn, 1983), and a plasmid carrying the cloned

K. pneurnoniae glnF(ntrA) gene (De Bruijn and Ausu-
bel, 1983). However, in the case of the cloned
hisDG0 region of K. pneumoniae, hisG : : Tn5 inser-
tion mutations were shown not to exert a polar effect
on the hisD gene located immediately downstream
on the same plasmid (De Bruijn et al., 1983). This
result was unexpected in the light of previous find-
ings regarding the genetic organization of the his
operon in enteric bacteria, as determined by trans-
poson mutagenesis (Kleckner et al., 1975; 1977;
1979; Ciampi et al., 1982; Schmid and Roth, 1983).
In addition, in the case of the Tn5 mutagenesis ana-
lysis of the cloned dnaG region of E. coli (Lupski
et al., 1982), a number of Tn5 insertions that were
subsequently shown to map in the region between the
dnaG promoter and its structural gene (Lupski et al.,
1983~) also failed to abolish dnaG expression, sug-
gesting lack of Tn5 polarity. In both of these cases,
the lack of polarity effect was shown to be indepen-
dent of the relative orientation of the Tn5, unlike the
orientation-dependent polarity effects observed with
IS2 (Saedler et al., 1974; Pilacinski et al., 1977;
Sommer et al., 1979). Also, insertions of TnlO into
an E. coli ribosomal RNA operon (Morgan, 1980)
and TnlO insertions in the his operon (Ciampi et al.,
1982; Schmid and Roth, 1983) have been shown to
be incompletely polar.
These examples of lack of Tn5 polarity, or expres-
sion of genes within an operon downstream of the
Tn5 insertion site, can be explained in a variety of
different ways. First, one can invoke the presence of
a low-level promoter, responsible for the expression
of the gene(s) downstream of the Tn5 insertion site,
whose action is not observed when the operon is
present in single copy in the cell. Alternatively, the
supercoiled nature of multicopy plasmids carrying
cloned regions of interest may result in elevated,
randomly initiated transcription of plasmid DNA
sequences, allowing sufficient expression of a gene
carried on the plasmid, to observe genetic comple-
mentation. Third, the absence of polarity associated
with TnS-induced insertion mutations in operons
carried on multicopy plasmids may be the result of
a paucity of factors such as rho that are essential for
Tn-mediated termination of transcription (see
Kleckner, 198 1). Fourth, in some operons there may
exist recognition sequences for an endogenous anti-
termination factor. Finally, Tn5 itself may carry
minor promoters within its inverted repeats (Berg
141
et al., 1980), or insertion of Tn5 DNA sequences
may result in some cases in the creation of a
fortuitous promoter at the Tn5 : : target junction
site, which would result in expression of the down-
stream gene(s). The fact that transcription originat-
ing from “outward-reading” promoters carried by
the Tn element is usually not detectable, may be due
to termination of such messages by the rho factor;
only when the transposon is located so near to a
ribosome-binding site that rho-dependent termi-
nation is prevented will expression of downstream
genes be observed (Schmid and Roth, 1983).
Regardless of the explanation for the lack of Tn5
polarity in some of the cases, caution ought to be
exercised when using Tn5 mutagenesis of operons
carried on multicopy plasmids to determine the
genetic organization of such cloned regions. Tn5-
induced insertion mutations will generally exert a
strongly polar effect on downstream genes within the
same operon, but this effect may be masked by
genetic “artifacts” introduced by the use of multicopy
plasmids.
(f) Stability of TnS-induced insertion mutations
Tn elements frequently induce DNA rearrange-
ments, such as deletions, duplications and inver-
sions, usually with one endpoint at the site of inser-
tion (for a review see Kleckner, 1981). This pheno-
menon would complicate any interpretation of re-
sults obtained with Tn-induced insertion mutations.
In addition, secondary transposition events, result-
ing in the insertion of a copy of the Tn element at a
new site in the genome of the cell or into any plasmid
it may harbour, would also complicate the interpre-
tation of results obtained from transposon muta-
genesis experiments. It is therefore important to con-
sider the stability of TnS-induced insertion mutations
in our experimental system. Regarding TnS-induced
DNA rearrangements, either upon insertion or sub-
sequently, we have observed gross alterations of the
target plasmid DNA sequences in less than 1% of
the TnS-containing plasmids from any given Tn5-
mutagenesis experiment, as determined by restric-
tion analysis. When strains harboring TnS-mutated
plasmids are stored in agar stabs in the presence of
kanamycin, l-5% of the plasmid DNAs undergo
DNA rearrangements. We therefore usually store
our plasmid-containing strains as glycerinated cul-
tures at - 20” C or - 70” C and use the antibiotic ing out gene-replacement experiments (Ruvkun and
resistance marker carried by the plasmid vector for Ausubel, 1981) and even incorporating TnS-
selection when growing up fresh cultures for genetic containing DNA sequences into the genome of
complementation tests or plasmid DNA prepara- higher plants via the A. turnefaciens infection system
tions. This reduces the frequency of rearrangements (Gartinkel et al., 1981; Matzke and Chilton, 1981;
of plasmid DNA sequences to less than 1’;. Kahl and Schell, 1982).
Restriction analysis of TnS-containing plasmids of
course does not rule out the possibility of small DNA (g) Use of TnS-induced insertion mutations to identify
rearrangements at the site of insertion. However, proteins and to determine the direction of transcrip-
DNA sequencing of the TnS-target junction se- tion of cloned genes
quences (Lupski et al., 1983a) in nine independent
isolates has shown that not one single base was Tn5 contains translational stop signals in all three
deleted at the insertion and in each case 9 bp of reading frames within the terminal 30 bp of its
target sequence were duplicated as has been pre- inverted repeats (Auerswald et al., 1981; Giphart-
viously shown (Auerswald et al.. 1981; Bossi and Gassler et al., 1982). Therefore, Tn5 insertions can
Ciampi, 198 1). be used to identify gene products and to determine
Regarding the occurrence of secondary Tn5 trans- the direction of transcription. The polypeptides
position events, or the presence of two independent synthesized by plasmids can be examined for
Tn5 insertions into a single plasmid, we have rarely example in maxicells (Sankar et al., 1979) or mini-
encountered any such cases. However, when analys- cells (Reeve, 1979). Since TnS-mutated genes will
ing TnS-induced chromosomal insertion mutations direct the synthesis of truncated polypeptides, a cor-
in K. pneunzoniae, strains carrying two or more relation between the physical location of Tn5 inser-
copies of Tn5 have been encountered after con- tions and the observed IV,. of the truncated poly-
tinuing selection for Km’ (Meade, H., Brown, S., De peptides yields information regarding the direction of
Bruijn, F. and Ausubel, F., unpublished observa- transcription of the cloned gene and the approximate
tions). This phenomenon may be due to increased translational start site. This methodology has been
Tn5 transposition frequencies and/or altered regu- employed to identify the products of genes such as
lation of Tn5 transposition in this organism (see also the bacteriophage PI yes gene (Heilmann et al.,
Merrick et al., 1980). A slight tendency of Tn5 to 1980), the phage Mu S and Ugenes (Giphart-Gassler
transpose from its original site of insertion to a new et al., 1982) the nifgene clusters of K. pneumoniae.
locus has also been observed in E. coli (Harayama R. rneliloti and R. japonicum (Puhler and Klipp,
et al., 1979). However, generally Tn5 appears to reg- 1981; Pi.ihler et al., 1983; Fuhrmann and Hennecke,
ulate its own transposition with an efficient, trans- 1982) the CloDF13 Hgene (Hakkaart et al., 1981)
acting negative regulator of transposition of the resi- and the K. pneumoniae glnF(ntrA) gene (De Bruijn
dent, or an incoming Tn5 (Biek and Roth, 1980; and Ausubel, 1983). The only problem encountered
Reznikoff, 1982; Berg, 1983). appears to be the relative instability of truncated
Thus, even though some reports of instability of polypeptides, which is observed especially in maxi-
TnS-induced insertion mutations above the observed cells (De Bruijn, F., unpublished observations).
reversion frequency of lo’-10” (Berg, 1977) exist,
such mutations are generally stable, nonleaky, null (h) Tn.5 mutagenesis of cosmid derivatives
mutations and usually do not generate extensive
DNA rearrangements or secondary transpositions. We have observed some anomalous results while
The stability is sufficient to allow transductional mutagenizing derivatives of the pLAFR1 cosmid
mapping in a variety of bacterial species (see vector (Friedman et al., 1982) with Tn5 using the
Kleckner et al., 1977; Forrai et al., 1983; Shana- /1: : Tn5 phage. The pLAFR1 vector consists of the
bruch and Walker, 1981; Harayama et al., 1979; low-copy-number pRK290 replicon (Tc’; Ditta
Fouts and Barbour, 1981; Kuner and Kaiser, 1981) et al., 1980) and a DNA fragment carrying the A cos
re-cloning of Tn5 containing DNA sequences by site. When RecA- E. co/i cells harboring a
Km’ selection (see EXPERIMENTAL. section i), carry- pLAFRl-derivative plasmid were infected with

I : : Tn5, the resulting Km’ colonies were shown to
contain mostly “cointegrate” plasmids between the
pLAFR1 derivative and the defective I : : Tn5
genome, as opposed to simple, Tn.5 containing,
pLAFR1 derivatives. Plasmid DNA prepared from
such colonies failed to transform competent cells of
E. coli to Tc’ and Km’; moreover, extensive killing of
the recipient cells was observed (De Bruijn, F. and
Ausubel, F., unpublished observations). We postu-
lated that the apparent recombination event between
the pLAFR1 derivative and the 1: : Tn5 phage
genome, taking place in a RecA - strain, might be
due to ATer-mediated recombination at the I cos
sites of the pLAFR1 and the 1: : Tn5. The observed
inability to obtain Tc’ Km’ transformants with the
resulting cointegrate plasmid DNA could be ex-
plained by zygotic induction of the 2 upon entry into
the recipient cell, resulting in killing of such cells. We
tested this hypothesis by doing the initial 2 : : Tn5
infection experiment in an E. coli strain lysogenic for
2, and observed a dramatic increase in genuine Tn5
transpositions (resulting in a pLAFR1 : : Tn5 plas-
mid), presumably due to repression of ATer-
mediated cos-cos recombination. Thus when muta-
genizing cosmid derivatives using the ,? : : Tn5 phage
vector, it may be necessary to carry out the initial
infection experiment using a strain capable of sup-
pression of A functions, and harboring the cosmid.
(i) Use of Tn5 mutagenesis to clone genes of interest
that do not have a readily selectable phenotype
When a gene is difficult to clone, either because the
only mutants available are conditional lethal mutants
that display leaky phenotypes or the phenotype is
one for which there is no easy screening procedure,
then transposition mutagenesis followed by drug
marker selection may be useful. One limitation is that
it must be performed in bacteria in which Tn5 trans-
position occurs. The organism is subjected to
random mutagenesis with Tn5. After mutagenesis a
Km’ colony is isolated which has lost the phenotype
of the gene of interest or in which the Km’ marker
is found closely linked to the gene. The DNA from
this strain is then shotgun-cloned into a plasmid
vector, such as pBR322, competent cells are trans-
formed, and Km’ transformants are selected. If the
original mutagenized strain contained a Tn5 insert
into the gene of interest then Km’ recombinant
143
clones should contain a portion of that gene. This
recombinant is then used as a probe to screen a
library that contains wild-type sequences and geno-
mic DNA fragments carrying the gene of interest are
isolated.
The absence of an EcoRI restriction site in Tn5
facilitates this procedure. EcoRI restriction frag-
ments containing TnS-mutagenized genes are easily
cloned either by insertional inactivation of the Cm’
gene of pBR325 or by using one of the more recently
developed direct-selection vectors. In particular, we
have found pKY2700 (Ozaki et al., 1982) very useful.
This plasmid contains the colicin El structural gene
with a unique EcoRI site within the gene. The colicin
El gene is inducible by the SOS system (Ebina et al.,
1982). Therefore, transformation into a LexA-
strain kills all cells except those which contain a
plasmid carrying a defective colicin El gene, due to
insertional inactivation by a cloned EcoRI restriction
fragment (Ozaki, L., personal communication).
The above approach has already been used
successfully by a number of investigators. A Cuufo-
batter crescentus gene cluster specifying flagellum
production has been isolated by Purucker et al.
(1982), using nonmotile TnS-insertion mutants. Iso-
lation of the E. coli tolC gene was performed by
Morona and Reeves (1982) in a similar manner using
a TnlO-insertion mutant and selecting for Tc’.
(j) Tn5 mutagenesis analysis of eukaryotic genes
In vivo transposition mapping with Tn5 is not
limited to the study of prokaryotic gene structure and
function. The only requirement for this methodology
is that a piece of DNA expresses a measurable
phenotype in bacteria. With the advent of cDNA
cloning and nucleotide synthesis many eukaryotic
genes have been cloned and expressed in E. coli. One
way of detecting phenotypic expression involves the
use of monoclonal antibodies directed against the
protein product of these cDNA recombinant clones.
Transposition mapping with Tn5 may be used to
further localize on a cloned DNA fragment the
genetic information encoding an enzymatic activity
or antibody reactivity. Thus, this may be a way of
rapidly determining an immunodominant region of a
protein.
This is particularly applicable to synthetic vac-
cines, either bacterial, viral or parasitic. Here the
144

gene for a surface antigen protein involved in the their ability to complement the mammalian cell
infectious process is first cloned, sequenced, and marker. Theoretically, clusters of Tn5 insertions into
expressed in bacteria. Then, rather than synthesize a the cloned fragment should be found which are still
number of small peptides, the immunodominant able to complement the defective mammalian cells,
region of this surface antigen protein is identified and these would correspond to Tn5 insertions within
using transposition mapping and a monoclonal anti- intervening sequences that were spliced out during
body to the natural constituent as a probe. Cells RNA processing.
harboring plasmids with different Tn5 insertion
mutations are lysed and the lysates are screened for (k) Physical mapping of plasmid-borne transposon
their ability to bind monoclonal antibody in a RIA. Tn5 insertions
Insertion mutations which destroy the ability of the
monoclonal antibody to bind lysates identify a region A number of convenient restriction sites are avail-
of DNA that may encode antigenic portions of the able for the physical mapping of Tn5 insertions in
protein molecule. multicopy plasmids (see Fig. 2). When mapping Tn5
This approach has been applied to delineate an insertions in a large segment of cloned DNA, it is
immunodominant region of a surface antigen of the often convenient to first determine the approximate
malarial parasite Plasmodium knowlesi which is location of the insertions. For this purpose the
responsible for causing malaria in monkeys (Lupski restriction endonucleases which do not cleave Tn5
et al., 1983b). A cDNA clone, which expressed a are useful (see Fig. 2). For example, EcoRI was used
plactamase fusion polypeptide in E. coli that reacted to determine the approximate location of the Tn5
with a monoclonal antibody to the parasite surface insertions shown in Fig. 4. Since EcoRI does not
antigen (Ellis et al., 1983), was mutagenized with cleave Tn5, the presence ofTn5 in any of the pFB5 14
Tn5, and lysates were screened for their ability to EcoRI fragments leads to the disappearance of the
bind antibody. A few insertion mutations destroyed corresponding band on the gel (X bp) and the
the ability of the lysates to bind antibody, and appearance of a new band corresponding to a DNA
restriction analysis of these Tn.5-mutated plasmids size of X + 5700 bp (De Bruijn and Ausubel,
identified within the cDNA the approximate region 1981). For precise mapping, the restriction endo-
coding for the epitope. Thus, the above experiment nucleases that cleave within the IRS of Tn5 are most
demonstrates that any gene, prokaryotic or useful (see Fig. 2). These sites allow orientation-
eukaryotic, can be mapped by this method provided independent localization of Tn5. A restriction endo-
a phenotype is readily detectable in bacteria. nuclease from this group which cleaves the original
The insertion of Tn5 into cloned eukaryotic DNA plasmid once (or twice), provides a reference point
sequences has also been used to introduce mutations relative to which the Tn5 insertions will be mapped.
whose effects are observable as an altered phenotype The TnS-containing plasmids are cleaved with the
in mammalian cells. A cloned DNA fragment desired restriction endonuclease and the fragments
containing the human adenovirus type 5 transform- run on a gel. An example is shown in Fig. 3, where
ing gene was subjected to Tn5 mutagenesis in E. coli Hind111 was used to map the Tn5 insertions. It
(McKinnon et al., 1982). Individual insertion mu- cleaves the parental plasmid once and Tn5 twice,
tants were transfected into mammalian cells and within the IRS, 1195 bp away from the ends of the
assayed for their ability to transform. In this manner transposon (see Fig. 2; Jorgenson et al., 1979;
adenovirus sequences essential for DNA-mediated Auerswald et al., 1981). Thus, three fragments on the
transformation were identified. TnS-containing plasmids are generated by cleavage
One might also be able to map intervening sequen- with Hind111 (see Fig. 3). By measuring the sizes of
ces of eukaryotic genes by Tn5 mutagenesis. A the smallest junction fragments and subtracting
cloned genomic DNA fragment able to complement 1195 bp (Tn5 DNA sequence), the distance of that
some mutation in mammalian cells would be sub- Tn5 insertion to the reference site can be calculated,
jected to Tn5 mutagenesis in E. coli. The individual and a physical map constructed (see Fig. 4 for an
insertion mutations would then be transfected into example). To determine the relative orientation of
the appropriate mammalian cells and checked for Tn5 on the plasmid, the Sal1 and SmaI sites within

Tn5 can be used since these sites are located asym-
metrically within the transposon (see Fig. 2).
The physical location of a gene as determined by
transposon T&mutagenesis mapping correlates
very well with the actual location as determined by
nucleotide sequence analysis. For example, the
E. coli dnaG gene was physically mapped by (1) the
isolation of multicopy plasmids containing restric-
tion fr~ents that could complement a condi-
tionally lethal dnaGts chromosomal marker at non-
permissive 4O”C, followed by (2) TnS-mutagenesis
mapping of these recombinant plasmids and (3)
genetic analysis of individual Tn5 inserts into dnaG-
cont~~g plasmids harbored in a dvraGts
background at 30°C and 40°C (Lupski et al., 1982).
This correlated physical and genetic map was sub-
sequently shown to be within 50 bp of the actual
location for dnaG as determined by nucleotide
sequence analysis (Smiley et al., 1982).
(I) Use of transposon Tn5 mutagenesis in directed
nucleotide sequencing
Tn5 insertion mutations obtained in thegeneration
of correlated physical and genetic maps can also be
very useful for determining the nucleotide sequence
of a particular DNA segment. Having mapped an
individual Tn5 insertion one can take a directed, as
opposed to a “shotgun”, approach to nucleotide
sequencing. The presence of a unique HpaI site
185 bp from the end of Tn5 (Auerswald et al., 198 1)
is very helpful for this purpose. A plasmid containing
Tn5 in a known position is digested with @aI and
another enzyme, X (X = BarnHI, SalI, Pst I, IiindIII
or any enzyme that leaves blunt ends), which cuts the
cloned DNA. This fragment mixture is ligated into
M13mp9 (Messing et al., 1981), digested with SmaI
plus enzyme X, competent cells are transformed and
transformants screened by hybridization with a
plasmid probe that contains only the terminal 185 bp
of Tn5 to identify subclones of interest. DNA
sequence analysis using the Sanger et al. (1977)
dideoxy technique is then carried out by priming of
the Ml3mp9 recombinants with synthetic Ml3
primer. Extension of the primer with DNA poly-
merase I (Klenow fra~ent) yields DNA sequences
that extend through known Tn5 sequence into
unknown sequence. This “unknown sequence” cor-
responds to the DNA into which the Tn5 originally
145
inserted. Thus, since the physical location of the Tn.5
is known, this locates within the Tn.5 mapped DNA
fragment, the nucleotide sequence obtained. The
method can be further generalized with the synthesis
of a synthetic primer consisting of the terminal 15-20
nucleotides of Tn5.
CONCLUSIONS
In this paper we have reviewed and examined a
number of parameters of transposon Tn5 muta-
genesis as applied to cloned DNA segments carried
by multicopy plasmids, such as transposition fre-
quency and insertional specificity of Tn5, stability
and polarity of TnS-induced insertion mutations and
Tn.5 carrying vectors.
Based on the physical mapping of at least 500
independent Tn5 insertions in seven different cloned
DNA segments we have carried out, and similar
experiments carried out in a number of other labora-
tories, we conclude that the insertional specificity of
Tn.5 is very low; no gross “insertional hot-spots”
were found, and we did not encounter difficulties
generating large collections of well-distributed Tn5
insertions in any of the genes and operons examined.
We have found the A: : Tn5 vector (Berg et al.,
197.5; Berg, 1977; Kleckner, N., unpublished obser-
vations) to be extremely useful in the generation of
Tn5-induced insertion mutations in cloned DNA
segments carried by multicopy plasmids, due to the
efficient recovery of cells in which Tn5 transposition
and concomitant loss of the 1: : Tn5 vector DNA
sequences have occurred.
We have found the physical mapping of Tn5-
induced insertion mutations in multicopy plasmids
to be greatly facilitated by the occurrence of a number
of convenient restriction sites in Tn5, useful for the
determination of the physical location of Tn5
(F 100 bp), the determination of the relative orien-
tation of the Tn5 and the determination of the
nucleotide sequence of the TnS-target junction sites.
Moreover, the availability of rapid, small-scale plas-
mid DNA preparation protocols makes possible the
physical mapping of large numbers of plasmid-bone
Tn5 insertions with relative ease.
Lastly, we have found plasmid-borne TnS-induced
insertion mutations to be essentially stable and there-
fore useful for a number of purposes other than leucine-valine auxotrophs in Escherichia coli K-12: Evidence

physical and genetic mapping, such as subcloning for an internal promoter in the ilvOGEDA operon. Genetics
93 (1979) 309-319.
and gene-replacement experiments, and eventually
Berg, D.E.: Insertion and excision of the transposable kanamycin
for directed DNA sequencing of cloned genes. determinant Tn5, in: Bukhari, AI., Shapiro, J.A. and Adhya,
Thus we conclude that the Tn5 mutagenesis proto- S.L. (Eds.), DNA Insertion Elements, Plasmids and Epi-
col, described in MATERIALS AND METHODS, somes, Cold Spring Harbor Laboratory, Cold Spring Harbor.
section d, and discussed above, is extremely useful in NY, 1977. pp. 205-212.
Berg. D.E.: Control of gene expression by a mobile recombi-
the rapid generation of correlated physical and
national switch. Proc. Natl. Acad. Sci. USA 77 (1980)
genetic maps of cloned DNA segments carried by
4880-4884.
multicopy plasmids. As such, it constitutes an Berg. D.E.: Structural requirement for ISSO-mediated gene
important step in the analysis of cloned DNA seg- transposition. Proc. Natl. Acad. Sci. USA 80 (1983) 792-796.
ments by allowing a rapid identification of the loca- Berg, D.E. and Berg. CM.: The prokaryotic transposable ele-

tion, approximate boundaries and genetic as well as ment Tn5. Biotechnology 1 (1983) 417-435.
Berg, D.E.. Davies, J., Allet, B. and Rochaix, J.D.: Transposition
transcriptional organization of cloned genes and
of R factor genes to bacteriophage lambda. Proc. Natl. Acad.
operons. Sci. USA 72 (1975) 3628-3632.
Berg, DE., Egner, C. and Lowe, J.B.: Mechanism of F factor-
enhanced excision of transposon Tn5. Gene 22 ( 1983) 1-7.
Berg, D.E., Jorgensen, R. and Davies. J.: Transposable kanamy-
tin-neomycin resistance determinants. in Schlessinger, 0.
ACKNOWLEDGEMENTS
(Ed.). Microbiology - 1978, American Society for Microbiol-
ogy. Washington, DC, 1978, pp. 13-15.
We would like to thank all those individuals who Berg, D.E., Weiss, A. and Crossland, L.: Polarity ofTnS insertion
have helped us by discussing and partaking in our mutations in Escherichiu coli. J. Bacterial. 142 (1980) 439346.
Tn.5 mutagenesis experiments. In particular we Beringer. J.E., Beynon, J.L.. Buchanan-Wollaston, A.V. and
Johnston, A.W.B.: Transfer of the drug-resistance trsns-
thank Drs. Fred Ausubel, Peter D’Eustachio, G.
poson Tn5 to Rhkobium. Nature 276 (1978) 633-634.
Nigel Godson and Luiz Ozaki for helpful dis- Bevan, M.W. and Chilton, M.-D.: T-DNA of the ilgwbac.kWm
cussions and Dr. Ahmad I. Bukhari for discussing Ti and Ri plasmids. Annu. Rev. Genct. 16 (1982) 357-384.
and reviewing the m~usc~pt. This work was sup- Biek, D. and Roth. J.R.: RegulatioI~ of TnS transpositi~~xl in
ported by NSF Grant PAM-8104193 to Fred M. ~u~~~o~~ell~ f~phiFrzii~iuf~7. Proc. Natl. Acad. Sci. USA 77

Ausubel and NIAID Grant 7-l 142-996 to G. Nigel (1080)6047-6051.

Birnboim. H.C. and Doiy, J.: A rapid extraction procedure for
Godson. JRL is a recipient of an N.I.H, Medical
screening recombinant plasmid DNA. Nucl. Acids Res. 7
Scientist in Training Grant. (1979) lS13-1523.
Bolivar, F.. Rodriguez, R.L.. Greene. P.J.. Betlach. M.C.,
Heynecker, H.L., Boyer, H.W., Cross, J.H. and Falkow, S.:
Construction and characterization of new cloning vehicles,
II. A multipurpose cloning system. Gene 2 ( 1977) 9%113.
REFERENCES Bossi, L. and Ciampi. M.S.: DNA sequence of the sites of three
insertions of the transposable element Tn5 into the histidine
Auerswald, E.A., Ludwig, G. and Schaller, H.: Structural ana- operon of ~alrn~~ze~lu. Mol. Gen. Genet. 183 (1981) 406-408.
lysis of Tn5. Cold Spring Harbor Symp. Quant. Biol. 45 Bricker. J., Mulks, M.H., Plaut, A.G., Moxon. E.R. and Wright,
(1981) 107-I 13. A.: IgAl proteases of ~ae~zo~~jl~s j~?~~e~~ffe: cloning and
Ausubel, F.M.: Molecular genetics ofsymbiotic nitrogen fixation. characterization in Escherichia coli K-12. Proc. Natl. iicad.
Cell 29 (1982) l-2. Sci. LISA 80 (1983) 2681-2685.
Beck. E., Ludwig, G., Auerswald, E.A., Reiss. B. and Schaller, Buchanan-Wollaston, A.V.. Beringer, J.E.. Brewin, N.J.. Hirsch.
H.: Nucleotide sequence and exact localization of the P.R. and Johnston, A.W.B.: Isolation of symbiotically dcfec-
neomycin phosphotransferase gene from transposon Tn5. tive mutants in Rhirobium leguminosanmz by insertion of the
Gene 19 (1982) 327-336. transposon Tn5 into a transmissable plasmid. Mol. Gcn.
Belas, R., Mileham, A.. Cohn, D., Hilmer, M., Simon, M. and Genet. 178 (1980) 185-190.
Silverman, M.: Bacterial bioluminescence: isolation and Ciampi. M.S., S&mid, M.B. and Roth. J.R.: Transposon Tnfo
expression of the luciferase genes from Vibrio Harvey?. Science provides a promoter for transcription of adjacent sequences.
218 (1982) 791-793. Proc. Natl. Acad. Soi. US4 79 (1982) 5016-5020.
Berg. C.M., Shaw, K.J., Vender, J. and Borucha-.M~kie~vicz. Cleweli. D.B. and Hehnski. D.R.: Supercoiled circular DNA-
M.: Physiological characterization of polar TnZ-induced iso- protein complex in ~~che~~~,hj~ coli: Purification and induced
 147

conversion to an open circular form. Proc. Nat]. Acad. Sci. cloning vector and its use in the genetic analysis of Rhizobium
USA 62 (1969) 1159-1166. mutants. Gene 18 (1982) 289-296.
Corbin, D., Barran, L. and Ditta, G.: Organization and expres- Fuhrmann, M. and Hennecke, H.: Coding properties of cloned
sion of Rhizobium meliloti nitrogen fixation genes. Proc. Natl. nitrogenase structural genes from Rhizobium japonicum. Mol.
Acad. Sci. USA 80 (1983) 3005-3009. Gen. Genet. 187 (1982) 419-425.
Corbin, D., Ditta, G. and Helinski, D.R.: Clustering of nitrogen Gantotti, V.B., Kindle, K.L. and Beer, S.V.: Transfer of the
fixation (nif) genes in Rhizobium meliloti. J. Bacterial. 149 drug-resistance transposon Tn5 to Erwinia herbicola and the
(1982) 221-228. induction of insertion mutations. Curr. Microbial. 6 (1981)
De Bruijn, F.J.: Regulation of the Klebsiellupneumoniae nitrogen 377-381.
fixation (nr$) genes by the ntr genes. Ph.D. Thesis, Harvard Gartinkel, D.J. and Nester, E.W.: Agrobacterium tumefaciens
University, Cambridge, MA, 1983. mutants affected in crown gall tumorigenesis and octopine
De Bruijn, F.J. and Ausubel, F.M.: The cloning and transposon catabolism. J. Bacterial. 144 (1980) 732-743.
Tn5 mutagenesis of the gbtA region of Klebsiella pneumoniae: Gartinkel, D.J., Simpson, R.B., Ream, L.W., White, F.F.,
Identification ofglnR, a gene involved in the regulation of the Gordon, M.P. and Nester, E.W.: Genetic analysis of crown
nifand hut operons. Mol. Gen. Genet. 183 (1981) 289-297. gall: tine structure map of the T-DNA by site directed muta-
De Bruijn, F.J. and Ausubel, F.M.: The cloning and charac- genesis. Cell 27 (1981) 143-153.
terization of the glnF(ntrA) gene of Klebsiella pneumoniae: Giphart-Gassler, M., Plaskerk, R.H.A. and Van de Putte, P.:
Role of glnF[ntrA) in the regulation of nitrogen fixation (nil) G-inversion in bacteriophage Mu: a novel way of gene splic-
and other nitrogen assimilation genes. Mol. Gen. Genet. ing. Nature 297 (1982) 339-342.
192(1983)342-353. Godson, G.N.. Ellis, J., Svec, P., Schlessinger, D.H. and Nussen-
De Bruijn, F.J., Stroke, I.L., Marvel, D.J. and Ausubel, F.M.: zweig, V.: Identification and chemical synthesis of an epitope
Construction of a correlated physical and genetic map of the of Plusmodium knowlesi circumsporozoite protein. Evidence
Klebsiella pneumoniae hisDG0 region using transposon Tn5 for its tandemly repeated nature. Nature 305 (1983) 29-33.
mutagenesis. EMBO J. 2 (1983) 1831-1838. Guarente, L.P., Isberg, R.R., Sylvanen, M. and Silhavy, T.J.:
Ditta, G., Stantield, S., Corbin, D. and Helinski, D.R.: Broad Conferral of transposable properties to a chromosomal gene
host range DNA cloning system for gram-negative bacteria. in Escherichiu coli. J. Mol. Biol. 141 (1980) 235-248.
Construction of a gene bank of Rhizobium meliloti. Proc. Natl. Hakkaart, M.J.J., Veltkamp, E. and Nijkamp, H.J.J.: Protein H
Acad. Sci. USA 77 (1980) 7347-7351. encoded by plasmid CloDF 13 involved in lysis ofthe bacterial
Duncan, M.J.: Properties of T&induced carbohydrate mutants host, I. Localization of the gene and identification and sub-
in Rhizobium meliloti. J. Gen. Microbial. 122 (1981) 61-67. cellular localization of the gene H product. Mol. Gen. Genet.
Ebina, Y., Kishi, F. and Nakasawa, A.: Direct participation of 183 (1981) 318-325.
LexA protein in repression of colicin El synthesis. J. Bac- Harayama, S., Palva, E.T. and Hazelbauer, G.L.: Transposon
teriol. 150 (1982) 1479-1481. insertion mutants of Eschenchia coli K-12 defective in a com-
Egner, C. and Berg, D.E.: Excision of transposon Tn.5 is depen- ponent common to galactose and ribose chemotaxis. Mol.
dent on the inverted repeats but not on the transposase Gen. Genet. 171 (1979) 193-203.
function of Tn5. Proc. Nat]. Acad. Sci. USA 78 (1981) Heilmann, H., Burkardt, H.J., Puhler, A. and Reeve, J.N.: Trans-
459-463. poson mutagenesis of the gene encoding the bacteriophage
Ellis, J., Ozaki, L.S., Gwadz, R.W., Cochrane, A.H., Nussen- Pl restriction endonuclease. J. Mol. Biol. 144 (1980) 387-396.
zweig, V., Nussenzweig, R.S. and Godson, G.N.: Cloning and Holmes, D.S. and Quigley, M.: A rapid boiling method for the
expression in E. coliofthe malarial sporozoite surface antigen preparation of bacterial plasmids. Anal. Biochem. 114 (1981)
gene from Plasmodium knowlesi. Nature 302 (1983) 536-538. 193-197.
Ely, B. and Croft, R.H.: Transposon mutagenesis in Caulobacter Hooykaas, P.J.J., Van Brussel, A.A.N., Den Dulk-Ras, H., Van
crescentus. J. Bacterial. 149 (1982) 620-625. Slogteren, G.M.S. and Schilperoort, R.A.: Sym plasmid of
Engebrecht. J., Nealson, K. and Silverman, M.: Bacterial bio- Rhizobium trifoliiexpressed in different rhizobial species and
luminescence: isolation and genetic analysis of functions Agrobacterium tumefaciens. Nature 291 (1981) 35 l-353.
from VibrioJischeri. Cell 32 (1983) 773-781. Ish-Horowitz, D. and Burke, J.F.: Rapid and efficient cosmid
Fitts, R.A. and Taylor, A.L.: Integration of bacteriophage Mu at cloning. Nucl. Acids Res. 9 (1981) 2989-2998.
host chromosome replication forks during lytic development. Johnson, R.C., Yin, J.C.P. and Reznikoff, W.S.: Control of Tn5
Proc. Natl. Acad. Sci. USA 77 (1980) 2801-2805. transposition in Escherichiu coli is mediated by protein from
Forrai. T., Vincze, E., Banfalvi, Z., Kiss, G.B., Randhawa, G.S. the right repeat. Cell 30 (1982) 873-882.
and Kondorosi, A.: Location of symbiotic mutations in Rhizo- Johnston, A.W.B., Beynon, J.L., Buchanan-Wollaston, A.V.,
bium meliloti. J. Bacterial. 153 (1983) 635-643. Setchell, S.M., Hirsch, P.R. and Beringer, J.E.: High fre-
Fouts, K.E. and Barbour, S.D.: Transductional mapping of ksgB quency transfer of nodulating ability between strains and
and a new T&induced kasugamycin-resistance gene, ksgD, species of Rhizobium. Nature 276 (1978) 634-636.
in Escherichiu coh K-12. J. Bacterial. 145 (1981) 914-919. Jorgensen, R.A., Rothstein, S.J. and Reznikoff, W.S.: A restric-
Friedman, A.M., Long, S.R., Brown, S.E., Buikema, W.J. and tion enzyme cleavage map of Tn5 and location of a region
Ausubel, F.M.: Construction of a broad host range cosmid encoding neomycin resistance. Mol. Gen. Genet. 177 (1979)
65-72.
14x

Kahl, G. and &hell, J.S.: Molecular Biology of Plant Tumors, McKinnon, R.D., Bacchetti, S. and Graham, F.L.: Tn5 muta-
Academic Press. New York, 1982. genesis of the transforming genes of human adenovirus type
Kleckner, N.: Transposable elements in prokaryotes. Annu. Rev. 5. Gene 19 (1982) 33-42.
Genet. 15 (1981) 341-404. Meade, H.M., Long, S.R.. Ruvkun, G.B., Brown, S.E. and
Kleckner, N., Chan, R.K., Tye, B.K. and Botstein, D.: Muta- Ausubel, F.M.: Physical and genetic characterization of
genesis by insertion of a drug resistance element carrying an symbiotic and auxotrophic mutant of Rhizobium meliloti
inverted repetition. J. Mol. Biol. 97 (1975) 561-575. induced by transposon Tn5 mutagenesis. J. Bacterial. 149
Kleckner, N., Roth, J. and Botstein, D.: Genetic engineering in (1982) 114-122.
vivo using translocatable drug-resistance elements: New Merrick, M., Filser. M., Dixon, R.. Elmerich, C., Sibold. L. and
methods in bacterial genetics. J. Mol. Biol. 116 (1977) Houmard, J.: The use of translocatable genetic elements to
125-159. construct a line-structure map of the Klebsiellu pneumoniae
Kleckner, N., Steele, D.A., Reichardt, K. and Botstein, D.: nitrogen fixation (nif) gene cluster. J. Gen. Microbial. 117
Specificity of insertion by the translocatable tetracycline (1980) 509-520.
resistance element TnlO. Genetics 92 (1979) 1023-1040. Merrick, M., Filser, M., Kennedy, C. and Dixon, R.: Polarity of
Klee. H.J., White, F.F.. Iyer, V.N., Gordon, M.P. and Nester. mutations induced by insertion of transposons Tn5, Tn7. and
E.W.: Mutational analysis of the virulence region of an TnlO into the nifgene cluster of Klebsiella pneumoniue. Mol.
Agrobacterium fumefaciens Ti plasmid. J. Bacterial. 153 (1983) Gen. Genet. 165 (1978) 103-111.
878-883. Messing, J., Crea, R. and Seeburg, P.H.: A system for shotgun
Kuner, K.M. and Kaiser. D.: Introduction of transposon Tn.5 DNA sequencing. Nucl. Acids Res. 9 (1981) 309-321.
into M~.XOCOCCUS
for analysis of developmental and other Meyer, R., Both, G. and Shapiro, J.A.: Transposition of DNA
nonselectable mutants. Proc. Natl. Acad. Sci. USA 78 (1981) inserted into deletions of the Tn5 kanamycin resistance
425429. element. Mol. Gen. Genet. 171 (1979) 7-13.
Laird, A.J. and Young, LG.: Tn5 mutagenesis of the enterocholin Miller, J.H.: Experiments in Molecular Genetics. Cold Spring
gene cluster of Escherichia coli. Gene 11 (1980) 359-366. Harbor Laboratory, Cold Spring Harbor, NY. 1972.
Lamb, J.W., Hombrecher. G. and Johnson, A.W.B.: Plasmid- Miller, J.H., Calos, M.P., Galas. D., Hofer. M., Biichel. D.E. and
determined nodulation and nitrogen-fixation abilities in Rhi- Miiller-Hill. B.: Genetic analysis of transposition in the luc
zobium phaseoli. Mol. Gen. Genet. 186 (1982) 449-452. region of Escherichiu coli. J. Mol. Biol. 144 (1980) 1-18.
Leong, S.A., Ditta. G.S. and Helinski. D.: Heme biosynthesis in Morgan, E.A.: Insertions ofTnJ0 into an E. coli ribosomal RNA
Rhbobium identification of a cloned gene coding for &amino operon are incompletely polar. Cell 21 (1980) 257-265.
levulinic acid synthetase from Rhizobium meliloti. J. Biol. Morona, R. and Reeves, P.: Molecular cloning of the tolC locus
Chem. 257 (1982) 8724-8730. ofEscherichia coli K-l 2 with the use of transposon TnlO. Mol.
Lupski. J.R.. Gershon, P., Ozaki, L.S. and Godson, G.N.: Gen. Genet. 184 (1981) 430-434.
Specificity of Tn5 insertions into a 36 nucleotide DNA Ozaki, L.S.. Kimura. A.. Shimada, K. and Takagi. Y.: ColEl
sequence repeated in tandem eight times. (1983a), submitted. vectors for direct selection of cells carrying a hybrid plasmid.
Lupski, J.R., Ozaki, L.S.. Ellis, J. and Godson. G.N.: Locali- J. Biochem. 91 (1982) 1155-1162.
zation of the Plusmodium surface antigen epitope by Tn5 Palva. E.T., Liljestrom. P. and Harayama, S.: Cosmid cloning
mutagenesis mapping of a recombinant cDNA clone. Science and transposon mutagenesis in Salmonellu typhimurium using
220 (1983b) 1285-1288. phage i vehicles. Mol. Gen. Genet. 181 (1981) 153-157.
Lupski, J.R.. Smiley, B.L. and Godson, G.N.: Regulation of the Pannekoek. H.. Hille, J. and Noordermeer, I.: Relief of polarity
rpslJ-dmzG-rpoD macromolecular synthesis operon and the caused by transposon Tn5: application in mapping a cloned
initiation of DNA replication in Escherichiu coli K-12. Mol. region of the Escherichiu coli uvrB locus essential for UV
Gen. Genet. 189 (1983~) 48-57. resistance. Gene 12 (1980) 51-61.
Lupski, J.R., Smiley, B.L., Blattner. F.R. and Godson, G.N.: Phua. S.H., Bergquist, P.L. and Lane. H.E.D.: Effects of Tn.5
Cloning and characterization of the Escherichiu coli chromo- insertion in the incD region on mini-F maintenance and poly-
somal region surrounding the dnaC gene, with a correlated peptide synthesis. Mol. Gen. Genet. 188 (1982) 353-355.
physical and genetic map of dnaG generated via transposon Pilacinski, W., Mosharrafa. E., Edmundson. R.. Zissler, J..
Tn5 mutagenesis. Mol. Gen. Genet. 185 (1982) 120-128. Fiandt, M. and Szybalski. W.: Insertion sequence IS2 asso-
Ma, Q.S., Johnston, A.W.B.. Hombrecher. G. and Downie, J.A.: ciated with int-constitutive mutants ofbacteriophage lambda.
Molecular genetics of mutants of Rhkobium leguminosarum Gene 2 (1977) 61-74.
which fail to fix nitrogen, Mol. Gen. Genet. 187 (1982) Portier. C.: Isolation of a polynucleotide phosphorylase mutant
166-171. using a kanamycin resistant determinant. Mol. Gen. Genet.
Maniatis. T., Fritsch. E.F. and Sambrook, J.: Molecular Cloning, 178(1980)343-349.
A Laboratory Manual. Cold Spring Harbor Laboratory, Cold Piihler, A. and Klipp. W.: Fine structure analysis of the gene
Spring Harbor, NY, 1982. region for N,-fixation (nif ofKlebsiellupneumoniae. in Bothe.
Matzke, A.J.M. and Chilton, M.-D.: Site specific insertion of H. and Trebst, A. (Eds.). Biology of Inorganic Nitrogen and
genes into the T-DNA of the Agrobacterium tumor-inducing Sulfur. Springer Verlag, Heidelberg, 1981, pp. 276-286.
plasmid: An approach to genetic engineering of higher plant Ptihler. A.. Klipp, W. and Weber. G.: Mapping and regulation of
cells. J. Mol. Appl. Genet. 1 (1981) 39-49. the structural genes niJK. nifD and tlifH of Rhirobium meliloti.

in: Ptihler, A. (Ed.), Molecular Genetics ofthe Bacteria-Plant
Interaction. Springer-Verlag, Heidelberg, in press.
Purcell, B.K. and Clegg, S.: Construction and expression of
recombinant plasmids encoding type 1 limbriae of a urinary
Klebsiellu pneumoniue isolate. Infect. Immun. 39 (1983)
1122-l 127.
Purucker, M., Bryan, R., Amemiya, K., Ely, B. and Shapiro, L.:
Isolation of a Cuulobacter gene cluster specifying flagellum
production by using nonmotile Tn5 insertion mutants. Proc.
Natl. Acad. Sci. USA 79 (1982) 6797-6801.
Putnoky, P., Kiss, G.B., Ott, I. and Kondorosi, A.: Tn5 carries
a streptomycin-resistance determinant downstream from the
kanamycin-resistance gene. Mol. Gen. Genet. 191 (1983)
288-294.
Reeve, J.: Use ofminicells for bacteriophage directed polypeptide
synthesis, in Wu, R. (Ed.), Methods in Enzymology, Vol. 68,
Academic Press, New York, 1979, pp. 493-503.
Reznikoff, W.S.: Tn5 transposition and its regulation. Cell 34
(1982) 362-368.
Riedel, G.E.: The use of molecular cloning to study the
organization and expression of the nitrogen fixation (aif)
genes of Klebsiella pneumoniae. Ph.D. Thesis, Harvard Uni-
versity, Cambridge, MA, 1980.
Riedel, G.E., Ausubel, F.M. and Cannon, F.C.: Physical map of
chromosomal nitrogen fixation (nif genes of Klebsiellu
pneumoniae. Proc. Natl. Acad. Sci. USA 76 (1979) 2866-2870.
Rolfe, B.C., Gresshoff, P.M. and Shine, J.: Rapid screening for
symbiotic mutants of Rhizobium and white clover. Plant Sci.
Lett. 19 (1980) 277-284.
Ruvkun, G.B. and Ausubel, F.M.: A general method for site-
directed mutagenesis in prokaryotes. Nature 289 (1981)
85-88.
Ruvkun, G.B., Sundaresan, V. and Ausubel, F.M.: Directed
transposon Tn5 mutagenesis and complementation analysis
ofRhizobium melilotisymbiotic nitrogen fixation genes. Cell 29
(1982) 551-559.
Saedler, H., Reif, H.J., Hu, S. and Davidson, N.: IS2, a genetic
element for turn-off and turn-on ofgene activity in Escherichiu
coli. Mol. Gen. Genet. 132 (1974) 265-269.
Sanger, F., Nicklen, S. and Coulson, A.R.: DNA sequencing with
chain-terminating inhibitors. Proc. Natl. Acad. Sci. USA 74
(1977) 5643-5647.
Sankar, A., Hack, A.M. and Rupp, W.D.: Simple method for
identification of plasmid-coded proteins. J. Bacterial. 137
(1979) 692-693.
Sasakawa, C. and Yoshikawa, M.: Transposon (Tn_5)-mediated
suppressive integration of ColEl derivatives into the chromo-
some ofEscherichia coli K-12 (dnaA). Biochem. Biophys. Res.
Commun. 96 (1980) 1364-1370.
149
Schell, M.A.: Cloning and expression in Escherichiu coli of the
naphthalene degradation genes from plasmid NAH7. J.
Bacterial. 153 (1983) 822-829.
Schmid, M.B. and Roth, J.R.: Internal promoters of the his
operon in Salmonella typhimurium. J. Bacterial. 153 (1983)
1114-1119.
Scott, K.F., Hughes, J.E., Gresshoff, P.M., Beringer, J.E., Rolfe,
B.C. and Shine, J.: Molecular cloning of Rhizobium tn$lii
genes involved in symbiotic nitrogen fixation. J. Mol. Appl.
Genet. 1 (1982) 315-326.
Shanabruch, W.G., Behlau, I. and Walker, G.C.: Spontaneous
mutations of Salmonella typhimurium LT2 generated by
insertion of transposable elements. J. Bacterial. 147 (1981)
827-835.
Shanabruch, W.G. and Walker, G.C.: Localization of the plas-
mid (pKMlO1) gene(s) involved in recA ‘1exA +-dependent
mutagenesis. Mol. Gen. Genet. 179 (1980) 289-297.
Shaw, K.J. and Berg, C.M.: Escherichia coli K-12 auxotrophs
induced by insertion of the transposable element Tn5.
Genetics 92 (1979) 741-744.
Shaw, K.J., Berg, CM. and Sobol, T.: Salmonella typhimurium
mutants defective in acetohydroxy acid synthetase I and II.
J. Bacterial. 141 (1980) 1258-1263.
Shimkets, L.J., Gill, R.E. and Kaiser, D.: Developmental cell
interactions in ~Uyxococcusxunthus and the spoC locus. Proc.
Natl. Acad. Sci. USA 80 (1983) 1406-1410.
Silverman. M. and Simon, M.: Phase variation: genetic analysis
of switching mutants. Cell 19 (1980) 845-854.
Simon, R., Priefer, U. and Ptihler, A.: Vector plasmids for in vivo
and in vitro manipulations of Gram-negative bacteria, in
Ptihler, A. (Ed.), Molecular Genetics of the Bacteria-Plant
Interaction. Springer-Verlag, Heidelberg, 1983, pp. 98-106.
Smiley, B.L., Lupski, J.R., Svec, P.S., McMacken. R. and
Godson, G.N.: Sequences of the Escherichia coli dnaG
primase gene and regulation of its expression. Proc. Natl.
Acad. Sci. USA 79 (1982) 4550-4554.
Sommer, H., Cullum, J. and Saedler, H.: ISZ-43 and ISZ-44: New
alleles of the insertion sequence IS2 which have promoter
activity. Mol. Gen. Genet. 175 (1979) 53-56.
Sprent, J.I.: The Biology ofNitrogen Fixing Organisms. McGraw-
Hill, London, 1979.
Van Vliet, F., Silva, B., Van Montagu, M. and Schell, J.: Transfer
of RP4 : : Mu plasmids to Agrobacterium tumefuciens. Plasmid
1 (1978) 446-455.
Walton, D.A. and Moseley,B.E.B.: Inducedmutagenesis inRhizo-
bium frifolii. J. Gen. Microbial. 124 (1981) 191-195.
Communicated by A.I. Bukhari.
ANTIMICROBIAL AGENTS AND CHEMOTHERAPY, Apr. 2006, p. 1282–1286 Vol. 50, No. 4
0066-4804/06/$08.00⫹0 doi:10.1128/AAC.50.4.1282–1286.2006
Copyright © 2006, American Society for Microbiology. All Rights Reserved.

In Vitro Analysis of ISEcp1B-Mediated Mobilization of Naturally

Occurring ␤-Lactamase Gene blaCTX-M of Kluyvera ascorbata
Marie-Frédérique Lartigue, Laurent Poirel, Daniel Aubert, and Patrice Nordmann*
Service de Bactériologie-Virologie, Hôpital de Bicêtre, Assistance Publique/Hôpitaux de Paris, Faculté de Médecine Paris-Sud,
Université Paris XI, 94275 K-Bicêtre, France
Received 14 November 2005/Returned for modification 13 December 2005/Accepted 11 January 2006

ISEcp1B has been reported to be associated with and to mobilize the emerging expanded-spectrum ␤-lac-
tamase blaCTX-M genes in Enterobacteriaceae. Thus, the ability of this insertion sequence to mobilize the
blaCTX-M-2 gene was tested from its progenitor, Kluyvera ascorbata. Insertions of ISEcp1B upstream of the
blaCTX-M-2 gene in K. ascorbata reference strain CIP7953 were first selected with cefotaxime (0.5 and 2 ␮g/ml).
In those cases, ISEcp1B brought promoter sequences enhancing blaCTX-M-2 expression in K. ascorbata. Then,

Downloaded from http://aac.asm.org/ on June 5, 2015 by guest

ISEcp1B-mediated mobilization of the blaCTX-M-2 gene from K. ascorbata to Escherichia coli J53 was attempted.
The transposition frequency of ISEcp1B-blaCTX-M-2 occurred at (6.4 ⴞ 0.5) ⴛ 10ⴚ7 in E. coli. Cefotaxime,
ceftazidime, and piperacillin enhanced transposition, whereas amoxicillin, cefuroxime, and nalidixic acid did
not. Transposition was also enhanced when studied at 40°C.

Increasing worldwide reports of expanded-spectrum ␤-lac-
tamases of the CTX-M type in Enterobacteriaceae and mostly
in Escherichia coli raise the question of their way of acquisition
(4, 31). These enzymes are now widespread not only in noso-
comial but also in community-acquired pathogens (4, 23). The
40 CTX-M-type ␤-lactamases may be grouped into five main
subgroups according to amino acid sequence identity (CTX-
M-1, -M-2, -M-8, -M-9, and -M-25) (1, 4, 13, 15, 27, 29). Most
CTX-M enzymes hydrolyze cefotaxime better than ceftazi-
dime. However, the latest reported enzymes, including CTX-
M-15 (13), hydrolyze ceftazidime better than cefotaxime and
are also widespread (32). It has been shown that different
genetic elements are associated with blaCTX-M genes. ISEcp1-
like insertion sequences are most frequently reported (5, 7, 11,
13, 29). This insertion sequence element has been found to be
associated with four out of the five blaCTX-M gene clusters
(CTX-M-1, -M-2, -M-9, and -M-25 clusters) (1, 4, 13, 15, 27,
29). Nevertheless, the DNA sequence that separates the ␤-lac-
tamase gene from ISEcp1 varies within a given cluster of CTX-M
genes, indicating that different insertion events may have oc-
curred (16). Moreover, several plasmid-encoded cephalospori-
nase genes, such as the blaCMY- or blaACC-type genes, may be
associated also with the same ISEcp1-like element (2, 18).
ISEcp1 is weakly related to other IS elements and belongs to
the IS1380 family (IS Database home page [http://www-is
.biotoul.fr/page-is.html]) (8). Since ISEcp1-like elements are
located upstream of several ␤-lactamase genes, analysis of the
variable sequences separating these IS elements from initiation
codons of these genes allowed us to determine its boundaries.
ISEcp1B possesses two imperfect inverted repeats (IR) likely
made of 14 bp, with 12 of these 14 bp being complementary
(Table 1), and a gene encoding a 420-amino-acid transposase.
ISEcp1B brings promoter sequences for high-level expression
* Corresponding author. Mailing address: Service de Bactériologie-
Virologie, Hôpital de Bicêtre, 78 rue du Général Leclerc, 94275 K.-
Bicêtre, France. Phone: 33-1-45-21-36-32. Fax: 33-1-45-21-63-40. E-mail:
nordmann.patrice@bct.ap-hop-paris.fr.
of the blaCTX-M-14/-18, blaCTX-M-17, and blaCTX-M-19 ␤-lacta-
mase genes (4, 6, 24). Recently, we have shown that ISEcp1B
is able to mobilize the adjacent blaCTX-M-19 gene by a trans-
positional mechanism in Escherichia coli by recognizing a va-
riety of DNA sequences as right inverted repeats (IRR) (26).
Chromosome-encoded ␤-lactamases of several Kluyvera spe-
cies have been identified as progenitors of CTX-M-derived en-
zymes. The CTX-M-1 and CTX-M-2 subgroups are derived from
Kluyvera ascorbata (12, 28), whereas the CTX-M-8 and CTX-M-9
subgroups are derived from Kluyvera georgiana (21, 25).
The aim of this study was to experimentally evaluate the
ability of ISEcp1B to mobilize a chromosome-encoded ␤-lac-
tamase gene from its reservoir, K. ascorbata, to a plasmid
location in Escherichia coli. The effects of addition of different
antibiotics (mostly ␤-lactams) and of growth at various tem-
peratures were also tested.
MATERIALS AND METHODS
Bacterial strains. Clinical strain Klebsiella pneumoniae ILT-3 (expressing the
blaCTX-M-19 gene associated with ISEcp1B) has been described previously (24).
Kluyvera ascorbata CIP7953 reference strain, the recombination-deficient strain
E. coli DH5␣ (harboring pOX38-Gen, a self-conjugative, IS-free, and gentami-
cin-resistant plasmid), and the azide-resistant E. coli J53 were used for transpo-
sition and conjugation experiments (10, 17). The low-copy-number cloning vec-
tor pBBR1MCS.3 was used for cloning experiments (14). Bacterial cells were
grown in Trypticase soy (TS) broth or onto TS agar plates (Sanofi Diagnostics
Pasteur, Marnes-La-Coquette, France) with antibiotics when required.
Antimicrobial agents and susceptibility testing. Routine antibiograms were
determined by the disk diffusion method on Mueller-Hinton agar (Sanofi Diagnos-
tics Pasteur). The antimicrobial agents and their sources have been referenced
elsewhere (22). The antibiotic concentrations used for selection were as follows:
cefotaxime (CTX; 0.5 and 2 ␮g/ml), amoxicillin (AMX; 100 ␮g/ml), tetracycline
(TET; 15 ␮g/ml), kanamycin (KAN; 30 ␮g/ml), and gentamicin (GEN; 7 ␮g/ml).
Nucleic acid extraction. Recombinant plasmids and pOX38-Gen derivative
plasmids were extracted using QIAGEN Plasmid Midi kits and the very-low-copy
plasmid purification protocol, respectively (QIAGEN, Courtaboeuf, France).
Extraction of whole-cell DNA was done as described elsewhere (22).
PCR experiments. PCR experiments were performed as previously described
(30). The entire ISEcp1B gene was amplified using the primers preTnCTXM-1
(5⬘-CTAACAGAGCTTAAGCTTCC-3⬘) and preISEcp1-2 (5⬘-CTCCCAATAC
GGTCAATCCG-3⬘) and subsequently cloned into the SmaI site of plasmid
pBBR1MCS.3.

1282
VOL. 50, 2006 ISEcp1B-MEDIATED MOBILIZATION OF THE blaCTX-M GENE 1283

TABLE 1. Sequences identified as IRR boundaries after Cloning experiments and sequencing. T4 DNA ligase and restriction endo-
ISEcp1B transposition nucleases were used according to the recommendations of the manufacturer
(Amersham Biosciences, Orsay, France). The recombinant plasmid pISE was
No. of base constructed by inserting the PCR product of the ISEcp1B gene into the SmaI site
Size of
pairs
Description of Nucleotide sequence transposed of plasmid pBBR1MCS.3, which was then electroporated into electrocompetent
identical to
sequence (5⬘33⬘)a fragment Kluyvera ascorbata CIP7953 cells, as previously described (22), and selection was
perfect
(bp) performed on TET (15 ␮g/ml)-containing plates (Fig. 1). In order to study the
IRR
transposition of ISEcp1B, an omega fragment (⍀Km) from plasmid pHP45⍀-
IRL of ISEcp1B GATTCTACGTCAGT Km, made of a kanamycin resistance gene [aph(3⬘)-IIa] flanked by transcrip-
Deduced perfect ACGTAGAATCTAGG tional and translational termination sequences, was introduced into ISEcp1B.
IRR of The recombinant plasmid pISE was digested by NsiI enzyme (into ISEcp1B
ISEcp1B between the stop codon of the transposase gene and the IRR). The digested
IRR of ISEcp1B ACGTGGAATTTAGG 12 plasmid was mixed with an EcoRI-restricted ⍀Km fragment (2.2 kb) in order to
IRR-1 CGTCATATAGCTGG 4 5,464
create the tagged insertion sequence ISEcp1B.Kan, yielding the plasmid
IRR-2 ATATGGATAAGGAG 5 3,450
pISEcp1B.Kan.
IRR-3 CTTTGTAAAGAACG 5 3,951
Sequencing of the insert was performed using laboratory-designed primers on
IRR-4 GAGAAGAAAATGGG 8 3,582
an ABI PRISM 3100 automated sequencer (Applied Biosystems, Les Ulis,
IRR-5 GCTCTTTTTTCTGG 4 2,667
France).
a
Underlined nucleotides correspond to those identified at the same positions Transposition experiments. Several transposition experiments were per-

Downloaded from http://aac.asm.org/ on June 5, 2015 by guest

in the deduced IRR of ISEcp1B. formed to determine (i) the mobility of ISEcp1B alone (i.e., without the ␤-lac-

FIG. 1. Schematic representation of ISEcp1B-mediated mobilization of the naturally occurring CTX-M-2 ␤-lactamase gene of Kluyvera
ascorbata CIP7953. 1. Introduction of ISEcp1B in K. ascorbata by electroporation. 2. Selection of K. ascorbata strains with ISEcp1B inserted
upstream of the blaCTX-M-2 gene. ISEcp1B enhances the gene expression by bringing promoter sequences. 3. Transfer of recipient plasmid in
cefotaxime-resistant K. ascorbata strains. 4. ISEcp1B mobilizes the blaCTX-M-2 gene on the plasmid under various conditions: with or without
amoxicillin, piperacillin, cefuroxime, cefotaxime, ceftazidime, or nalidixic acid at 22, 30, 37, or 40°C. 5. Plasmid-located events of transposition are
isolated. White arrow with black dots, ISEcp1B; black arrow with white dots, blaCTX-M.
1284 LARTIGUE ET AL.
tamase gene), (ii) the ability of ISEcp1B to mobilize a chromosome-encoded
␤-lactamase gene from K. ascorbata to a plasmid location in E. coli by transpo-
sition, and (iii) the effects of antibiotics and of temperature on the transposition
events. The transposition of ISEcp1B.Kan onto the conjugative plasmid
pOX38-Gen was investigated with a mating-out technique in liquid medium
(22). The recombinant plasmid pISEcp1B.Kan was electroporated into E. coli
DH5␣(pOX38-Gen) for transposition experiments. Transfer of the recombinant
plasmids with the pOX38 backbone into the E. coli J53 azide-resistant (AZr)
strain was then performed by conjugation. One colony obtained after 24 h of
growth was cultured under weak agitation in 1-ml of TS broth at 37°C for 3 h and
was used as a donor for the mating assay with E. coli J53 as recipient. Conjuga-
tion was done by incubating 800 ␮l of recipient and 200 ␮l of donor under low
agitation at 37°C for an additional 3-h step. Mating was stopped by vigorous
vortexing and cooling on ice. The transconjugants were selected on agar plates
containing 7 ␮g per ml of GEN (plasmid marker), 30 ␮g per ml of KAN
(transposon marker), and 100 ␮g per ml of azide (chromosomal marker). The
transposition frequency was calculated by dividing the number of transconju-
gants by the number of donors.
The MIC of cefotaxime for the wild-type K. ascorbata strain is 0.06 ␮g/ml. To
select for ISEcp1B upstream of blaCTX-M, K. ascorbata harboring the recombi-
nant plasmid pISE was screened on agar plates containing 0.5 or 2 ␮g of
cefotaxime per ml after 24 h of growth in TS broth (Fig. 1). This strategy was
based on the previous observations demonstrating ISEcp1B mediated high-level
expression of blaCTX-M ␤-lactamases genes, with the IS element providing pro-
moter sequences. The transposition frequency was calculated by dividing the
number of CTXr transformants by the total number of bacteria. The second step
consisted of mobilization by ISEcp1B of the blaCTX-M-2 gene to the E. coli J53
recipient strain. Plasmid pOX38-GEN was used as a target for transposition
events. Transfer of pOX38-GEN into the K. ascorbata transformants was per-
formed, and transconjugants were selected on agar plates containing 7 ␮g per ml
of GEN (plasmid marker) and 100 ␮g per ml of AMX (chromosomal marker)
(Fig. 1). Transposition events were searched for between the chromosomal
blaCTX-M-2 gene and the recipient plasmid pOX38-GEN after overnight growth
in TS broth with and without antibiotics at different subinhibitory concentrations
and after 3 h of growth at various temperatures (22°C, 30°C, 37°C, and 40°C).
Several structurally unrelated ␤-lactams were studied, since it is unknown
whether any ␤-lactam might enhance mobilization of the blaCTX-M-2 gene. Na-
lidixic acid was also studied, since quinolone resistance is frequently associated
with extended-spectrum ␤-lactamase (ESBL)-mediated resistance in Entero-
bacteriaceae (23) and since quinolones are known to induce antibiotic resistance
through the SOS response (3). Transfer of the recombinant plasmids with the
pOX38-GEN backbone into the E. coli J53AZr strain was then performed by
conjugation, and transconjugants were selected on agar plates containing 7 ␮g
per ml of GEN (plasmid marker), 100 ␮g per ml of AMX (transposon marker),
and 100 ␮g per ml of azide (chromosomal marker) (Fig. 1). The transposition
frequency was calculated by dividing the number of transconjugants by the
number of donor bacteria. All the GENr AMXr AZr colonies were screened for
tetracycline susceptibility to exclude those that may have resulted from nontrans-
position events.
Insertion site determination. The regions where for insertions of ISEcp1B
upstream of the blaCTX-M-2 gene were amplified by PCR and sequenced. Plasmid
pOX38-GEN carrying various transposed structures was extracted and se-
quenced in part.
RESULTS AND DISCUSSION
Transposition of ISEcp1B upstream of the chromosome-
located blaCTX-M-2 gene. Selection of the K. ascorbata CIP7953
reference strain, in which ISEcp1B was inserted upstream of
the blaCTX-M-2 gene, was obtained at a frequency of (1 ⫾ 0.5) ⫻
10⫺7 per donor, whereas the overall transposition frequency of
ISEcp1B.Kan was 10⫺6 (data not shown). Thus, transposition
of ISEcp1B upstream of the chromosomal ␤-lactamase gene
occurred at a high frequency (10% of overall transposition
events of ISEcp1B.Kan). Among the K. ascorbata strains over-
expressing their blaCTX-M-2 gene, the ISEcp1B element was
inserted in such a manner that its transposase gene was tran-
scribed in the same orientation as the ␤-lactamase gene. As
observed in clinical isolates, ISEcp1B brought promoter
ANTIMICROB. AGENTS CHEMOTHER.
sequences enhancing blaCTX-M expression. The transposition
of ISEcp1B generated a 5-bp duplication that was located at
various insertion sites in the chromosome of K. ascorbata
CIP7953 (TACTA, TAATA, and AATAC). Twenty transfor-
mants were analyzed, including 11 obtained on agar plates
containing 2 ␮g/ml of cefotaxime and 9 on plates with 0.5
␮g/ml of cefotaxime. On one hand, detailed analysis of the
target sites of transformants selected on CTX (2 ␮g/ml) re-
vealed a preferential location 22 bp upstream of blaCTX-M-2
(64%), but other insertions were observed located 19 bp and 43
bp upstream of the blaCTX-M-2 gene. The insertion sites of most
of the transformants (five of nine) selected on CTX at 0.5
␮g/ml were located 19 bp upstream of the blaCTX-M-2 gene as
the insertion site of ISEcp1B upstream of blaCTX-M-5, de-
scribed on a natural plasmid (12). Moreover, ISEcp1 insertions
located 43 bp upstream of blaCTX-M have been identified up-
Downloaded from http://aac.asm.org/ on June 5, 2015 by guest
stream of blaCTX-M-9, blaCTX-M-14, blaCTX-M-17, and blaCTX-M-19
(6, 7, 24, 29). Nevertheless, no ISEcp1 insertion has been
identified 22 bp upstream of the blaCTX-M gene in clinical
isolates.
ISEcp1B-mediated transposition of blaCTX-M. A conjugation
was realized with the gentamicin-resistant plasmid pOX38-GEN
(GENr) (a transfer-proficient F plasmid derivative) as donor and
the K. ascorbata strain harboring ISEcp1B-blaKLUA as recipient
(Fig. 1). Susceptibility to tetracycline was systematically observed
in transconjugant strains, ruling out full integration or cointegra-
tion (TET resistance) of the recombinant plasmids derived from
pBBR1MCS.3 into plasmid pOX38. The transposition ability of
ISEcp1B-blaKLUA was investigated by conjugation with the K.
ascorbata strain, harboring ISEcp1B-blaKLUA and the plasmid
pOX38-GEN as donor, and E. coli J53 AZr as recipient (Fig. 1).
Transposition of the ISEcp1B-blaCTX-M-2 fragment occurred at a
frequency of (6.4 ⫾ 0.5) ⫻ 10⫺7 in E. coli. Analysis of the genetic
environment of several transposition events leading to insertions
of ISEcp1B-blaCTX-M-2 revealed in all cases a 5-bp duplication
that very likely confirmed the acquisition of that structure through
a transposition mechanism. These observations also indicated
that ISEcp1B was able to mobilize by itself the blaCTX-M-2 gene;
this was consistent with previous results (26).
Target site preference of the ISEcp1B-blaCTX-M-2 transposon.
Several GENr AMXr AZr E. coli isolates from independent ex-
periments were analyzed. The insertion sites of ISEcp1B-blaKLUA
were determined by sequencing the neighboring regions of the
inverted repeats. A 5-bp target site duplication, consistent with a
transposition event, was observed in all the studied insertion
events.
To determine whether the ISEcp1B-blaKLUA transposon had
a target site preference, the locations of five insertion events
were mapped onto recombinant plasmid pOX38-GEN. Inser-
tions of ISEcp1B-blaKLUA sequences had occurred into five
different sites which were distantly located on the plasmid.
Alignment of the insertion site sequences revealed variable se-
quences in recombinant plasmids (TATGA, TATCA, TACAT,
TATAC, and TTCAT). No consensus sequence was identified
among the 5-bp duplicated sites, whereas an AT-rich content
that may target ISEcp1B-mediated transposition was identified
again (26).
Characterization of ISEcp1B-blaCTX-M-2 transposons. Five
different transposons were analyzed (Table 1). Their sizes var-
ied from 2,667 to 5,464 bp. ISEcp1B possesses two imperfect
VOL. 50, 2006 ISEcp1B-MEDIATED MOBILIZATION OF THE blaCTX-M GENE
TABLE 2. IS transposition frequency with and without antibioticsa
Transposition frequency
Antibiotic (␮g/ml)
(mean ⫾ SD)
None ............................................................................ (4.3 ⫾ 2.3) ⫻ 10⫺7
Amoxicillin (256)........................................................ (3.4 ⫾ 2.3) ⫻ 10⫺7
Amoxicillin (128)........................................................ (4.2 ⫾ 3.8) ⫻ 10⫺7
Amoxicillin (52).......................................................... (5.2 ⫾ 2.5) ⫻ 10⫺7
Cefuroxime (128) ....................................................... (6.5 ⫾ 8.2) ⫻ 10⫺7
Cefuroxime (64) ......................................................... (4.6 ⫾ 3.9) ⫻ 10⫺7
Cefuroxime (26) ......................................................... (2.0 ⫾ 0.4) ⫻ 10⫺7
Cefotaxime (8)............................................................ (6.6 ⫾ 3.8) ⫻ 10⫺6
Cefotaxime (4)............................................................ (1.9 ⫾ 1.4) ⫻ 10⫺6
Cefotaxime (1.5)......................................................... (5.4 ⫾ 1.4) ⫻ 10⫺7
Ceftazidime (0.5) ....................................................... (5.2 ⫾ 2.7) ⫻ 10⫺5
Ceftazidime (0.25) ..................................................... (5.6 ⫾ 3.2) ⫻ 10⫺6
Ceftazidime (0.1) ....................................................... (6.2 ⫾ 2.7) ⫻ 10⫺7
Piperacillin (64).......................................................... (2.9 ⫾ 2.8) ⫻ 10⫺6
Piperacillin (32).......................................................... (6.5 ⫾ 7.4) ⫻ 10⫺7
Piperacillin (16).......................................................... (2.8 ⫾ 2.8) ⫻ 10⫺7
Nalidixic acid (2)........................................................ (7.4 ⫾ 1.3) ⫻ 10⫺7
Nalidixic acid (1)........................................................ (2.9 ⫾ 1.4) ⫻ 10⫺7
Nalidixic acid (0.4)..................................................... (2.9 ⫾ 1.4) ⫻ 10⫺7
a
In each case, three independent experiments were performed and the mean
and standard deviation were calculated.
IRs likely made of 14 bp, with 12 of these 14 bp being com-
plementary (Table 1) (26). A detailed analysis of the bound-
aries of the transposed fragments identified five different IRR-
like sequences downstream of the ␤-lactamase gene (Table 1).
These IRR-like sequences had been recognized by the trans-
posase of ISEcp1B during the mobilization process. The num-
ber of identical base pairs among the sequences defined as
IRR boundaries varied from 4 to 8 bp, corresponding to less
than 50% identity with the IRR of ISEcp1B. These results
indicated that ISEcp1B might use different sequences as IRRs
downstream of the ␤-lactamase gene of the K. ascorbata chro-
mosome. No consensus sequence could be determined by com-
paring the 14-bp-long IRR sequences identified in recombi-
nant plasmids (Table 1). Nevertheless, a guanosine residue
located at the 3⬘ end of these IRRs was always found, likely
indicating that this nucleotide was necessary in the transposi-
tion process, as already reported (26).
Transposition frequencies under different growth condi-
tions. Since changes of growth conditions may affect transpo-
sition efficiency of several mobile elements (19), the transpo-
sition of ISEcp1B-blaKLUA was examined in the presence of
several concentrations of different antibiotics at different sub-
inhibitory concentrations and under different temperatures.
Studied antibiotic concentrations were 1/2, 1/4, and 1/10 of the
MICs. For K. ascorbata strain CIP7953, harboring ISEcp1B-
blaKLUA, the MICs were as follows: amoxicillin, 512 ␮g/ml; piperacil-
lin, 128 ␮g/ml; cefuroxime, 512 ␮g/ml; cefotaxime, 16 ␮g/ml;
ceftazidime, 1 ␮g/ml; nalidixic acid, 4 ␮g/ml. No significant dif-
ference of transposition frequency was found with or without
amoxicillin, cefuroxime, and nalidixic acid at the studied concen-
trations (Table 2). In contrast, the transposition frequency of the
ISEcp1B-blaKLUA element was 100-fold higher when ceftazidime
was added at 0.5 ␮g/ml (half of the MIC), 10-fold higher with
ceftazidime at 0.25 ␮g/ml (one-quarter of the MIC) or cefotaxime
at 8 ␮g/ml (half of the MIC), 7-fold higher with piperacillin at 64
␮g/ml (half of the MIC), and 4-fold higher with cefotaxime at 4
␮g/ml (one-quarter of the MIC) (Table 2). Under these experi-
1285
TABLE 3. IS transposition frequency at different temperaturesa
Transposition frequency
Temp (°C)
(mean ⫾ SD)
22 ........................................................................... (4.9 ⫾ 2.8) ⫻ 10⫺7
30 ........................................................................... (6.8 ⫾ 5.8) ⫻ 10⫺7
37 ........................................................................... (6.4 ⫾ 0.5) ⫻ 10⫺7
40 ........................................................................... (2.6 ⫾ 2.2) ⫻ 10⫺6
a
In each case, three independent experiments were performed and the aver-
age and standard deviation were calculated.
mented conditions, ceftazidime seemed to enhance the transpo-
sition of ISEcp1B at a higher level than cefotaxime. No difference
of transposition frequency was found with ceftazidime (0.1 ␮g/
ml), cefotaxime (1.5 ␮g/ml), or piperacillin (32 and 13 ␮g/ml).
Although amoxicillin and cefuroxime are widely prescribed for
Downloaded from http://aac.asm.org/ on June 5, 2015 by guest
community-acquired infections and cefuroxime has been de-
scribed as a risk factor in selection of ESBLs in the community
(9), our results indicated that, under the experimental conditions
described, those ␤-lactams might not select for those transposi-
tion events. By contrast, cefotaxime and ceftazidime may induce
those transposition events, but they are not given frequently for
treating community-acquired gram-negative infections.
Nalidixic acid, which may induce the SOS repair system and
recombination events (3), did not play a role in ISEcp1B-
related transposition events. Thus, although there is somewhat
of a strong association between quinolone resistance and
ESBL phenotype in Enterobacteriaceae (23), this result would
indicate that quinolone might not enhance the transposition of
the blaCTX-M gene. The association between quinolone resis-
tance and ESBLs may rather result from clonal selection of
specific strains by quinolone or expanded-spectrum cephalo-
sporins.
The transposition frequency was increased at 40°C com-
pared to that determined at 37°C, whereas no difference was
observed at 22°C and 30°C (Table 3). These results may em-
phasize the induction of ISEcp1B transposition under high-
temperature conditions, as recently observed for four IS ele-
ments, IS401, IS402, IS406, and ISBmu3 in Burkholderia
multivorans ATCC 17616 (20). This is the first report demon-
strating increased transposition frequencies at a high temper-
ature for an IS belonging to the IS1380 family. This result
contrasts with the decreased transposition activities of several
mobile elements (e.g., Tn3, IS1, IS30, and IS911) in E. coli at
this temperature (8, 19). The temperature sensitivity of trans-
position in E. coli has been considered to be a feature of each
transposase (8, 19). It is tempting to speculate that global
warming may increase transposition of ISEcp1B. In addition,
although growth curves did not differ significantly for E. coli
and K. ascorbata, a slight increase of growth of both donor
(Kluyvera sp.) and recipient (E. coli) at temperatures ranging
from 22°C to 37°C may account also for dissemination of
blaCTX-M genes.
The reservoir of most blaCTX-M genes has been identified in
Kluyvera spp. (12, 21, 25, 28). Their translocation process has been
studied here, which represents the first evidence of in vivo mobi-
lization of a clinically important antibiotic resistance gene from its
natural reservoir. Thus, there is now an urgent need for identifi-
cation of the environmental location (if any) of Kluyvera spp.
1286 LARTIGUE ET AL.
ACKNOWLEDGMENTS
This work was financed by a grant from the Ministère de l’Education
Nationale et de la Recherche (UPRES-EA3539), Université Paris XI,
Paris, France, and mostly by the European Community (6th PCRD,
LSHM-CT-2003-503-335). L.P. is a researcher from the INSERM,
Paris, France.
We thank T. Naas for precious advice.
REFERENCES
1. Baraniak, A., J. Fiett, W. Hryniewicz, P. Nordmann, and M. Gniadkowski.
2002. Ceftazidime-hydrolysing CTX-M-15 extended-spectrum ␤-lactamase
(ESBL) in Poland. J. Antimicrob. Chemother. 50:393–396.
2. Bauernfeind, A., I. Schneider, R. Jungwirth, H. Sahly, and U. Ullmann.
1999. A novel type of AmpC ␤-lactamase, ACC-1, produced by a Klebsiella
pneumoniae strain causing nosocomial pneumonia. Antimicrob. Agents Che-
mother. 43:1924–1931.
3. Beaber, J. W., B. Hochhut, and M. K. Waldor. 2004. SOS response promotes
horizontal dissemination of antibiotic resistance genes. Nature 427:72–74.
4. Bonnet, R. 2004. Growing group of extended-spectrum ␤-lactamases: the
CTX-M enzymes. Antimicrob. Agents Chemother. 48:1–14.
5. Bou, G., M. Cartelle, M. Tomas, D. Canle, F. Molina, R. Moure, J. M. Eiros,
and A. Guerrero. 2002. Identification and broad dissemination of the CTX-
M-14 ␤-lactamase in different Escherichia coli strains in the northwest area
of Spain. J. Clin. Microbiol. 40:4030–4036.
6. Cao, V., T. Lambert, and P. Courvalin. 2002. ColE1-like plasmid pIP843 of
Klebsiella pneumoniae encoding extended-spectrum ␤-lactamase CTX-M-17.
Antimicrob. Agents Chemother. 46:1212–1217.
7. Chanawong, A., F. H. M’Zali, J. Heritage, J. H. Xiong, and P. M. Hawkey.
2002. Three cefotaximases, CTX-M-9, CTX-M-13, and CTX-M-14, among
Enterobacteriaceae in the People’s Republic of China. Antimicrob. Agents
Chemother. 46:630–637.
8. Chandler, M., and J. Mahillon. 2002. Insertion sequences revisited, p. 305–
366. In N. L. Craig, R. Craigie, M. Gellert, and A. M. Lambowitz (ed.),
Mobile DNA II. ASM Press, Washington, D.C.
9. Colodner, R., W. Rock, B. Chazan, N. Keller, N. Guy, W. Sakran, and R. Raz.
2004. Risk factors for the development of extended-spectrum ␤-lactamase-
producing bacteria in nonhospitalized patients. Eur. J. Clin. Microbiol. In-
fect. Dis. 23:163–167.
10. Derbyshire, K. M., L. Hwang, and N. D. Grindley. 1987. Genetic analysis of
the interaction of the insertion sequence IS903 transposase with its terminal
inverted repeats. Proc. Natl. Acad. Sci. USA 84:8048–8053.
11. Dutour, C., R. Bonnet, H. Marchandin, M. Boyer, C. Chanal, D. Sirot, and
J. Sirot. 2002. CTX-M-1, CTX-M-3, and CTX-M-14 ␤-lactamases from
Enterobacteriaceae isolated in France. Antimicrob. Agents Chemother. 46:
534–537.
12. Humeniuk, C., G. Arlet, V. Gautier, P. Grimont, R. Labia, and A. Philippon.
2002. ␤-Lactamases of Kluyvera ascorbata, probable progenitors of some
plasmid-encoded CTX-M types. Antimicrob. Agents Chemother. 46:3045–
3049.
13. Karim, A., L. Poirel, S. Nagarajan, and P. Nordmann. 2001. Plasmid-medi-
ated extended-spectrum ␤-lactamase (CTX-M-3-like) from India and gene
association with insertion sequence ISEcp1. FEMS Microbiol. Lett. 201:237–
241.
14. Kovach, M. E., R. W. Phillips, P. H. Elzer, R. M. Roop II, and K. M.
Peterson. 1994. pBBR1MCS: a broad-host-range cloning vector. BioTech-
niques 16:800–802.
15. Lartigue, M. F., L. Poirel, C. Héritier, V. Tolün, and P. Nordmann. 2003.
ANTIMICROB. AGENTS CHEMOTHER.
First description of CTX-M-15-producing Klebsiella pneumoniae in Turkey.
J. Antimicrob. Chemother. 52:315–316.
16. Lartigue, M. F., L. Poirel, and P. Nordmann. 2004. Diversity of genetic
environment of the blaCTX-M genes. FEMS Microbiol. Lett. 234:201–207.
17. Martinez-Martinez, L., A. Pascual, and G. A. Jacoby. 1998. Quinolone re-
sistance from a transferable plasmid. Lancet 351:797–799.
18. Nadjar, D., M. Rouveau, C. Verdet, L. Donay, J. Herrmann, P. H. Lagrange,
A. Philippon, and G. Arlet. 2000. Outbreak of Klebsiella pneumoniae pro-
ducing transferable AmpC-type ␤-lactamase (ACC-1) originating from
Hafnia alvei. FEMS Microbiol. Lett. 187:35–40.
19. Nagy, Z., and M. Chandler. 2004. Regulation of transposition in bacteria.
Res. Microbiol. 155:387–398.
20. Ohtsubo, Y., H. Genka, H. Komatsu, Y. Nagata, and M. Tsuda. 2005. High-
temperature-induced transposition of insertion elements in Burkholderia
multivorans ATCC 17616. Appl. Environ. Microbiol. 71:1822–1828.
21. Olson, A. B., M. Silverman, D. A. Boyd, A. McGeer, B. M. Willey, V. Pong-
Porter, V. Daneman, and M. R. Mulvey. 2005. Identification of a progenitor
of the CTX-M-9 group of extended-spectrum ␤-lactamases from Kluyvera
georgiana isolated in Guyana. Antimicrob. Agents Chemother. 49:2112–2115.
22. Philippon, L. N., T. Naas, A. T. Bouthors, V. Barakett, and P. Nordmann.
1997. OXA-18, a class D clavulanic acid-inhibited extended-spectrum ␤-lac-
tamase from Pseudomonas aeruginosa. Antimicrob. Agents Chemother. 41:
Downloaded from http://aac.asm.org/ on June 5, 2015 by guest
2188–2195.
23. Pitout, J. D., P. Nordmann, K. B. Laupland, and L. Poirel. 2005. Emergence
of Enterobacteriaceae producing extended-spectrum ␤-lactamases (ESBLs)
in the community. J. Antimicrob. Chemother. 56:52–59.
24. Poirel, L., J. W. Decousser, and P. Nordmann. 2003. Insertion sequence
ISEcp1B is involved in the expression and mobilization of a blaCTX-M ␤-lac-
tamase gene. Antimicrob. Agents Chemother. 47:2938–2945.
25. Poirel, L., P. Kämpfer, and P. Nordmann. 2002. Chromosome-encoded
Ambler class A ␤-lactamase of Kluyvera georgiana, a probable progenitor of
a subgroup of CTX-M extended-spectrum ␤-lactamases. Antimicrob. Agents
Chemother. 46:4038–4040.
26. Poirel, L., M. F. Lartigue, J. W. Decousser, and P. Nordmann. 2005.
ISEcp1B-mediated transposition of blaCTX-M in Escherichia coli. Antimicrob.
Agents Chemother. 49:447–450.
27. Poirel, L., T. Naas, I. Le Thomas, A. Karim, E. Bingen, and P. Nordmann.
2001. CTX-M-type extended-spectrum ␤-lactamase that hydrolyzes ceftazi-
dime through a single amino acid substitution in the omega loop. Antimi-
crob. Agents Chemother. 45:3355–3361.
28. Rodriguez, M. M., P. Power, M. Radice, C. Vay, A. Famiglietti, M. Galleni,
J. A. Ayala, and G. Gutkind. 2004. Chromosome-encoded CTX-M-3 from
Kluyvera ascorbata: a possible origin of plasmid-borne CTX-M-1-derived
cefotaximases. Antimicrob. Agents Chemother. 48:4895–4897.
29. Saladin, M., V. T. Cao, T. Lambert, J. L. Donay, J. L. Herrmann, Z. Ould-
Hocine, C. Verdet, F. Delisle, A. Philippon, and G. Arlet. 2002. Diversity of
CTX-M ␤-lactamases and their promoter regions from Enterobacteriaceae
isolated in three Parisian hospitals. FEMS Microbiol. Lett. 209:161–168.
30. Sambrook, J., and D. W. Russell. 2001. Molecular cloning: a laboratory
manual, 3rd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor,
N.Y.
31. Tzouvelekis, L. S., E. Tzelepi, P. T. Tassios, and N. J. Legakis. 2000. CTX-
M-type ␤-lactamases: an emerging group of extended-spectrum enzymes.
Int. J. Antimicrob. Agents 14:137–140.
32. Woodford, N., M. E. Ward, M. E. Kaufmann, J. Turton, E. J. Fagan, D.
James, A. P. Johnson, R. Pike, M. Warner, T. Cheasty, A. Pearson, S. Harry,
J. B. Leach, A. Loughrey, J. A. Lowes, R. E. Warren, and D. M. Livermore.
2004. Community and hospital spread of Escherichia coli producing CTX-M
extended-spectrum ␤-lactamases in the UK. J. Antimicrob. Chemother. 54:
735–743.
Appl Microbiol Biotechnol (1999) 51: 1±12
MINI-REVIEW
H.-M. Tan
Bacterial catabolic transposons
Received: 13 July 1998 / Received revision: 22 September 1998 / Accepted: 26 September 1998
Abstract The introduction of foreign organic hydro-
carbons into the environment in recent years, as in the
widespread use of antibiotics, has resulted in the evo-
lution of novel adaptive mechanisms by bacteria for the
biodegradation of the organic pollutants. Plasmids have
been implicated in the catabolism of many of these
complex xenobiotics. The catabolic genes are prone to
undergo genetic rearrangement and this is due to their
presence on transposons or their association with
transposable elements. Most of the catabolic transpo-
sons have structural features of the class I (composite)
elements. These include transposons for chlorobenzoate
(Tn5271), chlorobenzene (Tn5280), the newly discovered
benzene catabolic transposon (Tn5542), and transpo-
sons encoding halogenated alkanoates and nylon-oligo-
mer-degradative genes. Transposons for the catabolism
of toluene (Tn4651, Tn4653, Tn4656) and naphthalene
(Tn4655) belong to class II (Tn3 family) elements. Many
catabolic genes have been associated with insertion se-
quences, which suggests that these gene clusters could be
rapidly disseminated among the bacterial populations.
This greatly expands the substrate range of the micro-
organisms in the environment and aids the evolution of
new and novel degradative pathways. This enhanced
metabolic versatility can be exploited for and is believed
to play a major part in the bioremediation of polluted
environments.
Introduction
The role microorganisms play in the recycling of ele-
ments in the biosphere is a crucial one, albeit not new.
H.-M. Tan
Department of Microbiology,
National University of Singapore,
Lower Kent Ridge Road, 119260, Singapore
e-mail: mictanhm@nus.edu.sg
Tel.: +65-874-6407
Fax: +65-776-6872
Ó Springer-Verlag 1999
However, the degradation of xenobiotic chemicals by
microorganisms has attracted more recent attention. A
wide variety of microorganisms in the environment can
adapt to use xenobiotic chemicals as new sources of
growth and energy. Despite being isolated from geo-
graphically separated areas of the world, these microor-
ganisms are often found to possess similar specialized
enzyme systems and metabolic pathways for the degra-
dation of synthetic organic compounds such as toluene,
naphthalene and chlorinated biphenyls. At the molecular
level, similarities in genetic organisation and nucleotide
sequence suggest that discrete modules of catabolic genes
may have been combined in di€erent fashions to generate
new catabolic pathways in bacteria. The tendency for
catabolic genes to undergo genetic rearrangements, such
as insertions, deletions, duplications and inversions, is
attributable to the presence of elements that possess the
ability to mobilise the catabolic genes.
Although many of the well-characterized bacterial
transposons were originally isolated from clinical sam-
ples (Galas and Chandler 1989; Sheratt 1989), new ele-
ments and transposons associated with catabolic
operons are being characterized from environmental
isolates, suggesting that bacterial catabolic transposons
(Nakatsu et al. 1991; Springael et al. 1993; Tsuda et al.
1989; van der Meer et al. 1991b) may be widespread and
diverse in nature. It is clear that transposons are in-
volved in the rapid adaptation of bacterial communities
to aromatic compounds, just as is the case for antibi-
otics, and their frequent location on transmissible plas-
mids may facilitate their dissemination (Dahlberg and
Hermansson 1995). Such plasmids are usually very large
(70±500 kb) and, where investigated, are often trans-
missible (Table 1).
Transposons are discrete DNA segments that are able
to move in the absence of genetic homology from one
genetic location (donor site) to another (target site)
(Berg and Howe 1989). This process requires a trans-
posase that is encoded by the genetic element itself. The
transposase interacts with the ends of the transposon in
a site-speci®c manner, cuts the DNA at both ends of the
 2

Table 1 A selection of catabolic plasmids from bacteria

Primary substrate Plasmid Size (kb) Organism Conjugative Reference

Benzene pHMT112 112 P. putida ML2 Tan and Mason 1990

Camphor PpG1 (CAM) 500 P. putida + Rheinwald et al. 1973; Frantz and Chakrabarty 1986
Chlorobenzene pP51 110 Pseudomonas sp. P51 + van der Meer et al. 1991a
3-Chlorobenzoate pAC25 117 P. putida AC858 + Chatterjee et al. 1981
pB13 (pWR1) 111 Pseudomonas sp. B13 + Chatterjee and Chakrabarty 1983
pBRC60 88 Alcaligenes sp. BR60 + Fulthorpe and Wyndham 1991
p-Cresol pRA4000 85 P. putida NCIB 9886 + Dean et al. 1989
Haloacetates pUO1 65 Moraxella sp. B + Kawasaki et al. 1981
Isopropylbenzene pRE4 105 P. putida RE204 + Eaton and Timmis 1986
pBD2a 210 R. erythropolis BD2 + Dabrock et al. 1994; Kesseler et al. 1996
Naphthalene NAH7 83 P. putida PpG7 + Dunn and Gunsalus 1973; Gunsalus et al. 1982
pNPL-41 P. putida + Boronin et al. 1980
pND 140 P. putida UB1-1 + Dunn et al. 1980
Nylon pOAD2 44 Flavobacterium sp. K172 Kato et al. 1994
2,4-Dichlorophenoxyacetate pJP4 80 A. eutrophus JMP134 + Don et al. 1985
Phenol pPGH1 200 P. putida H ) Hermann et al. 1987
Salicylate SAL 1 90 P. putida R1 + Gunsalus et al. 1982; Chakrabarty 1972
p-Sulphobenzoate pPSB 85 C. testosteroni PSB-4 + Junker and Cook 1997
p-Toluene sulphonate pTSA 85 C. testosteroni T-2 + Junker and Cook 1997
Toluene pWW0 (TOL) 117 P. putida mt-2 + Williams and Worsey 1976; Duggleby et al. 1977
pDKI 180 P. putida HS1 (PpCl) + Kunz and Chapman 1981
pTKO 150 P. putida PPK1 ) Keshavarz et al. 1985
pKJ1 225 Pseudomonas sp. TA8 + Yano and Nishi 1980;
1,2,4-Trimethylbenzene pGB 85 P. putida TMB Besteti and Galli 1987
3,5-Xylenol pRA500 500 P. putida NCIB9869 + Hooper and Kemp 1980; Jain et al. 1984
a
This is a linear plasmid that also encodes trichloroethane catabolism

element and proceeds with the strand-transfer reaction
(Hallet and Sherratt 1997; Mizuuchi 1992).
Most transposable elements can be grouped into
three classes depending on the genetic organisation,
DNA sequence homologies and mechanistic properties
(Grindley and Reed 1995).
Class I elements include the insertion sequences (IS)
containing the genetic determinants for transposition
only, and composite transposons that have an inter-
vening sequence with ¯anking IS constituents (Syvanen
1988). IS elements, which are normal constituents of
many bacterial chromosomes and plasmids, are distin-
guished from transposons in that they do not carry any
accessory genes such as drug or heavy-metal resistance,
or catabolic markers. IS elements from both gram-neg-
ative and gram-positive bacteria have been described
and their genetic properties have been extensively re-
viewed (Syvanen 1988; Galas and Chandler 1989). These
include the duplication of 2±20 bases at the target site of
insertion of the IS element (Syvanen1988). These direct
repeats are believed to have arisen during the transpo-
sition process when the transposase, encoded by the IS
element itself and acting in cis, makes a staggered cut in
the DNA molecule, which is eventually ®lled-in by a
DNA-repair system of the host cell (Grindley and Reed
1985). The length of the target duplication is charac-
teristic for each IS element (Craig 1996; Galas and
Chandler 1989). IS elements are generally ¯anked by
inverted repeats, which are believed to be the recognition
and binding site for the transposase during transposition
(Gamas et al. 1987; Syvanen 1988). Class II groups the
set of ancestrally related elements that are known as
the Tn3 family of transposons, while class III covers the
transposing bacteriophage Mu and related phages.
There are, however, a number of transposable elements
that do not ®t into any of these classes. Nevertheless, re-
gardless of their classi®cation, all transposable elements
consist of a unique DNA sequence ¯anked by short (8±
40 bp) inverted nucleotide sequence repeats. Almost all
the bacterial catabolic elements reported to date are
classi®ed as either class I or class II transposons (Table 2).
Transposition of catabolic transposons can either be
replicative or non-replicative depending on whether the
mobile genetic element is duplicated or not during the
transposition process. In general, replicative transposi-
tion proceeds through the formation of a cointegrate
and involves the enzymatic activities of a transposase
that acts on the ends of the original transposon, and a
resolvase that acts on the duplicated copies. Class-II-
type elements transpose in this way. On the other hand,
non-replicative transposition requires only a transposase
and allows the transposon to move as a physical entity
directly from a donor to a recipient site, as seen in the
transposition of Tn5 and Tn10. Other mobile elements,
such as IS1, use both non-replicative and replicative
pathways (Biel and Berg 1984; Bender and Kleckner
1986; Berg and Howe 1989).
This review describes bacterial catabolic transposons
that encode enzymes involved in the catabolism of aro-
3
matic hydrocarbons and other organic compounds.
Newly isolated and other potential catabolic transpo-
sons are also presented. The ecological potential of these
elements and their in¯uence on bioremediation are
brie¯y discussed. There have been a number of recent
related reviews of the structure and function of trans-
posable elements (van der Meer et al. 1992; Wyndham
et al. 1994a; Tsuda 1996).
Class I catabolic transposons
Several examples of class I transposable elements that
encode enzymes for catabolic pathways involved in the
biodegradation of organic compounds are known.
Transposition of class I elements is via the conservative
model, in which the transposing element translocates as
a physical entity from one site to another. These ele-
ments exhibit great diversity at the sequence level in both
their inverted repeats and their transposase genes.
The chlorobenzoate catabolic genes on the plasmid
pBRC60 from Alcaligenes sp. strain BR60 have been
reported to be localised on a composite transposon,
designated Tn5271 (Nakatsu et al. 1991). Tn5271 is
17 kb in length and is ¯anked by two copies of IS1071,
which are 3.2-kb direct repeats. Tn5271 resembles a class
I composite transposon but the ¯anking IS1071 element
is a class II transposon (see below).
The catabolic operon of Tn5271 carries cbaABC,
which encodes a 3-chlorobenzoate 3,4-(4,5)-dioxygenase
and the corresponding reductase and dehydrogenase in-
volved in the conversion of 3-chloro- and 3,4-dichloro-
benzoate to protocatechuate and chloroprotocatechuate
respectively (Nakatsu and Wyndham 1993; Wyndham
et al. 1994b; Nakatsu et al. 1997). Tn5271 was observed
to be able to transpose to the chromosomes of Coma-
monas acidovorans (ATCC 15668) and Comamonas test-
osteroni (ATCC 11996), always as a portion of a larger
encompassing element, which suggests that Tn5271 may
be part of a larger transposon (Wyndham et al. 1994a).
The ®rst two dioxygenase/dehydrogenase enzymes of
the chlorobenzene degradation pathway of Pseudomonas
sp. strain P51 are encoded on a plasmid-located trans-
poson Tn5280 (van der Meer et al. 1991b). The trans-
poson comprises the 5.1-kb tcbAB catabolic gene cluster
¯anked by 1142-bp insertion sequences, IS1066 and
IS1067, which di€er at only 1 base pair within the 13-bp
inverted repeat on the pP51 plasmid. The 1068-bp open
reading frame within the insertion elements could en-
code a putative transposase, with features similar to
other IS-encoded transposases, such as a basic isoelectric
point and a helix-turn-helix motif (Galas and Chandler
1989). The end of IS1066 also contained an outwardly
facing putative )35 promoter sequence, which is char-
acteristic of other bacterial insertion elements. The in-
sertion elements IS1066 and IS1067 were found to be
homologous to a class of repetitive elements of Brady-
rhizobium japonicum and distantly related to IS630 of
Shigella sonnei (van der Meer et al. 1991b). When
 4

Table 2 Catabolic transposons and transposon-like elements from bacteria

Transposon Size (kb) Catabolic marker (genes) IS Location Organism Reference

Class I transposons
Tn5271 17 Chlorobenzoate (cbaABC) IS1071 pBRC60 Alcaligenes sp. BR60 Nakatsu et al. 1991
Tn5280 8.5 Chlorobenzene (tcbAB) IS1066 pP51 Pseudomonas sp. P51 van der Meer et al. 1991b
IS1067
Tn5542 12 Benzene (bedDC1C2BA) IS1489 pHMT112 P. putida ML2 Fong et al. 1997
Chlorinated aliphatic acids (dehH2) IS1071 pUO1 Moraxellai sp. B Kawasaki et al. 1992
Nylon oligomers (nylABC) IS6100 pOAD2 Flavobacterium sp. K172 Kato et al. 1994
Transposons of unknown class and transposon-like elements
Tn4371 59 Chlorobiphenyl (bph/cbpABCD) ND Chromosome A. eutrophus A5 Springael et al. 1993
90 Biphenyl/salicylate (bph/sal) ND Chromosome P. putida KF715 Nishi et al. 1998
26 Aniline (tdnQTA1A2BR) IS1071 pTDN1 P. putida UCC22 Fukumori and Saint 1997
p-Toluenesulphonate/sulphobenzoate IS1071 pTSA C. testosteroni T2 Junker and Cook 1997
(tsaMBCDR/psbAC)
4-Carboxydiphenyl ether (pobAB) IS1071 pPOB P. pseudoalcaligenes POB310 Dehmel et al. 1995
Monobromoacetate (dhlB) IS1247 Chromosome X. autotrophicus GJ10 van der Ploeg et al. 1995
Class II transposons
Tn4651 56 Toluene (xyl) 46a pWW0 P. putida mt-2 Tsuda and Iino 1987
Tn4652 17 None Chromosome P. putida PaW85 Horak and Kivisaar 1998
a
Tn4653 70 Toluene (xyl) 38 pWW0 P. putida mt-2 Tsuda and Iino 1988
Tn4655 38 Naphthalene (nah) 38a NAH7 P. putida G7 Tsuda and Iino 1990
Tn4656 39 Toluene (xyl) 38a pWW53 P. putida MT53 Williams et al. 1992
a
Size of inverted repeats in bp
ND not determined

marked with a kanamycin-resistance gene and intro-
duced into Pseudomonas putida KT2442, Tn5280 was
found to be functional, transposing into the genome at
random and in a single copy.
The structure of a new catabolic transposon in
P. putida ML2 carrying the ``upper'' pathway genes for
benzene degradation has been described (Fong et al.
1997). Two directly repeated sequences of the IS element
IS1489 are found to ¯ank the benzene dioxygenase and
dehydrogenase genes on the catabolic plasmid
pHMT112 in PpML2 (Tan and Mason 1990; Tan et al.
1993; Tan and Fong 1993; Fong et al. 1996) forming a
composite transposon Tn5542. Only the left-hand IS is
bound by a pair of 40-bp imperfect, inverted repeats
while the right-hand copy lacks the right-hand inverted
repeat. Both IS elements, termed variants 1 and 2 re-
spectively, contain an identical 1371-bp open reading
frame, tnpA, which was shown to encode a 53-kDa
polypeptide. The deduced amino acid sequence and pI
are typical of other transposases. In addition, seven dam
methylation and four possible chi sites, involved in
regulating transposition and homologous recombination
respectively, were identi®ed throughout the sequence of
tnpA. An outwardly facing )10 and )35 promoter was
detected at the right-hand end of IS1489 v1 and overlaps
with the bedD gene encoding cis-benzene dehydrodiol
dehydrogenase (Fong et al. 1996, 1997). Tn5542, how-
ever, was not shown to be functional.
The presence of a catabolic transposon is also im-
plicated in the nylon-oligomer-degrading Flavobacterium
sp. strain K172 (Kato et al. 1994). Three of the degra-
dative genes (nylABC) are borne on the 43.6-kb plasmid
pOAD2 ¯anked by two types of repeated sequences. The
nucleotide sequence of the type of repeated sequences
¯anking ny1A revealed a 765-bp open reading frame
encoding a putative transposon and 14-bp IR that
showed identity to IS6100, which is part of Tn610 in
Mycobacterium fortuitum encoding sulphonamide resis-
tance (Martin et al. 1990).
There are indications that the haloacid dehalogenase
genes may be located on a transposon. In Morexella sp.
strain B, the dehH2 gene encoding one of the two ha-
loacetate dehalogenases, which acts on monochloro-,
monobromo- and monoiodoacetate was found to be
sandwiched between two repeated sequences of 1.8 kb
on a 65-kb conjugative plasmid, pUO1 (Kawasaki et al.
1981). The repeat sequences are IS1071-like IS, encoding
the tnpA transposase, and are thought to play a part in
the frequent spontaneous deletion of a 5.4-kb fragment
containing dehH2 from pUO1 (Kawasaki 1992; Wynd-
ham et al. 1994a).
P. putida strain PP3 contains the gene for dehaloge-
nase (deh1) involved in the hydrolytic dehalogenation of
2-chloro- and 2,3,-dichloroacetic, propionic and but-
anoic acids on a mobile genetic element designated
DEH. Integration of the DEH element into a number of
plasmids and the chromosome of the host has been re-
ported (Beeching et al. 1983; Slater et al. 1979; Thomas
et al. 1992a, b).
5
IS1071
The IS1071 element deserves special mention. Although
a class-II-type insertion element it has been associated
with class-I transposons. Increasingly a number of
transposable catabolic gene clusters that are ¯anked by
IS1071 have been found. Involvement of IS1071 is
thought to be responsible for transferring degradative
phenotypes into the environment (Fulthorpe and
Wyndham 1992; Fukumori and Saint 1997).
IS1071, part of the chlorobenzoate catabolic trans-
poson Tn5271, is 3201 bp long and contains inverted
110-bp repeat sequences. The outer 38 bp of these re-
peats are related to the class II inverted repeats that
comprise the transposase recognition sequence. There is
an integration host factor (IHF) consensus binding se-
quence within the IR sequences. However, the 5-bp di-
rect repeats of DNA following transposition of all class
II elements (Sherratt 1989) are missing in Tn5271
(Nakatsu et al. 1997). Between the inverted repeats is a
tnpA open reading frame. There are no tnpR gene or res
sites within IS1071. Intramolecular transposition/inver-
sion derivatives of the right-hand IS1071 end of Tn5271
have been detected in an actual groundwater bioreme-
diation system and indicate the functionality of IS1071
(Wyndham et al. 1994b).
Besides ¯anking the cba genes in Alcaligenes sp. strain
BR60, isolated from the Niagara River watershed,
IS1071 has been found to ¯ank the dehH2 dehalogenase
gene of plasmid pU01 from Moraxella sp. strain B,
which has been isolated from industrial wastewater in
Japan (Wyndham et al. 1994a; Kawasaki et al. 1992). In
addition, partial sequence information reveals a copy of
IS1071 upstream of the pobA gene that is involved in the
metabolism of carboxydiphenyl ethers in Pseudomonas
pseudoalcaligenes POB310 (Dehmel et al. 1995), while
two copies of the IS are present on plasmid pTSA car-
rying the sulphobenzoate- and p-toluenesulphonate-
degradative genes in C. testosteroni strain T-2 (Nakatsu
et al. 1997; Junker et al. 1996; Junker and Cook 1997).
Class II transposons
The class II or Tn3-like transposons are characterised by
having short (fewer than 50 bp) terminal inverted repeat
(IR) sequences and the involvement of a transposase
(TnpA) and resolvase (TnpR) in a two-step transposi-
tion process that leads to a 5-bp duplication of the target
sequence (Sherratt 1989; Grinsted et al. 1990). In the
®rst step of transposition, TnpA recognises the IR and
catalyses the formation of a cointegrate of the donor and
target DNA molecules connected by two directly re-
peated copies of the transposon. In the second step, the
cointegrate between the two sites of resolution (res) is
resolved in a site-speci®c manner by the resolvase.
The ®rst report of a transposable catabolic operon
was that of the xyl genes for toluene catabolism from
6
P. putida mt-2 (Jacoby et al. 1978; Nakazawa et al. 1978;
Chakrabarty et al. 1978). The genetic organization and
regulation of the xyl operon is perhaps the most well-
characterized one among aromatic catabolic pathways
in bacteria. The xyl genes are encoded on the pWW0
plasmid and located on a 56-kb transposon, Tn4651,
which is contained within another 70-kb transposon,
Tn4653. Transposition of both toluene-catabolic trans-
posons has been shown to occur via cointegrate forma-
tion and resolution, and both have properties
characteristic of class II elements (Tsuda and Iino 1987;
1988; Tsuda et al. 1989). Tn4651 carries a complete set
of transposition functions not interchangeable with
those of other class II transposons. These include the cis-
acting res region and two uncharacterised divergently
transcribed trans-acting tnpS and tnpT gene products.
These are present on a 2.2-kb DNA fragment located
some 48 kb away from the tnpA gene. Located within
the 48-kb intervening region is a 39-kb fragment con-
taining the xyl operon ¯anked by two directly repeated
1275-bp sequences, designated IS1246. Reciprocal re-
combination between the IS elements is believed to be
responsible for the high frequency of the 39-kb deletion
observed (Assinder and Williams 1990; Reddy et al.
1994). Tn4651 is considered a novel subgroup of class II
transposons (Grinsted et al. 1990). The 46-bp IR of
Tn4651 contain sequences associated with TnpA binding
and nuclease activity (Ichikawa et al. 1990).
A 17-kb derivative of the 56-kb Tn4651, designated
Tn4652, has been reported in the chromosome of P. putida
PaW85 and shown to belong to the Tn3 family on the basis
of its transposition properties (Tsuda and Iino 1987).
Genetic analysis has localised the tnpA gene to the right
arm of Tn4652 (Tsuda and Iino 1987). The transposase
was most homologous to Tn5041, a mercury-resistance
transposon. Functional analysis showed the presence of
an IHF-binding site upstream of the tnpA promoter,
which is only active in P. putida. Expression of the trans-
posase gene was found to be enhanced by IHF, which
binds to the terminal sequences of Tn4652 just adjacent to
the inverted repeats (HoÄrak and Kivisaar 1998).
Tn4653, found on pWW0 in P. putida mt-2, encom-
passes Tn4651 and two other transposition genes, tnpR
and tnpA (Tsuda and Iino 1988). The transposase of
Tn4653 ensures the independent cointegration of the
two toluene transposons. It is interchangeable with those
of Tn1722-type transposons, encoding resistance to tet-
racycline, but not the Tn4651 transposase. The IR of
Tn4653 are also closely related to the those of Tn1722-
type transposons. Furthermore, the resolvases of
Tn4653 and Tn1722 are identical in primary sequence
(Grinsted et al. 1990) with the tnpR gene from Tn4653
being able to complement the tnpR mutation in Tn1722.
Resolution of the Tn4653-mediated cointegrate, how-
ever, requires the resolution system of Tn4651. This is
due to the fact that the Tn4653 res region, located up-
stream of tnpR, is defective, lacking a 30-bp sequence
that contains the essential crossover site for resolution of
the cointegrate (Tsuda 1996; Allmeier et al. 1992).
Both Tn4651 and Tn4653 are responsible for the
many reported transpositions of the xyl gene cluster into
the chromosome and various resistance and catabolic
plasmids of bacteria reported (Sinclair et al. 1986; Sin-
clair and Holloway 1991; Assinder and Williams 1990;
Jahnke et al. 1993).
Another toluene-catabolic transposon, Tn4656, has
been described which contains the xyl gene cluster and a
res-tnpR-tnpA region (Tsuda 1996). This transposon is
present in P. putida strain MT53 (Assinder and Williams
1990; Williams et al. 1992) and is believed to have arisen
through DNA site-speci®c recombination. It is likely
that more such catabolic transposons will be uncovered
in other bacteria capable of degrading toluene.
Similar to the xyl operon of pWW0, the nah gene
cluster for naphthalene catabolism was found to be part
of a 38-kb class II transposon, Tn4655, on the 83-kb
plasmid NAH7 from P. putida PpG7 (Yen and Serdar
1988; Tsuda and Iino 1990). Tn4655 contains the tnpR
gene and res region but is defective in the cointegration
step requiring a complementing transposase from other
Tn1722-type transposons in order to transpose. The 38-
bp IR of Tn4655 are almost identical to those of Tn4653
and Tn1722. The resolution function of Tn4655 is un-
ique as it cannot complement other class II transposons
nor be replaced by their resolvases. The 1.8-kb tnpR
region of Tn4655 is three times longer than the typical
class II resolvase gene sequence. Analysis and experi-
mental observation suggest the Tn4655 TnpR protein to
be a site-speci®c integrase (Tsuda 1996) able to catalyse
both integration and resolution reactions (Berg and
Howe 1989; Abremski and Hoess 1992).
Transposons of unknown class
and transposon-like elements
C. testosteroni T-2 degrades p-toluenesulphonate (TSA)
via p-sulphobenzoate (PSB) and protocatechuate. The
TSA operon (tsaMBCD) with its regulatory gene (tsaR),
the PSB genes (psbAC) and possibly the transport genes
for both substrates are carried on an 85-kb plasmid pTSA
in strain T-2 (Junker and Cook 1997). In another strain
C. testosteroni PSB-4, which catabolises PBS, another
plasmid pPSB was shown to carry the psb genes. Both
pTSA and pPSB contain two copies of IS1071. Conju-
gation experiments that yielded PSB+ transconjugants
that carried two copies of IS1071 in the chromosome
suggest the presence of a composite PSB transposon
analogous to Tn5271 (Junker and Cook 1997).
Xanthobacter autotrophicus GJ10 was found to be
able to adapt to monobromoacetate by overexpressing a
haloacid dehalogenase encoded by the dhlB gene, similar
to the enzyme encoded by deh2. This overexpression is
due to the insertion of IS1247 34 bp upstream of dhlB.
IS1247 was found to encode a putative 464-amino-acid
transposase and to have terminal IR of 17 bp and 16 bp
with three mismatches (van der Ploeg et al. 1995). In-
sertion of the IS generated a 4 bp duplication of the

target site. One copy of IS1247 together with dhlB is able
to transpose to another replicon in the pPJ20. This
phenomenon has been termed one-ended transposition
and occurs in Tn3, Tn21 and Tn1721, which lack one
end of the inverted repeats (Arthur et al. 1984; Avila
et al. 1984; Motsch and Schmitt 1984).
The aromatic-amine-catabolising plasmid pTDN1
was discovered in a derivative of P. putida mt-2 fol-
lowing adaptation to growth on aromatic amines. The
catabolic genes are contained within a 26-kb region
bounded by 1.8-kb direct-repeat sequences which are
readily deleted when the host cells are cultured on non-
selective carbon sources (Saint et al. 1990; McClure and
Venables 1987). Six of the aniline-degradative genes
(tdnQTA1A2BR) are located downstream of a copy of
IS1071, which di€ers in a single base from that involved
in the transposition of the chlorobenzoate genes (Nak-
atsu et al. 1991). The partial sequence of the other direct-
repeat sequence on pTDN1 is reported to be identical to
IS1071 (Fukumori and Saint 1997).
Alcaligenes eutrophus A5 can degrade chlorobiphenyl
and biphenyl. It contains a 75-kb plasmid that carries
4-chlorobenzoate-catabolic genes (cbp). Upon repeated
subculturing, A. eutrophus A5 lost the ability to min-
eralize 4-chlorobenzoate to CO2 as a result of a loss of
dehalogenase activity, and consequently had a 51-kb
plasmid pSS50 that did not contain any chlorobiphenyl-
or chlorobenzoate-catabolic genes. However, in conju-
gation experiments, pSS50 was found to acquire a
59-kb transposon Tn4371 from the A. eutrophus A5
chromosomal DNA. The transposon carries the genes
involved in the degradation of biphenyl and 4-chlor-
obiphenyl to benzoate and 4-chlorobenzoate (bphA-
BCD/cbpABCD) respectively, and is able to transpose
to pSS50 and other plasmids. Tn4371 awaits further
characterization.
The biphenyl- and salicylate-catabolic pathways of P.
putida KF715 are encoded by bph and sal gene clusters
arranged 10 kb apart on the chromosome. These gene
clusters were found to be highly prone to deletion when
KF715 was grown in rich medium and could be trans-
ferred by conjugation to P. putida AC30 at a frequency
of 10)6/cell. The integrated 90-kb DNA fragment con-
taining bph and sal genes in AC30 could undergo further
transposition and deletion and is believed to be located
on a large catabolic transposon (Nishi et al. 1997, 1998).
The presence of IS and IS-like elements on catabolic
plasmids or in association with chromosomally encoded
catabolic genes suggests that these elements may be
precursors to the development of catabolic transposons.
By recruiting genes from a variety of host bacteria, IS
elements play an important role in the evolution and
patchwork assembly of novel catabolic genes and ope-
rons.
IS1412, a 1656-bp IS element, contains terminally
partially matched 17-bp and 18-bp inverted repeats and
a putative 457-amino-acid transposase. IS1412 has been
cloned from plasmid DNA in the carbofuran-degrading
Sphingomonas sp. CF06 (Feng et al. 1997a, b).
7
In P. putida TMB, which is involved in the catabolism
of methyl-substituted aromatic hydrocarbons, a region
of the chromosome was found to contain the tmb ope-
ron, which is functionally and genetically homolgous to
the xyl genes of pWW0. Located within the tmb operon
is a DNA fragment that has homology to IS1001 (Fa-
varo et al. 1996).
There are other examples of catabolic genes associ-
ating with mobile elements but many of these genes have
not been observed to transpose as discrete elements. The
(2,4,5-trichlorophenoxy)acetic-acid-catabolic genes of
Burkholderia cepacia AC1100 are ¯anked by the 1477-bp
IS931 (Tomasek et al. 1989; Haugland et al. 1990),
which is self-transposable.
Non-aromatic catabolic transposons
In addition to the aromatic catabolic transposons,
there are other catabolic transposable elements that
carry genes encoding the metabolism of sugars and
citric acid. The lactose operon in Yersinia enterocolitica
is carried on a 16.6-kb transposon, Tn951, encoded on
the plasmid pGC1 (Cornelis et al. 1976). The lacIZY
genes of Tn951 are homologous to the E. coli K12
lactose operon (Cornelis et al. 1978) as well as to those
of 11 other plasmids of di€erent origins, although none
of these was found to contain the entire Tn951
(Cornelis 1981). Tn951 is ¯anked by two perfect 41-bp
IR containing within them the 38-bp IR of Tn3. Tn951
also contains a single copy of the insertion sequence,
IS1, upstream of lacI. This IS has been shown to
generate deletions and inversions within the transposon
resulting in Lac) phenotypes (Cornelis and Saedler
1980). Internal to the transposon and downstream of
lacY is another transposon, Tn2501, which is related to
the Tn21 family of class II transposable elements.
Tn2501 is ¯anked by 48-bp imperfect IR and does not
carry any transposon marker. The right-hand IR of
Tn2501 is found to lie next to the end of the defective
tnpA gene of Tn951 (Michels and Cornelis 1984).
Tn951 does not transpose on its own. However, its
transposition can be complemented by the tnpA gene of
other class II or Tn3-type transposons such as Tn801
or Tn3, but not Tn501, Tn1721 and c-d, suggesting
that the complementation is a speci®c process (Cornelis
et al. 1981). Tn951 is able to transpose into multiple
sites on the broad-host-range plasmid RP1 in Pseudo-
monas aeruginosa as well as into other replicons
(Cornelis et al. 1978, 1979, 1981). As is the case for
Tn3, the transposition of Tn951, in the presence of
Tn801 or Tn3, is a thermosensitive process occurring
readily at 30 °C, but not at 37 °C (Cornelis 1980).
Transposition also results in a 5-bp duplication of the
target site. However, unlike Tn3, Tn951 has not been
shown to contain any tnpR gene (Cornelis et al. 1981),
but is thought to rely on the resolution system of
Tn2501 (Michels and Cornelis 1984).
8
The genes for the catabolism of sucrose (sacA, en-
coding sucrose-6-phosphate hydrolase) and the produc-
tion of the peptide antibiotic nisin (nisA) are located on
conjugative transposons, Tn5276 and Tn5301, in Lac-
tococcus lactis (Dodd et al. 1990; Rauch and de Vos
1992). The better-characterised Tn5276 is a 70-kb con-
jugative transposon in the chromosome of L. lactis. It
contains (A + T)-rich termini ¯anked by a direct
hexanucleotide repeat without any clear IR. The inte-
gration of Tn5276 into the chromosome of a plasmid-
free donor strain displays orientation and site speci®city
(Rauch and de Vos 1992).
In Escherichia coli the plasmid-borne genes encoding
citrate metabolism are carried on a 7.4-kb transposon,
Tn3411 (Ishiguro et al. 1982). Tn3411, a class I com-
posite element, contains two directly repeated copies of
an insertion sequence, IS3411, and generates a 3-bp re-
peat upon integration. IS3411 is 1309 bp long, has 27-bp
imperfect IR and a putative 240-amino-acid transposase
(Ishiguro and Sato 1988). Intramolecular recombination
between the two copies of IS3411 occurs frequently in a
recA strain resulting in Cit) deletion mutants (Ishiguro
and Sato 1984). Tn3411 has been shown to transpose to
k bb phage and then from the Cit+-transducing k bb
phage to pBR3222 in recA-de®cient strains (Ishiguro
et al. 1982). Other plasmids specifying citrate utilization
have also been reported (Ishiguro et al. 1981), but none
of these could transpose its citrate-utilization genes
(Ishiguro et al. 1982).
Ecological signi®cance
The nature of waste being generated has changed dras-
tically over the years. In recent decades a large number
of xenobiotics have been released into the environment.
While many of these chemicals are rapidly degraded by
microorganisms in the environment, some resist attack
and remain recalcitrant. Given time, however, most
microorganisms, in particular bacteria, are able to adapt
to using these compounds as energy and carbon sources.
This biochemical versatility is largely due to the plas-
ticity of the microbial genomes. By modifying the ex-
isting genes, a novel metabolic capacity can be
developed that allows xenobiotics to be metabolised.
This requires the alteration and exchange of genetic in-
formation, and recombination processes such as gene
conversion, duplication and transposition play crucial
roles in the reassortment of discrete genetic modules and
their expression (van der Meer et al. 1992).
Catabolic plasmids and transposons allow the hori-
zontal spreading of degradative genes among microbial
communities and this is important for genetic ¯exibility
and adaptation. It has been reported that greater num-
bers of plasmid-containing bacteria are often isolated
from polluted areas than are found in unpolluted areas
(Hardman et al. 1986; Focht et al. 1996). Genetic ex-
change between strains of P. aeruginosa and P. putida
containing plasmids and indigenous microbiota has re-
sulted in the appearance of a bacterial population able
to degrade 3-chlorobenzoate (Pertsova et al. 1984). The
conjugative transfer of the catabolic transposon Tn5271
allows the utilization of 3-chlorobenzoate in a freshwa-
ter microcosm, and its constituent IS1071 plays a sig-
ni®cant role in the adaptation to 4-chloroaniline
(Fulthorpe and Wyndham 1992).
Soil having indigenous biphenyl degraders but not
chlorobenzoate degraders and inoculated with P. aeru-
ginosa JB2, a chlorobenzoate degrader, was found to
produce two recombinant strains, P. aeruginosa JB2-3
and Pseudomonas sp. JB2-M, which could metabolise
the polychlorinated biphenyls of Aroclor 1242. Results
from repetitive extragenic palindromic polymerase chain
reaction analysis suggest that the recombinants arose
from the transfer of bph genes from indigenous biphenyl
degraders to the inoculant JB2, probably via a trans-
poson (Focht et al. 1996). The recombinant strains iso-
lated were, however, unstable. This instability presents
an interesting paradox, for while it may be dicult to
maintain stable inoculants for bioremediation applica-
tion, this phenomenon may be advantageous in mini-
mizing the perceived risks of recombinant persistance in
the environment (Focht et al. 1996).
Plasmid-encoded pathways, such as those for
chlorocatechol degradation (e.g. on pAC27, pJP4, pP51),
are expected to evolve more rapidly than chromosomal
genes (Eberhard 1990). It is believed that the ancient
chlorocatechol-degradation operon, which evolved long
ago to give rise to the clc, tfd and tcb variants, probably
existed originally for the catabolism of naturally occur-
ring halogenated compounds such as chlorosubstituted
and bromosubstituted aromatics (SchloÈmann 1994). Al-
though gene transfers may occur within a very limited
taxonomic range of hosts (Williams and Sayers 1994), the
genetic characteristics of the bedDC1C2BA operon in
PpML2 carried on Tn5542 (Tan et al. 1993; Fong et al.
1996) point to a juxtaposition of catabolic genes on the
catabolic plasmid pHMT112 from hosts of di€erent ge-
nera. This would also have been true of other catabolic
operons (Shanley et al. 1994; van der Meer et al. 1991b;
Werlen et al. 1996). Furthermore, horizontal transfer of
genetic material encoding toluene or biphenyl-degrada-
tion enzymes from gram-negative species (most probably
Pseudomonas) to gram-positive Rhodococcus globerulus
P6 is thought to be an example of how microorganisms
gain novel catabolic activities for xenobiotics (Asturias
et al. 1995). The co-location of these di€erent catabolic
gene clusters on a single plasmid or in a single organism
ensures the proper regulation of the pathways involved
(van der Meer et al. 1991b). Indeed, the transposition of
mobile genetic elements can e€ect the expression, by
activation, of adjacent catabolic genes (Kivisaar et al.
1990; Haugland et al. 1990; van der Meer et al. 1992; van
der Ploeg et al. 1995; Fong et al. 1996).
The combination of transposons with plasmids plays
a major role in the rapid dissemination and evolution of
antibiotic and heavy-metal-resistance markers in the
chemical environment following widespread use of an-

tibiotics (Amabile-Cuevas and Chicuvel 1992). The same
is true for the ability of microbial communities to adapt
to arthropogenic xenobiotic compounds. This genetic
adaptability is aided by an accumulation of single-site
mutations followed by gene conversion or slipped-strand
mispairing and by gene duplication followed by muta-
tions in one of the gene copies as well as DNA rear-
rangements and transposition (van der Meer et al. 1992;
Williams and Sayers 1994). These mechanisms greatly
expand the substrate range of the microorganisms and
contribute to the evolution of catabolic pathways in
general. Currently there is a paucity of ®eld data for the
rate of evolution in the natural environment (Liu and
Su¯ita 1993; van der Meer et al. 1992; Williams and
Sayers 1994). Nevertheless, the involvement of catabolic
transposons in such genetic processes would inevitably
enhance evolutionary changes in metabolic pathways,
rendering microorganisms the ability to evolve novel
and productive pathways.
Biotechnological implications
Bioremediation, which exploits the catabolic versatility
of microorganisms to accelerate the degradation of en-
vironmental pollutants, is an important industry in
mitigating environmental contamination. Microorgan-
isms that are able to utilize the polluting chemicals as
potential energy sources survive in the hostile environ-
ment. In a sense, bioremediation of pollutants is an ex-
tension of the metabolism that occurs within the
microorganism and its e€ectiveness depends largely on
the extent to which the desired microbial community can
be selected and maintained in the contaminated envi-
ronment. In a way, catabolic transposons serve as units
of genetic ¯uidity in the environment. Although the rates
of genetic rearrangement may not be high under normal
conditions, it can be envisaged that, upon the intro-
duction of xenobiotic substrates into the environment,
selective pressure would increase to the point where high
growth rates, resulting in suciently high microbial
populations that favour such genetic interactions, can be
established. The location of catabolic genes on trans-
posons and plasmids increases the possibilities of re-
combination events and accelerates the metabolic
potential of the microbial population best adapted to
biodegrading the xenobiotic compound.
Many bacterial strains have been engineered to pos-
sess additional catabolic genes for added metabolic
prowess (Rojo et al. 1987; Lehrbach et al. 1984; Ramos
et al. 1987; reviewed by Timmis et al. 1994). While such
engineered strains may perform under laboratory con-
ditions and be considered for application in bioremedi-
ation, these bacteria in general are too fastidious to
survive in a complex multisubstrate, multiorganism-in-
teracting environment without speci®c selection pres-
sure. Coupled with the need for physical containment,
genetic tracking and attendant biosafety issues and
public pressure, it is debatable whether such engineered
9
bacteria would be employed in the natural environment
on a large scale. On the other hand, an increasing
number of natural isolates have been reported that
possess unique metabolic genes and new catabolic
plasmids and transposons. Some of these, selected under
speci®c substrates during enrichment, have been shown
to possess a combination of catabolic genes encoding
metabolic functions as versatile if not more so, than
those designed using molecular techniques in the labo-
ratory (Mars et al. 1997; Pettigrew et al. 1991; Eaton
1996; Williams and Sayers 1994). These microorganisms
that have evolved naturally tend to be more competitive
(van der Meer et al. 1992). The extensive bioremediation
of the Alaskan oil spill represents the most successful use
of indigenous microorganisms to date (Pritchard and
Costa 1991) although numerous other successful appli-
cations for the remediation of sites polluted by chemical
spills, leaking underground storage tanks and industrial-
process wastes also exist (Caplan 1993). Clearly, optimal
physical and chemical conditions, superimposed on the
genetic exchange among the microbial community, me-
diated by catabolic transposons, among other mecha-
nisms, will allow the eventual evolution of the most
adapted microbial population, the growth and activity
of which ensures the recycling of aromatic compounds in
the biosphere and a balanced global ecosystem.
Acknowledgements The author appreciates the helpful comments
and suggestions of M. Tsuda. Some of the results presented here
come from the contributions of K. Fong, C. Goh and other
members of the laboratory. The support of the National University
of Singapore is gratefully acknowledged.
References
Abremski KE, Hoess RH (1992) Evidence for a second conserved
arginine residue in the integrase family of recombination pro-
teins. Protein Eng 5: 87±91
Allmeier H, Cresnar B, Greck M, Schmitt R (1992) Complete nu-
cleotide sequence of Tn1721: gene organization and a novel
gene product with features of a chemotaxis protein. Gene 111:
11±20
Amabile-Cuevas CF, Chicurel ME (1992) Bacterial plasmids and
gene ¯ux. Cell 70: 189±199
Arthur A, Nimmo E, Hettle S, Sherratt D (1984) Transposition and
transposition immunity of transposon Tn3 derivatives having
di€erent ends. EMBO J 3: 1723±1729
Assinder SJ, Williams PA (1990) The TOL plasmids: determinants
of the catabolism of toluene and the xylenes. Adv Microb
Physiol 31: 1±69
Asturias JA, Diaz E, Timmis KN (1995) The evolutionary rela-
tionship of biphenyl dioxygenase from gram-positive Rho-
dococcus globerulus P6 to multicomponent dioxygenases from
gram-negative bacteria. Gene 156: 11±18
Avila P, Cruz F de la, Ward E, Grinsted J (1984) Plasmids con-
taining one inverted repeat of Tn21 can fuse with other plas-
mids in the presence of Tn21 transposase. Mol Gen Genet 195:
288±293
Beeching JR, Weightman AJ, Slater JH (1983) The formation of an
R prime carrying the fraction I dehalogenase from Pseudo-
monas putida PP3 using the IncP plasmid R68.44. J Gen Mi-
crobiol 129: 2071±2078
Bender J, Kleckner N (1986) Genetic evidence that Tn10 transposes
by a nonreplicative mechanism. Cell 45: 801±815
10
Berg DE, Howe MM (eds) (1989) Mobile DNA. American Society
for Microbiology, Washington DC
Besteti G, Galli E (1987) Characterization of a novel TOL-like
plasmid from Pseudomonas putida involved in 1,2,4-trimethyl-
benzene degradation. J Bacteriol 169: 1780±1783
Biel SW, Berg DE (1984) Mechanism of IS1 transposition in E. coli:
choice between simple insertion and cointegration. Genetics
108: 319±330
Boronin AM, Kochetkov VV, Skryabin GK (1980) Incompatibility
groups of naphthalene degradative plasmids in Pseudomonas.
FEMS Microbiol Lett 7: 249±252
Caplan JA (1993) The worldwide bioremediation industry: pros-
pects for pro®t. Trends Biotechnol 11: 320±323
Chakrabarty AM (1972) Genetic basis of the biodegradation of
salicylate in Pseudomonas. J Bacteriol 112: 815±823
Chakrabarty AM, Friello DA, Bopp LH (1978) Transposition of
plasmid DNA segments specifying hydrocarbon degradation
and their expression in various microorganisms. Proc Natl
Acad Sci USA 75: 3109±3112
Chatterjee DK, Kellogg ST, Hamada S, Chakrabarty AM (1981)
Plasmid specifying total degradation of 3-chlorobenzoate by a
modi®ed ortho pathway. J Bacteriol 146: 639±646
Chatterjee DK, Chakrabarty AM (1983) Genetic homology be-
tween independently isolated chlorobenzene-degradative plas-
mids. J Bacteriol 153: 532±534
Cornelis G (1980) Transposition of Tn951 (Tnlac) and cointegrate
formation are thermosensitive processes. J Gen Microbiol 117:
243±247
Cornelis G (1981) Sequence relationships between plasmids carry-
ing genes for lactose utilization. J Gen Microbiol 124: 91±97
Cornelis G, Saedler H (1980) Deletions and an inversion induced
by a resident IS1 of the lactose transposon Tn951. Mol Gen
Genet 178: 367±374
Cornelis G, Bennett PM, Grinsted J (1976) Properties of pGL1, a
lac plasmid originating in Yersinia enterocolitica 842. J Bacteriol
127: 1058±1062
Cornelis G, Ghosal D, Saedler H (1978) Tn951: a new transposon
carrying a lactose operon. Mol Gen Genet 160: 215±224
Cornelis G, Ghosal D, Saedler H (1979) Multiple integration sites
for the lactose transposon Tn951 on plasmid RP1 and estab-
lishment of a coordinate system for Tn951. Mol Gen Genet 168:
61±67
Cornelis G, Sommer H, Saedler H (1981) Transposon Tn951
(Tnlac) is defective and related to Tn3. Mol Gen Genet 184:
241±248
Craig NL (1996) Transposition. In: Neidhart FC (ed) Escherichia
coli and Salmonella: cellular and molecular biology. American
Society for Microbiology, Washington DC, pp 2239±2362
Dabrock B, Kesseler M, Averho€ B, Gottschalk RG (1994) Iden-
ti®cation and characterization of a transmissible linear plasmid
from Rhodococcus erythropolis BD2 that encodes isopropyl-
benzene and trichoroethane catabolism. Appl Environ Micro-
biol 60: 853±860
Dahlberg C, Hermansson M (1995) Abundance of Tn3, Tn21, and
Tn501 transposase (tnpA) sequences in bacterial community
DNA from marine environments. Appl Environ Microbiol 61:
3051±3056
Dean HF, Cheevadhanarak S, Skurray Ra, Bayly RC (1989)
Characterization of a degradative plasmid in Pseudomonas
putida that controls the expression of 2,3-xylenol degradative
genes. FEMS Microbiol Lett 61: 153±158
Dehmel U, Engesser KH, Timmis KN, Dwyer DF (1995) Cloning,
nucleotide sequence, and expression of the gene encoding a
novel dioxygenase involved in metabolism of carboxydiphenyl
ethers in Pseudomonas pseudoalcaligenes POB310. Arch Mi-
crobiol 163: 35±41
Dodd HM, Horn N, Gasson MJ (1990) Analysis of the genetic
determinant for production of the peptide antibiotic nisin. J Gen
Microbiol 136: 555±566
Don RH, Weightman AJ, Knackmuss HJ, Timmis KN (1985)
Transposon mutagenesis and cloning analysis of the pathways
for degradation of 2,3-dichlorophenoxy acetic acid and 3-
chlorobenzoate in Alcaligenes eutrophus JMP134 (pJP4).
J Bacteriol 161: 85±90
Duggleby CJ, Bayley SA, Worsey MJ, Williams PA, Broda P
(1977) Molecular sizes and relationships of TOL plasmids in
Pseudomonas. J Bacteriol 130: 1274±1284
Dunn NW, Dunn HM, Austen RA (1980) Evidence for the exis-
tence of two catabolic plasmids coding for the degradation of
naphthalene. J Gen Microbiol 117: 529±533
Dunn NW, Gunsalus IC (1973) Transmissible plasmid coding early
enzymes of naphthalene oxidation in Pseudomonas putida.
J Bacteriol 114: 974±979
Eaton RW (1996) p-Cumate catabolic pathway in Pseudomonas
putida F1: cloning and characterization of DNA carrying the
cmt operon. J Bacteriol 178: 1351±1362
Eaton RW, Timmis KN (1986) Characterization of a plasmid-
speci®ed pathway for catabolism of isopropylbenzene in Pseu-
domonas putida RE204. J Bacteriol 168: 123±131
Eberhard WG (1990) Evolution in bacterial plasmids and levels of
selection. Q Rev Biol 65: 3±22
Favaro R, Bernasconi C, Passini N, Bertoni G, Bestetti G, Galli E,
Deho G (1996) Organisation of the tmb catabolic operons of
Pseudomonas putida TMB and evolutionary relationship with
the xyl operons of the TOL plasmid pWWO. Gene 182:
189±193
Feng X, Ou LT, Ogram A (1997a) Plasmid mediated mineraliza-
tion of carbofuran by Sphingomonas sp. CF06. Appl Environ
Microbiol 63: 1332±1337
Feng X, Ou LT, Ogram A (1997b) Cloning and sequence analysis
of a novel insertion element from plasmids harbored by the
carbonfuran-degrading bacterium, Sphingomonas sp. CF06.
Plasmid 37: 169±179
Focht DD, Searles DB, Koh SC (1996) Genetic exchange in soil
between introduced chlorobenzoate degraders and indigenous
biphenyl degraders. Appl Environ Microbiol 62: 3910±3913
Fong KPY, Goh BH, Tan HM (1996) Characterization and ex-
pression of the plasmid-borne bedD gene from Pseudomonas
putida ML2, which codes for a NAD+-dependent cis-benzene
dihydrodiol dehydrogenase. J Bacteriol 178: 5592±5601
Fong PY, Goh C, Tan G, Tan HM (1997) Identi®cation and ge-
netic analysis of Tn5542, a transposable element carrying the
bedD and bedC1C2BA genes in Pseudomonas putida ML2. In:
Abstracts of the Sixth International Congress on Pseudomonas:
molecular biology and biotechnology 1997. Madrid, Spain, p 53
Frantz B, Chakrabarty AM (1986) Degradative plasmids in Pseu-
domonas. In: Sokatch JR (ed) The Bacteria, vol X. Academic
Press, London, pp 295±323
Fukumori F, Saint CP (1997) Nucleotide sequence and regulational
analysis of genes involved in conversion of aniline to catechol in
Pseudomonas putida UCC22 (pTDN1). J Bacteriol 179: 399±408
Fulthorpe RR, Wyndham RC (1991) Transfer and expression of
the catabolic plasmid pBRC60 in wild bacterial recipients in a
fresh water ecosystem. Appl Environ Microbiol 57: 1546±1553
Fulthorpe RR, Wyndham RC (1992) Involvement of a chloro-
benzoate-catabolic transposon, Tn5271, in a community ad-
aptation to chlorobiphenyl, chloroaniline and 2,4-
dichlorphenoxyacetic acid in a freshwater ecosystem. Appl
Environ Microbiol 58: 314±325
Galas DJ, Chandler (1989) Bacterial insertion sequences. In: Berg
DE, Howe MM (eds) Mobile DNA. American Society for
Microbiology. Washington, DC, pp 109±62
Gamas P, Chandler MG, Prentki P, Galas DJ (1987) Escherichia
coli integration host factor binds speci®cally to the ends of the
insertion sequence IS1 and to its major insertion hot-spot in
pBR322. J Mol Biol 195: 261±272
Grindley NDF, Reed PR (1985) Transpositional recombination in
prokaryotes. Annu Rev Biochem 54: 863±896
Grinsted J, De La Cruz F, Schmitt R (1990) The Tn21 subgroup of
bacterial transposable elements. Plasmid 24: 163±189
Gunsalus IC, Grund A, Yen KM (1982) Aromatic oxygenation,
genetics and regulation. In: Nozaki M, Yamamoto S, Ishimura
Y, Coon MJ, Ernster L, Eastbrook RW (eds) Oxygenases and
oxygen metabolism. Academic Press, London, pp 79±91

Hallet B, Sherratt DJ (1997) Transposition and site-speci®c re-
combination: adapting DNA cut-and-paste mechanisms to a
variety of genetic rearrangements. FEMS Microbiol Rev 21:
157±178
Hardman D, Gowland PC, Slater JH (1986) Large plasmids from
soil bacteria enriched on halogenated alkanoic acids. Appl
Environ Microbiol 51: 41±51
Haugland RA, Sangodkar UMX, Chakrabarty AM (1990) Re-
peated sequences including RS1100 from Pseudomonas cepacia
AC1100 function as IS elements. Mol Gen Genet 220:
222±228
Hermann H, Janke D, Krejsa S, Kunze I (1987) Involvement of the
plasmid pPGH1 in the phenol degradation of Pseudomonas
putida strain H. FEMS Microbiol Lett 43: 133±137
Hooper DJ, Kemp PD (1980) Regulation of enzymes of the 3,5-
xylenol degradative pathway in Pseudomonas putida: evidence
for a plasmid. J Bacteriol 142: 21±26
HoÄrak R, Kivisaar M (1998) Expression of the transposase gene
tnpA of Tn4652 is positively a€ected by integration host factor.
J Bacteriol 180: 2822±2829
Ichikawa H, Ikeda K, Amemura J, Ohtsubo E (1990) Two domains
in the terminal inverted-repeat sequence of transposon Tn3.
Gene 86: 11±17
Ishiguro N, Sato G (1984) Spontaneous deletion of citrate-utilizing
ability promoted by insertion sequences. J Bacteriol 160:
642±652
Ishiguro N, Sato G (1988) Nucleotide sequence of IS3411, which
¯anks the citrate utilization determinant of transposon Tn3411.
J Bacteriol 170: 1902±1906
Ishiguro N, Hirose K, Asagi M, Sato G (1981) Incompatibility of
citrate utilization plasmids isolated from Escherichia coli. J Gen
Microbiol 123: 193±196
Ishiguro N, Sato G, Sasakawa C, Danbara H, Yoshikawa M
(1982) Identi®cation of citrate utilization transposon Tn3411
from a naturally occurring citrate utilization plasmid. J Bac-
teriol 149: 961±968
Jacoby GA, Rogers JE, Jacob AE, Hedges RW (1978) Transposi-
tion of Pseudomonas toluene-degrading genes and expression in
Escherichia coli. Nature 274: 179±180
Jahnke M, Lehmann F, Schoebel A, Auling G (1993) Transposi-
tion of the TOL catabolic genes (Tn4651) into the degradative
plasmid pSAH of Alcaligenes sp. 0-1 ensures simultaneous
mineralisation of sulpho- and methyl-substituted aromatics.
J Gen Microbiol 139: 1959±1966
Jain RK, Bayly RC, Skurray Ra (1984) Characterization and
physical analysis of a 3,5-xylenol degradative plasmid in Pseu-
domonas putida. J Gen Microbiol 130: 3019±3028
Junker F, Cook AM (1997) Conjugative plasmids and the degra-
dation of arylsulfonates in Comamonas testosteroni. Appl En-
viron Microbiol 63: 2403±2410
Junker F, Saller E, Schla® Oppenberg HR, Kroneck PHM, Lei-
singer T, Cook AM (1996) Degradative pathways for p-toluate
and p-toluenesulfonate and their multicomponent oxygenases in
Comamonas testosteroni strains PSB-4 and T-2. Microbiology
142: 2419±2427
Kato K, Ohtsuki K, Mitsuda H, Yomo T, Negoro S, Urabe I
(1994) Insertion sequence IS6100 on plasmid pOAD2, which
degrades nylon oligomers. J Bacteriol 176: 1197±1200
Kawasaki H, Yahara N, Tonomura K (1981) Isolation and char-
acterization of plasmid pU01 mediating dehalogenation of ha-
loacetate and mercury resistance in Moraxella sp. B. Agric Biol
Chem 45: 1477±1481
Kawasaki H, Tsuda K, Matsushita I, Tunomura K (1992) Lack of
homology between two haloacetate dehalogenase genes encod-
ed on a plasmid from Moraxella sp. strain B. J Gen Microbiol
138: 1317±1323
Keshavarz T, Lilly MD, Clarke PH (1985) Stability of a catabolic
plasmid in continuous culture. J Gen Microbiol 131: 1193±1203
Kesseler M, Dabbs ER, Averho€ B, Gottschalk G (1996) Studies
on the isopropylbenzene 2,3-dixoygenase and 3-isopropyl cat-
echol 2,3-dioxygenase genes encoded by the linear plasmid of
Rhodococcus erythropolis BD2. Microbiology 142: 3241±3251
11
Kivisaar M, Horak R, Kasak L, Heinaru A, Habicht J (1990) Se-
lection of independent plasmids determining phenol degrada-
tion in Pseudomonas putida and expression of genes encoding
phenol monoxygenase and catechol-1,2-dioxygenase. Plasmid
24: 25±36
Kunz DA, Chapman PJ (1981) Isolation and characterization of
spontaneously occurring TOL plasmid mutants of Pseudomonas
putida HS1. J Bacteriol 146: 952±964
Lehrbach PR, Zeyer J, Reineke W, Knackmuss HJ, Timmis KN
(1984) Enzyme recruitment in vitro: use of cloned genes to ex-
tend the range of haloaromatics degraded by Pseudomonas sp.
strain B13. J Bacteriol 158: 1025±1032
Liu S, Su¯ita JM (1993) Ecology and evolution of microbial pop-
ulations for bioremediation. Trends Biotechnol 11: 344±352
Mars AE, Kasberg T, Kaschabek SR, Agtenen MH van, Janssen
DB, Reineke W (1997) Microbial degradation of chloroaro-
matics: Use of the meta-cleavage pathway for mineralization of
chlorobenzene. J Bacteriol 179: 4530±4537
Martin C, Timm J, Rauzier J, Gomez-Lus R, Davies J, Gicquel B
(1990) Transposition of an antibiotic resistance element in
mycobacteria. Nature 345: 739±743
Meer JR van der, Zehnder AJB, Vos WM de (1991b) Identi®cation
of a novel composite transposable element, Tn5280, carrying
chlorobenzene dioxygenase genes of Pseudomonas sp. strain
P51. J Bacteriol 173: 7077±7083
Meer JR van der, Vos WM de, Harayama S, Zehnder AJB (1992)
Molecular mechanisms of genetic adaptation to xenobiotic
compounds. Microbiol Rev 56: 677±694
Meer JR van der, Neerven ARW van, Vries EJ de, Vos WM de,
Zehnder AJB (1991a) Cloning and characterization of plasmid-
encoded genes for the degradation of 1,2-dichloro-,1,4-dichlo-
ro-, and 1,2,4-trichlorobenzene of Pseudomonas sp. strain P51.
J Bacteriol 173: 6±15
McClure NC, Venables WA (1987) pTDN1, a catabolic plasmid
involved in aromatic amine catabolism in Pseudomonas putida
mt-2. J Gen Microbiol 133: 2073±2077
Michels T, Cornelis G (1984) Detection and characterization of
Tn2501, a transposon included within the lactose transposon
Tn951. J Bacteriol 158: 866±871
Mizuuchi K (1992) Transpositional recombination: mechanistic
insights from studies of Mu and other elements. Annu Rev
Biochem 61: 1011±1051
Motsch S, Schmitt R (1984) Replicon fusion mediated by a single-
ended derivative of transposon Tn1721. Mol Gen Genet 195:
281±287
Nakatsu CH, Wyndham RC (1993) Cloning and expression of the
transposable chlorobenzoate-3,4-dioxygenase genes of Alcali-
genes sp. strain BR60. Appl Environ Microbiol 59: 3625±3633
Nakatsu C, Ng J, Singh R, Straus N, Wyndham C (1991)
Chlorobenzoate catabolic transposon Tn5271 is a composite
class I element with ¯anking class II insertion sequences. Proc
Natl Acad Sci USA 88: 8312±8316
Nakatsu CH, Providenti M, Wyndham RC (1997) The cis-diol de-
hydrogenase cbaC gene of Tn5271 is required for growth on 3-
chlorobenzoate but not 3,4-dichlorobenzoate. Gene 196: 209±218
Nakazawa T, Hayashi E, Yokota T, Ebina Y, Nakazawa A (1978)
Isolation of TOL and RP4 recombinants by integrative sup-
pression. J Bacteriol 134: 270±277
Nishi A, Tominaga K, Furukawa K (1997) Deletion and self-mo-
bilization of the chromosomal gene clusters coding for biphenyl
and salicylate metabolism in Pseudomonas putida KF715. In:
Abstracts of Sixth International Congress on Pseudomonas:
molecular biology and biotechnology 1997. Madrid, Spain,
p 125
Nishi A, Tominaga K, Furukawa K (1998) Horizontal transfer of
the chromosomal gene clusters coding for biphenyl and salicy-
late metabolism in Pseudomonas putida KF715. In: Abstracts of
First International Conference of the Federation of Asia-Paci®c
Microbiology Societies 1998. Singapore, p 92
Pertsova RN, Kunc F, Golovleva LA (1984) Degradation of 3-
chlorobenzoate in soil by pseudomonads carrying biodegrada-
tive plasmids. Folia Microbiol 29: 242±247
12
Pettigrew CA, Haigler BE, Spain JC (1991) Simultaneous biodeg-
radation of chlorobenzene and toluene by a Pseudomonas
strain. Appl Environ Microbiol 57: 157±162
Ploeg J van der, Willemsen M, van Hall G, Janssen DB (1995)
Adaptation of Xanthobacter autotrophicus GJ10 to bromoace-
tate due to activation and mobilization of the haloacetate de-
halogenase gene by insertion element IS1247. J Bacteriol 177:
1348±1356
Pritchard PH, Costa CF (1991) EPA's Alaska oil spill bioremedi-
ation project. Environ Sci Technol 25: 372±379
Ramos JL, Wasserfallen A, Rose K, Timmis KN (1987) Rede-
signing metabolic routes: manipulation of TOL plasmid path-
way for catabolism of alkylbenzoates. Science 235: 593±596
Rauch PJG, Beerthuyzen MM, Vos VM de (1990) Nucleotide se-
quence of IS904 from Lactococcus lactis subsp. lactis strain
NIZO R5. Nucleic Acids Res 18: 4253±4254
Rauch PJG, Vos VM de (1992) Characterization of the novel nisin-
sucrose conjugative transposon Tn5276 and its insertion in
Lactococcus lactis. J Bacteriol 174: 1280±1287
Reddy BR, Shaw LE, Sayers JR, Williams PA (1994) Two identical
copies of IS1246, a 1275 base pair sequence related to other
bacterial insertion sequences, enclose the xyl genes on TOL
plasmid pWWO. Microbiology 140: 2305±2307
Rheinwald JG, Chakarabarty AM, Gunsalus IC (1973) A trans-
missable plasmid controlling camphor oxidation in Pseudo-
monas putida. Proc Natl Acad Sci USA 70: 885±889
Rojo F, Pieper DH, Engesser KH, Knackmuss HJ, Timmis KN
(1987) Assemblage of ortho cleavage route for simultaneous
degradation of chloro- and methylaromatics. Science 238:
1395±1398
Saint CP, McClure NC, Venables WA (1990) Physical map of the
aromatic amine and m-toluate catabolic plasmid pTDN1 in
Pseudomonas putida: location of a unique meta-cleavage path-
way. J Gen Microbiol 136: 615±625
SchloÈmann M (1994) Evolution of chlorocatechol catabolic path-
ways. Biodegradation 5: 301±321
Shanley MS, Harrison A, Parales RE, Kowalchuk G, Mitchell DJ,
Ornston LN (1994) Unusual G + C content and codon usage
in catIJF, a segment of the ben-cat supra-operonic cluster in the
Acinetobacter calcoaceticus chromosome. Gene 138: 59±65
Sherratt D (1989) Tn3 and related transposable elements: site-
speci®c recombination and transposition. In: Berg DE, Howe
MM (eds) Mobile DNA. American Society for Microbiology,
Washington DC, pp 163±184
Sinclair MI, Holloway BW (1991) Chromosomal insertion of TOL
transposons in Pseudomonas aeruginosa PAO. J Gen Microbiol
137: 1111±1120
Sinclair MI, Maxwell PC, Lyon BR, Holloway BW (1986) Chro-
mosomal location of TOL plasmid DNA in Pseudomonas put-
ida. J Bacteriol 168: 1302±1308
Slater JH, Lovatt D, Weightman AJ, Senior E, Bull AT (1979) The
growth of Pseudomonas putida on chlorinated aliphatic acids
and its dehalogenase activity. J Gen Microbiol 114: 125±136
Springael D, Kreps S, Mergeay M (1993) Identi®cation of a cata-
bolic transposon, Tn4371, carrying biphenyl and 4-chlorobi-
phenyl degradation genes in Alcaligenes eutrophus A5.
J Bacteriol 175: 1674±1681
Syvanen M (1988) Bacterial insertion sequences. In: Kucherlapati
R, Smith GR (eds) Genetic recombination. American Society
for Microbiology, Washington DC, pp 331±356
Tan HM, Mason JR (1990) Cloning and expression of the plasmid-
encoded benzene dioxygenase genes from Pseudomonas putida
ML2. FEMS Microbiol Lett 72: 259±264
Tan HM, Fong KPY (1993) Molecular analysis of the plasmid-
borne bed gene cluster from Pseudomonas putida ML2 and
cloning of the cis-benzene dihydrodiol dehydrogenase gene.
Can J Microbiol 39: 357±362
Tan HM, Tang HY, Joannou CL, Abdel Wahab NH, Mason JR
(1993) The Pseudomonas putida ML2 plasmid-encoded genes
for benzene dioxygenase are unusual in codon usage and low in
G + C content. Gene 130: 32±39
Thomas AW, Slater JH, Weightman AJ (1992a) The dehalogenase
gene dehI from Pseudomonas putida PP3 is carried on an un-
usual mobile genetic element designated DEH. J Bacteriol 174:
1932±1940
Thomas AW, Topping AW, Slater JH, Weightman AJ (1992b)
Localization and functional analysis of structural and regula-
tory dehalogenase genes carried on DEH from Pseudomonas
putida PP3. J Bacteriol 174: 1941±1947
Timmis KN, Ste€an RJ, Unterman R (1994) Designing microor-
ganisms for the treatment of toxic wastes. Annu Rev Microbiol
48: 525±557
Tomasek PH, Frantz B, Sangodkar UMX, Haugland RA,
Chakrabarty AM (1989) Characterization and nucelotide se-
quence determination of a repeat element isolated from a 2,4,5-
T degrading strain of Pseudomonas cepacia. Gene 76: 227±238
Tsuda M (1996) Catabolic transposons in pseudomonads. In:
Nakazawa T, Furukawa K, Haas D, Silver S (eds) Molecular
biology of pseudomonads. American Society for Microbiology,
Washington DC, pp 219±228
Tsuda M, Iino T (1987) Genetic analysis of a transposon carrying
toluene degrading genes on a TOL plasmid pWWO. Mol Gen
Genet 210: 270±276
Tsuda M, Iino T (1988) Identi®cation and characterization of
Tn4653, a transposon convering the toluene transposon Tn4651
on TOL plasmid pWWO. Mol Gen Genet 213: 72±77
Tsuda M, Iino T (1990) Naphthalene degrading genes on plasmid
NAH7 are on a defective transposon. Mol Gen Genet 223:
33±39
Tsuda M, Minegishi KI, Iino T (1989) Toluene transposons
Tn4651 and Tn4653 are class II transposons. J Bacteriol 171:
1386±1393
Werlen C, Kohler HPE, Meer JR van der (1996) The broad sub-
strate chlorobenzene dioxygenase and cis-chlorobenzene dihy-
drodiol dehydrogenase of Pseudomonas sp. strain P51 are linked
evolutionarily to the enzymes for benzene and toluene degra-
dation. J Biol Chem 271: 4009±4016
Williams PA, Sayers JR (1994) The evolution of pathways for ar-
omatic hydrocarbon oxidation in Pseudomonas. Biodegradation
5: 195±217
Williams PA, Worsey MJ (1976) Ubiquity of plasmids in coding for
toluene and xylene metabolism in soil bacteria: evidence for the
existence of new TOL plasmids. J Bacteriol 125: 818±828
Williams PA, Assinder SJ, Marco P, O'Donnell KJ, Poh CL, Shaw
LE, Winson MK (1992) Catabolic gene duplications in TOL
plasmids. In: Galli E, Silver S, Witholt B (eds) Pseudomonas:
molecular biology and biotechnology. American Society for
Microbiology, Washington DC, pp 341±352
Wyndham RC, Cashore AE, Nakatsu CH, Peel MC (1994a) Cat-
abolic transposons. Biodegradation 5: 323±342
Wyndham RC, Nakatsu C, Peel M, Cashore A, Ng J, Szilagyi F
(1994b) Distribution of the catabolic transposon Tn5271 in a
groundwater bioremediation system. Appl Environ Microbiol
60: 86±93
Yano K, Nishi T (1980) pKJ1, a naturally occurring conjugative
plasmid coding for toluene degradation and resistance to
streptomycin and sulphonamides. J Bacteriol 143: 552±560
Yen KM, Serdar CM (1988) Genetics of naphthalene catabolism in
pseudomonads. Crit Rev Microbiol 15: 247±268
 Virology 434 (2012) 210–221

Contents lists available at SciVerse ScienceDirect

Virology
journal homepage: www.elsevier.com/locate/yviro

Review

Pirates of the Caudovirales

Gail E. Christie a,n, Terje Dokland b
a
Department of Microbiology and Immunology, Virginia Commonwealth University School of Medicine, P.O. Box 980678, 1011 E Marshall St., Richmond, VA 23298-0678, USA
b
Department of Microbiology, University of Alabama at Birmingham, 845 19th Street South, BBRB 311, Birmingham, AL 35294, USA

a r t i c l e i n f o a b s t r a c t

Available online 3 November 2012 Molecular piracy is a biological phenomenon in which one replicon (the pirate) uses the structural
Keywords: proteins encoded by another replicon (the helper) to package its own genome and thus allow its
Molecular piracy propagation and spread. Such piracy is dependent on a complex web of interactions between the helper
Bacteriophage 80a and the pirate that occur at several levels, from transcriptional control to macromolecular assembly.
Staphylococcus aureus pathogenicity island The best characterized examples of molecular piracy are from the E. coli P2/P4 system and the S. aureus
mobilization SaPI pathogenicity island/helper system. In both of these cases, the pirate element is mobilized and
Bacteriophage P2 packaged into phage-like transducing particles assembled from proteins supplied by a helper phage
Satellite phage P4 that belongs to the Caudovirales order of viruses (tailed, dsDNA bacteriophages). In this review we will
Capsid assembly
summarize and compare the processes that are involved in molecular piracy in these two systems.
Size determination
& 2012 Elsevier Inc. All rights reserved.
Derepression
Transactivation
Interference
DNA packaging
SaPI

Contents

Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
Overview of the P2/P4 system . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
Mobilization of S. aureus pathogenicity islands (SaPIs) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
Derepression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
P4 immunity and derepression of P4 by P2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 212
P2 immunity and reciprocal derepression of P2 by P4 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
Derepression of SaPIs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
Transactivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
Assembly and capsid size determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 214
DNA packaging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
Interference . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
Conclusion and perspectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 218
Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 219

Introduction describe the P2/P4 system of bacteriophages, where the ‘‘pirate’’

phage/plasmid replicon P4 usurps the structural gene products
Inspired by stories of piracy in the South China Sea (Rosenberg, of an unrelated ‘‘helper’’ bacteriophage for its own propagation
2009), the term ‘‘molecular piracy’’ was coined by Bjørn Lindqvist (Christie and Calendar, 1990; Lindqvist et al., 1993). In fact, the
while he was on sabbatical in the Dokland lab in Singapore, to term ‘‘molecular piracy’’ was not entirely new, but has been
used previously to refer, variously, to the way by which viruses
take control of their hosts’ biosynthetic machinery (Flaitz and
n Hicks, 1998; Fujimuro et al., 2007), and to the acquisition of host
Corresponding address: Fax: þ1 804 828 9946.
E-mail addresses: christie@vcu.edu (G.E. Christie), genes during viral evolution (Ahuja and Murphy, 1993; Choi et al.,
dokland@uab.edu (T. Dokland). 2001; Sinkovics et al., 1998). Of course, by the first definition, all

0042-6822/$ - see front matter & 2012 Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.virol.2012.10.028
 G.E. Christie, T. Dokland / Virology 434 (2012) 210–221
Table 1
Comparison of the steps involved in molecular piracy by SaPIs and P4.
SaPI/80a
Derepression 80a Sri, Dut, gp15 bind and inactivate different SaPI Stl repressors P2 Cox activates immunity-insensitive transcription from P4 PLL promoter
Reciprocal No known mechanism
derepression
Mutual transactivation No known mechanism
Excision SaPI encoded. Derepression by helper required
Replication SaPI encoded Derepression by helper required
Capsid size redirection Internal scaffold; SaPI CpmA and CpmB
DNA packaging Headful packaging. SaPI-encoded TerS redirects specificity
SaPI Ppi blocks phage DNA packaging
viruses are pirates, since all viruses require functions supplied by
the host cell for their own propagation, and exchange of genetic
material is a fundamental mechanism in the evolution of viruses
and other organisms. Some viruses, such as HIV, vaccinia or
herpesviruses even incorporate host proteins into their virions
(Maxwell and Frappier, 2007; Ott, 2008), but such incorporation
tends to be incidental or play an auxiliary role, rather than serving
as an integral part of the viral structure.
In our definition, molecular piracy refers specifically to the
case in which one infectious genetic element (the ‘‘pirate’’) uses
the structural proteins encoded by a viral replicon (the ‘‘helper’’)
for assembly of its own virion. This characteristic distinguishes
the pirate/helper systems from the satellite viruses commonly
found in eukaryotes (Hu et al., 2009), or the recently described
‘‘virophage’’, which depends on and interferes with the replica-
tion of mimivirus (La Scola et al., 2008). Although these satellites
depend upon the helpers for their propagation, they encode their
own capsid proteins. Even hepatitis delta virus, which packages
its genome-containing nucleocapsids within a viral envelope
formed by glycoproteins encoded by a Hepatitis B virus helper,
encodes its own nucleocapsid protein (Sureau, 2006).
In the P2/P4 system, not only does the pirate depend on the
helper for structural proteins, but has the ability to redirect the
capsid assembly process to suit its own needs. As it turns out,
the P2/P4 system is not the only example of such a phenomenon.
More recently, a similar system was discovered in Staphylococcus
aureus, where genetic elements called pathogenicity islands
(SaPIs) are mobilized by specific helper phages (Lindsay et al.,
1998; Novick et al., 2010) and are packaged into phage-like
transducing particles using structural proteins supplied by the
helper phage (Poliakov et al., 2008; Tallent et al., 2007; Tormo
et al., 2008). These two molecular pirates are not degenerate
versions of their helpers, but rather independent replicons that
have evolved a highly specialized machinery to exploit helper
bacteriophages for their own benefit. An additional example of
molecular piracy has been described in Sulfolobus, where two
nonconjugative plasmids have been shown to exploit archaeal
fuselloviruses for packaging and spread. However, little is known
about the underlying molecular mechanisms in this system
(Arnold et al., 1999; Wang et al., 2007).
The focus of this review will be on the mechanisms used by
the P4-related elements and the SaPIs to manipulate their respec-
tive helper phages, which are members of the order Caudovirales—
tailed, double-stranded DNA (dsDNA) bacteriophages. The biology
of the P2/P4 system has been described in great detail in the
decades since its discovery (Christie and Calendar, 1990; Deho
and Ghisotti, 2005; Lindqvist et al., 1993). Reports elucidating
SaPI biology have a much briefer history, but there have been
significant recent advances in our understanding of the interac-
tions between these pathogenicity islands and their helpers in
S. aureus (Novick et al., 2010). The molecular piracy that takes
place in these systems involves several steps, typically including
211
P4/P2
P4 Epsilon binds and inactivates the P2 master repressor C
P2 Ogr and P4 Delta each activate both P2 and P4 late promoters
Independent of helper; P4 encoded. Derepression by helper enhances
Independent of helper, P4 encoded Derepression by helper enhances
External scaffold; P4 Sid P4 Psu provides additional stability
Cos-site packaging. No known P4-encoded functions
transcriptional activation, excision and replication of the pirate
DNA, and finally assembly and packaging of pirate DNA into
virus-like particles made from helper proteins. A variety of
interactions between the pirates and their helpers modulate
these processes, ranging from gene regulation to morpho-
genetic control (Table 1). In the following sections, we will
discuss each of these interactions separately and also outline
where the systems differ.
Overview of the P2/P4 system
Bacteriophage P2 was originally isolated from the Lisbonne &
Carre re strain of Escherichia coli by Bertani in 1951 and is a
member of the Myoviridae family of viruses, having an icosahedral
head (capsid) and a contractile tail, and a 33.6 kb double-stranded
DNA genome (Bertani, 1951; Bertani and Six, 1988; Nilsson
and Haggard Ljungquist, 2005). P2 is a so-called ‘‘non-inducible’’
phage; unlike l and many other prophages P2 is not mobilized by
UV light. Several other P2-related phages have also been shown to
function as helpers for P4, including PK (the original helper in the
strain from which P4 was isolated) (Six, 1963; Six and Klug,
1973), P3 (Lin, 1984) and coliphage 186 (Sauer et al., 1982). P2-
like prophages are common in the environment (Breitbart et al.,
2002) and are present in about 30% of strains in the E. coli
reference collection (Nilsson et al., 2004), in enterohemorrhagic
E. coli, and in a variety of other g-proteobacteria, including strains
of Salmonella, Pseudomonas, Serratia, Haemophilus, Vibrio, Yersinia
and the Burkholderia cepacia complex (Garcia et al., 2008;
Lynch et al., 2010; Nilsson and Haggard Ljungquist, 2005).
Most of the characterization of helper exploitation by P4 has
been carried out using P2, however, which will be the focus of
the discussion here.
P4 is an 11.6 kb replicon that can exist either as a plasmid or
integrated into the host genome like a prophage (Briani et al.,
2001; Deho and Ghisotti, 2005; Lindqvist et al., 1993). P4 is
genetically unrelated to P2, and while it has been described as a
satellite phage it is probably more appropriate to consider it as an
integrative plasmid that has acquired functions for helper phage
piracy. P4 lacks genes encoding major structural proteins and
requires all of the morphogenetic genes of its helper phage
(Six, 1975). A second P4-like element found in E. coli, retronphage
fR73, can also exploit P2 as a helper (Inouye et al., 1991). The
exploitation of P2 by P4 can take place under a variety of circum-
stances, including P2 infection of a strain carrying P4 in either the
immune-integrated or multicopy plasmid state, P4 infection of
a P2 lysogen, and coinfection by both phages. In each of these
scenarios, P4 responds to the presence of the helper phage by
interacting with certain phage-encoded functions and by activat-
ing P4 functions that allow it to manipulate the helper phage
appropriately. The nature and timing of the regulatory crosstalk
212 G.E. Christie, T. Dokland / Virology 434 (2012) 210–221
between P4 and its P2 helper depends on the infection conditions,
and appears to be designed to optimize P4 reproduction.
Mobilization of S. aureus pathogenicity islands (SaPIs)
SaPIs are a family of 14–27 kb genetic elements that are
integrated into the S. aureus host genome and contain phage-
like repressor, integrase and terminase genes. Different SaPIs also
express a variety of superantigen toxins and other virulence
and antibiotic resistance factors. Ten of the 17 identified to date
in staphylococcal genomes have been shown to be inducible by
either known or endogenous prophages (Novick et al., 2010). The
two best characterized SaPIs are SaPI1 and SaPIbov1, found in S.
aureus strains RN4282 and RF122, respectively (Novick et al.,
2010). SaPI1 carries genes for the toxic shock syndrome toxin 1
(tst) and enterotoxins K (sek) and Q (seq) (Ruzin et al., 2001).
SaPIbov1 carries tst as well as genes encoding enterotoxins C (sec)
and L (sel), and is associated with bovine pathogenic S. aureus
(Fitzgerald et al., 2001). Some SaPIs (SaPI2, SaPI5) are found in
clinically important MRSA strains, including USA200 and USA300
(Highlander et al., 2007). Related elements have also been
identified in other staphylococcal species (Kuroda et al., 2005,
Takeuchi et al., 2005) and in streptococci (Scott et al., 2012).
While normally repressed and stably integrated in the host
genome, SaPIs become activated and mobilized when a compa-
tible helper bacteriophage enters the lytic cycle, whether through
infection or by induction of an endogenous prophage. Upon
mobilization, the SaPI genomes are packaged into transducing
particles formed by structural proteins encoded by the helper
phage (Tallent et al., 2007, Tormo et al., 2008). Helper phages for
SaPI mobilization belong to a large family of temperate transdu-
cing phages found in S. aureus that are members of the family
Siphoviridae, with dsDNA genomes ranging from 39.6 to 45.9 kb
(Kwan et al., 2005). Several helper phages for different SaPIs have
been described, including f11, 53, 80, 80a, and fNM1 (Christie
et al., 2010; Dearborn and Dokland, 2012; Lindsay et al., 1998).
The phage-induced mobilization of SaPIs is specific: thus, SaPI1
can be mobilized by phage 80a but not by the closely related f11,
nor the more distantly related phage 80; SaPI2 can be mobilized
both by phage 80 and by 80a. SaPIbov1 can be mobilized by f11
and 80a, but not by 80 (Christie et al., 2010). f13 is able to induce
SaPI1 excision and replication, but not to package the genome
into phage particles (Lindsay et al., 1998). The helper/SaPI
specificity is manifested at several levels, which will be discussed
in greater detail in the following sections.
Derepression
SaPIs are normally integrated stably in their host genome in a
repressed state, as is the case with P4 in the immune-integrated
state. For both of these elements, an essential first step in
mobilization is relief of repression by a helper phage-encoded
function. Each has evolved a mechanism that exploits helper
phage genes which also perform other roles in the phage life
cycle. In addition, P4 has the ability to derepress a resident helper
prophage (mutual derepression).
P4 immunity and derepression of P4 by P2
P4 immunity involves a unique mechanism in which a short,
stable RNA (CI RNA) regulates transcription termination through
sequence-specific binding. Leftward transcription of the P4 a
operon (Fig. 1), which encodes functions required for both
plasmid and lytic growth, initiates from two different promoters
Fig. 1. Mutual derepression in P2 and P4. Maps depict the portions of the P2 (top)
and P4 (bottom) genomes encoding the transcripts involved in lysogeny control
and DNA replication. Transcripts and their promoters are indicated by arrows
below the map; the dotted line shows the part of the early leftward transcript
that is no longer synthesized when P4 establishes lysogeny. The repressors and
their targets in both genomes are shown by the red lines. Genes encoding the
derepression functions are designated with a yellow circle, and the targets on
which they act are denoted by green lines to indicate activation of transcription.
The arrowhead indicates that Cox acts at the PLL promoter to stimulate transcrip-
tion, while the perpendicular line indicates that Epsilon blocks the action of P2 C,
thereby turning on Pe transcription.
(Deho et al., 1988). One of these is constitutive and the other is
subject to complex regulatory control. In a P4 lysogen, transcription
from the constitutive early leftward promoter, PLE, yields a transcript
of  300 nucleotides that terminates upstream of the P4 replication
genes (Briani et al., 2000; Deho et al., 1992) and is subsequently
processed to generate the CI RNA (Forti et al., 2002). Termination
depends upon RNA–RNA interactions between the CI RNA and two
specific target sequences in the untranslated leader region of the
nascent transcript. (Sabbattini et al., 1995). In order to overcome P4
immunity, this RNA-mediated termination must be circumvented.
This is accomplished by initiation of transcription about 400 bp
upstream of PLE from a second promoter, PLL. Translation of two
nested genes in this longer transcript leads to translational suppres-
sion of the CI RNA-mediated transcription termination and therefore
expression of the P4 replication functions (Forti et al., 1999).
Derepression of immune-integrated P4 by a P2 helper phage
requires activation of transcription from PLL, which bypasses the
P4 immunity system and leads to transcription of the P4 replica-
tion genes (Saha et al., 1989). This requires the product of the
P2 cox gene (Saha et al., 1989). Activation of PLL by Cox also leads
to the induction of P4 prophage excision that is observed upon
P2 infection of a P4 lysogenic strain (Six and Lindqvist, 1978).
Excision requires the P4 Vis protein, which is encoded by the first
reading frame in the PLL transcript (Cali et al., 2004). Derepression
of P4 following P2 infection appears to be primarily a mechanism
for allowing P4 to survive P2- mediated host cell death rather
than a mode for efficient horizontal transfer, since the yield of P4
when P2 infects a P4 lysogen is normally quite low, less than 1 P4
per cell (Six and Lindqvist, 1978). However, if P2 replication is
blocked (by infection of a bacterial host that lacks the rep gene, or
by mutation in a phage replication function) the yield of P4
increases to a level comparable to that seen during P4 infection of
a P2 lysogen (Six and Lindqvist, 1978).
 G.E. Christie, T. Dokland / Virology 434 (2012) 210–221
Fig. 2. SaPI derepression by multiple helper phage genes. The known derepression functions are identified by yellow circles on a map of the 80a genome. Shown below are
the non-homologous immunity regions from three different SaPIs, with the rightward str promoter repressed by the product of the stl gene, as indicated by the red lines.
Each of the phage derepression functions activates str expression from a different SaPI by interfering with Stl repression, as indicated by the green lines.
P2 immunity and reciprocal derepression of P2 by P4
When the P2 helper is present as a prophage, P4 is able to
derepress it to activate expression of required helper functions.
P4 infection of P2 lysogens gives rise to about 100 P4 and about
10  3 P2 per infected cell (Six and Klug, 1973). As in other temperate
phages, P2 early transcription initiates from a pair of divergent
promoters encoding competing repressors that regulate the lysogeny
functions (Fig. 1). The leftward transcript encodes the P2 immunity
repressor, C, and the phage integrase, while the rightward transcript
includes genes encoding the repressor of the lysogenic promoter
(Cox), as well as the replication functions. C regulates its own
promoter and blocks expression of Cox, while Cox blocks expression
of C (Saha et al., 1987). P2 Cox is a remarkable protein with multiple
roles; it functions not only as the repressor of the lysogenic promoter
but also as the recombination directionality factor for prophage
excision (Yu and Haggard-Ljungquist, 1993) and, as discussed above,
positively regulates the P4 PLL promoter to derepress P4.
The derepression of P2 by P4 requires the P4 e gene product
(Geisselsoder et al., 1981; Liu et al., 1997). Epsilon binds to the P2
immunity repressor and interferes directly with binding of the
repressor to its operator (Liu et al., 1998). This leads to expression
of the helper early genes and to in situ replication of the P2
prophage, which does not excise (Six and Lindqvist, 1978). The e
gene is essential for P4 growth in a P2 lysogen, but not during a
P2þP4 co-infection of a nonlysogenic cell. However, Epsilon does
appear to contribute to interference with growth of the helper
phage during a coinfection (Diana et al., 1978). The interaction
between Epsilon and the phage repressor determines whether P4
can use a lysogenic helper phage. The P2-related phage 186,
which has morphogenetic genes similar to those of P2 but an
unrelated repressor (Kalionis et al., 1986), cannot be derepressed
by P4 and can only serve as a P4 helper if it is growing lytically
(Sauer et al., 1982).
Derepression of SaPIs
In the absence of helper phage, SaPIs are maintained in a
stable repressed state by a master repressor, Stl. Like prophage
repressors, Stl binds to a region between two divergent promoters
where it inhibits most SaPI gene expression. Inactivation of stl by
mutation leads to SaPI excision and replication (Ubeda et al.,
2008). Thus, derepression by the helper phage is a key regulatory
step in SaPI mobilization. Remarkably, the Stl proteins of different
SaPIs are widely divergent, and the ability of a particular helper
phage to derepress a given SaPI appears to be a primary determi-
nant of helper phage-SaPI specificity (Tormo-Mas et al., 2010).
213
Three different derepression proteins encoded by phage 80a have
been identified, and each targets a different SaPI (Fig. 2). All of
these proteins share a common mechanism; they act as anti-
repressors by direct binding and inhibition of their respective
Stl proteins (Harwich, 2009; Tormo-Mas et al., 2010). SaPI1 is
derepressed by Sri, the product of 80a ORF22, SaPIbov1 is
derepressed by Dut, the product of 80a ORF32, and SaPIbov2 is
derepressed by the product of 80a ORF15. Each of these genes is
nonessential for phage growth but required for mobilization of
the respective SaPI (Tormo-Mas et al., 2010). Two of these phage-
encoded antirepressors have other known functions. Sri was
previously identified in the related phage 77 as a protein that
inhibited host DNA replication by binding to DnaI (Liu et al.,
2004), while Dut is a dUTPase (Tormo-Mas et al., 2010). Like P4,
SaPIs have apparently evolved to sense the presence of a helper
phage by exploiting genes that play another role in the biology of
the phage.
Transactivation
Both P4 and the SaPIs depend upon their helper phages for
gene products needed for virion assembly, DNA packaging, and
cell lysis. During lytic growth of the helper phages, these func-
tions are expressed late in infection as part of the normal
temporal regulation of the phage morphogenetic genes. In the
P2/P4 system, at least, there is a second set of reciprocal interac-
tions that regulate late gene transcription, allowing P4 to opti-
mize exploitation of the helper phage under the different
conditions it might encounter. This is accomplished by a pair of
related transcriptional activators encoded by P2 and P4 that
recognize the same promoters on both genomes but differ in
the efficiencies with which they activate gene expression.
The P2 morphogenetic genes, encoding the head, tail, packa-
ging and lysis functions, lie in four operons expressed late in
infection (Fig. 3). P2 late gene transcription requires the product
of the phage ogr gene, a transcriptional activator that binds to a
site about 55 bp upstream of the initiation sites for the four P2
late promoters (Christie and Calendar, 1985; Christie et al., 2003)
and interacts with the C-terminal domain of the a subunit(s) of E.
coli RNA polymerase (Ayers et al., 1994; Sunshine and Sauer,
1975; Wood et al., 1997). Ogr belongs to a family of zinc-binding
transcription factors found almost exclusively among P2-related
phages and their satellites, with a C2C2 motif essential for metal
binding and activity (Julien et al., 1998; Pountney et al., 1997).
P4 has two operons that are expressed during lytic growth
(Fig. 3), and the two P4 late promoters have the same conserved
214 G.E. Christie, T. Dokland / Virology 434 (2012) 210–221
Fig. 3. Mutual transactivation in P2 and P4. The four P2 and two P4 late transcription units are indicated by black arrows below their respective genetic maps. The genes
for the late transcription factors ogr (P2) and d (P4) are identified by yellow circles on the maps. Both proteins activate transcription from the same P2 and P4 late
promoters, as indicated by the green arrows.
sequence element found upstream of the P2 late promoters.
The leftward PLL promoter is the same promoter that is dere-
pressed by P2 Cox to initiate P4 excision and replication from
the prophage state. The second late promoter, Psid, regulates
rightward transcription of three genes involved in helper exploi-
tation: sid, d and psu. Sid and Psu play roles in P4 capsid assembly
(see below). The third gene product, Delta, is an Ogr homolog that
activates transcription from the two P4 late promoters and the
four P2 late promoters. Likewise, Ogr activates transcription from
the two P4 late promoters as well as the four P2 late promoters
(Dale et al., 1986; Deho et al., 1988; Halling and Calendar, 1990).
Although there is extensive similarity among proteins in the
P2 Ogr family, they fall into two functionally discrete classes.
Members of the ‘‘helper’’ class, exemplified by Ogr, activate the P4
late promoters better than the P2 late promoters. Members of the
‘‘satellite class,’’ exemplified by the Delta proteins of P4 and fR73,
activate the P2 late promoters better than the P4 late promoters
and are able to cause transcription in the absence of replicating P2
DNA (Julien and Calendar, 1996; McAlister et al., 2003). These
differences contribute to earlier expression of P4 late genes in the
presence of a P2 helper and maximize expression of the P2 late
genes in the presence of P4. They also allow P4 to activate directly
the transcription of the P2 morphogenetic genes required for
packaging and lysis, bypassing their normal requirement for P2
DNA replication.
There is at this point no evidence to suggest a similar set of
complex, reciprocal interactions as a general mechanism regulat-
ing helper phage exploitation by SaPIs. In contrast to the P2/P4
system, infection of a helper phage lysogen by a SaPI-containing
particle has not been reported to lead to a burst of progeny SaPI
virions. Helper phage late transcription, studied in most detail for
80a and f11, appears to initiate from a single late promoter that
is activated by the RinA transcription factor, which is encoded by
a gene that lies immediately upstream of the late operon. Deletion
of rinA eliminates phage production and essentially eliminates
SaPI1 transduction by 80a (Ferrer et al., 2011). This argues that
SaPI1 does not encode a function that can replace rinA in helper
phage late gene transcription. Consistent with this, no increase in
80a late transcription was detected following prophage induction
in the presence of SaPI1 (Harwich, 2009). However, there is still
significant residual transduction of SaPIbov1 by both 80aDrinA
and f11DrinA (Ferrer et al., 2011), suggesting that unlike SaPI1,
SaPIbov1 may encode a function that can activate helper phage
late transcription to some extent.
RinA does not appear to have a reciprocal influence on SaPI1
transcription. The SaPI genes involved in capsid size determination
and packaging specificity (discussed below) lie in a six-gene
operon designated as operon 1, which is preceded by a LexA-
regulated promoter (Ubeda et al., 2007). During 80a infection,
transcription of these genes in SaPI1 requires derepression of the
SaPI and initiates farther upstream, at a promoter that has not yet
been identified (Harwich, 2009). Transcription from the Lex-A
regulated promoter would lead to a burst of operon 1 expression
during SOS induction of a resident helper prophage, which might
improve SaPI yield but is not essential for mobilization. A f11DrinA
mutant did not show any impairment in transcription of SaPIbov1
operon 1, even under conditions where transcription from the
LexA-regulated promoter was blocked by mutation (Ferrer et al.,
2011). This indicates that the helper phage RinA transcription
factor does not play a direct role in controlling SaPI operon
1 expression.
Assembly and capsid size determination
Tailed, dsDNA bacteriophages of the Caudovirales assemble
their capsids (or heads) as empty precursors—procapsids—from
the major capsid protein (CP), typically requiring a scaffolding
protein (SP) that acts as a chaperone for the assembly process
(Fig. 4) (Dokland, 1999; Fane and Prevelige, 2003). The main
exception to the requirement for SP is the HK97-like phages, in
which an N-terminal sequence in CP appears to serve this role
(Conway et al., 1995; Duda et al., 1995). During DNA packaging,
the capsid undergoes expansion accompanied by major confor-
mational changes in CP (Johnson, 2010). Tail structures (and
sometimes ‘‘decoration’’ proteins) are added to the finished
capsid. Capsids are either icosahedral or elongated with icosahe-
dral caps, and—in spite of weak or undetectable sequence
homology—all members of the Caudovirales studied to date share
a characteristic and unique CP fold, called the HK97-like fold
(Johnson and Chiu, 2007, Wikoff et al., 2000).
One of the most striking features about the piracy both in the
P2/P4 system and in the mobilization of SaPIs is the redirection
of the helper phage assembly pathway to form capsids that are
about 1/3 the size (45 nm, T¼ 4) of those normally made by the
phage itself (60 nm, T¼7), commensurate with the difference in
size of the genomes (Figs. 4 and 5A) (Dearborn et al., 2011;
Dearborn et al., 2012; Dokland et al., 1992; Ruzin et al., 2001;
Spilman et al., 2011). The small capsids are unable to package
complete phage P2 genomes, thus this redirection of the assembly
pathway strongly interferes with P2 multiplication.
 G.E. Christie, T. Dokland / Virology 434 (2012) 210–221 215

Fig. 4. Comparison of assembly pathways for P2/P4 (A) and 80a/SaPI1 (B). In each panel the top pathway is the helper phage and the bottom pathway is the pirate. Phage
procapsids are assembled from the major capsid protein (gpN in P2, gp47 in 80a; yellow), scaffolding protein (P2 gpO, 80a gp46; red), and portal protein (P2 gpQ, 80a
gp42; green). 80a also incorporates a minor capsid protein (gp44; cyan) that may play a possible role in stabilizing the DNA in the capsid. The presence of the P4 Sid (panel
A) or SaPI CpmA and CpmB (panel B) proteins (orange) lead to the formation of small procapsids. Sid forms an external scaffold while CpmB forms an internal one; the
location of CpmA in the small SaPI1 procapsids is currently unknown. DNA is packaged into procapsids by terminase complexes (gpP and gpM for P2, TerS and TerL for
80a), concomitant with removal of the scaffolding proteins and expansion of the capsid. In P2, only the C-terminal half of gpO (DO) is removed, leaving the O* protease
domain inside the capsid. P4 Psu is added to small capsids as a decoration protein.

How do the pirate elements carry out this size change? In P4,
the size redirection depends on a P4 size determination gene,
sid (Barrett et al., 1976), which encodes an external scaffolding
protein that forms an external dodecahedral cage around the
procapsid (Fig. 4A) (Marvik et al., 1995). Sid is an elongated
protein made up of bundles of a-helices (Fig. 5B) (Dearborn et al.,
2012). Trimers of Sid connect hexamers of the gpN capsid protein
across the threefold axes, forcing the shell into a T¼ 4 architecture
(Fig. 6A). P4 sid mutants fail to form small capsids, and while P4
DNA can still get packaged as dimers or trimers into large capsids,
the efficiency is low (Shore et al., 1978). Mutants in gpN, called sir
(sid responsiveness) render the capsid protein resistant to the Sid-
induced size redirection and thus do not form small capsids (Six
et al., 1991). These mutations are clustered in an external loop in
the gpN CP, where they presumably interfere with gpN–Sid
interactions (Fig. 5B) (Dearborn et al., 2012). Conversely, muta-
tions in Sid, named super-sid or nms (N mutation sensitive) (Kim
et al., 2001), which are clustered in a C-terminal a-helix,
(Dearborn et al, 2012) recover the ability of Sid to form small
capsids even on a P2 Nsir background.
Expression of gpN and Sid alone is sufficient to efficiently form
small procapsids (Dokland et al., 2002), although the gpO SP is
incorporated when both proteins are present (Fig. 6A) (Wang
et al., 2006). However, gpO is required for the formation of viable
P4 phage (Six, 1975), presumably due to other functions of gpO, in
particular the protease activity that resides in its N-terminal
domain, On, which remains inside the mature capsids (Fig. 4A)
(Chang et al., 2008, 2009; Dokland, 2012). Indeed, the mutant
Oam279, which lacks the C-terminal 47 amino acids and is
defective in scaffolding activity, retains protease activity and is
viable in the presence of Sid (Agarwal et al., 1990).
The P4-encoded psu (polarity suppression) gene product,
which acts as a suppressor of rho-dependent transcription termi-
nation (Pani et al., 2009; Sauer et al., 1981), also serves a role as a
decoration protein that is added to the outside of the completed
capsid (Fig. 4A) (Dokland et al., 1993). Psu apparently stabilizes
the inherently less stable P4 capsids against environmental stress
(Isaksen et al., 1993).
Size determination by SaPIs works differently. In the most well
described system—SaPI1 mobilized by phage 80a—two SaPI1
proteins, gp6 and gp7, are both required for efficient small capsid
formation (Fig. 4B) (Damle et al., 2012; Poliakov et al., 2008).
Homologous proteins are found in most, but not all, SaPIs, and the
corresponding capsid morphogenesis genes have been named
cpmA (gp7) and cpmB (gp6) (Damle et al., 2012; Dearborn and
Dokland, 2012; Ram et al., 2012). These two proteins, CpmA and
CpmB, are sufficient to induce small capsid formation when
expressed during phage infection or upon co-expression with just
CP and SP in a S. aureus co-expression system (Damle et al., 2012;
Ram et al., 2012; Spilman et al., in press).
Unlike the P2/P4 system, there is no Sid-like external scaffold-
ing protein. Instead, SaPI1 procapsids contain internal fingerlike
216 G.E. Christie, T. Dokland / Virology 434 (2012) 210–221

Fig. 5. (A) Isosurface representations of cryo-EM reconstructions of P2, P4, 80a and SaPI1 procapsids, and radially colored from red (inside) to blue (outside) (Spilman et al,
2011; Dearborn et al, 2011; Dearborn et al, 2012). For 80a and SaPI1, the right halves show a cutaway view of the interior of the procapsids, revealing the internal
protrusions in SaPI1 corresponding to CpmB. (B) Closeup view of the P4 hexamer (ivory isosurface with three copies of gpN fitted in, shown as blue, red, and yellow
ribbons) with the density corresponding to Sid shown in red. The Nsir mutations are indicated as purple balls on one gpN monomer. (C) Ribbon representation of the NMR
structure of two SaPI1 gp6 (CpmB) dimers (orange and yellow; Dearborn et al., 2011) fitted into the internal protrusions in the SaPI1 procapsid reconstruction, shown as a
solid isosurface. The predicted C-terminal a-helices are shown in pink for one subunit of each CpmB dimer. The gp47 capsid protein model is shown in green. The 3D
reconstructions were generated using AUTO3DEM (Yan et al, 2007). Rendering and fitting of the maps was done in UCSF Chimera (Pettersen et al, 2004).

Fig. 6. Models for capsid size determination. (A) The P2 internal scaffolding protein gpO (red) promotes assembly of the gpN capsid protein (yellow) through dimerization
and specific interactions with gpN (top panel). The cylinders represent the C-terminal a-helical domain, while the bullets indicate the N-terminal protease domain of gpO.
In the presence of P4 Sid (orange), gpN is tethered at the threefold (triangle) and twofold (ovals) symmetry axes, forcing the formation of a smaller capsid. (B) In the 80a/
SaPI1 system, the gp46 scaffolding protein, which forms an internal core, is also believed to interact with the capsid protein (gp47) through a predicted C-terminal a-helix.
The C-terminal a-helices of the SaPI1-encoded CpmB protein dimers (orange) compete with gp46 for the same binding site on gp47. CpmA (pink) may be required to
remove gp46 in order to provide access for CpmB.

projections absent from the helper phage procapsids (Fig. 5A).
CpmB, which has a structure similar to that of the SP of bacterio-
phage f29, acts as an internal scaffolding protein (Dearborn et al.,
2011, Morais et al., 2003). CpmB binds as a dimer to the inside of
the SaPI1 shell (Fig. 5C) and most likely competes with the
cognate 80aSP for the same binding site on the 80a CP (Fig. 6B).
The role of CpmA in size determination is less clear. CpmA is
only present in procapsids in small amounts, suggesting that its
action is transient in nature (Poliakov et al., 2008). Deletion of
cpmB in SaPI1 led to the formation of a large number of non-
isometric ‘‘monsters’’ (Damle et al., 2012; Dearborn et al., 2011),
and while CpmB alone could promote small capsid formation in
 G.E. Christie, T. Dokland / Virology 434 (2012) 210–221 217

the absence of SP, CpmA had an inhibitory effect on capsid
assembly (Spilman et al., in press). Small procapsids lack the
internal scaffolding core that can be seen in reconstructions of
large procapsids (Spilman et al., 2011). The role of CpmA may be
to reorganize the scaffolding core that would otherwise prevent
small capsid formation, or to bind SP to allow access to binding
sites on CP by CpmB (Fig. 6B).
Size redirection depends on compatibility between CpmA/
CpmB and the helper capsid proteins. SaPI2, for example, forms
small capsids when mobilized by phage 80a, but not by phage 80,
presumably due to incompatibility with the phage 80 CP, which
shares only 16% sequence identity with that of 80a (Christie et al.,
2010). Size determination of SaPIbov1 by 80a also appears to be
less efficient than for SaPI1 even though the CpmA and CpmB
proteins are almost identical (Dearborn and Dokland, 2012).
Other factors, including relative protein expression levels, may
also play a role in this process.
It should be pointed out that capsid size redirection is not
absolutely essential in either of these systems. P4 sid mutants are
viable, although reduced in burst size (Diana et al., 1978; Shore
et al., 1978). SaPI1 cpmAB mutants are also viable, and appear to be
transduced at normal frequency (Damle et al., 2012; Ram et al.,
2012). Furthermore, in some phage/SaPI systems, size redirection
does not occur. For example, SaPIbov2 (27 kb) and SaPIbov5 do
not contain cpmAB homologs, and do not produce small capsids
(Novick et al., 2010; Ram et al., 2012). However, the fact that the
cpmAB genes are highly conserved when present and always come
together (Novick et al., 2010) suggest that they do confer an
evolutionary advantage—presumably by interfering with helper
phage growth. Both sid and cpmAB mutants have lost the ability to
interfere with their helper phages (Damle et al., 2012; Diana et al.,
1978; Ram et al., 2012), and SaPIs that lack size redirection have
other interference mechanisms, as discussed below.
DNA packaging
In the Caudovirales, DNA is packaged into the procapsids through
a ring-shaped portal at one fivefold vertex in an ATP-dependent
process that requires a terminase complex, which consists of small
(TerS) and large (TerL) subunits (Black, 1989; Feiss and Rao, 2012;
Fujisawa and Morita, 1997). The large terminase subunit is respon-
sible for prohead binding, DNA translocation and DNA cleavage,
while the small subunit is involved in DNA recognition and binding
(Catalano, 2005; Feiss and Rao, 2012; Roy et al., 2012; Teschke,
2012). P4 and SaPIs have evolved different strategies to exploit the
DNA packaging machinery of their helper phages. P4 has simply co-
opted the P2 packaging machinery by incorporating the same
packaging signals into its own genome. P2 and P4 contain identical
55 bp cos site sequences that include the 19 bp cohesive ends found
in virion DNA (Ziermann and Calendar, 1990). DNA packaging and
cos site-specific cleavage requires the small (gpM) and large (gpP)
terminase subunits as well as procapsids (Pruss et al., 1975; Bowden
and Modrich, 1985). For both genomes, covalently closed circular
DNA molecules are the preferred packaging substrate, unlike the
linear concatemers preferred by most bacteriophages (Black, 1989;
Bowden and Modrich, 1985; Fujisawa and Morita, 1997; Pruss and
Calendar, 1978).
SaPIs, in contrast, redirect the specificity of the DNA packaging
machinery of their helpers (Fig. 7). Like their helper phages, SaPIs
replicate as linear concatemers, and are packaged as headfuls,
resulting in virion DNA that is terminally redundant and partially
circularly permuted (Ruzin et al., 2001). The phage TerS protein
recognizes a pac site on the phage genome that lies within the terS
coding sequence (KD Lane, EK Read, GEC; unpublished), as is the case
for the pac site of several other phages that use headful packaging,
including P22 (Wu et al., 2002) and PY100 (Schwudke et al., 2008). An
initial cut is then followed by several rounds of headful packaging.
In the presence of the SaPI, a SaPI-encoded TerS subunit
together with the phage-encoded TerL directs the specific clea-
vage and packaging of SaPI DNA by binding to a SaPI-specific
pac sequence that lies in an intergenic region upstream of
the operon that encodes SaPI terS (JC Bento, KD Lane, EK Read,
GEC; unpublished). The SaPI-encoded TerS is required for high
frequency transduction for both SaPI1 and SaPIbov1, while the
phage-encoded TerS is required for packaging of helper phage
DNA (Ubeda et al., 2009).
The compatibility of the SaPI-encoded TerS with the helper
phage TerL likely accounts for some of the observed SaPI-helper
specificity. For example, phage f13, a cos site phage, can induce
SaPI1 excision and replication but fails to produce SaPI1 transdu-
cing particles (Ruzin et al., 2001). This is presumably due to an
inability to form a functional hybrid between the cos-site based
DNA packaging machinery of the phage and the pac site-based
TerS subunit of SaPI1.
Some SaPIs also influence DNA packaging at another level,
by interfering directly with the packaging of helper phage DNA.
This novel mechanism requires the SaPI ppi (phage packaging

Fig. 7. Model for SaPI packaging redirection. Specific pac sites on the concatemeric phage DNA (top) are recognized by the phage-encoded small terminase (TerSphage; pink)
and packaged into procapsids through the action of the phage-encoded large terminase subunit (TerL). The DNA is cleaved when the capsid is full and the DNA is ready for
another round of packaging. Phage DNA can also be packaged into small capsids, but since the DNA will be only a fragment of the genome, the resulting virions are not
viable. The pac sites on SaPI DNA (bottom) are recognized specifically by the SaPI-encoded TerSSaPI subunit (blue), which also interacts with the phage-encoded TerL for
DNA packaging. While for most SaPIs the majority of capsids formed will be small, any SaPI DNA packaged into large capsids will be multimeric and able to be transduced.
Some SaPIs also encode an interference factor, Ppi, which specifically prevents packaging of phage DNA. Coloring of capsid proteins is as in Fig. 4B.
218 G.E. Christie, T. Dokland / Virology 434 (2012) 210–221
interference) gene (originally called pif; (Tormo-Mas et al., 2010)),
which encodes a protein that binds directly to the phage TerS
protein and blocks packaging of phage DNA (Ram et al., 2012). The
known Ppi proteins fall into two conserved subsets, each
of which appears to target a different phage small terminase
superfamily (Ram et al., 2012).
Interference
Both P4 and SaPIs interfere with the multiplication of their
helper phages. In the case of P4, capsid size determination
appears to be the primary interference mechanism. Interference
with P2 by P4 can range from 5- to 10-fold in a simultaneous co-
infection to greater than 500-fold if P4 is given a ten minute
head start or is present as a multicopy plasmid (Diana et al.,
1978; Deho and Ghisotti, 2005). Although there is some evidence
that a still unidentified P4 function may augment P4 Sid for full
interference with P2 growth, the degree of interference seen
when both phages are growing lytically correlates with the
percentage of small capsids formed (Nilssen et al., 1996).
Furthermore, P2 sir mutants, do not form small capsids, are also
resistant to interference and exhibit normal phage growth (Six
et al., 1991).
In the case of SaPIs, the situation is considerably more
complex. There are at least three strategies for interference, not
all of which are used in the interactions between a particular SaPI
and a specific helper phage. While small capsid formation
certainly prevents packaging of a complete helper phage genome
and thereby interferes with phage growth, the loss of the ability
to form small capsids by mutation of either SaPI1 cpmA or cpmB
alone does not relieve SaPI1 interference with 80a (Damle et al.,
2012). The interference retained by cpmA or cpmB mutants does
not appear to depend on any SaPI1 functions other than cpmA or
cpmB, suggesting a second direct role for these gene products in
helper interference. The effect of the size determination genes
also differs for different helper phages.
A second level at which interference has been documented is
the inhibition of phage DNA packaging by the SaPI-encoded ppi
genes (Ram et al., 2012). These genes fall into two different but
related families, each of which appears to target different helper
phages depending which family the phage small terminase
subunit belongs to. For example, the SaPI1 ppi gene does not
interfere with the growth of 80a, but does block f12, while
the SaPIbov2 ppi gene strongly interferes with 80a growth
(Ram et al., 2012). Different allelic variants of both cpmAB and
ppi confer differing levels of interference, which in some cases
are additive and in others redundant. An additional SaPI gene
involved in interference has also recently been identified.
This gene, ORF17 in SaPI2, blocks growth of phage 80 (which is
not affected by the SaPI2 ppi or cpmAB genes) but not 80a,
and has homologs in other SaPIs as well (Ram et al., 2012).
The mechanism for this third interference function remains to be
elucidated.
Conclusion and perspectives
Horizontal gene transfer (HGT) is now commonly accepted
to play a major role in prokaryotic evolution (Koonin and Wolf,
2008; Toussaint and Chandler, 2012). The vehicles that drive
this ongoing exchange of genetic material, the so-called mobilome,
includes viruses, plasmids, transposons, and a variety of other
selfish elements. Bacteriophages play multiple roles as agents of
HGT. They mediate the exchange of fragments of chromosomal
DNA via generalized and specialized transduction. Temperate
phage integration and excision contributes to the remodeling
of bacterial chromosomes, and can interrupt genes or bring in
new phage-encoded functions via lysogenic conversion. The pirate
elements we have described add a new dimension to phage-
mediated HGT. They differ from other phage-like elements in that
they do not encode their own capsids. They differ from other types
of mobile DNA in that they have found a way to directly
manipulate bacteriophages, through changes in gene expression
and morphogenesis, as vehicles for their own specific high fre-
quency transduction. P4 and the SaPIs both exhibit specialized
adaptations to the lifestyles of their helper phages that allow them
to exploit these phages to their advantage.
How did these elements arise? One possibility is that the
pirates evolved from temperate phages, retaining just those
phage-like functions required for integration/excision, replication
and helper exploitation. Alternatively, they may have been
independent extrachromosomal replicons that have acquired
genes conferring the ability to manipulate phage gene expression
and utilize phage proteins for their own purpose. The answer may
depend on the specific element, since the lifestyle of P4 differs
greatly from the SaPIs. P4 can exist and replicate as a plasmid
independently of P2, and it has been proposed that P4 evolved
from an ancestral plasmid replicon by acquisition of independent
modules for site-specific integration and for helper exploitation
(Deho and Ghisotti, 2005). The complex web of mutual interac-
tions between P2 and P4 suggests that this is a finally tuned
and highly evolved relationship. The SaPI lifestyle more closely
resembles that of a prophage; it has a phage-like repressor and
integration functions and it does not exist as an independent
extrachromosomal replicon. Accordingly, it has been suggested
that SaPIs may have evolved from prophages (Novick et al., 2010).
However, the absence of genes encoding any virion structural
proteins and the acquisition of functions allowing exploitation of
helper phages indicates that SaPIs are not merely some kind of
defective prophage, but like P4 have co-evolved with their helpers
in a highly specific manner.
Despite differences in lifestyle and regulatory circuitry, P4 and
the SaPIs share certain common features (Table 1). Both encode
integration/excision and replication functions and do not depend
on helper functions for these processes. Both have the ability to
sense lytic multiplication of their respective helper phages
and respond by excising and escaping from the bacterial host.
This provides a clear evolutionary advantage, since lytic infection
by a phage would mean the death of the host cell and the loss of
the pirate element. Remodeling of the helper phage capsid is
another conserved feature, and it is striking that these two pirates
have evolved quite different mechanisms to accomplish this
outcome. While not obligatory for transduction of the pirate
genome, capsid size redirection leads to the packaging of sub-
genomic fragments of the helper phage DNA and thereby inter-
feres with phage propagation. This is likely of evolutionary benefit
to the host, since fewer cells in the surrounding population would
be lysed—and would also benefit the pirate, since it would
increase the likelihood that bacteria infected by the transducing
particles carrying the pirate element would not also be infected
by a phage. The importance of interference in the pirate-helper
relationship is underscored by the fact that the SaPIs have evolved
at least three independent mechanisms for helper phage inter-
ference. One remaining unresolved question is what the helper
phages get out of the three way relationship between the
bacterial host, the helper phage, and the pirate. Why have the
helper phages not evolved resistance to this interference by losing
the functions required to derepress the pirates or altering the
genes targeted by the interference functions? The finely tuned
relationship between these pirates and their helpers suggests
that these elements are highly co-evolved in a way that must be
of mutual benefit.
 G.E. Christie, T. Dokland / Virology 434 (2012) 210–221
Molecular piracy, once thought to be a curiosity seen only in
the P2/P4 system, has turned out to be considerably more wide-
spread. Many S. aureus genomes contain one or more SaPIs, and
the presence of similar elements in other Gram positive genera
has led to their general designation as ‘‘phage-related chromoso-
mal islands’’ (Novick et al., 2010). A BLAST search reveals P4-like
elements in the genomes of a number of enterobacteria, including
members of the genera Escherichia, Shigella and Salmonella.
Identification of the helper phages for these related elements,
and further study of the interactions between them, is likely to
reveal additional mechanisms by which these pirates can exploit
their helpers. With the explosion of available genome sequences,
we anticipate the discovery of similar elements in other systems
which will provide fertile ground for further study.
Acknowledgments
This work was supported by the National Institutes of Health
grants R21 AI067654 and R56 AI081837 to G.E.C.; R21 AI071982
and R01 AI083255 to T.D.
References
Agarwal, M., Arthur, M., Arbeit, R.D., Goldstein, R., 1990. Regulation of icosahedral
virion capsid size by the in vivo activity of a cloned gene product. Proc. Natl.
Acad. Sci. U.S.A 87, 2428–2432.
Ahuja, S.K., Murphy, P.M., 1993. Molecular piracy of mammalian interleukin-8
receptor type B by herpesvirus saimiri. J. Biol. Chem. 268, 20691–20694.
Arnold, H.P., She, Q., Phan, H., Stedman, K., Prangishvili, D., Holz, I., Kristjansson,
J.K., Garrett, R., Zillig, W., 1999. The genetic element pSSVx of the extremely
thermophilic crenarchaeon Sulfolobus is a hybrid between a plasmid and a
virus. Mol. Microbiol. 34, 217–226.
Ayers, D.J., Sunshine, M.G., Six, E.W., Christie, G.E., 1994. Mutations affecting two
adjacent amino acid residues in the alpha subunit of RNA polymerase block
transcriptional activation by the bacteriophage P2 Ogr protein. J. Bacteriol 176,
7430–7438.
Barrett, K.J., Marsh, M.L., Calendar, R., 1976. Interactions between a satellite
bacteriophage and its helper. J. Mol. Biol. 106, 683–707.
Bertani, G., 1951. Studies on lysogenesis. I. The mode of phage liberation by
lysogenic Escherichia coli. J. Bacteriol 62, 293–300.
Bertani, L.E., Six, E.W., 1988. The P2-like phages and their parasite, P4. In:
Calendar, R. (Ed.), The Bacteriophages. Plenum Press, New York, pp. 73–143.
Black, L.W., 1989. DNA packaging in dsDNA bacteriophages. Annu. Rev. Microbiol.
43, 267–292.
Bowden, D.W., Modrich, P., 1985. In vitro maturation of circular bacteriophage P2
DNA. Purification of ter components and characterization of the reaction.
J. Biol. Chem 260, 6999–7007.
Breitbart, M., Salamon, P., Andresen, B., Mahaffy, J.M., Segall, A.M., Mead, D., Azam,
F., Rohwer, F., 2002. Genomic analysis of uncultured marine viral commu-
nities. Proc. Natl. Acad. Sci. U.S.A 99, 14250–14255.
Briani, F., Deho, G., Forti, F., Ghisotti, D., 2001. The plasmid status of satellite
bacteriophage P4. Plasmid 45, 1–17.
Briani, F., Ghisotti, D., Deho, G., 2000. Antisense RNA-dependent transcription
termination sites that modulate lysogenic development of satellite phage P4.
Mol. Microbiol. 36, 1124–1134.
Cali, S., Spoldi, E., Piazzolla, D., Dodd, I.B., Forti, F., Deho, G., Ghisotti, D., 2004.
Bacteriophage P4 Vis protein is needed for prophage excision. Virology 322,
82–92.
Catalano, C.E., 2005. Viral Genome Packaging Machines: Genetics, Structure, and
Mechanism. Landes Bioscience/Eurekah.com and Kluwer Academic/Plenum
Publishers, Georgetown, TX and New York.
Chang, J.R., Poliakov, A., Prevelige, P.E., Mobley, J.A., Dokland, T., 2008. Incorpora-
tion of scaffolding protein gpO in bacteriophages P2 and P4. Virology 370,
352–361.
Chang, J.R., Spilman, M.S., Rodenburg, C.M., Dokland, T., 2009. Functional domains
of the bacteriophage P2 scaffolding protein: identification of residues involved
in assembly and protease activity. Virology 384, 144–150.
Choi, J., Means, R.E., Damania, B., Jung, J.U., 2001. Molecular piracy of Kaposi’s
sarcoma associated herpesvirus. Cytokine Growth Factor Rev. 12, 245–257.
Christie, G.E., Anders, D.L., McAlister, V., Goodwin, T.S., Julien, B., Calendar, R.,
2003. Identification of upstream sequences essential for activation of a
bacteriophage P2 late promoter. J. Bacteriol. 185, 4609–4614.
Christie, G.E., Calendar, R., 1990. Interactions between satellite bacteriophage P4
and its helpers. Annu. Rev. Genet. 24 (465-90), 465–490.
Christie, G.E., Calendar, R., 1985. Bacteriophage P2 late promoters. II. Comparison
of the four late promoter sequences. J. Mol. Biol. 181, 373–382.
219
Christie, G.E., Matthews, A.M., King, D.G., Lane, K.D., Olivarez, N.P., Tallent, S.M.,
Gill, S.R., Novick, R.P., 2010. The complete genomes of Staphylococcus aureus
bacteriophages 80 and 80alpha—implications for the specificity of SaPI
mobilization. Virology 407, 381–390.
Conway, J.F., Duda, R.L., Cheng, N., Hendrix, R.W., Steven, A.C., 1995. Proteolytic
and conformational control of virus capsid maturation: the bacteriophage
HK97 system. J. Mol. Biol. 253, 86–99.
Dale, E.C., Christie, G.E., Calendar, R., 1986. Organization and expression of the
satellite bacteriophage P4 late gene cluster. J. Mol. Biol. 192, 793–803.
Damle, P.K., Wall, E.A., Spilman, M.S., Dearborn, A.D., Ram, G., Novick, R.P.,
Dokland, T., Christie, G.E., 2012. The roles of SaPI1 proteins gp7 (CpmA) and
gp6 (CpmB) in capsid size determination and helper phage interference.
Virology 432, 277–282.
Dearborn, A.D., Dokland, T., 2012. Mobilization of pathogenicity islands by
Staphylococcus aureus strain Newman bacteriophages. Bacteriophage 2, 70–78.
Dearborn, A.D., Laurinmaki, P., Chandramouli, P., Rodenburg, C.M., Wang, S.,
Butcher, S.J., Dokland, T., 2012. Structure and size determination of bacter-
iophage P2 and P4 procapsids: function of size responsiveness mutations.
J. Struct. Biol. 178, 215–224.
Dearborn, A.D., Spilman, M.S., Damle, P.K., Chang, J.R., Monroe, E.B., Saad, J.S.,
Christie, G.E., Dokland, T., 2011. The Staphylococcus aureus pathogenicity island
1 protein gp6 functions as an internal scaffold during capsid size determina-
tion. J. Mol. Biol. 412, 710–722.
Deho, G., Ghisotti, D., 2005. The Satellite Phage P4. In: Calendar, R.L. (Ed.), The
Bacteriophages. Oxford University Press, New York, pp. 391–408.
Deho, G., Zangrossi, S., Ghisotti, D., Sironi, G., 1988. Alternative promoters in the
development of bacteriophage plasmid P4. J. Virol. 62, 1697–1704.
Deho, G., Zangrossi, S., Sabbattini, P., Sironi, G., Ghisotti, D., 1992. Bacteriophage P4
immunity controlled by small RNAs via transcription termination. Mol.
Microbiol. 6, 3415–3425.
Diana, C., Deho, G., Geisselsoder, J., Tinelli, L., Goldstein, R., 1978. Viral interference
at the level of capsid size determination by satellite phage P4. J. Mol. Biol. 126,
433–445.
Dokland, T., 1999. Scaffolding proteins and their role in viral assembly. Cell Mol.
Life Sci 56, 580–603.
Dokland, T., 2012. gpO peptidase (Enterobacteria phage P2). In: Rawlings, N.D. and
Salvesen, G. (Eds.) Handbook of proteolytic enzymes, 3rd ed. Elsevier, Oxford,
U.K., ISBN: 9780123822192.
Dokland, T., Isaksen, M.L., Fuller, S.D., Lindqvist, B.H., 1993. Capsid localization of
the bacteriophage P4 Psu protein. Virology 194, 682–687.
Dokland, T., Lindqvist, B.H., Fuller, S.D., 1992. Image reconstruction from cryo-
electron micrographs reveals the morphopoietic mechanism in the P2–P4
bacteriophage system. EMBO J 11, 839–846.
Dokland, T., Wang, S., Lindqvist, B.H., 2002. The structure of P4 procapsids
produced by coexpression of capsid and external scaffolding proteins. Virology
298, 224–231.
Duda, R.L., Martincic, K., Hendrix, R.W., 1995. Genetic basis of bacteriophage HK97
prohead assembly. J. Mol. Biol. 247, 636–647.
Fane, B.A., Prevelige Jr., P.E., 2003. Mechanism of scaffolding-assisted viral
assembly. Adv. Protein Chem. 64, 259–299.
Feiss, M., Rao, V.B., 2012. The bacteriophage DNA packaging machine. Adv. Exp.
Med. Biol. 726, 489–509.
Ferrer, M.D., Quiles-Puchalt, N., Harwich, M.D., Tormo-Mas, M.A., Campoy, S.,
Barbe, J., Lasa, I., Novick, R.P., Christie, G.E., Penades, J.R., 2011. RinA controls
phage-mediated packaging and transfer of virulence genes in Gram-positive
bacteria. Nucleic Acids Res. 39, 5866–5878.
Fitzgerald, J.R., Monday, S.R., Foster, T.J., Bohach, G.A., Hartigan, P.J., Meaney, W.J.,
Smyth, C.J., 2001. Characterization of a putative pathogenicity island from
bovine Staphylococcus aureus encoding multiple superantigens. J. Bacteriol.
183, 63–70.
Flaitz, C.M., Hicks, M.J., 1998. Molecular piracy: the viral link to carcinogenesis.
Oral Oncol. 34, 448–453.
Forti, F., Dragoni, I., Briani, F., Deho, G., Ghisotti, D., 2002. Characterization of the
small antisense CI RNA that regulates bacteriophage P4 immunity. J. Mol. Biol.
315, 541–549.
Forti, F., Polo, S., Lane, K.B., Six, E.W., Sironi, G., Deho, G., Ghisotti, D., 1999.
Translation of two nested genes in bacteriophage P4 controls immunity-
specific transcription termination. J. Bacteriol. 181, 5225–5233.
Fujimuro, M., Hayward, S.D., Yokosawa, H., 2007. Molecular piracy: manipulation
of the ubiquitin system by Kaposi’s sarcoma-associated herpesvirus. Rev. Med.
Virol. 17, 405–422.
Fujisawa, H., Morita, M., 1997. Phage DNA packaging. Genes Cells. 2, 537–545.
Garcia, E., Chain, P., Elliott, J.M., Bobrov, A.G., Motin, V.L., Kirillina, O., Lao, V.,
Calendar, R., Filippov, A.A., 2008. Molecular characterization of L-413C, a
P2-related plague diagnostic bacteriophage. Virology 372, 85–96.
Geisselsoder, J., Youdarian, P., Deho, G., Chidambaram, M., Goldstein, R., Ljung-
quist, E., 1981. Mutants of satellite virus P4 that cannot derepress their
bacteriophage P2 helper. J. Mol. Biol. 148, 1–19.
Halling, C., Calendar, R., 1990. Bacteriophage P2 ogr and P4 delta genes act
independently and are essential for P4 multiplication. J. Bacteriol. 172,
3549–3558.
Harwich, M.D., 2009. Transcriptional Profiling of Staphylococcal Bacteriophage
80a and Regulatory Interactions with Pathogenicity Island SaPI1. Ph.D.
Dissertation, Virginia Commonwealth University, Richmond, VA.
Highlander, S.K., Hulten, K.G., Qin, X., Jiang, H., Yerrapragada, S., Mason . Jr, E.O.,
Shang, Y., Williams, T.M., Fortunov, R.M., Liu, Y., Igboeli, O., Petrosino, J.,
220 G.E. Christie, T. Dokland / Virology 434 (2012) 210–221
Tirumalai, Y., Uzman, A., Fox, G.E., Cardenas, A.M., Muzny, D.M., Hemphill, L.,
Ding, Y., Dugan, S., Blyth, P.R., Buhay, C.J., Dinh, H.H., Hawes, A.C., Holder, M.,
Kovar, C.L., Lee, S.L., Liu, W., Nazareth, L.V., Wang, Q., Zhou, J., Kaplan, S.L.,
Weinstock, G.M., 2007. Subtle genetic changes enhance virulence of methi-
cillin resistant and sensitive Staphylococcus aureus. BMC Microbiol 7, 99.
Hu, C.C., Hsu, Y.H., Lin, N.S., 2009. Satellite RNAs and Satellite Viruses of Plants.
Viruses 1, 1325–1350.
Inouye, S., Sunshine, M.G., Six, E.W., Inouye, M., 1991. Retronphage phi R73: an
E. coli phage that contains a retroelement and integrates into a tRNA gene.
Science 252, 969–971.
Isaksen, M.L., Dokland, T., Lindqvist, B.H., 1993. Characterization of the capsid
associating activity of bacteriophage P4’s Psu protein. Virology 194, 674–681.
Johnson, J.E., 2010. Virus particle maturation: insights into elegantly programmed
nanomachines. Curr. Opin. Struct. Biol. 20, 210–216.
Johnson, J.E., Chiu, W., 2007. DNA packaging and delivery machines in tailed
bacteriophages. Curr. Opin. Struct. Biol. 17, 237–243.
Julien, B., Calendar, R., 1996. Bacteriophage PSP3 and phiR73 activator proteins:
analysis of promoter specificities. J. Bacteriol. 178, 5668–5675.
Julien, B., Pountney, D., Christie, G.E., Calendar, R., 1998. Mutational analysis of a
satellite phage activator. Gene 223, 129–134.
Kalionis, B., Dodd, I.B., Egan, J.B., 1986. Control of gene expression in the P2-related
template coliphages. III. DNA sequence of the major control region of phage
186. J. Mol. Biol. 191, 199–209.
Kim, K.J., Sunshine, M.G., Lindqvist, B.H., Six, E.W., 2001. Capsid size determination
in the P2-P4 bacteriophage system: suppression of sir mutations in P2’s capsid
gene N by supersid mutations in P4’s external scaffold gene sid. Virology 283,
49–58.
Koonin, E., Wolf, Y., 2008. Genomics of bacteria and archaea: the emerging
dynamic view of the prokaryotic world. Nucleic Acids Res. 36, 6688–6719.
Kuroda, M., Yamashita, A., Hirakawa, H., Kumano, M., Morikawa, K., Higashide, M.,
Maruyama, A., Inose, Y., Matoba, K., Toh, H., Kuhara, S., Hattori, M., Ohta, T.,
2005. Whole genome sequence of Staphylococcus saprophyticus reveals the
pathogenesis of uncomplicated urinary tract infection. Proc. Natl. Acad. Sci.
U.S.A 102, 13272–13277.
Kwan, T., Liu, J., DuBow, M., Gros, P., Pelletier, J., 2005. The complete genomes and
proteomes of 27 Staphylococcus aureus bacteriophages. Proc. Natl. Acad. Sci.
U.S.A 102, 5174–5179.
La Scola, B., Desnues, C., Pagnier, I., Robert, C., Barrassi, L., Fournous, G., Merchat,
M., Suzan-Monti, M., Forterre, P., Koonin, E., Raoult, D., 2008. The virophage as
a unique parasite of the giant mimivirus. Nature 455, 100–104.
Lin, C.S., 1984. Nucleotide sequence of the essential region of bacteriophage P4.
Nucleic Acids Res. 12, 8667–8684.
Lindqvist, B.H., Deho, G., Calendar, R., 1993. Mechanisms of genome propagation
and helper exploitation by satellite phage P4. Microbiol. Rev. 57, 683–702.
Lindsay, J.A., Ruzin, A., Ross, H.F., Kurepina, N., Novick, R.P., 1998. The gene for
toxic shock toxin is carried by a family of mobile pathogenicity islands in
Staphylococcus aureus. Mol. Microbiol. 29, 527–543.
Liu, J., Dehbi, M., Moeck, G., Arhin, F., Bauda, P., Bergeron, D., Callejo, M., Ferretti, V.,
Ha, N., Kwan, T., McCarty, J., Srikumar, R., Williams, D., Wu, J.J., Gros, P.,
Pelletier, J., DuBow, M., 2004. Antimicrobial drug discovery through bacter-
iophage genomics. Nat. Biotechnol. 22, 185–191.
Liu, T., Renberg, S.K., Haggard-Ljungquist, E., 1998. The E protein of satellite phage
P4 acts as an anti-repressor by binding to the C protein of helper phage P2.
Mol. Microbiol. 30, 1041–1050.
Liu, T., Renberg, S.K., Haggard-Ljungquist, E., 1997. Derepression of prophage P2 by
satellite phage P4: cloning of the P4 epsilon gene and identification of its
product. J.Virol 71, 4502–4508.
Lynch, K.H., Stothard, P., Dennis, J.J., 2010. Genomic analysis and relatedness of
P2-like phages of the Burkholderia cepacia complex. BMC Genomics 11, 599.
Marvik, O.J., Dokland, T., Nokling, R.H., Jacobsen, E., Larsen, T., Lindqvist, B.H., 1995.
The capsid size-determining protein Sid forms an external scaffold on phage
P4 procapsids. J. Mol. Biol. 251, 59–75.
Maxwell, K.L., Frappier, L., 2007. Viral proteomics. Microbiol. Mol. Biol. Rev. 71,
398–411.
McAlister, V., Zou, C., Winslow, R.H., Christie, G.E., 2003. Purification and in vitro
characterization of the Serratia marcescens NucC protein, a zinc-binding
transcription factor homologous to P2 Ogr. J. Bacteriol. 185, 1808–1816.
Morais, M.C., Kanamaru, S., Badasso, M.O., Koti, J.S., Owen, B.A., McMurray, C.T.,
Anderson, D.L., Rossmann, M.G., 2003. Bacteriophage phi29 scaffolding protein
gp7 before and after prohead assembly. Nat. Struct. Biol. 10, 572–576.
Nilssen, O., Fossdal, C.G., Johansen, B.V., Lindqvist, B.H., 1996. Bacteriophage P4
capsid-size determination and its relationship to P2 helper interference.
Virology 219, 443–452.
Nilsson, A.S., Haggard Ljungquist, E., 2005. The P2-Like Bacteriophages. In:
Calendar, R.L. (Ed.), The Bacteriophages. Oxford University Press, New York,
pp. 365–390.
Nilsson, A.S., Karlsson, J.L., Haggard-Ljungquist, E., 2004. Site-specific recombina-
tion links the evolution of P2-like coliphages and pathogenic enterobacteria.
Mol. Biol. Evol. 21, 1–13.
Novick, R.P., Christie, G.E., Penades, J.R., 2010. The phage-related chromosomal
islands of Gram-positive bacteria. Nat. Rev. Microbiol. 8, 541–551.
Ott, D.E., 2008. Cellular proteins detected in HIV-1. Rev. Med. Virol. 18, 159–175.
Pani, B., Ranjan, A., Sen, R., 2009. Interaction surface of bacteriophage P4 protein
Psu required for complex formation with the transcription terminator Rho.
J. Mol. Biol. 389, 647–660.
Pettersen, E.F., Goddard, T.D., Huang, C.C., Couch, G.S., Greenblatt, D.M., Meng, E.C.,
Ferrin, T.E., 2004. UCSF Chimera—a visualization system for exploratory
research and analysis. J. Comput. Chem. 25, 1605–1612.
Poliakov, A., Chang, J.R., Spilman, M.S., Damle, P.K., Christie, G.E., Mobley, J.,
Dokland, T., 2008. Capsid size determination by Staphylococcus aureus patho-
genicity island SaPI1 involves specific incorporation of SaPI1 proteins into
procapsids. J. Mol. Biol. 380, 465–475.
Pountney, D.L., Tiwari, R.P., Egan, J.B., 1997. Metal- and DNA-binding properties
and mutational analysis of the transcription activating factor, B, of coliphage
186: a prokaryotic C4 zinc-finger protein. Protein Sci. 6, 892–902.
Pruss, G.J., Calendar, R., 1978. Maturation of bacteriophage P2 DNA. Virology 86,
454–467.
Pruss, G.J., Wang, J.C., Calendar, R., 1975. In vitro packaging of covalently closed
circular monomers of bacteriophage DNA. J. Mol. Biol. 98, 465–478.
Ram, G., Chen, J., Kumar, K., Ross, H.F., Ubeda, C., Damle, P.K., Lane, K.D., Penades,
J.R., Christie, G.E., Novick, R.P., 2012. SaPI interference with helper phage
reproduction is a paradigm of molecular parasitism. Proc. Natl. Acad. Sci. U.S.A
109, 16300–16305.
Rosenberg, D., 2009. The political economy of piracy in the South China Sea. Naval
War Coll. Rev. 62, 43–58.
Roy, A., Bhardwaj, A., Datta, P., Lander, G.C., Cingolani, G., 2012. Small terminase
couples viral DNA binding to genome-packaging ATPase activity. Structure 20,
1403–1413.
Ruzin, A., Lindsay, J., Novick, R.P., 2001. Molecular genetics of SaPI1—a mobile
pathogenicity island in Staphylococcus aureus. Mol. Microbiol. 41, 365–377.
Sabbattini, P., Forti, F., Ghisotti, D., Deho, G., 1995. Control of transcription
termination by an RNA factor in bacteriophage P4 immunity: identification
of the target sites. J. Bacteriol. 177, 1425–1434.
Saha, S., Haggard-Ljungquist, E., Nordstrom, K., 1989. Activation of prophage P4 by
the P2 Cox protein and the sites of action of the Cox protein on the two phage
genomes. Proc. Natl. Acad. Sci. U.S.A 86, 3973–3977.
Saha, S., Haggard-Ljungquist, E., Nordstrom, K., 1987. The cox protein of bacter-
iophage P2 inhibits the formation of the repressor protein and autoregulates
the early operon. EMBO J. 6, 3191–3199.
Sauer, B., Calendar, R., Ljungquist, E., Six, E., Sunshine, M.G., 1982. Interaction of
satellite phage P4 with phage 186 helper. Virology 116, 523–534.
Sauer, B., Ow, D., Ling, L., Calendar, R., 1981. Mutants of satellite bacteriophage P4
that are defective in the suppression of transcriptional polarity. J. Mol. Biol.
145, 29–46.
Schwudke, D., Ergin, A., Michael, K., Volkmar, S., Appel, B., Knabner, D., Konietzny, A.,
Strauch, E., 2008. Broad-host-range Yersinia phage PY100: genome sequence,
proteome analysis of virions, and DNA packaging strategy. J. Bacteriol. 190,
332–342.
Scott, J., Nguyen, S.V., King, C.J., Hendrickson, C., McShan, W.M., 2012. Phage-like
streptococcus pyogenes chromosomal islands (SpyCI) and mutator phenotypes:
control by growth state and rescue by a SpyCI-encoded promoter. Front.
Microbiol 3, 317.
Shore, D., Deho, G., Tsipis, J., Goldstein, R., 1978. Determination of capsid size by
satellite bacteriophage P4. Proc. Natl. Acad. Sci. U.S.A 75, 400–404.
Sinkovics, J., Horvath, J., Horak, A., 1998. The origin and evolution of viruses
(a review). Acta Microbiol. Immunol. Hung. 45, 349–390.
Six, E.W., 1975. The helper dependence of satellite bacteriophage P4: which gene
functions of bacteriophage P2 are needed by P4? Virology 67, 249–263.
Six, E.W., 1963. A defective phage depending on phage P2. Bacteriol. Proc. 1963,
138.
Six, E.W., Klug, C.A., 1973. Bacteriophage P4: a satellite virus depending on a
helper such as prophage P2. Virology 51, 327–344.
Six, E.W., Lindqvist, B.H., 1978. Mutual derepression in the P2–P4 bacteriophage
system. Virology 87, 217–230.
Six, E.W., Sunshine, M.G., Williams, J., Haggard-Ljungquist, E., Lindqvist, B.H., 1991.
Morphopoietic switch mutations of bacteriophage P2. Virology 182, 34–46.
Spilman, M.S., Damle, P.K., Dearborn, A.D., Rodenburg, C.M., Chang, J.R., Wall, E.A.,
Christie, G.E., Dokland, T. Assembly of bacteriophage 80alpha capsids in a
Staphylococcus aureus expression system. Virology, http://dx.doi.org/10.1016/j.
virol.2012.08.031, in press.
Spilman, M.S., Dearborn, A.D., Chang, J.R., Damle, P.K., Christie, G.E., Dokland, T.,
2011. A conformational switch involved in maturation of Staphylococcus
aureus bacteriophage 80alpha capsids. J. Mol. Biol. 405, 863–876.
Sunshine, M.G., Sauer, B., 1975. A bacterial mutation blocking P2 phage late gene
expression. Proc. Natl. Acad. Sci. U.S.A 72, 2770–2774.
Sureau, C., 2006. The role of the HBV envelope proteins in the HDV replication
cycle. Curr. Top. Microbiol. Immunol. 307, 113–131.
Takeuchi, F., Watanabe, S., Baba, T., Yuzawa, H., Ito, T., Morimoto, Y., Kuroda, M.,
Cui, L., Takahashi, M., Ankai, A., Baba, S., Fukui, S., Lee, J.C., Hiramatsu, K., 2005.
Whole-genome sequencing of Staphylococcus haemolyticus uncovers the
extreme plasticity of its genome and the evolution of human-colonizing
staphylococcal species. J. Bacteriol. 187, 7292–7308.
Tallent, S.M., Langston, T.B., Moran, R.G., Christie, G.E., 2007. Transducing particles
of Staphylococcus aureus pathogenicity island SaPI1 are comprised of helper
phage-encoded proteins. J. Bacteriol. 189, 7520–7524.
Teschke, C.M., 2012. Themes and variations of viral small terminase proteins.
Structure 20, 1291–1292.
Tormo, M.A., Ferrer, M.D., Maiques, E., Ubeda, C., Selva, L., Lasa, I., Calvete, J.J.,
Novick, R.P., Penades, J.R., 2008. Staphylococcus aureus pathogenicity island
DNA is packaged in particles composed of phage proteins. J. Bacteriol. 190,
2434–2440.
 G.E. Christie, T. Dokland / Virology 434 (2012) 210–221
Tormo-Mas, M.A., Mir, I., Shrestha, A., Tallent, S.M., Campoy, S., Lasa, I., Barbe, J.,
Novick, R.P., Christie, G.E., Penades, J.R., 2010. Moonlighting bacteriophage
proteins derepress staphylococcal pathogenicity islands. Nature 465, 779–782.
Toussaint, A, Chandler, M., 2012. Prokaryote genome fluidity: toward a system
approach of the mobilome. Methods Mol. Biol. 804, 57–80.
Ubeda, C., Maiques, E., Barry, P., Matthews, A., Tormo, M.A., Lasa, I., Novick, R.P.,
Penades, J.R., 2008. SaPI mutations affecting replication and transfer and
enabling autonomous replication in the absence of helper phage. Mol.
Microbiol. 67, 493–503.
Ubeda, C., Maiques, E., Tormo, M., Campoy, S., Lasa, I., Barbe, J., Novick, R.P.,
Penades, J.R., 2007. SaPI operon I is required for SaPI packaging and is
controlled by LexA. Mol. Microbiol. 65, 41–50.
Ubeda, C., Olivarez, N.P., Barry, P., Wang, H., Kong, X., Matthews, A., Tallent, S.M.,
Christie, G.E., Novick, R.P., 2009. Specificity of staphylococcal phage and SaPI
DNA packaging as revealed by integrase and terminase mutations. Mol.
Microbiol. 72, 98–108.
Wang, S., Chang, J.R., Dokland, T., 2006. Assembly of bacteriophage P2 and P4
procapsids with internal scaffolding protein. Virology 348, 133–140.
Wang, Y., Duan, Z., Zhu, H., Guo, X., Wang, Z., Zhou, J., She, Q., Huang, L., 2007. A
novel Sulfolobus non-conjugative extrachromosomal genetic element capable
221
of integration into the host genome and spreading in the presence of a
fusellovirus. Virology 363, 124–133.
Wikoff, W.R., Liljas, L., Duda, R.L., Tsuruta, H., Hendrix, R.W., Johnson, J.E., 2000.
Topologically linked protein rings in the bacteriophage HK97 capsid. Science
289, 2129–2133.
Wood, L.F., Tszine, N.Y., Christie, G.E., 1997. Activation of P2 late transcription by
P2 Ogr protein requires a discrete contact site on the C terminus of the alpha
subunit of Escherichia coli RNA polymerase. J. Mol. Biol. 274, 1–7.
Wu, H., Sampson, L., Parr, R., Casjens, S., 2002. The DNA site utilized by
bacteriophage P22 for initiation of DNA packaging. Mol. Microbiol. 45,
1631–1646.
Yan, X., Sinkovits, R., Baker, T., 2007. AUTO3DEM—an automated and high through-
put program for image reconstruction of icosahedral particles. J. Struct. Biol. 157,
73–82.
Yu, A., Haggard-Ljungquist, E., 1993. The Cox protein is a modulator of direction-
ality in bacteriophage P2 site-specific recombination. J. Bacteriol. 175,
7848–7855.
Ziermann, R., Calendar, R., 1990. Characterization of the cos sites of bacteriophages
P2 and P4. Gene 96, 9–15.
 CHAPTER TEN

THE FORCES BEHIND THE DISSEMINATION

OF BACTERIAL VIRULENCE
AND ANTIBIOTIC RESISTANCE

GABRIEL TRUEBA

It is intriguing that a large number of genes coding for virulence and antibiotic resistance are associated with
genes involved in intra-genomic mobility and horizontal gene transfer. In other cases, similar gene
associations have been observed to carry genes coding for functions such as photosynthesis, plant symbiosis,
degradation different substrates including pollutants, etc. A great variety of these structures, called mobile
genetic elements (MGEs), are widely disseminated in nature, and I will present the possible scenarios and
events in which natural selection may favour the genesis and propagation of these genetic structures.

Introduction
The horizontal transference of Mobile Genetic Elements (MGEs) is a major factor contributing to the
emergence of bacterial pathogens and antibiotic resistant bacteria (Ho Sui et al, 2009; Baquero, 2008).
Examples include enterohemorrhagic E. coli O104:H4 (Rasko et al, 2011), Cronobacter sakasaki (Kucerova,
2010) and highly virulent-methicillin resistant strains of Staphylococcus aureus (Diep, 2006). In addition to
clinically relevant examples, MGEs are also implicated in other important functions such as photosynthetic
carbon fixation in the oceans (Rohwer and Thurber, 2009), bacteria-plant symbiosis (Dobrindt, 2004), and
degradation of pollutants (Top, 2003).

MGEs are genetic structures that may contain both genes conferring selective advantages to the bacterial cells
and genes involved in horizontal gene transfer (hereinafter referred to as inter-genomic mobility) (Dobrindt,
2004; Jackson 2011), and translocation or transposition (hereinafter referred to as intra-genomic mobility).
Horizontal transference of MGEs is experimentally observable, in other cases it is assumed because of
molecular discrepancies (GC content, molecular phylogeny and codon usage) found when MGE and
housekeeping genes are compared. Additional evidences are: the existence of ancestral versions of MGEs in
distantly related bacteria (Gillings, 2008; Rowe-Magnus, 2001); the presence of remnants of mobility genes in
some MGEs, and irregular distribution of MGEs in members of the same bacterial species (Diep, 2006).

MGEs contain core and cargo regions: MGE core regions have genes coding for intra-genomic gene mobility
(integration, transposition) and intercellular transfer (conjugation and viral packaging); the cargo or moron
region contains genes coding diverse bacterial adaptive functions (Seth-Smith, 2009). MGEs could be
grouped according to the nature of the core and cargo regions. Core regions containing integrases and
relaxosomes (conjugation machinery) genes are named integrative conjugal elements (ICEs), those that have a
relaxosome and an origin of replication are known as plasmids, the ones containing phage-like integrases and
integration sites are known as integrons, and finally those containing many phage (viral) genes are
converting-phages (Rankin, 2011). Similarly, some MGEs are classified according to the cargo region in:
pathogenicity islands, metabolic islands, symbiotic islands, antibiotic resistance islands, etc.

The evolution of MGEs is complex and it involves the association of genes from different origins. The core
genes likely derive from selfish genetic elements (such as plasmids, phages and transposons) whose success
often depends on their aggressive replication without necessarily benefiting the bacterial host (Rankin, 2011;
Eberhard, 1990; Werren, 2011; Doolittle, 1980; Dionisio, 2005). Integrases in MGEs are genetically related to
phage tyrosine integrases; however, some MGE integrases form a separate phylogenetic cluster, which may
indicate that they emerged from an ancient phage-like structure (Napolitano, 2011). Relaxosome genes in
plasmid and ICEs are genetically related, however it is not possible to establish which structure is more
ancestral (Guglielmini, 2011). The origins of some virulence and antibiotic resistance genes in MGEs have
been linked to environmental bacteria and are reviewed extensively in other manuscripts (Martinez, 2012;
Wright, 2012).

Selfish genetic elements may prosper at the expense of the bacterial fitness; in addition, transposons and
phages may impose a cost on their host by disrupting the coding sequences of essential genes (Doolittle and
Sapienza 1980). Conversely, cargo genes alone could provide competitive advantage (in specific niches) to
the bacterial cell (Eberhard, 1990; Werren, 2011) while the combination of both (cargo and core genes) may
result in successful alliances which disseminate easily in nature (Werren, 2011).

The structural characteristics of MGEs (abundance of ISs and other recombinases) may provide evidence of
their possible origins, including: 1) gene capture; 2) integration into genetic platforms capable of moving
from cell to cell (plasmids or phages); 3) MGE gene (moron) exchange; and 4) integration into bacterial
chromosome and loss of unnecessary genes (Fig. 1). The present review summarizes current views and
presents the similarities in the origin of diverse MGEs.

Gene capture
Gene capture involves the joining of a DNA fragment coding for a meaningful function to a recombinase
gene, which could be transposase, tyrosine recombinase, or serine recombinase. Transposase is the most
abundant gene in nature (Aziz et al, 2010) and insertion sequences (ISs) may be the most common
transposable element in bacteria. Composite transposons arise when two ISs insert on the flanks of a gene,
leading to a region containing the two ISs elements with an intervening DNA fragment (Baquero, 2008;
Jackson et al, 2011; Iida, et al; Mahillon, 1998); in this way a gene could gain the ability to move from one
place to another within the genome (Figure 1). Similar events can take place when a gene becomes associated
with phage recombinases during specialized transduction (Jackson et al 2011; Gillings et al 2008; Wagner and
Waldor, 2002) or other illegitimate recombination events involving viral or plasmid integrases. If the
composite transposon or integrative gene inserts into a higher copy-number replicon (such as a plasmids) it
could increase its burden on bacterial metabolism. However if the captured gene codes for an adaptive feature
this event could improve the odds for the bacterial success (due to higher expression of the gene) and could
increase the abundance of this gene association in nature. Adaptive genes are very common in environments
wherever there is a selective pressure over a bacterial population (such as the presence of antibiotics).
Selection results in large numbers of bacteria carrying these genes which increases the chances for the
transposon gene capture.

Gene integration into platforms capable

of cell to cell transference
Insertion of the captured gene (composite transposon or integrative gene) into a conjugal plasmid or into a
viral genome would create a MGE able to mobilize horizontally between bacterial cells; those bacteria
benefiting from cargo genes will prosper, and so will these MGEs (Eberhard 1990; Dionisio et al, 2005).
There are many examples of such combinations that allow a pathogen to emerge and occupy a unique
ecological niche such as Vibrio cholerae, Bacillus anthracis, Yersinia pestis, etc (Hacker and Kaper, 2000). A
gene coding for a meaningless function, or a non-coding DNA, can be also associated with recombinases and
integrated into mobile platforms; however, in the absence of adaptive characteristics, these associations are
probably destined for extinction (Rowe-Magnus, 2001). This phenomenon may explain the abundance of
MGEs carrying genes that code for significant functions. Horizontal transference of MGEs into different
bacterial backgrounds (genomes) may result in gene combinations that offer compatible characteristics, for
instance MGEs coding for a virulence features may prosper in pathogenic bacteria, but not in non-pathogenic
bacteria.
 MGE gene (moron) exchange
When more than one selective pressure has been maintained for long time over a bacterial population, there
may be a large number of bacterial cells carrying different MGEs. The existence of two or more MGEs inside
bacterial cells sets the stage for the next event, which is the exchange of adaptive genes (modules) between
MGEs (Baquero, 2008) and the formation of more complex MGEs such as integrons (Baquero, 2008; Rowe-
Magnus et al, 2002 ).

MGE Integration into bacterial chromosomes

and gene loss
Inactivating mutations in core genes (integrases, relaxosome genes, transposases, replicons, viral structural
genes, genes involved in bacterial lysis, etc.) could affect MGE fitness by reducing its ability to multiply or
move (Hacker and Kaper, 2000). Similarly, mutations in beneficial cargo genes could push the bacterial hosts
to extinction.

In some cases, the cargo genes provide such valuable benefits to the host that the loss of mobility genes does
not seem to diminish the success in chromosomally anchored MGEs (Eberhard, 1990; Wang et al, 2010;
Osborn and Böltner, 2002). Some anchored MGEs still contain mobility-associated genes and exhibit lateral
transference when the cell is infected with phages or plasmids capable of complementing this function
(Jackson et al, 2011; Napolitano Napolitano et al, 2011; Guglielmini et al, 2011; Wang et al, 2010). In other
cases there seems to be a pressure in MGEs to maintain the core genes: for instance, photosynthetic genes are
very common in phages infecting most abundant photosynthetic bacteria in oceans (Rohwer and Thurber,
2009), however these genes didn’t seem to be associated with anchored MGEs (Coleman et al, 2006). In this
case, viral genes may allow MGEs to disseminate in environments with low bacterial concentration such as
oceans (Whitman et al, 1998). Unlike conjugation, viral dissemination does not require direct cell-to-cell
contact; in this environment, phages could move genes long distances searching for new bacterial hosts.

Conclusions
The synergy of the cargo-core gene association could explain the evolutionary abundance of MGEs
(Guglielmini et al, 2011; Aziz et al, 2010); and the more successful, the more abundant. But it is important to
understand the environmental context that allows the survival and dissemination of MGEs. Those carrying
metabolic, virulence or antibiotic resistance genes, for example, could disseminate very fast among different
bacterial species in the presence of selective pressure (Peters et al, 1997). Hence, a cargo gene that is
advantageous under specific environmental conditions may be detrimental if the bacterial population expands
beyond this niche (Eberhard, 1990)

From the infectious disease point of view it is important to understand the environmental factors involved in
the persistence and dissemination of MGEs carrying virulence or antibiotic resistance genes. Environmental
stressors, such as antibiotic exposure, not only select for bacteria carrying antibiotic resistance genes, but
could also stimulate intra-genomic and inter-genomic mobility of MGEs carrying resistance genes
(McGannon et al, 2010; Guerin et al, 2009; Prudhomme et al, 2006; Beaber et al, 2004). A recent report
shows that intestines of antibiotic treated mice have far more MGEs carrying antibiotic resistance genes than
non-treated counterparts (Modi et al, 2012).

The understanding of the interactions between antibiotics and antibiotic resistance (and virulence) genes
should inform professionals about the risks of using sub-therapeutic doses of antibiotics as growth promoters
in food animals (Silbergeld et al, 2008). Finally this knowledge may lead to the development of new
approaches (such as eco-evo drugs) to control antibiotic resistance and the emergence of new bacterial
pathogens (Baquero et al, 2011).

Acknowledgments
The author thanks Paul Keim and Ana Trueba for their valuable suggestions.
 References
Aziz, Ramy, Mya Breitbart, and Robert Edwards. “Transposases are the most abundant, most ubiquitous
genes in nature.” Nucleic Acids Research 38 (2010): 4207-4217.
Baquero, Fernado, Teresa Coque, and Fernando-de-la-Cruz F. “Ecology and evolution as targets: the need for
novel eco-evo drugs and strategies to fight antibiotic resistance.” Antimicrobial Agents Chemotherapy 55
(2011): 3649-3660.
Baquero, Fernando. “Modularization and Evolvability in Antibiotic Resistance.” In Evolutionary biology of
bacterial and fungal pathogens, 233-247. Washington: ASM Press, 2008.
Beaber, John, Blanca Hochhut, and Mathew Waldor. “SOS Response Promotes Horizontal Dissemination of
Antibiotic Resistance Genes.” Nature 427 (2004): 72-74.
Coleman, Maureen, Matthew Sullivan, Adam Martiny, Steglich Claudia, Kerrie Barry, Edward Delong, and
Sallie Chisholm. “Genomic Islands and the Ecology and Evolution of Prochlorococcus.” Science 311
(2006): 1768-1770.
Diep, Binh, Heather Carleton, Richard Chang, George Sensabaugh, and Françoise Perdreau-Remington.
“Roles of 34 Virulence Genes in the Evolution of Hospital- and Community-Associated Strains of
Methicillin-Resistant Staphylococcus aureus” The Journal of Infectious Diseases 193 (2006): 1495-1503.
Dionisio, F., I Conceição, A Marques, L Fernandes, and I Gordo, “The Evolution of a Conjugative Plasmid
and its Ability to Increase Bacterial Fitness.” Biology Letters 1 (2005): 250-252.
Dobrindt, Ulrich, Bianca Hochhut, Ute Hentschel, Jörg Hacker. “Genomic Islands in Pathogenic and
Environmental Microorganisms.” Nature Reviews Microbiology 2 (2004): 414-424.
Doolittle, W. Ford, and Carme Sapienza. “Selfish Genes, the Phenotype Paradigm and Genome Evolution.”
Nature 284 (1980): 601-603.
Eberhard, William. “Evolution in Bacterial Plasmids and Levels of Selection.” The Quarterly Review of
Biology 65 (1990): 3-22.
Gillings, Michael, Yan Boucher, Maurizio Labbate, Andrew Holmes, Samyuktha Krishnan, Marita Holley,
and H.W. Stokes. “The Evolution of Class 1 Integrons and the Rise of Antibiotic Resistance.” Journal of
Bacteriology 190 (2008): 5095-5100.
Guerin, Emilie, Guillaume Cambray, Neus Sanchez-Alberola, Susana Campoy, Ivan Erill, Sandra Da-Re,
Bruno Gonzalez-Zorn, Jordi Barbé, Marie-Cécile Ploy, and Didier Mazel. “The SOS Response Controls
Integron Recombination.” Science 324 (2009): 1034.
Guglielmini, Julien, Leonor Quintais, Maria-Pilar Garcillán-Barcia, Fernando de-la-Cruz, and Eduardo
Rocha. “The Repertoire of ICE in Prokaryotes Underscores the Unity, Diversity, and Ubiquity of
Conjugation” PLoS Genetics 7 (2011): e1002222.
Hacker, Jörg, and James Kaper, “Pathogenicity Islands and the Evolution of Microbes.” Annual Reviews of
Microbiology 54 (2000): 641-679.
Ho Sui, Shannan, Amber Fedynak, William Hsiao, Morgan Langille, and Fiona Brinkman. “The Association
of Virulence Factors with Genomic Islands.” PLoS One 4 (2009): e8094.
Iida, S, J. Meyer, and W. Arber. “Genesis and Natural History of IS-Mediated Transposons.” Cold Spring
Harbor Symposia of Quantitative Biology 45 (1980): 27-37.
Jackson, Robert, Boris Vinatzer, Dawn Arnold, Steve Dorus, and Jesús Murillo. “The Influence of the
Accessory Genome on Bacterial Pathogen Evolution.” Mobile Genetic Elements 1 (2011): 55-65.
Kucerova, Eva, Sandra Clifton, Xiao-Qin Xia, Fred Long, Steffen Porwollik, Lucinda Fulton, Catrina
Fronick, Patrick Minx, Kim Kyung, Wesley Warren, Robert Fulton, Dongyan Feng, Aye Wollam, Neha
Shah, Veena Bhonagiri, William Nash, Kymberlie Hallsworth-Pepin, Richard Wilson, Michael
McClelland, and Stephen Forsythe. “Genome Sequence of Cronobacter sakazakii BAA-894 and
Comparative Genomic Hybridization Analysis with other Cronobacter species.” PLoS One 5 (2010):
e9556.
Mahillon, J. “Transposons as Gene Haulers.” Acta Pathologica, Microbiologica et Immunologica
Scandinavica Suppl 84 (1998): 29-36.
Martínez, José. “Bacterial Pathogens: From Natural Ecosystems to Human Hosts.” Environmental
Microbiology 15 (2012): 325–333.
McGannon, Colleen, Cynthia Fuller, and Alison Weiss. “Different Classes Of Antibiotics Differentially
Influence Shiga Toxin Production.” Antimicrobrial Agents and Chemotherapy 54 (2010): 3790-3798.
Modi, Sheetal, Henry Lee, Catherine Spina, and James Collins. “Antibiotic Treatment Expands The
Resistance Reservoir And Ecological Network Of The Phage Metagenome.” Nature 499 (2013): 219-222.
Napolitano, Michael, Salvador Almagro-Moreno, Fidelma Boyd. “Dichotomy In The Evolution Of
Pathogenicity Island And Bacteriophage Encoded Integrases From Pathogenic Escherichia Coli Strains.”
Infection, Genetics and Evolution 11 (2011): 423-436.
Osborn, Mark, and Dietmar Böltner. “When Phage, Plasmids, And Transposons Collide: Genomic Islands,
And Conjugative- And Mobilizable-Transposons As A Mosaic Continuum.” Plasmid 48 (2002): 202-212.
Peters, M, E. Heinaru, E. Talpsep, H. Wand, U. Stottmeister, A. Heinaru, and A. Nurk. “Acquisition Of A
Deliberately Introduced Phenol Degradation Operon, Pheba, By Different Indigenous Pseudomonas
Species.” Applied and Environmental Microbiology 63 (1997): 4899-4906.
Prudhomme, Marc, Laetitia Attaiech, Guillaume Sanchez, Bernard Martin, and Jean-Pierre Claverys.
“Antibiotic Stress Induces Genetic Transformability In The Human Pathogen Streptococcus Pneumoniae.”
Science 313 (2006): 89-92.
Rankin, D., E. Rocha, and S. Brown, “What traits are carried on mobile genetic elements, and why?”
Heredity106 (2011): 1-10.
Rasko, David, Dale Webster, Jason Sahl, Ali Bashir, Nadia Boisen, Flemming Scheutz, Ellen Paxinos, Robert
Sebra, Chen-Shan Chin, Dimitris Iliopoulos, Aaron Klammer, Paul Peluso, Lawrence Lee, Andrey
Kislyuk, James Bullard, Andrew Kasarskis, Susanna Wang, John Eid, David Rank, Julia Redman, Susan
Steyert, Jakob Frimodt-Møller, Carsten Struve, Andreas Petersen, Karen Krogfelt, James Nataro, Eric
Schadt, and Matthew Waldor. “Origins Of The E. Coli Strain Causing An Outbreak Of Hemolytic-Uremic
Syndrome In Germany.” New England Journal of Medicine 365 (2011): 709-717.
Rohwer, Forest, and Rebbeca Vega Thurber, “Viruses manipulate the marine environment.” Nature 459
(2009): 207-212.
Rowe-Magnus, Dean, Anne-Marie Guerout, and Didier Mazel. “Bacterial Resistance Evolution By
Recruitment Of Super-Integron Gene Cassettes.” Molecular Microbiology 43 (2002): 1657-1669.
Rowe-Magnus, Dean, Anne-Marie Guerout, Pascaline Ploncard, broderick Dychinco, Julian Davies, and
Didier Mazel. “The evolutionary history of chromosomal super-integrons provides an ancestry for
multiresistant integrons.” Proceedings of the National Academy of Sciences of the United States of
America 98 (2001): 652-657.
Seth-Smith, Helena, and Nicholas Crouche. “Genome watch: breaking the ICE.” Nature Reviews
Microbiology 7 (2009): 328-329.
Silbergeld, Ellen, Jay Graham, and Lance Price. “Industrial food animal production, antimicrobial resistance,
and human health.” Annual Review of Public Health 29 (2008):151-69.
Top, Eva, and Dirk Springael. “The role of mobile genetic elements in bacterial adaptation to xenobiotic
organic compounds” Current Opinion in Biotechnolology 14 (2003): 262-269.
Wagner, Patrick, and Matthew Waldor. “Bacteriophage control of bacterial virulence.” Infection and
Immunity 70 (2002): 3985-3993.
Wang, Xiaoxue, Younghoon Kim, Qun Ma, Seok Hong, Karina Pokusaeva, Joseph Sturino, and Thomas
Wood. “Cryptic Prophages Help Bacteria Cope With Adverse Environments.” Nature Communications 1
(2010): 147.
Werren, JH. “Selfish Genetic Elements, Genetic Conflict, And Evolutionary Innovation.” Proceedings of the
National Academy of Sciences of the United States of America 108 Suppl 2 (2011): 10863-10870.
Whitman, William, David Coleman, and William Wiebe. “Prokaryotes: The Unseen Majority.” Proceedings
of the National Academy of Sciences of the United States of America 95 (1998): 6578-6583.
Wright, Gerard. “The origins of antibiotic resistance.” Handbook of Experimental Pharmacology 211 (2012):
13-30.
Figure 1. Steps in the evolution of mobile genetic elements (MGEs). Two ISs insert and capture a chromosomal gene, 1;
Composite transposons integrate into genetic platforms capable of moving from cell to cell (plasmids or phages), 2;
Successful MGEs exchange genes coding for adaptive functions, 3; MGEs with multiple morons are formed 4; Mobile
MGEs integrate into bacterial chromosome and lose unnecessary genes (chromosomally
 Plasmid 48 (2002) 202–212
www.academicpress.com

When phage, plasmids, and transposons

collide: genomic islands, and conjugative- and
mobilizable-transposons as a mosaic continuum
A. Mark Osborn* and Dietmar Bo
€ltner
Department of Biological Sciences, University of Essex, Colchester, CO4 3SQ, UK
Received 29 August 2002

Abstract

Plasmids and bacteriophage represent the classical vectors for gene transfer within the horizontal gene
pool. However, the more recent discovery of an increasing array of other mobile genetic elements (MGE)
including genomic islands (GIs), conjugative transposons (CTns), and mobilizable transposons (MTns)
which each integrate within the chromosome, offer an increasingly diverse assemblage contributing to
bacterial adaptation and evolution. Molecular characterisation of these elements has revealed that they are
comprised of functional modules derived from phage, plasmids, and transposons, and further that these
modules are combined to generate a continuum of mosaic MGE. In particular, they are comprised of any
one of three distinct types of recombinase, together with plasmid-derived transfer and mobilisation gene
functions. This review highlights both the similarities and distinctions between these integrating trans-
ferable elements resulting from combination of the MGE toolbox.
Ó 2002 Elsevier Science (USA). All rights reserved.

Keywords: Gene transfer; Genomic islands; Conjugative transposons; Conjugation; Tyrosine recombinase; Resolvase;
Transposase

1. Mobile genetic elements and the horizontal gene
pool
The prokaryotic horizontal gene pool (HGP)
represents a rich tapestry of adaptive phenotypes
conveyed within and between bacterial (and ar-
chaeal) cells, by an increasingly diverse assem-
blage of mobile genetic elements (MGE). Fifty
years on from the discovery of transduction by
*
Corresponding author. Fax: +44-1206-872592.
E-mail address: osborn@essex.ac.uk (A. Mark Os-
born).
Salmonella bacteriophage (Zinder and Lederberg,
1952) and the introduction of the term plasmid
(Lederberg, 1952), following the earlier identifi-
cation of bacterial conjugation, research has re-
vealed numerous combinations of genetic modules
derived from phage, plasmid (and transposons) to
generate an array of MGE that include conjuga-
tive and mobilizable transposons, genomic is-
lands, and integrons (Toussaint and Merlin,
2002).
Whilst plasmids and bacteriophage have well-
established genetic and phenotypic identities, the
emerging families of mosaic MGE pose a con-
siderable challenge for the systematicist as to how

0147-619X/02/$ - see front matter Ó 2002 Elsevier Science (USA). All rights reserved.
PII: S 0 1 4 7 - 6 1 9 X ( 0 2 ) 0 0 1 1 7 - 8
 A. Mark Osborn, D. B€oltner / Plasmid 48 (2002) 202–212 203

Fig. 1. The mosaic continuum of mobilizable and conjugative genetic elements. MGE (plasmids, phage, and transpo-
sons) that contribute key functional components are circled (dotted lines). MGE functional modules are as given in the
key. Black triangles inserted within the functional modules of the Anchored Genomic Island indicate that molecular
evidence exists for remnants or mutant forms of such modules (see text for further details).

best to classify such elements. Such complications
arise not least from the utilisation of three distinct
enzyme families catalysing the vital process of
recombination between DNA molecules (Tous-
saint and Merlin, 2002). These families are named
after key conserved residues in their active site: (i)
tyrosine recombinases (reviewed in Esposito and
Scocca, 1997; Nunes-D€ uby et al., 1998), (ii) serine
recombinases (resolvase/invertase) (reviewed in
Smith and Thorpe, 2002), and (iii) DD-E recom-
binases (transposases), named following the con-
served aspartate and glutamate residues (Polard
and Chandler, 1995). Moreover, the situation is
complicated further by the additional inclusion on
MGE of genes encoding plasmid-related mobili-
sation and transfer functions.
In this review we focus in particular on two
groups of MGE; the genomic islands (GIs) and
conjugative transposons (CTns), which from their
names at least suggest vastly differing entities. Yet
molecular comparison of their backbone modules
identifies key similarities, namely the presence of
phage-related integrases belonging to the tyrosine
recombinase family, and second, the occurrence of
plasmid-related transfer genes on CTns, and
which are being discovered increasingly on GIs.
However, whilst comparative analysis of these
elements reveals fascinating evolutionary insights,
it is apparent on further inspection that they
represent just part of a continuum of mosaic
MGE (Fig. 1) that are either self-transmissible or
mobilizable between bacterial cells. Thus this re-
view will focus also on the emerging groups of
mobilizable transposons comprised of different
combinations of recombinases and plasmid-re-
lated mobilisation functions.
2. Genomic islands
Genomic islands, which are found in some
bacterial strains but are absent from otherwise
very closely related strains, are now recognised as
important contributors to bacterial adaptation
and evolution. GIs, which vary in size between 10
and 500 kb, were first identified as chromosomally
204 A. Mark Osborn, D. B€oltner / Plasmid 48 (2002) 202–212
located virulence genes in uropathogenic Escheri-
chia coli, differing in G + C content and codon
usage from the surrounding DNA (Hacker et al.,
1983), and subsequently termed pathogenicity is-
lands (PAIs) by Hacker et al. (1990). More re-
cently PAIs have been identified in a diverse range
of animal pathogens from both Gram-negative
(Yersinia, Salmonella, Vibrio, Helicobacter, and
Neisseria) and Gram-positive genera (Staphylo-
coccus, Listeria, and Clostridium) (Hacker and
Kaper, 2000).
In addition, PAIs have also been identified in a
number of plant pathogens including Erwinia,
Xanthomonas (No€el et al., 2002), and Pseudomo-
nas syringae (Jackson et al., 1999). PAIs carry
genes encoding a variety of phenotypes including
adhesins, secretion systems and iron uptake sys-
tems critical to pathogenicity (reviewed in Hacker
and Kaper, 2000). Typically PAIs are integrated
into, or near to, tRNA genes and are flanked by
short direct repeats resembling phage attachment
sites (Hacker and Kaper, 2000). In addition to the
virulence genes they typically carry an integrase
gene related to that of phage lambda and be-
longing to the larger family of tyrosine recom-
binases. Recent comparison of tRNA attachment
sites for the broader group of elements utilising
tyrosine recombinases, has revealed three sub-lo-
cations for integration within these genes. Signif-
icantly, a phylogeny of integrase sequences has
also identified three groups that are consistent
with the tRNA sub-location classification (Wil-
liams, 2002).
In some GIs the integrase gene may be deleted
or non-functional, resulting in permanently an-
chored islands (see below). Other integrases,
however, have been demonstrated to be func-
tional, and thus, potentially capable of horizontal
gene transfer, especially if the island also carries
conjugal transfer genes, which is increasingly ob-
served (see below). Phenotypes carried by genomic
islands are not limited to those encoding virulence
but also include antibiotic resistance e.g., the SRL
PAI from Shigella flexneri (Turner et al., 2001),
degradation of xenobiotic compounds e.g., the clc
element encoding chlorocatechol degradation
(Ravatn et al., 1998) and symbiosis (Sym islands)
(Sullivan and Ronson, 1998). Acquisition of such
traits as functional units allows bacteria to re-
spond rapidly to environmental challenges and
explore new ecological niches. As a consequence
such islands have been termed ÔecologicalÕ or Ôfit-
nessÕ islands (Hacker and Carniel, 2001). Ac-
cordingly, given that the phenotype of bacteria
receiving such islands is often altered dramatically
(e.g., from non-pathogenic to pathogenic, or from
non-symbiotic to symbiotic), this phenomenon
has been described as Ôevolution in quantum leapsÕ
(Groisman and Ochman, 1996).
A number of the ecological islands have been
demonstrated to be conjugative. For example the
clc element (Ravatn et al., 1998) was initially
identified following conjugative transfer from
Pseudomonas strain B13 to P. putida Fl, and
subsequently found to integrate using a tyrosine
recombinase into either of two tRNA gene at-
tachment sites. Similarly, the bph-sal element,
encoding biphenyl and salicylate metabolism
functions, from Pseudomonas putida KF715 also
transfers by conjugation and is believed to inte-
grate within a specific insertion hot-spot (Nishi
et al., 2000), whilst a third conjugative element
from Pseudomonas aeruginosa JB2 encoding deg-
radation of hydroxy- and halo-aromatic com-
pounds is also believed to integrate within the
chromosome (Hickey et al., 2001). Conjugative
transfer has also been demonstrated for the 500 kb
Sym island from Mesorhizobium loti (Sullivan and
Ronson, 1998), and subsequently in the related
islands from M. loti R7A and MAFF303099
(Sullivan et al., 2002). Other likely members of
this group of Ôconjugative genomic islandsÕ include
the related conjugative integrating elements R391
and SXT that carry metal and/or antibiotic resis-
tance genes (see below).
In Gram-positive bacteria an increasing num-
ber of potentially transmissible GIs are identified.
In Streptococcus thermophilus, partial sequence
analysis of the integrative element ICEStl (Burrus
et al., 2000) has shown this to be a mosaic con-
taining a tyrosine recombinase related to that of
the CTns Tn5276 and Tn5252, together with
plasmid pSK41- and Tn916-related transfer genes.
Integration of this element has been shown to be
site-specific within its host, whilst at present it is
not known whether this element is capable of in-
dependent transfer. More recently, a 153 kb
pathogenicity island has been identified in En-
terococcus faecalis MMH594 PAI (Shankar et al.,
2002), that includes an integrase most closely re-
lated to that from Mesorhizobium loti GIs, and
also carries transfer genes related to the Entero-
coccus plasmids pAD1 and pAM373. Although
mobilisation of the MMH594 PAI has been
demonstrated, it is not currently known whether
the element is also self-transmissible.
Whilst the presence of plasmid-related transfer
genes on GIs offers conjugation as a dissemination
 A. Mark Osborn, D. B€oltner / Plasmid 48 (2002) 202–212
mechanism, GIs can also utilise other MGE and
mechanisms to facilitate their dispersion. The
15.2 kb PAI SaPI1 and the related SaPI2 from
Staphylococcus aureus which both carry the tst
gene for toxic shock syndrome toxin-1 (Lindsay
et al., 1998) both contain a putative integrase
(tyrosine recombinase) and are flanked by direct
repeats. Whilst these may facilitate integration
into the chromosome, UV-inducible excision has
not been demonstrated, suggesting an absence of
an excisionase on these elements. However, the
PAIs can be excised and circularised by certain
Staphylococcal phage (80a and /13) to form cir-
cular forms of the PAIs that can then re-integrate
in a site-specific manner. Alternatively, replication
of the circular form can lead to transduction of
the PAIs at high frequencies, and in the case of
SaPI1 this is a direct consequence of SaPI1 repli-
cation interfering with that of the phage, in a
semi-parasitic manner. Following encapsulation
of SaPI1 and the subsequent transfer to another
cell, SaPI1 uses its own integrase to integrate into
the recipient hostÕs chromosome (Ruzin et al.,
2001). Thus these elements, whilst lacking their
own transfer system, are readily transferred intact
via phage transduction, offering another example
of the numerous mechanistic variations found in
the HGP.
3. R391, SXT, and the IncJ elements
Two notable recent examples of the combina-
tion of conjugative plasmid transfer genes with
phage-related integration systems are offered by
two related elements; R391 from a South African
isolate of Providencia rettgeri (Coetzee et al.,
1972) and the SXT element from Vibrio cholerae,
isolated in India (Waldor et al., 1996). They form
part of a larger series of elements, including R997
(Matthew et al., 1979) and pMERPH (Peters et al.,
1991), classically, though now inappropriately,
referred to as IncJ elements (following plasmid
incompatibility nomenclature). The recent deter-
mination of the DNA sequences of R391 (B€ oltner
et al., 2002) and SXT (Beaber et al., 2002) shows
them to share conserved backbone regions, with
at least 95% identify to each other over 65 kb. In
particular, these backbone regions include a
phage-related integrase and associated regulatory
genes, together with a conjugative transfer system
related to that from the plasmid R27 (seen also in
SGI1 from S. enterica DT104—see below). How-
ever, in contrast to SGI1, both R391 and SXT are
205
capable of excision and subsequent conjugative
transfer to recipient cells.
Prior to the determination of the sequences of
these elements and the earlier demonstration of
the site-specific integration of SXT and R391 into
the prfC gene of E. coli (Hochhut and Waldor,
1999), the nature of R391, had represented
something of an enigma, being variously de-
scribed as a transmissible resistance factor, an
integrating plasmid, and subsequently as a con-
jugative transposon (Murphy and Pembroke,
1995). The absence of any identifiable plasmid
replicon (B€oltner et al., 2002) is consistent with
repeated failures to isolate ccc DNA and now
confirms that these elements are not plasmids.
Thus, this proposed classification of R391 and
SXT as conjugative transposons might at first
seem appropriate, at least at the level of pheno-
type, i.e., recA-independent chromosomal inte-
gration combined with conjugative transfer.
However, where these elements differ from ar-
chetypal conjugative transposons, such as Tn916,
is in the absence of random chromosomal inte-
gration (see below), with R391 and SXT only in-
tegrating into a single site on the E. coli
chromosome (see Fig. 1). In this respect, they
share greater similarities with the GIs that simi-
larly integrate into just one or two sites (typically
tRNA genes) on the chromosome. In retrospect,
on the basis of sequence data and phenotypic and
molecular characterisation, R391, isolated 30
years ago, may in fact have been the first genomic
island to be isolated, albeit a self-transmissable
one.
4. Anchored genomic islands and integrative plas-
mids
Analysis of the 43 kb genomic island (SGI1)
carrying multi-drug resistances, in Salmonella
enterica serovar Typhimurium DT104 (Boyd et al.,
2001) indicates a complex evolutionary history.
With a failure to detect excision of the SGI1 island
(Boyd et al., 2000) and the inability to demon-
strate transfer of the multi-drug resistances
(Threlfall et al., 1994), this island would now ap-
pear to be permanently anchored within the
chromosome (Fig. 1). However, the DNA se-
quence of SGI1 suggests a possible previous ex-
istence as an integrative conjugative plasmid, as
indicated by both the presence of a number of
ORFs related to genes from the IncHI1 plasmid
R27 encoding mating pair stabilisation and pilus
206 A. Mark Osborn, D. B€oltner / Plasmid 48 (2002) 202–212
assembly proteins; together with an ORF related
to the replication gene (repA) from the Rhodo-
pseudomonas plasmid pMG101.
Past traces of a conjugative lifestyle may also
be indicated in the Neisseria gonorrhoeae Gono-
coccal Genetic Island (GGI) by the presence of
traG and traH homologues (Dillard and Seifert,
2001), though subsequent insertion mutagenesis
suggests the fascinating possibility that these
genes function instead as a Type IV secretion
system acting as a potential DNA donation sys-
tem to facilitate DNA transformation, without
requiring autolysis (Hamilton et al., 2001). In-
deed, the similarities between plasmid conjugative
transfer systems and the type IV protein secretion
systems including those carried by the cag PAI of
Helicobacter pylori or, responsible for extracellu-
lar transport of the pertussis toxin in Bordetella
pertussis, suggest common evolutionary origins
for these systems (Christie, 2001). Given that the
incorporation of plasmid-derived conjugative
transfer genes into the chromosome is readily
demonstrated, it is tempting to speculate on the
possibility for divergence and subsequent evolu-
tion over time of these genes to fulfil new func-
tional roles. Alternatively, such similarities
between DNA and protein transporters may rep-
resent divergent evolution of ancestral macro-
molecule secretion systems.
Chromosomal integration of plasmids has long
been recognised, most notably of the F plasmid in
the formation of E. coli Hfr strains. Integration of
F typically occurs by homologous recombination
between IS elements (IS2 or IS3a and 3b) present
on the F plasmid and chromosome. Whilst inte-
gration by homologous recombination enables
plasmids to form stable cointegrates, an increas-
ing number of plasmids are reported to carry a
lambda-related integrase that facilitates recA-in-
dependent chromosomal integration and excision.
For example, the Streptomyces ambofaciens
11.2 kb conjugative circular plasmid pSAM2
(Boccard et al., 1989) carries a tyrosine recom-
binase in addition to a rolling circle type replicon
and conjugative transfer genes (Fig. 1). Similarly,
a lambda-like integrase mediates integration of
the 11.3 kb plasmid pSE101 into a threonine
tRNA gene in its host Saccharoployspora eryth-
raea. In contrast, when pSE101 is, introduced into
Streptomyces lividans the element demonstrates
integration into multiple sites (Brown et al., 1994)
more akin to the relatively random integration of
CTns. In Gram-negative bacteria tyrosine re-
combinase-mediated integration of plasmids has
also been reported, with reversible integration of
the plasmid pKLK106 into two lysine tRNA
genes in Pseudomonas aeruginosa (Kiewitz et al.,
2000). We have also identified ORFs encoding
potential tyrosine recombinases using the PFAM
database within archaeal plasmids (Osborn and
B€oltner, unpublished data) namely, ORF439 in
the 41.2 kb conjugative plasmid pNOB8 from
Sulfolobus sp. NOB8H2 (She et al., 1998), and the
related ORF457 in the pING plasmid from Sulf-
olobus islandicus (Stedman et al., 2000). Whilst the
function of their gene products is unknown, we
speculate that these plasmids may represent the
first examples of integrative plasmids in the ar-
chaea.
5. Plasmid-located genomic islands
Localisation of GIs is not limited to the chro-
mosome, with a number of elements found lo-
cated on plasmids, in particular the PAI on the
Shigella flexneri virulence plasmid pWR201
(Venkatesan et al., 2001) but also notably in the
discovery of a PAI carrying the three toxin genes
on the plasmid pXO1 from Bacillus anthracis
(Okinaka et al., 1999). Whereas this PAI includes
an integrase gene, it is likely that its insertion into
pXO1 is mediated by insertion sequences, as op-
posed to a tyrosine recombinase (Fig. 1), as evi-
denced by the presence of copies of IS1627, which
encode a DD-E type recombinase (transposase) at
the flanking ends of the 44.3 kb element. Similarly,
sequence analysis of a 27.5 kb PAI found on the
80 kb plasmid p33071 in Rhodococcus equi again
suggest involvement of transposons in PAI inser-
tion, as evidenced by the presence of two trans-
poson-related resolvases at either end of the PAI
(Takai et al., 2000) These examples of PAIs that
utilise transposons for insertion represent yet an-
other intriguing illustration of the versatility of
mosaic MGE.
6. Conjugative transposons
Conjugative transposons were first identified as
chromosomally-borne transposons that were ca-
pable of conjugative transfer. Two such elements
were originally identified: Tn916 from Enterococ-
cus faecalis (Franke and Clewell, 1981), which is
now the accepted archetype of the CTns, and
Tn5253 from Streptococcus pneumoniae (Shoe-
maker et al., 1980). Whilst numerous CTns have
 A. Mark Osborn, D. B€oltner / Plasmid 48 (2002) 202–212
subsequently been identified, albeit predominantly
from Gram-positive bacteria, much of our mo-
lecular understanding of CTns stems from analy-
ses of Tn916 and the Bacteroides element
CTnDOT. A full review of the CTns is beyond the
scope of this current paper, however, we refer
readers to the excellent reviews by Salyers et al.
(1995) and Scott and Churchward (1995).
Integration of the majority of CTns utilises a
tyrosine recombinase, as is similarly the case for
most GIs. Classical CTns such as Tn916 demon-
strate semi-random integration in a manner more
redolent of classical transposons. However, Tn916
does demonstrate distinct preferences for AT-rich
regions with a consensus target of 50 -TT/
ATTTT(N6 )AAAAAA/TA-30 (Lu and Church-
ward, 1995), although no individual base within
this generic sequence is conserved amongst the
numerous integration sites identified. In contrast,
CTnDOT shows greater specificity and is found to
integrate into only about seven sites in the Bac-
teroides chromosome (Bedzyk et al., 1992).
Less is known about the transfer system of
Tn916, although FASTA analysis (Osborn and
B€oltner, unpublished data) of ORF21 of Tn916
shows it to be a member of the FtsK/SpoIIIE
family. SpoIIIE genes are believed to drive DNA
transport from the mother cell into the prespore
during sporulation in Bacillus (Bath et al., 2000)
and the increasing number of SpoIIIE homo-
logues suggest that this may be a widespread
mechanism for DNA transport. Moreover, Spo-
IIIE homologues are also responsible for conju-
gative transfer of the Streptomyces plasmids
pIJ101 (Pettis and Cohen, 2001) and pSAM2
(Hagege et al., 1993). FASTA analysis of ORF20
(Osborn and B€ oltner, unpublished data) identifies
relationships to certain Vibrio phage replication
proteins (Nasu et al., 2000) and the rep gene from
the Bacillus plasmid pGI3 (Hoflack et al., 1997),
suggesting a possible role for this ORF for DNA
nicking prior to transfer. A functional origin of
transfer has been located between ORFs 20 and
21, and includes sequences related to both IncP
and F plasmid nic sites (Jaworski and Clewell,
1995).
The transfer system of CTnDOT is unrelated
to that of Tn916-like elements, and is thought to
represent a new class of conjugal transfer system
(Bonheyo et al., 2001). Following our own FAS-
TA analysis, we again found that many of the
putative ORFs described show no relationship to
known transfer genes. However, the putative TraP
protein shows homology to DNA primases, whilst
207
the TraG protein is homologous to the BctA
protein involved in conjugative transfer of the
Bacteroides pBF4 plasmid (Morgan and Macrina,
1997), and also to a number of proteins related to
the TraC protein from the F plasmid, which is
involved in pilus assembly. Similarly the TraJ
protein from CTnDOT is related to the TrwI
protein from the transfer operon of R388. The
CTnDOT TraA protein is related to a number of
ParA proteins involved in plasmid partitioning.
Of two additional ORFS upstream of the putative
tra genes on the CTnDOT sequence (accession
number: AF289050), the putative protein encoded
by ORF1 is homologous to relaxase proteins from
both conjugative plasmids and mobilizable trans-
posons. Thus it is probable that the CTnDOT
transfer operon represents a distant relative of
Gram-negative conjugative plasmid transfer sys-
tems.
For both Tn916 and CTnDOT, it is possible
that the difficulties in identifying relationships to
well-known plasmid conjugation systems may be
a consequence that these CTns have been isolated
from organisms for which our knowledge of
plasmid biology is relatively sparse, notwith-
standing the more detailed characterisation of the
Enterococcus pheromone-sensing plasmids pAD1
(Francia et al., 2001) and pAM373 (De Boever et
al., 2000). Thus, as more plasmids are sequenced
from these organisms, and in particular from
Bacteroides, this may identify a possible closer
linkage between the transfer systems of classical
CTns and those from conjugative plasmids.
Over recent years a number of elements iso-
lated from members of the proteobacteriaceae
have been proposed as CTns, including R391, and
SXT (see above), CTnscr94 from Salmonella
senftenberg (Hochhut et al., 1997) and the M. loti
strain R7A symbiosis island (Sullivan et al., 2002),
yet many of these elements would appear to have
more in common with genomic islands, albeit
conjugative ones, on the basis of their preference
for specific integration sites, in particular tRNA
genes, as opposed to the more random integration
exhibited by classical CTns. The 55 kb catabolic
biphenyl degradation transposon Tn4371 from
Ralstonia eutropha CH34 does, however, offer a
better candidate for a conjugative transposon
within the proteobacteriaceae. This element that
utilises a tyrosine recombinase demonstrates semi-
random integration into a number of sites present
on both the CH34 chromosome and the co-resi-
dent pMOL plasmid. However, whilst Tn4371
carries RP4/Ti-related transfer genes, conjugative
208 A. Mark Osborn, D. B€oltner / Plasmid 48 (2002) 202–212
transfer of the element has not been demon-
strated, although the transfer genes have been
demonstrated to be functional via mobilisation of
a second transposable element Tn-bph (Merlin
et al., 1999).
7. Mobilizable transposons
Classical transposons utilise one of two types
of recombinase. Firstly transposons such as Tn3
and cd utilise a serine recombinase (resolvase/in-
vertase). Alternatively, a second group including
Tn7, Tn10, and the IS3 family utilise the so-called
DD-E recombinase (transposase). Upon compar-
ison of such classical transposons with their
namesakes, the conjugative transposons (that
utilise a tyrosine recombinase), it is attractive to
conjecture upon the existence of truly randomly
integrating, conjugative MGE, perhaps combin-
ing a DD-E or serine recombinase and a complete
plasmid-related conjugative transfer system. At
present such systems await discovery, yet an in-
creasing number of mobilizable transposons
(MTns), consisting of a transposon that includes
an oriT site, enabling mobilisation by a co-resi-
dent conjugative plasmid, have been reported.
The majority of these have been identified in
Bacteroides isolates, with the Non-replicating
Bacteroides Units (NBUs), NBU1 and NBU2,
and the mobilizable transposon Tn4555 all iden-
tified as utilising a tyrosine recombinase-based
integration system (Fig. 1) to integrate into either
tRNA genes (NBU1 and NBU2) (Shoemaker
et al., 1996a; Wang et al., 2000), or semi-randomly
(Tn4555) [as is the case for Tn916, (Tribble et al.,
1997)]. Intriguingly, NBU1, as described above
for pSE 101, when transferred into E. coli, dem-
onstrates semi-random integration (Shoemaker
et al., 1996b), suggesting that site-specificity
of tyrosine recombinase-based systems is a
host-specific phenomenon. Mobilisation of these
elements utilises oriT sequences and mob genes
resembling those from Gram-positive plasmids
(Shoemaker et al., 2000; Smith and Parker, 1998;
Wang et al., 2000).
In contrast, Tn4451 and Tn4453 from Clos-
tridium spp. represent a second type of MTn that
instead utilise a serine recombinase (the TnpX
resolvase) (Lyras and Rood, 2000) to mediate
excision and insertion, and additionally carry a
plasmid-related mob gene (Fig. 1). These trans-
poson-encoded mob determinants permit mobili-
sation of Tn4551—carrying plasmids lacking their
own mob system by a helper plasmid, such as RP4
(Crellin and Rood, 1992).
A third type of mobilizable element, Tn5398
from C. difficile 630 represents yet another fasci-
nating variation on this theme, being mobilizable
and capable of semi-random integration upon
transfer to Bacillus subtilis, yet following analysis
of the DNA sequence of this element, Tn5398 was
found to lack both a gene encoding any recogni-
sable recombinase, required for integration, as
well as potential mobilisation genes. It does,
however, carry a putative oriT site related to that
of the conjugative transposon Tn916 (Fig. 1). It is
postulated that Tn5398 utilises both resolvases
and mob gene functions carried on the conjugative
transposon Tn5397 which is co-resident in the
same host (Farrow et al., 2001).
8. Conjugative transposons or conjugative genomic
islands—that is the question
The state of MGE nomenclature is a direct
consequence of their highly mosaic composition,
with indistinct boundaries between a number of
groups of elements. Obviously, historical prece-
dence in the naming of individual elements will
lead to future researchers trying to classify newly
discovered elements as part of existing groupings.
However, as new DNA sequences are generated
and molecular mechanisms unravelled, it is clear
that some earlier classifications are inappropriate,
whilst others have led to highly similar MGE be-
ing classified with very different names. Arguably,
this is most apparent with respect to the conju-
gative transposons (e.g., Tn916 and CTn-DOT)
and the increasing number of GIs that encode
their own conjugation systems (e.g., the clc ele-
ment, the M. loti R7A symbiosis island, and
R391). On molecular comparison of these ele-
ments, they clearly share a tyrosine recombina-
tion-type integration mechanism, and both utilise
conjugative transfer systems that are related, al-
beit to differing degrees, to those from conjugative
plasmids. Arguably the clearest distinction be-
tween CTns and GIs is with respect to integration
site preference, namely, whether the element in-
tegrates into multiple sites (as is the case for
Tn916), or into one, or occasionally two, sites, as
is typical for the GIs. Perhaps on this basis alone
CTns and a second group, for which we propose
the name Conjugative Genomic Islands, can be
delineated. We would suggest that CTn is retained
for those classical CTns exhibiting random or
 A. Mark Osborn, D. B€oltner / Plasmid 48 (2002) 202–212
semi-random integration, whilst those conjugative
integrating elements which have distinct site-
preference, which in many cases will be for a
tRNA gene, or in the case of R391 and SXT for
prfC, are termed conjugative genomic islands (a
sub-family of the GIs). However, even here the
waters are muddied by elements such as the mo-
bilizable transposons NBU1 and NBU2, or the
integrative plasmid pSE101 (see above) that show
single-site specificity in one host, but integration
into multiple sites in other recipients. As a third
group we propose the adoption of the generic
term mobilizable transposon, to cover MGE that
integrate (irrespective of their type of recombin-
ase) into the chromosome, and can be mobilised
via conjugation to other cells, but lack a conju-
gation system of their own. This term has been
adopted already for a number of MGE (see
above), but a number of additional elements,
would clearly fall within this group.
Whilst some may ask ÔWhat’s in a name? That
which we call a rose by any other name would smell
as sweet,Õ classification remains an important dis-
cipline within biology. For MGE, one approach is
to accept their mosaic nature, and classify all
MGE on the presence of key functional modules.
Whilst such an approach is extremely valuable
(Toussaint and Merlin, 2002), and will increas-
ingly further benefit from ongoing developments
of a number of MGE-module specific websites,
the desire to also assign generic names to new and
existing elements remains compelling. We hope
that the MGE research community will face these
challenges and be willing to embrace appropriate
changes to nomenclature, as needed. As a lead, we
point towards the numerous examples of reclas-
sification of bacterial genera and species that have
resulted from ribosomal RNA-based phylogenetic
studies over the last 15 years.
Acknowledgments
The authors wish to thank Professor Julian
Rood for an introduction to the serine-recom-
binase-based mobilizable transposons.
References
Bath, J., Wu, L.J., Errington, J., Wang, J.C., 2000. Role
of Bacillus subtilis SpoIIIE in DNA transport across
the mother cell-prespore division septum. Science
290, 995–997.
209
Beaber, J.W., Hochhut, B., Waldor, M.K., 2002. Geno-
mic and functional analyses of SXT, an integrating
antibiotic resistance gene transfer element derived
from Vibrio cholerae. J. Bacteriol. 184, 4259–4269.
Bedzyk, L.A., Shoemaker, N.B., Young, K.E., Salyers,
A.A., 1992. Insertion and excision of Bacteroides
conjugative chromosomal elements. J. Bacteriol. 174,
166–172.
Boccard, F., Smokvina, T., Pernodet, J.L., Friedmann,
A., Guerineau, M., 1989. The integrated conjugative
plasmid pSAM2 of Streptomyces ambofaciens is
related to temperate bacteriophages. EMBO J. 8,
973–980.
B€
oltner, D., MacMahon, C., Pembroke, J.T., Strike, P.,
Osborn, A.M., 2002. R391: a conjugative integrating
mosaic comprised of phage, plasmid and transposon
elements. J. Bacteriol. 184, 5158–5169.
Bonheyo, G., Graham, D., Shoemaker, N.B., Salyers,
A.A., 2001. Transfer region of a Bacteroides conju-
gative transposon, CTnDOT. Plasmid 45, 41–51.
Boyd, D., Peters, G.A., Cloeckaert, A., Boumedine,
K.S., Chaslus-Dancla, E., Imberechts, H., Mulvey,
M.R., 2001. Complete nucleotide sequence of a 43-
kilobase genomic island associated with the multi-
drug resistance region of Salmonella enterica serovar
Typhimurium DT104 and its identification in phage
type DT120 and serovar Agona. J. Bacteriol. 183,
5725–5732.
Boyd, D.A., Peters, G.A., Ng, L.-K., Mulvey, M.R.,
2000. Partial characterization of a genomic island
associated with the multidrug resistance region of
Salmonella enterica Typhimurium DT104. FEMS
Microbiol. Lett. 189, 285–291.
Brown, D.P., Idler, K.B., Backer, D.M., Donadio, S.,
Katz, L., 1994. Characterization of the genes and
attachment sites for site-specific integration of plas-
mid pSE101 in Saccharopolyspora erythraea and
Streptomyces lividans. Mol. Gen. Genet. 242, 185–
193.
Burrus, V., Roussel, Y., Decaris, B., Guedon, G., 2000.
Characterization of a novel integrative element,
ICEStl, in the lactic acid bacterium Streptococcus
thermophilus. Appl. Environ. Microbiol. 66, 1749–
1753.
Christie, P.J., 2001. Type IV secretion: intracellular
transfer of macromolecules by systems ancestrally
related to conjugation machines. Mol. Micro. 40,
294–305.
Coetzee, J.N., Datta, N., Hedges, R.W., 1972. R factors
from Proteus rettgeri. J. Gen. Microbiol. 72, 543–552.
Crellin, P.K., Rood, J.I., 1992. Tn4451 from Clostridium
perfringens is a mobilizable transposon that encodes
the functional Mob protein, TnpZ. Mol. Microbiol.
28, 631–642.
De Boever, E.H., Clewell, D.B., Fraser, C.M., 2000.
Enterococcus faecalis conjugative plasmid pAM373:
complete nucleotide sequence and genetic analyses of
sex pheromone response. Mol. Microbiol. 37, 1327–
1341.
210 A. Mark Osborn, D. B€oltner / Plasmid 48 (2002) 202–212
Dillard, J.P., Seifert, H.S., 2001. A variable genetic
island specific for Neisseria gonorrhoeas is involved
in providing DNA for natural transformation and is
found more often in disseminated infection isolates.
Mol. Microbiol. 41, 263–277.
Esposito, D., Scocca, J.J., 1997. The integrase family of
tyrosine recombinases: evolution of a conserved
active site domain. Nucl. Acids Res. 25, 3605–3614.
Farrow, K.A., Lyras, D., Rood, J.I., 2001. Genomic
analysis of the erythromycin resistance element
Tn5398 from Clostridium difficile. Microbiology
147, 2717–2728.
Francia, M.V., Haas, W., Wirth, R., Samberger, E.,
Muscholl-Silberhorn, A., Gilmore, M.S., Ike, Y.,
Weaver, K.E., An, F.Y., Clewell, D.B., 2001. Com-
pletion of the nucleotide sequence of the Enterococ-
cus faecalis conjugative virulence plasmid pAD1 and
identification of a second transfer origin. Plasmid 46,
111–127.
Franke, A.E., Clewell, D.B., 1981. Evidence for a
chromosome-borne resistance transposon (Tn916)
in Streptococcus faecalis that is capable of ‘‘conju-
gal’’ transfer in the absence of a conjugative plasmid.
J. Bacteriol. 145, 494–502.
Groisman, E.A., Ochman, H., 1996. Pathogenicity
islands: bacterial evolution in quantum leaps. Cell
87, 791–794.
Hacker, J., Bender, L., Ott, M., Wingender, J., Lund, B.,
Marre, R., Goebel, W., 1990. Deletions of chromo-
somal regions coding for fimbriae and hemolysins
occur in vivo and in vitro in various extra-intestinal
Escherichia coli isolates. Microb. Pathog. 8, 213–
225.
Hacker, J., Carniel, E., 2001. Ecological fitness, genomic
islands and bacterial pathogenicity. A Darwinian
view of the evolution of microbes. EMBO Reports 2,
376–381.
Hacker, J., Kaper, J.B., 2000. Pathogenicity islands and
the evolution of microbes. Ann. Rev. Microbiol. 54,
641–679.
Hacker, J., Knapp, S., Goebel, W., 1983. Spontaneous
deletions and flanking regions of the chromosomal
inherited hemolysin determinant of an Escherichia
coli O6 strain. J. Bacteriol. 154, 1145–1152.
Hagege, J., Pernodet, J.L., Sezonov, G., Gerbaud, C.,
Friedmann, A., Guerineau, M., 1993. Transfer
functions of the conjugative integrating element
pSAM2 from Streptomyces ambofaciens: character-
ization of a kil-kor system associated with transfer. J.
Bacteriol. 175, 5529–5538.
Hamilton, H.L., Schwartz, K.J., Dillard, J.P., 2001.
Insertion-duplication mutagenesis of Neisseria: use
in characterization of DNA transfer genes in the
Gonococcal Genetic Island. J. Bacteriol. 183, 4718–
4726.
Hickey, W.J., Sabat, G., Yuroff, A.S., Arment, A.R.,
Perez-Lesher, J., 2001. Cloning, nucleotide sequenc-
ing, and functional analysis of a novel, mobile cluster
of biodegradation genes from Pseudomonas aerugin-
osa strain JB2. Appl. Environ. Microbiol. 67, 4603–
4609.
Hochhut, B., Jahreis, K., Lengeler, J.W., Schmid, K.,
1997. CTnscr94, a conjugative transposon found in
enterobacteria. J. Bacteriol. 179, 2097–2102.
Hochhut, B., Waldor, M.K., 1999. Site-specific integra-
tion of the conjugal Vibrio cholerae SXT element into
prfC. Mol. Micro. 32, 99–110.
Hoflack, L., Seurinck, J., Mahillon, J., 1997. Nucleotide
sequence and characterization of the cryptic Bacillus
thuringiensis plasmid pG13 reveals a new type of
rolling circle replicon. J. Bacteriol. 179, 5000–5008.
Jackson, R.W., Athanassopoulos, E., Tsiamis, G.,
Mansfield, J.W., Sesma, A., Arnold, D.L., Gibbon,
M.J., Murillo, J., Taylor, J.D., Vivian, A., 1999.
Identification of a pathogenicity island, which con-
tains genes for virulence and avirulence, on a large
native plasmid in the bean pathogen Pseudomonas
syringae pathovar phaseolicola. Proc. Natl. Acad.
Sci. (USA) 96, 10875–10880.
Jaworski, D.D., Clewell, D.B., 1995. A functional origin
of transfer (oriT) on the conjugative transposon
Tn916. J. Bacteriol. 177, 6644–6651.
Kiewitz, C., Larbig, K., Klockgether, J., Weinel, C.,
T€ummler, B., 2000. Monitoring genome evolution ex
vivo: reversible chromosomal integration of a 106 kb
plasmid at two tRNALys gene loci in sequential
Pseudomonas aeruginosa airway isolates. Microbiol-
ogy 146, 2365–2373.
Lederberg, J., 1952. Cell genetics and hereditary symbi-
oses. Physiol. Rev. 32, 403–430.
Lindsay, J.A., Ruzin, A., Ross, H.F., Kurepina, N.,
Novick, R.P., 1998. The gene for toxic shock toxin is
carried by a family of mobile pathogenicity islands in
Staphylococcus aureus. Mol. Microbiol. 29, 527–543.
Lu, F., Churchward, G., 1995. Tn916 target DNA
sequences bind the C-terminal domain of integrase
protein with different affinities that correlate with
transposon insertion frequency. J. Bacteriol. 177,
1938–1946.
Lyras, D., Rood, J.I., 2000. Transposition of Tn4451
and Tn4453 involves a circular intermediate that
forms a promoter for the large resolvase, TnpX.
Mol. Micro. 38, 588–601.
Matthew, M., Hedges, R.W., Smith, J.T., 1979. Types of
beta-lactamase determined by plasmids in Gram-
negative bacteria. J. Bacteriol. 138, 657–662.
Merlin, C., Springael, D., Toussaint, A., 1999. Tn4371: a
modular structure encoding a phage-like integrase, a
Pseudomonas-like catabolic pathway, and RP4/Ti-
like transfer functions. Plasmid 41, 40–54.
Morgan, R.M., Macrina, F.L., 1997. bctA: A novel
pBF4 gene necessary for conjugal transfer in Bacte-
roides spp. Microbiology 143, 2155–2165.
Murphy, D.B., Pembroke, J.T., 1995. Transfer of the
IncJ plasmid R391 to recombination deficient Esc-
herichia coli K12: evidence that R391 behaves as a
conjugal transposon. FEMS Microbiol. Lett. 134,
153–158.
 A. Mark Osborn, D. B€oltner / Plasmid 48 (2002) 202–212
Nasu, H., Iida, T., Sugahara, T., Yamaichi, Y., Park,
K.S., Yokoyama, K., Makino, K., Shinagawa, H.,
Honda, T., 2000. A filamentous phage associated
with recent pandemic Vibrio parahaemolyticus O3:K6
strains. J. Clin. Microbiol. 38, 2156–2161.
Nishi, A., Tominaga, K., Furukawa, K., 2000. A 90-
kilobase conjugative chromosomal element coding
for biphenyl and salicylate catabolism in Pseudomo-
nas putida KF715. J. Bacteriol. 182, 1949–1955.
No€el, L., Thieme, F., Nennstiel, D., Bonas, U., 2002.
Two novel type III-secreted proteins of Xanthomonas
campestris pv. vesicatoria are encoded within the hrp
pathogenicity island. J. Bacteriol. 184, 1340–1348.
Nunes-D€ uby, S.E., Kwon, H.J., Tirumalai, R.S., Ellen-
berger, T., Landy, A., 1998. Similarities and differ-
ences among 105 members of the Int family of
site-specific recombinases. Nucl. Acids Res. 26, 391–
406.
Okinaka, R.T., Cloud, K., Hampton, O., Hoffmaster,
A.R., Hill, K.K., Keim, P., Koehler, T.M., Lamke,
G., Kumano, S., Mahillon, J., Manter, D., Martinez,
Y., Ricke, D., Svensson, R., Jackson, P.J., 1999.
Sequence and organization of pX01, the large
Bacillus anthracis plasmid harboring the anthrax
toxin genes. J. Bacteriol. 181, 6509–6515.
Peters, S.E., Hobman, J.L., Strike, P., Ritchie, D.A.,
1991. Novel mercury resistance determinants carried
by IncJ plasmids pMERPH and R391. Mol. Gen.
Genet. 228, 294–299.
Pettis, G.S., Cohen, S.N., 2001. Unravelling the essential
role in conjugation of the Tra protein of Streptomy-
ces lividans plasmid pIJ101 Anton. V. Leeuwen. Int. J.
Gen. Mol. Microbiol. 79, 247–250.
Polard, P., Chandler, M., 1995. Bacterial transposases
and retroviral integrases. Mol. Microbiol. 15, 13–23.
Ravatn, R., Studer, S., Springael, D., Zehnder, A.J.B.,
van der Meer, J.R., 1998. Chromosomal integration,
tandem amplification, and deamplification in Pseu-
domonas putida F1 of a 105-kilobase genetic element
containing the chlorocatechol degradative genes
from Pseudomonas sp. strain B13. J. Bacteriol. 180,
4360–4369.
Ruzin, A., Lindsay, J., Novick, R.P., 2001. Molecular
genetics of SaPI1—a mobile pathogenicity island in
Staphylococcus aureus. Mol. Microbiol. 41, 365–377.
Salyers, A.A., Shoemaker, N.B., Stevens, A.M., Li, L.-
Y., 1995. Conjugative transposons: an unusual and
diverse set of integrated gene transfer elements.
Microbiol. Rev. 59, 579–590.
Scott, J.R., Churchward, G.G., 1995. Conjugative
transposition. Ann. Rev. Microbiol. 49, 367–397.
Shankar, N., Baghdayan, A.S., Gilmore, M.S., 2002.
Modulation of virulence within a pathogenicity
island in vancomycin-resistant Enterococcus faecalis.
Nature 417, 746–750.
She, Q., Phan, H., Garrett, R.A., Albers, S.V., Stedman,
K.M., Zillig, W., 1998. Genetic profile of pNOB8
from Sulfolobus: the first conjugative plasmid from
an archaeon. Extremophiles 2, 417–425.
211
Shoemaker, N.B., Smith, M.D., Guild, W.R., 1980.
DNAse-resistant transfer of chromosomal cat and let
insertions by filter mating in pneumococcus. Plasmid
3, 80–87.
Shoemaker, N.B., Wang, G.-R., Salyers, A.A., 1996a.
The Bacteroides mobilizable insertion element,
NBU1, integrates into the 30 end of a Leu-tRNA
gene and has an integrase that is a member of the
lambda integrase family. J. Bacteriol. 178, 3594–
3600.
Shoemaker, N., Wang, G., Salyers, A.A., 1996b. NBU1,
a mobilizable site-specific integrated element from
Bacteroides spp., can integrate nonspecifically in
Escherichia coli. J. Bacteriol. 178, 3601–3607.
Shoemaker, N.B., Wang, G.-R., Salyers, A.A., 2000.
Multiple gene products and sequences required
for the excision of the mobilizable integrated Bacte-
roides element NBU1. J. Bacteriol. 182, 928–
936.
Smith, C.J., Parker, A.C., 1998. The transfer origin for
Bacteroides mobilizable transposon Tn4555 is related
to a plasmid family from Gram-positive bacteria. J.
Bacteriol. 180, 435–439.
Smith, C.M., Thorpe, H.M., 2002. Diversity in the serine
recombinases. Mol. Microbiol. 44, 299–307.
Stedman, K.M., She, Q., Phan, H., Holz, I., Singh, H.,
Prangishvili, D., Garrett, R., Zillig, W., 2000. pING
family of conjugative plasmids from the extremely
thermophilic archaeon Sulfolobus islandicus: insights
into recombination and conjugation in Cre-
narchaeota. J. Bacteriol. 182, 7014–7020.
Sullivan, J.T., Ronson, C.W., 1998. Evolution of rhizo-
bia by acquisition of a 500 kb symbiosis island that
integrates into a phe-tRNA gene. Proc. Natl. Acad.
Sci. (USA) 95, 5145–5149.
Sullivan, J.T., Trzebiatowski, J.R., Cruickshank, R.W.,
Gouzy, J., Brown, S.D., Elliot, R.M., Fleetwood,
D.J., McCallum, N.G., Rossbach, U., Stuart, G.S.,
Weaver, J.E., Webby, R.J., de Bruijn, F.J., Ronson,
C.W., 2002. Comparative sequence analysis of the
symbiosis island of Mesorhizobium loti strain R7A. J.
Bacteriol. 184, 3086–3095.
Takai, S., Hines, S.A., Sekizaki, T., Nicholson, V.M.,
Alperin, D.A., Osaki, M., Takamatsu, D., Nakam-
ura, M., Suzuki, K., Ogino, N., Kakuda, T., Dan,
H., Prescott, J.F., 2000. DNA sequence and com-
parison of virulence plasmids from Rhodococcus equi
ATCC 33701 and 103. Infect. Immun. 68, 6840–
6847.
Threlfall, E.J., Frost, J.A., Ward, L.R., Rowe, B., 1994.
Epidemic in cattle and humans of Salmonella
typhimurium DT104 with chromosomally integrated
multiple drug resistance. Vet. Rec. 137, 577.
Toussaint, A., Merlin, C., 2002. Mobile elements as a
combination of functional modules. Plasmid 47, 26–
35.
Tribble, G.D., Parker, A.C., Smith, C.J., 1997. The
Bacteroides mobilizable transposon Tn4555 inte-
grates by a site-specific recombination mechanism
212 A. Mark Osborn, D. B€oltner / Plasmid 48 (2002) 202–212
similar to that of the Gram-positive bacterial element
Tn916. J. Bacteriol. 179, 2731–2739.
Turner, S.A., Luck, S.N., Sakellaris, H., Rajakumar, K.,
Adler, B., 2001. Nested deletions of the SRL
Pathogenicity Island of Shigella flexneri 2a. J.
Bacteriol. 183, 5535–5543.
Venkatesan, M.M., Goldberg, M.B., Rose, D.J., Grot-
beck, E.J., Burland, V., Blattner, F.R., 2001. Com-
plete DNA sequence and analysis of the large
virulence plasmid of Shigella flexneri. Infect. Immun.
69, 3271–3285.
Waldor, M.K., Tschape, H., Mekalanos, J.J., 1996. A
new type of conjugative transposon encodes resis-
tance to sulfamethoxazole, trimethoprim, and strep-
tomycin in Vibrio cholerae O139. J. Bacteriol. 178,
4157–4165.
Wang, J., Shoemaker, N., Wang, G.-R., Salyers, A.A.,
2000. Characterization of a Bacteroides mobilizable
transposon, NBU2, which carries a functional linco-
mycin resistance gene. J. Bacteriol. 182, 3559–3571.
Williams, K.P., 2002. Integration sites for genetic
elements in prokaryotic tRNA and tmRNA genes:
sublocation preference of integrase subfamilies.
Nucl. Acids Res. 30, 866–875.
Zinder, N.D., Lederberg, J., 1952. Genetic exchange in
Salmonella. J. Bacteriol. 64, 679–699.
 Integrons and the Origin of Antibiotic
Resistance Gene Cassettes
Super integrons with thousands of gene cassettes may have set the
stage for pathogens to develop antibiotic resistance very rapidly
Didier Mazel
ntegrons, which are natural genetic engi-
I neering platforms, can incorporate open
reading frames and convert them to
functional genes by ensuring correct ex-
pression. Although such integrons play a
major role in multidrug resistance phenomena
among gram-negative species, similar platforms
occur in numerous bacterial species, where they
can play other roles. Meanwhile, we have evi-
dence suggesting that they are a major source of
multidrug resistance integrons and their resis-
tance gene cassettes observed in clinical isolates.
Recruiting exogenous genes represents a
rapid adaptation against antimicrobial com-
pounds, and the integron functional platform
seems perfectly suited for capturing those spe-
cific genes that enable bacterial pathogens to
face challenges posed by multiple antibiotic
treatment regimes. With the discovery of the
superintegron and the thousands of cassettes
entrapped in integrons of environmental spe-
cies, we now have a better view of the immense
resources embedded within this system. How-
ever, several questions of importance about this
system are still without satisfactory answers,
including insights about specific recombination
processes, cassette genesis, and cassette ex-
change dynamics within complex bacterial pop-
ulations.
Didier Mazel is
head of the Unité
Antibiotic Resistance-Encoding Integrons
“Plasticité du Gé-
nome Bactérien”- With few exceptions, antibiotic resistance in
CNRS URA 2171, bacterial pathogens was identified soon after
dept Structure et particular drugs were introduced into clinical
Dynamique des Gé- practice, illustrating the genetic flexibility of
nomes, Institut Pas- bacteria. For instance, pathogens developed re-
teur, Paris, France. sistance to sulfonamides (Su) and penicillin in
520 Y ASM News / Volume 70, Number 11, 2004
the late 1930s, and in the late 1940s, mycobac-
teria developed resistance to streptomycin.
These single-resistance phenotypes were not
entirely unforeseen. For example, laboratory
studies had demonstrated penicillin-resistant
point mutants. In contrast, investigators did not
anticipate multidrug resistance because the co-
appearance of multiple mutations conferring
such phenotypes was considered to be beyond
the potential of a given bacterial population.
However, in 1956 in Japan, six years after
physicians began prescribing—and companies
began massively producing—the antibiotics
streptomycin (Sm), tetracycline (Tc), and chlor-
amphenicol (Cm), investigators were observing
isolates of Shigella dysenteriae resistant to up to
four antibiotics simultaneously (Tc, Cm, Sm,
Su).
This emergence of multiple resistant strains
could not be attributed to mutation alone. It was
soon established that bacteria were acquiring
genes that confer resistance, relying on that
means to escape antimicrobial activity rather
than on mutations arising in resident genes. At
that time, transposons and integrons began ap-
pearing among human gram-negative pathogens
as well as the commensal bacterial population,
playing a key role in disseminating resistance
genes while hitchhiking on conjugative plasmids
and interspecies transfer.
Nevertheless, integrons were not formally
identified as agents of resistance gene recruit-
ment until the late 1980s, following the obser-
vation that loci in transposons and resistance
(R)-plasmids expressing different antibiotic re-
sistance spectra share the same genetic back-
bone and differ only in the resistance genes that
they harbor. Regardless of this lag in FIGURE 1
awareness, however, those integrons
were certainly part of the first multi-
drug resistance outbreaks in the 1950s.
One proof is that Tn21, an integron-
containing transposon, was involved in
the resistance phenotype propagated by
plasmid NR1 (R100) in the very first
events in Japan.
In light of the six-year time scale for
the emergence of multidrug-resistant
Shigella strains in Japan in 1956, it was
hardly disputable that bacteria were al-
ready equipped with appropriate ge-
netic tools for meeting the challenge of
multidrug assaults. Indeed, molecular
studies during the last 15 years show
how powerful the bacterial integron re-
combination machinery is, despite its
relative functional simplicity.
Thus, all known integrons are com-
posed of three essential elements for
procuring exogenous genes: (i) a gene
coding for an integrase (intI), (ii) a pri-
mary recombination site (attI), and (iii)
a strong promoter (Pc). Integron inte-
grases recombine, in a recA-indepen-
dent manner, discrete units of circular-
ized DNA known as gene cassettes
downstream of the resident Pc pro- Integron-mediated gene capture and the model for cassette exchange. Outline of the
moter at the proximal attI site, permit- process by which circular antibiotic resistance gene cassettes (antR) are repeatedly
ting expression of their encoded pro- inserted at the specific attI site in a class1 integron downstream of the strong promoter
PC. intI1, integrase encoding gene; Int, integrase IntI1.
teins (Fig. 1). Even considering only
those that differ in nucleotide sequence
by more than 5%, more than 70 differ- from 57 bp to 141 bp in length, and their nucle-
ent resistance cassettes have been described, and otide sequence similarities are primarily re-
they confer resistance to all beta-lactams, all stricted to the inverse core-site and the core-site
aminoglycosides, chloramphenicol, tri- (Fig. 2).
methoprim, streptothricin, rifampin, erythro- Three classes of integrons (RI) involved in
mycin, and antiseptics of the quaternary ammo- multidrug resistance expression were defined on
nium compound family. the basis of homology of their integrase genes.
Moreover, all these integron-inserted cas- Each class appears capable of sharing and ac-
settes share several specific structural character- quiring the same gene cassettes. Several cassettes
istics. For instance, the boundaries of each inte- appear to belong in two different classes of
grated cassette are defined by two GTTRRRY integrons, and the class 1 integrase can recom-
(core-site) sequences in the same orientation, bine several structurally diverse attC sites. The
which are the targets of the recombination pro- IntI integrases belong to the catalytic family of
cess. The integrated cassettes also generally in- the tyrosine (Y) recombinases that are involved
clude a single gene and an imperfect inverted in the movement of numerous phages through
repeat located at the 3⬘ end of the gene called an site-specific recombination (such as the lambda
attC site (or “59-base element”), a diverse family phage integrase, ␭Int) or in fundamental cellular
of sequences that function as recognition sites processes such as chromosome dimer resolution
for the site-specific integrase. The attC sites vary in cell division (XerC/D). In spite of these rela-

Volume 70, Number 11, 2004 / ASM News Y 521

 FIGURE 2

Structural comparison of a “classical” multi-resistant integron and the V. cholerae N16961 super-integron. Top, schematic representation
of In40; the various resistance genes are associated with different attC sites (see text). Antibiotic resistance cassettes confer resistance
to the following compounds: aacA4, aminoglycosides; carb4, beta-lactams; cmlA2, chloramphenicol; qac, quarternary ammonium
compounds; oxa9, beta-lactams. The sul gene, which provides resistance to sulfonamides, is not a gene cassette. Bottom, the ORFs are
separated by highly homologous sequences, the VCRs.

tionships, the integron integrases have a unique
characteristic among the Y-recombinases, as they
recombine sequences that are only remotely re-
lated.
The RI platforms are defective for self-trans-
position, but this defect is often complemented
through association with IS, transposons,
and/or conjugative plasmids which can serve as
vehicles for the intra- and interspecies trans-
mission of genetic material. With this system,
bacteria are capable of stockpiling exogenous
genetic loci to establish an appreciable anti-
microbial armamentarium. Some bacteria habor
up to eight different resistance cassettes in a
single integron.
Another Type of Integron: the
Chromosomal Superintegrons
A pivotal question is, what are the origins of the
RI and their cassettes? The degree of homology
between the three RI integrases (45–58%) sug-
gests that their evolutionary divergence extends
longer than the 50 years since antibiotics came
into use. In the late 1990s, studies examining the
relationship between RI gene cassette arrays and
a cluster of repeated sequences identified in the
Vibrio cholerae genome, called VCRs (V. cholerae
repeats), led to discovery of another type of inte-
gron, a superintegron (SI) (Fig. 2). This distinct
type of integron is now known to be an integral
component of many g-proteobacterial genomes.
The integron discovered in chromosome 2 of
V. cholerae possesses a specific integrase, IntIA,
that is related to the RI integrases and also is
responsible for inserting coding sequences
(ORFs) into a unique chromosomal attI site.
However, this SI has two structural characteris-
tics that distinguish it from other RIs—namely,
the large number of cassettes that it gathers and
the high homology observed among the attC
sites of these cassettes, which in the case of V.
cholerae are known as VCRs.
These key features plus the sedentary nature
of the functional platform (IntIA ⫹ attI site)
define a superintegron. Such SI structures also
are found among the Vibrionaceae and their
close relatives, the xanthomonads, and a branch
of the pseudomonads. They share the same gen-
eral characteristics, such as their large number
of more than 20 cassettes and a high homology
between their endogenous cassette attC sites.
More importantly, they predate the antibiotic
era, and have been identified in isolates collected
from the 19th century.

522 Y ASM News / Volume 70, Number 11, 2004

 The SI carried in three Vibrio species whose
genomes are sequenced all show a large number
of cassettes, from 72 in V. parahaemolyticus to
more than 200 in V. vulnificus. In the case of the
V. cholerae strain El Tor N16961, the SI gathers
at least 216 mostly unidentified genes in an
array of 179 cassettes that starts from the
VchintIA gene and occupies about 3% of the
genome. The high level of identity shared by the
attC sites carried by the majority of these cas-
settes suggests that they are assembled in the
SI-carrying species through the physical associ-
ation of an attC site with an incoming DNA
fragment. However, the mechanics of this pro-
cess are not known.
Integrons that do not share all the character-
istics of the SI have been found in the genome of
two Shewanella species. These integrons gather
only a handful of cassettes with structurally
heterogeneous attC sites, a situation reminiscent
of the RI cassette arrays. Furthermore, integron
integrase-like genes have also been identified in
the genomes of other proteobacteria, including
Nitrosomonas europaea, Microbulbifer degrad-
ans, Geobacter sulfurreducens, and Treponema
denticola. Apart from the one carried in N.
europea, they have not been characterized.
Using PCR primers directed against con-
served regions of the integron-integrase genes
and attC sites, Harold Stokes of Macquarie Uni-
versity, Sydney, Australia, and Steven Schmidt
of the University of Colorado, Boulder, and
their respective collaborators amplified 19 new
integron integrases from markedly different en-
vironmental DNA samples, including from na-
tional parks or metal-contaminated soils. Un-
fortunately, in most cases their protocol does
not permit determining the source of these inte-
grons, be it the endogenous SI of a soil bacte-
rium or an integron located on a mobile struc-
ture. However, their findings support the
hypothesis that integrons are widespread among
bacterial populations either as components of
mobile DNA elements or the chromosome, and
that they are not confined to pathogenic or
multidrug-resistant bacteria.
Phylogeny Indicates Integrons Are
Ancient, Super Integrons Possibly
Ancestral
According to comparative analyses, the inte-
gron-integrases form a specific clade within the
Y-recombinase family. Moreover, the integron
platform appears to be ancient based on the
species-specific clustering of the respective SI
integrase genes in a pattern that mostly adheres
to the line of descent among the bacterial species
in which they are found. Thus, the establishment
of SIs likely predates speciation within respec-
tive genera, suggesting that integrons are ancient
structures that steered the evolution of bacterial
genomes for hundreds of millions of years.
However, it is possible that transfer of either a
part or all of a SI occurred long ago, perhaps
explaining some of the discrepancies observed
between the SI-integrase and 16S rRNA gene
trees, such as for V. fischeri.
Because the RIs and SIs have a common struc-
tural organization, the antiquity of SIs suggests
that they are the ancestors of RIs. Hence, we
proposed that RIs evolved from SIs through
entrapment of intI genes and their cognate attI
sites by mobile structures, such as easily and
randomly assembled compound transposons.
Furthermore, we noticed that 12 different resis-
tance cassettes carry an attC site almost identical
to those specifically found in and characteristic
of Xanthomonas and Vibrio SIs (Fig. 3). Once
mobile, one can imagine that the subsequent
harvesting of cassettes from various SI sources
led to contemporary RIs, including the great
diversity of attC sites associated with these gene
cassettes.
Two recent observations support this account
of how RIs could have arisen from SIs. The first
came from Henning Sørum of the Norwegian
School of Veterinary Science in Oslo and his
collaborators. While analyzing a plasmid asso-
ciated with trimethoprim resistance in V. salmo-
nicida, they found that resistance is due to a
dfrA1 cassette carried in an integron (Genbank
#AJ277063) that contains a “classical” dfrA1
cassette in second position and is followed by six
cassettes. This new integron has two features
that corroborate our model: (i) its integrase orig-
inated from a Vibrio species, or one from a
closely related genus, as attested to by its 74%
similarity to the V. cholerae and V. mimicus
IntIA; and (ii) the attC sites of three of the seven
cassettes are structurally homogeneous, a char-
acteristic until now only found in the SI cassette
arrays. Because this integron is surrounded by IS
sequences, we suspect that it corresponds to an
intermediate of what could possibly became a
true RI by harvesting other resistance cassettes,
Volume 70, Number 11, 2004 / ASM News Y 523
 FIGURE 3

A model for super-integrons as the source of the resistance gene cassettes of multi-resistant integrons. The pool of RI gene cassettes
is depicted as being derived from the reservoirs of SIs. The attC sites are represented as colored triangles. The structural relationships
between the attC sites of resistance cassettes found in RIs and those characteristic of identified, or unknown yet, SIs are indicated by
a larger triangle of the same color.

using the mobility of the assembled compound
transposon and multiple transfers.
The second observation from our group in-
volves SI cassettes that could be recruited by a
RI, and, moreover, confer a resistance pheno-
type. The majority of the examined SI cassettes
appear unique to the host species, and a major-
ity of their encoded genes have no counterparts
in the database or are the sole homologs of
unassigned ORFs. Nevertheless, the reservoir of
adaptive functions residing within SIs includes
cassettes with significant homology to known
antibiotic resistance genes. We have demon-
strated that class 1 integrons can randomly re-
cruit any gene cassettes harbored within a SI.
After we applied selective pressure for antibiotic
resistance, we discovered a chloramphenicol
acetyltransferase gene cassette, catB9, in the V.
cholerae SI.
Meanwhile, another resistance cassette, cod-
ing for a novel carbenicillinase, has been identi-
fied in the SI of another V. cholerae isolate. This
cassette, like the catB9 cassette, is structurally
identical to the other V. cholerae SI cassettes, as
its attC site is a canonical VCR. Along with
results demonstrating that SI cassettes are sub-
strates for the integrase of class 1 integrons
when present on a high-copy-number plasmid,
these results suggest that environmental condi-
tions, such as the presence of antibiotics, dictate
which of the randomly recruited cassettes are
retained within the RIs of clinical isolates.
It is likely that the assembly of complex RIs
having more than two resistance cassettes oc-
curred through recombination between differ-
ent RI, rather than through successive direct
recruitment from the SI cassette arrays of envi-
ronmental species. Niches exist which could fa-

524 Y ASM News / Volume 70, Number 11, 2004

vor the exchange of cassettes between RI on a
high scale. The recent characterization of a re-
markable collection of cassettes, integrons, and
plasmids circulating in a single wastewater
treatment plant, as well as the unpredicted dem-
onstration that aerobic, gram-positive Coryne-
bacteria can be a reservoir of class 1 RI in
poultry litter, indicates that environments other
than obvious clinical settings exist in which cas-
sette exchanges among RI may be common.
Trying To Trace Sources of Resistance
Genes in Integron Cassettes
As for antibiotic resistance genes that are not
carried in cassettes, key questions about their
origins and how they disseminate remain unan-
swered. Nonetheless, evidence suggests that
horizontal gene transfer among prokaryotes is a
perpetual activity. So ancient is this process that
determining a definitive source for any particu-
lar resistance marker will surely prove diffi-
cult—unless a recent event can be identified.
In 1973, Raoul Benveniste and Julian Davies
proposed that intrinsically resistant or antibiot-
ic-producing organisms routinely supply resis-
tance genes to clinical isolates, and some recent
findings substantiate their hypothesis. Their
proposal was based in part on the structural
relationship between the aminoglycoside phos-
photransferase (APH), which protects the acti-
nomycetes that produce aminoglycosides, and
the APH encoded in transposons such as Tn5.
The best example is certainly the strong similar-
ity in sequence and genetic organization ob-
served by Gerard Wright of McMaster Univer-
sity, Hamilton, Ontario, Canada, between the
vancomycin resistance gene cluster found in en-
terococci and those of the glycopeptide-produc-
ing actinomycetes. While no other direct evi-
dence is presently available, many antibiotic-
producing strains other than those used
industrially exist in nature and could be a source
of resistance genes.
The recruitment of general housekeeping
genes provides an alternate route by which re-
sistance determinants might evolve. For in-
stance, several examples of chromosomally en-
coded homologs whose functions are not related
to antibiotic resistance have been identified in
the genomes of Serratia marcescens, Providen-
cia stuartii, Streptomyces spp., and Mycobacte-
rium spp.
The marked differences in codon usage
among cassettes within the same integrons indi-
cate that these several antibiotic resistance de-
terminants are of diverse origins. However, evo-
lutionary rates and the passing of genes through
intermediates undoubtedly created descendants
that are appreciably divergent from the original
loci, meaning the original source of a given
resistance gene will remain elusive.

ACKNOWLEDGMENTS
I thank all the members, past and present, of my group. I especially thank D. Rowe-Magnus and M.-C. Ploy for their critical
reading of the manuscript. This work was supported by the Pasteur Institute and the CNRS.

SUGGESTED READING
Collis, C. M., G. D. Recchia, M. J. Kim, H. W. Stokes, and R. M. Hall. 2001. Efficiency of recombination reactions catalyzed
by class 1 integron integrase IntI1. J. Bacteriol. 183:2535–2542.
Hall, R. M., and C. M. Collis. 1998. Antibiotic resistance in gram-negative bacteria: the role of gene cassettes and integrons.
Drug Resist. Updates 1:109 –119.
Nandi, S., J. J. Maurer, C. Hofacre, and A. O. Summers. 2004. Gram-positive bacteria are a major reservoir of Class 1
antibiotic resistance integrons in poultry litter. Proc. Natl. Acad. Sci. USA 101:7118 –7122.
Nemergut, D. R., A. P. Martin, and S. K. Schmidt. 2004. Integron diversity in heavy-metal-contaminated mine tailings and
inferences about integron evolution. Appl. Environ. Microbiol. 70:1160 –1168.
Rowe-Magnus, D. A., A.-M. Guerout, P. Ploncard, B. Dychinco, J. Davies, and D. Mazel. 2001. The evolutionary history of
chromosomal super-integrons provides an ancestry for multi-resistant integrons. Proc. Natl. Acad. Sci. USA 98:652– 657.
Rowe-Magnus, D. A., A. M. Guerout, L. Biskri, P. Bouige, and D. Mazel. 2003. Comparative analysis of superintegrons:
engineering extensive genetic diversity in the Vibrionaceae. Genome Res 13:428 – 442.
Rowe-Magnus, D. A., A. M. Guerout, and D. Mazel. 2002. Bacterial resistance evolution by recruitment of super-integron
gene cassettes. Mol. Microbiol. 43:1657–1669.
Stokes, H. W., A. J. Holmes, B. S. Nield, M. P. Holley, K. M. Nevalainen, B. C. Mabbutt, and M. R. Gillings. 2001. Gene cassette
PCR: sequence-independent recovery of entire genes from environmental DNA. Appl. Environ. Microbiol. 67:5240 –5246.
Tennstedt, T., R. Szczepanowski, S. Braun, A. Puhler, and A. Schluter. 2003. Occurrence of integron-associated resistance
gene cassettes located on antibiotic resistance plasmids isolated from a wastewater treatment plant. FEMS Microbiol. Ecol.
45:239 –252.

Volume 70, Number 11, 2004 / ASM News Y 525

Molecular Microbiology (2002) 43(6), 1657–1669
Bacterial resistance evolution by recruitment of
super-integron gene cassettes
Dean A. Rowe-Magnus,† Anne-Marie Guerout and
Didier Mazel*
Unité de Programmation Moléculaire et Toxicologie
Génétique, CNRS URA 1444, Département des
Biotechnologies, Institut Pasteur, 25 rue du Dr Roux,
75724 Paris, France.
Summary
The capture and spread of antibiotic resistance deter-
minants by integrons underlies the rapid evolution of
multiple antibiotic resistance among diverse Gram-
negative clinical isolates. The association of multiple
resistance integrons (MRIs) with mobile DNA ele-
ments facilitates their transit across phylogenetic
boundaries and augments the potential impact of
integrons on bacterial evolution. Recently, ancestral
chromosomal versions, the super-integrons (SIs),
were found to be genuine components of the
genomes of diverse bacterial species. SIs possess
evolutionary characteristics and stockpiles of
adaptive functions, including cassettes related to
antibiotic resistance determinants previously charac-
terized in clinical isolates, which suggest that MRIs
and their resistance genes were originally recruited
from SIs and their pool of amassed genes. However,
the recombination activity of integrons has never
been demonstrated in a bacterium other than
Escherichia coli. We introduced a naturally occurring
MRI (TpR, SulR) on a conjugative plasmid into Vibrio
cholerae, a species known to harbour a SI. We show
that MRIs can randomly recruit genes directly from
the cache of SI cassettes. By applying a selective
constraint for the development of antibiotic resis-
tance, we demonstrate bacterial resistance evolution
through the recruitment a novel, but phenotypically
silent, chloramphenicol acetyltransferase gene from
the V. cholerae SI and its precise insertion into the
MRI. The resulting resistance profile (CmR, TpR, SulR)
could then be disseminated by conjugation to other
Accepted 24 December, 2001. *For correspondence. E-mail
mazel@pasteur.fr; Tel. (+33) 1 40 61 32 84; Fax (+33) 1 45 68 88 34.
†
Present address: Department of Microbiology and Division of Clini-
cal Integrative Biology, Sunnybrook and Women’s College Hospital
Health Sciences Centre Research Institute, 2075 Bayview Avenue,
Toronto, Canada M4N 3N5, and The Department of Laboratory
Medicine and Pathobiology, University of Toronto, Toronto, Canada.
© 2002 Blackwell Science Ltd
clinically relevant pathogens at high frequency. These
results demonstrate that otherwise phenotypically
sensitive strains may still be a genetic source for
the evolution of resistance to clinically relevant
antibiotics through integron-mediated recombination
events.
Introduction
The horizontal transfer of genetic material within micro-
bial communities has been instrumental in the emergence
of novel functions and species (de la Cruz and Davies,
2000; Ochman et al., 2000). The ‘antibiotic resistance
phenomenon’ is perhaps the most striking recent example
of the impact of horizontal transfer on microbial adapta-
tion. This phenomenon refers to the rapid and widespread
emergence of similar antibiotic resistance profiles among
phylogenetically diverse Gram-negative clinical isolates
over the last half-century (Davies, 1994). The source of
these antibiotic resistance determinants is often unknown,
but many are thought to have evolved or, more signifi-
cantly, been acquired in response to aeons of natural
bacterial competition (Davies, 1994; 1997). At the root of
antibiotic resistance gene acquisition is the unique
activity of integrons. Integrons are natural genetic engi-
neering systems that incorporate circularized open
reading frames, called gene cassettes, and convert them
to functional genes (Martinez and de la Cruz, 1990; Collis
and Hall, 1992; Collis et al., 1993; Hall and Collis, 1995).
The integron platform consists of an integrase gene of the
tyrosine recombinase family (Nunes-Duby et al., 1998)
and a primary recombination site, called the attI site,
located proximal to the integrase gene. The integrase
catalyses recombination between its cognate attI site
and a secondary target called an attC site (or 59 base
element). The attC site is normally found associated
with a single open reading frame (ORF) that encodes an
adaptive function, and the ORF–attC structure is termed
a gene cassette (Hall and Stokes, 1993; Recchia and
Hall, 1995; Stokes et al., 1997; Sundstrom, 1998).
Recombination between the attI and attC sites leads to
insertion of the gene cassette downstream of a resident
promoter, Pc, within the integron (Levesque et al., 1994;
Collis and Hall, 1995) that drives expression of the
encoded product. The strongest promoter variant of Pc
(TTGACAN17TAAACT) is six times more efficient than
the derepressed PTac promoter (Levesque et al., 1994).
1658 D. A. Rowe-Magnus et al.
The most notable gene cassettes identified within
integrons are those conferring resistance to antibiotics.
More than 70 different antibiotic resistance genes, cover-
ing most classes of antimicrobials presently in use, are
structured as gene cassettes (Hall and Collis, 1998;
Mazel and Davies, 1999). The stockpiling of these cas-
settes in integrons, to create multiresistance integrons
(MRIs), has contributed substantially to the current
dilemma in the treatment of infectious disease, as inte-
grons containing up to eight resistance cassettes have
been found in multiple-resistant clinical isolates (Naas
et al., 2001). Furthermore, integron platforms are often
found embedded within mobile DNA elements such as
transposons and/or conjugative plasmids that can serve
as vehicles for the intra- and interspecies transmission of
the resistance genes that have been amassed by inte-
grons. The proficiency of this partnership is confirmed by
the marked differences in codon usage among cassettes
within the same MRI, indicating that the antibiotic resis-
tance determinants are of diverse origin.
Five classes of MRIs have been defined based on the
homology of the integrase genes (Martinez and de la Cruz,
1990; Sundstrom et al., 1991; Arakawa et al., 1995;
Hochhut et al., 2001; Sorum et al., GenBank accession
no. AJ277063). They share between 39% and 58% amino
acid identity, which suggests that their evolutionary
divergence has extended over a much longer period of
time than the 50 years of the antibiotic era. The likely
ancestors of MRIs are the recently discovered chromo-
somal super-integrons (SIs) that have been identified in the
genomes of diverse Gram-negative proteobacteria (Mazel
et al., 1998; Clark et al., 2000; Rowe-Magnus et al., 2001;
Vaisvila et al., 2001). The first SI was discovered in the
V. cholerae genome (Mazel et al., 1998). Clustered in one
region of chromosome II, the V. cholerae SI spans
126 kb and harbours 214 open reading frames of mainly
unassigned function (Rowe-Magnus et al., 1999;
Heidelberg et al., 2000) in 179 cassettes. Its cassettes vary
extensively in base composition and codon usage, and
appear to be of bacterial, viral and eukaryotic origin. MRIs
and SIs share an identical structural organization, with an
intI gene divergently transcribed from an upstream cluster
of gene cassettes (Rowe-Magnus and Mazel, 1999).
Furthermore, the integrases of both MRIs and SIs are
functional for cassette recombination (Rowe-Magnus
et al., 2001), and their integrases form a group related to
but distinct from other members of the l family of site-
specific tyrosine recombinases (Nunes-Duby et al., 1998).
Despite these commonalities, four major differences
exist between MRIs and SIs. First, MRIs typically contain
fewer than six cassettes, with the largest described to
date containing eight cassettes. As is the case in V.
cholerae, SIs may contain hundreds of cassettes.
Second, the cassettes found within MRIs typically code
© 2002 Blackwell Science Ltd, Molecular Microbiology, 43, 1657–1669
for antibiotic resistance genes whereas those of SIs are
of mainly unknown function. Third, MRIs are commonly
associated with mobile DNA elements. Conversely, the
coevolution of SI integrases with their host genome
(Rowe-Magnus et al., 2001) strongly suggests that SIs
are sedentary. Finally, whereas the attC sites of the gene
cassettes of MRIs are highly variable in length and
sequence, the attC sites of SI gene cassettes are closely
related and species specific (Rowe-Magnus et al., 2001).
These relationships led to the proposal that MRIs evolved
from SIs through the entrapment of intI genes and their
cognate attI sites by mobile DNA elements (Rowe-
Magnus et al., 2001). The subsequent harvesting of
cassettes from various SI sources then led to the estab-
lishment of contemporary MRIs. However, the ability of
integrons to recruit gene cassettes has never been
demonstrated to occur in a bacterium other than E. coli.
Furthermore, the direct recruitment of cassettes from a SI
by a MRI has never been observed.
Functional analysis of the SI cassettes of V. cholerae
strain N16961 unveiled a variety of adaptive functions,
including metabolic activities, virulence traits and poten-
tial antibiotic resistance determinants (Rowe-Magnus
et al., 1999; Heidelberg et al., 2000). We sought to deter-
mine if any of these ORFs could confer resistance to
current antibiotics and become a source of the resistance
genes found in MRIs. Here, we show that the gene cas-
sette reservoirs of SIs are a source of the gene cassettes
identified within MRIs. We demonstrate that MRIs ran-
domly recruit gene cassettes directly from SIs. By then
applying a selection for the development of antibiotic
resistance, we followed the real-time evolution of bac-
terial resistance through the recruitment by a MRI of a
novel, but phenotypically silent, chloramphenicol acetyl-
transferase gene from the V. cholerae SI. The resulting
resistance profile (CmR, TpR, SulR) could then be dis-
seminated by conjugation to other clinically relevant
pathogens at high frequency.
Results
The evolution of a MRI through the recruitment of a
novel chloramphenicol acetyltransferase gene cassette
from the V. cholerae SI by a class 1 integron
The majority of integron gene cassettes are known to
be promoterless (Hall, 1997) and are probably poorly
expressed as a result. Using plasmids containing cloned
class 1 integron fragments that differed only in the order of
their inserted resistance gene cassettes, Collis and Hall
(1995) have shown that cassettes closest to the promoter,
PC, are more highly expressed than distal cassettes. Thus,
depending on the genetic context in which they are found,
the potential antimicrobial activity of silent determinants
encoded in SI gene cassettes may be overlooked. We
 Bacterial trafficking of integron gene cassettes 1659
Table 1. Potential antibiotic resistance cas-
ORF number on chromosome II b Proposed functionc E-valuec settes in the V. cholerae N16961 SIa.

VCA0300 (catB9) Chloramphenicol acetyltransferase 5 e–71

VCA0328, VCA0341, VCA0463d Fosfomycin glutathione-transferase 6 e–4
VCA0387 Phosphinothricin acetyltransferase 6 e–31
VCA0473 Streptogramin acetyltransferase 2 e–31

a. These four cassettes are also present in the 63.32 SI (see Experimental procedures); 10
other genes found in the N16961 SI cassettes (VCA0316, VCA0400, VCA0402, VCA0417,
VCA0436, VCA0442, VCA0474, VCA0479, VCA0496, VCA0505) code for potential acetyl-
transferases and a glutathione transferase.
b. The 214 ORFs in the 179 cassettes in the SI of V. cholerae strain N16961 are numbered
from VCA0292 to VCA0506 (Heidelberg et al., 2000).
c. According to BLAST analysis (http://www.ncbi.nlm.nih.gov/BLAST/).
d. Cassettes are identical in sequence.

determined the sensitivity profiles of V. cholerae isolates
obtained since 1935. Our profiles indicated that early V.
cholerae isolates, including strain N16961, were sensitive
to the classes of antibiotics active against Gram-negative
bacteria. To determine if any of the resistance gene
cassette homologues that we observed in the SI of V.
cholerae N16961 could confer resistance to their sug-
gested corresponding antibiotic (Table 1), we sought to
relocate SI gene cassettes to the insertion site of a class
1 integron (Fig. 1). This would place the SI cassettes

Fig. 1. The V. cholerae conjugative assay system. For clarity, only a portion of the V. cholerae SI on chromosome II is shown. The 214 ORFs
of the 179 SI gene cassettes are numbered from VCA0292 to VCA0506. Any of these (represented as VCA0xxx) may be recruited by In3.
The V. cholerae signature attC sites, the VCRs, are shown as black triangles; the unrelated attC sites typical of MRI cassettes are shown as
white and stripped triangles in In3. The dark wavy lines and black boxes represent chromosomal DNA and ORFs that are not part of the SI
structure. The class 1 integron–integrase protein, IntI1, is expressed from p112 (large white arrow) under the control of the IPTG-inducible trc
promoter. IntI1 catalyses site-specific recombination between the attC sites of gene cassettes, leading to random excision of the SI gene
cassettes (white and stripped boxes) in circular form. The subsequent integration of the free cassettes at the attI1 site (small white circle)
of In3, a naturally occurring class 1 integron on the conjugative plasmid R388, places the cassette downstream of an active promoter, Pc (bent
arrow), that drives expression of the cassette-encoded protein. R388 and its derivatives are then conjugated (thick black arrow) to a NalR
recipient and transconjugants are selected on the appropriate sets of antibiotics. VchintIA, the V. cholerae SI integrase gene; stripped box, a
potential antibiotic resistance gene cassette. The dfrB2 gene cassette in the In3 codes for trimethoprim resistance. The relative positions of
the attI1.1 and dfrfus primers are shown (small arrows).

© 2002 Blackwell Science Ltd, Molecular Microbiology, 43, 1657–1669

1660 D. A. Rowe-Magnus et al.

downstream of a fully active resident promoter to ensure

their expression and allow us to ascertain if resistance
could be attained by high-level expression of these antibi-
otic resistance gene cassette homologues.
In order to detect the trafficking of cassettes between
SIs and MRIs, we utilized a conjugation system involving
two plasmids. The first, p112, contains the intI1 gene,
which codes for the integrase of class 1 integrons under
the control of the inducible trc promoter. The second har-
bours a naturally occurring class 1 target integron, In3, on
the conjugative plasmid R388 (Avila and de la Cruz, 1988;
Martinez and de la Cruz, 1990). In3 contains the strong
promoter variant of PC, the dfrB2 gene cassette, which
confers resistance to trimethoprim, and an open reading
frame of unknown function, ORFA. We verified by poly-
merase chain reaction (PCR) that the resistance gene
cassette homologues shown in Table 1 were also present
in the SI of V. cholerae CIP 63.32 (see Experimental Pro-
cedures). Both plasmids were introduced into V. cholerae
strain CIP 63.32 to create strain 1275. As the complete
sequence of the SI of V. cholerae strain CIP 63.32 was
not known, we extended our selection to include not only Fig. 2. Thin-layer chromatography of the [14C]-Cm CAT assay
products. The CAT assay was performed as described in the
the antibiotics listed in Table 1 but also those for which
Experimental procedures. Lanes: 1, control strain UB5201 (R388);
resistance determinants structured as integron gene cas- 2, UB5201a (R388::catB9).
settes had already been described (Mazel and Davies,
1999). Bacterial resistance evolution was then assayed
by monitoring integrase-mediated integration of SI cas- of an arabinose-inducible promoter. Transformants were
settes into In3 (Fig. 1). found to be resistant to Cm in an induction-dependent
Following induction of intI1 gene expression, R388 was manner, whereas control strains containing vector without
transferred to E. coli and transconjugants were plated on insert remained susceptible regardless of the induction
a number of different antibiotics. Of the antibiotics tested, state (data not shown).
bacteria that were resistant to chloramphenicol (Cm)
were recovered (MIC of 128 mg ml–1). Sequence analysis
An integron-mediated switch of V. cholerae from a
of several transconjugants revealed that a SI cassette,
Cm-sensitive to a Cm-resistant state
corresponding to VCA0300 of V. cholerae strain N16961
(Table 1), had been precisely inserted into In3. The According to our antibiograms, the parent strain V.
VCA0300 cassette contained a single ORF of 630 cholerae 63.32 was sensitive to Cm (minimum inhibitory
nucleotides with significant homology to chloramphenicol concentration <1 mg ml–1). The fact that the catB9 gene
acetyltransferases (CATs) (64.9%, 65.1%% and 65.9% cassette did not contain a promoter potentially explained
amino acid identity to CATB6, CAT X82455 and CATB8 this sensitivity. In the CmR E. coli transconjugants, catB9
respectively; accession numbers AJ223604, X82455 and was precisely inserted into the first position of In3, which
AF227506; Fig. 2). Typical of integron gene cassettes, this placed the cassette downstream of the resident PC pro-
cassette did not contain a promoter element. Several moter. This suggested that in its native SI position the
E. coli transconjugants were selected and tested for the catB9 gene might not be sufficiently expressed to confer
presence of CAT activity. A standard CAT assay (Gorman resistance. To investigate this possibility, a 3.8 kb region
et al., 1982) verified 1-acetoxy- and 3-acetoxy-CAT activ- containing the attI4 site and the first seven cassettes up
ities in the transconjugants, whereas E. coli harbouring to the catB9 gene of the V. cholerae strain 63.32 SI was
native R388 showed no CAT activity (Fig. 3). This novel cloned downstream of the PLac promoter in the plasmid
CAT gene cassette was designated catB9. pSU38 (Fig. 4). In addition, sequential deletions of the cas-
To verify that the catB9 gene cassette alone was settes within the fragment were constructed in the same
responsible for the observed CmR phenotype and that no vector. Each construct was then tested for its ability to
other element carried by R388 participated in the CmR permit growth of E. coli transformants on media containing
phenotype, we amplified catB9 by PCR from V. cholerae 25 mg ml–1 Cm. Transformants with the catB9 gene in its
strain 63.32 and expressed it in E. coli under the control native seventh position were sensitive to Cm. Deletion
© 2002 Blackwell Science Ltd, Molecular Microbiology, 43, 1657–1669
 Bacterial trafficking of integron gene cassettes 1661

Fig. 3. Alignment (A) and phylogenetic relationships (B) of the known chloramphenicol (CAT), streptogramin (SatA) and virginiamycin
acetyltransferase gene (VatB). Accession numbers follow the corresponding designations. Resistance determinants that are structured as gene
cassettes within integrons are boxed. The novel CATB9 enzyme is marked with a grey box.

© 2002 Blackwell Science Ltd, Molecular Microbiology, 43, 1657–1669

1662 D. A. Rowe-Magnus et al.

Fig. 4. Cloning of the sequential deletions of the cassettes upstream of the catB9 gene and the corresponding resistance phenotype
conferred by the various constructs. The integrase gene of the V. cholerae SI, VchintIA (thick grey arrow), and the 3.8 kb fragment spanning
the attI4 site and the first seven cassettes, including the catB9 (stripped box) gene, is shown. Sequential cassette deletions were generated
by PCR and all products were cloned into the multiple cloning site (MCS) of the plasmid pSU38. E. coli transformants were tested for their
ability to grow on media containg 25 mg ml–1 Cm. The relative hybridization positions of the primers used to create the various constructs (I4,
I4Cm1 to I4Cm6 and cat2) are marked as arrows. Symbols: Top: horizontal lines represent the PCR products obtained with the corresponding
primer pairs; –, no growth; +, growth. Bottom: black triangles, VCRs; open boxes, ORFs; the PLac promoter of pSU38 is indicated with a bent
arrow.

of the first two cassettes to create transformants that
harboured the catB9 gene in the sixth and fifth positions
relative to the PLac promoter remained sensitive to the
antibiotic. Deletion of the third, fourth, fifth and sixth cas-
settes, which placed the catB9 gene in the fourth, third,
second and first positions relative to the PLac promoter,
gave rise to CmR transformants. Furthermore, the growth
rate of the CmR transformants in the presence of Cm
conformed with the position of the catB9 cassette relative
to the PLac promoter, i.e. the closer the catB9 cassette
was placed to the PLac promoter, the faster the growth rate
of the transformants in the presence of Cm (data not
shown). These results, along with those of the E. coli
transconjugants containing the catB9 cassette in the first
position of In3, suggested that the parent V. cholerae
strain 63.32 was sensitive to Cm because the catB9
gene was poorly expressed in its native seventh position
of the SI.
To determine if V. cholerae strain 63.32 could make a
direct switch from a Cm-sensitive to a Cm-resistant state
through integrase-mediated relocation of its SI gene cas-
settes to a class 1 integron, V. cholerae strain 1275 was
submitted to intI1 induction and then plated directly on
media containing 25 mg ml–1 Cm. V. cholerae isolates that
were resistant to Cm were obtained. PCR and sequenc-
ing verified that the catB9 gene had been relocated to the

© 2002 Blackwell Science Ltd, Molecular Microbiology, 43, 1657–1669


attI1 site of In3 in the resistant clones. These results
confirmed that the catB9 gene cassette was phenotypi-
cally silent in its native position in the SI of V. cholerae
strain CIP 63.32. The resulting MRI (CmR, TpR, SulR)
could then be readily conjugated, at high frequency
(>10–2), among clinically relevant bacterial species
such as Salmonella typhimurium, Shigella sonnei,
Pseudomonas aeruginosa and V. cholerae strain 569B,
which does not harbour the VCA0300 gene cassette
within its SI. The evolution of antibiotic resistance due to
the positional activation of a single traceable marker
allowed us to determine a frequency of recombination and
conjugation (RCF) of 7 ¥ 10–7 for a single-copy SI cas-
sette by a MRI. These results clearly demonstrated that
cassette-encoded phenotypes that might be quiescent
in certain genetic contexts can be presented through
integrase-mediated relocation events.
Integrons randomly recruit SI cassettes
To detect any V. cholerae SI cassette relocation event
to In3, we conducted allele-specific fusion primer PCR
directly on V. cholerae strain 1275 without further selec-
tion. In this procedure, a unique fusion primer was
designed to hybridize specifically to the VCR–dfrB2 cas-
sette junction that would arise from an IntI1-mediated
recombination event of any V. cholerae SI cassette at
the attI1 site of plasmid R388. After cloning of the bulk
PCR product, we selected and sequenced the inserts of
10 random clones. We identified 10 different cassettes
(Table 2), nine of which had homologues in the SI of V.
cholerae N16961 (Heidelberg et al., 2000). In addition, the
bulk PCR product was subjected to nested PCR using
primers specific for the gene cassettes listed in Table 1.
Sequencing of the amplified fragments confirmed the inte-
gration of the catB9, VCA0328, VCA0387 and VCA0473
gene cassettes at the attI1 site of In3. Thus, our detec-
tion of 14 independent recruitment events of SI gene
Table 2. Cassettes recruited from the SI of V. cholerae strain 63.32 by the class I integron, In3, of R388.
V. cholerae CIP 63.32 cassette clone Accession no.
1330 D82477
1335 F82451
1336 H82462
1337 H82465
1138 E82472
1339 G82462
1340 D82471
1341 AF462020
1342 D82474
1343 E82471
1344 H82471
© 2002 Blackwell Science Ltd, Molecular Microbiology, 43, 1657–1669
Bacterial trafficking of integron gene cassettes 1663
cassettes by a MRI, including the catB9 gene cassette
and three other potential antibiotic resistance determi-
nants, demonstrates that integrons randomly recruit cas-
sette-encoded functions that are later selected according
to environmental conditions, as demonstrated by our dis-
covery of catB9 by selective constraint.
Clonal expansion of V. cholerae isolates harbouring the
catB9 gene cassette
Thirty V. cholerae isolates, representing O:1, O:139 and-
non-O:1/non- O:139 serovars collected between 1937
and 1993, were tested for the presence of the catB9 gene
cassette by PCR (Table 3). In our sample pool, 22
isolates were negative for the catB9 cassette, while eight
isolates tested positive for the catB9 cassette. Sixteen of
the 30 isolates examined were from Asia and the Indian
subcontinent, and subsequent analysis focused on this
subset as it was the most statistically complete. Interest-
ingly, all the Asian isolates that tested negative for the
catB9 gene were collected before 1961 and all of the pos-
itive isolates were collected after 1961. Furthermore, the
data indicated that theses isolates could be divided into
clonal groups. For example, the O:1 isolates collected
from this region before 1961 (569B, 62.14, 62.15, 62.16)
were of the same clonal origin (A.-M. Guerout, D. A.
Rowe-Magnus and D. Mazel, manuscript in preparation)
and lacked the catB9 cassette. However, the catB9-
positive O:1 isolates collected from this region after 1961
(N16961, 63.32, 68.10, 254) represented the expansion
of a different O:1 clone that harboured the catB9 cassette.
Other differences in SI order and content were consistent
with this observation. Two late O:139 isolates (104151,
104152) of clonal origin also harboured the catB9 cas-
sette. There was no apparent correlation between serovar
type and the presence of the catB9 cassette, as it
was found among O:1, O:139 and-non-O:1/non- O:139
serovar types.
V. cholerae N16961 homologue
E value Function
–77
3e Chloramphenicol acetyltransferase
7e–86 Probable acetyltrafsferase
5e–33 Hypothetical protein
5e–66 Hypothetical protein
2e–89 Hypothetical protein
2e–41 Hypothetical protein
3e–67 Hypothetical protein
– Hypothetical protein
6e–35 Hypothetical protein
4e–18 Hypothetical protein
4e–42 Hypothetical protein
1664 D. A. Rowe-Magnus et al.
Table 3. V. cholerae strains examined for the catB9 gene cassette.

Strain name or CIP designation Date of isolation Place of isolation Serovar catB9 Reference

N16961 1975 Bangladesh O:1 + ATCC 39315

63.32 1961 Hong-Kong O:1 + CIP
104151 1992 Madras O:139 + CIP
104152 1993 Bangladesh O:139 + CIP
68.10 1968 Vietnam O:1 + CIP
1024 1961 Philippines ND + 930070
J. M. Fournier, IP
62.16 1959 Thailand O:1 – CIP
569B 1940 Indian O:1 – ATCC 25870
62.15 1959 Thailand O:1 – CIP
62.14 1959 Thailand O:1 – CIP
1144 1947 Egypt ND – 930051
J. M. Fournier, IP
1145 1953 India ND – 960640
J. M. Fournier, IP
1143 1959 India ND – 960652
J. M. Fournier, IP
1023 1937 Indonesia ND – 920189
J. M. Fournier, IP
723 1954 Thailand Non-O:1/O:139 – Vicente et al. (1997)
104153 1989 Bangladesh Non-O:1/O:139 – CIP
104154 1987 Bangladesh Non-O:1/O:139 + CIP

ND, not determined.

Discussion novel phenotypically silent cat gene cassette, catB9, in a

Bacteria have a remarkable genetic flexibility that almost
immediately permits the adaptive execution of events
resulting from alterations in their genome order and
content. In this regard, the directed gene capture activity
of integrons has played a pivotal role in the evolution
of antibiotic resistance among Gram-negative bacteria
(Martin et al., 1990; Jones et al., 1997; Hall and Collis,
1998; Mazel and Davies, 1999; Rowe-Magnus and Mazel,
1999). However, recent findings significantly expand the
impact of integrons on bacterial genome engineering.
The paradigm for integron-mediated gene acquisition has
been the class 1 integron. The common association of
class 1 integrons with mobile DNA elements such as the
Tn21 transposon family and conjugative plasmids has led
to the widespread occurrence of integrons among bacte-
rial populations (Martinez and de la Cruz, 1988; Nesvera
et al., 1998; Ploy et al., 1998; Liebert et al., 1999; Naas
et al., 1999; Poirel et al., 1999a,b; 2000; Dalsgaard et al.,
2000a). The perpetual phenomenon of lateral gene trans-
fer (Jain et al., 1999) will inevitably lead to an encounter
between a class 1 integron and a bacterial species har-
bouring a SI. This is supported by the recent finding of
class 1 integrons in many V. cholerae isolates (Dalsgaard
et al., 1999; 2000a,b; Falbo et al., 1999) and a new fourth
class of resistance integron in the SXTET constin of V.
cholerae (Hochhut et al., 2001). We have shown that
class 1 integrons can randomly recruit the gene cassettes
harboured within SIs. By applying a selective pressure for
the development of antibiotic resistance, we discovered a
© 2002 Blackwell Science Ltd, Molecular Microbiology, 43, 1657–1669
V. cholerae SI, clearly demonstrating that environmental
conditions dictate which of the randomly recruited
cassettes are retained and observed within the MRIs of
clinical isolates.
Each integron gene cassette is an individual mobile
genetic element, and the discovery of the chromosomal
SIs shifts their present numbers from the tens to the
thousands. To date, more than 20 SIs have been
discovered among diverse proteobacterial genera, and
just five Vibrio SIs contain a transferable genetic reservoir
that is equivalent in size to a small genome (Rowe-
Magnus et al., 1999; 2001; Vaisvila et al., 2001). Further-
more, several novel integron-integrases and hundreds
of unique gene cassettes have been readily amplified
from a variety of environmental samples (Nield et al.,
2001; Stokes et al., 2001). By tracking the movement of
a single traceable marker, we were able to determine
a RCF of 7 ¥ 10–7 for a V. cholerae SI cassette. Although
this rate may seem low, the potential number of inte-
gron gene cassettes and the widespread presence of
class 1 integrons in natural bacterial populations creates
a scenario in which bacterial chromosomal super-
integrons represent caches of dispersable, unpredicted
adaptive functions. The existence of other mobile DNA
elements such as transposons, conjugative plasmids,
phages and entire chromosomes, in Hfr strains for
example, along with the vast reservoirs of integron gene
cassettes supports the concept of a single massive
genetic pool that is available to and shared among
bacteria.

Although the class 1 integron is the most ubiquitous
among multidrug-resistant bacterial populations, its activ-
ity has never been demonstrated in a bacterium other
than E. coli and the establishment of a functional in vitro
integron system has proven difficult. This is the first
demonstration that the class 1 integron is fully functional
in bacteria other than E. coli. Furthermore, the fact that
the class 1 integron is active in V. cholerae indicates that
integrase function is either not dependent on a co-factor
or, if a co-factor is required for integrase function, it is not
specific to E. coli. Thus, MRIs have likely contributed sig-
nificantly to the capture and dissemination of V. cholerae
SI gene cassettes, as well as those of other bacterial SIs,
among bacterial populations.
Although most of their encoded genes have no homo-
logues in the database, at least some of the SI cassettes
encode adaptive functions including pathogenicity and
antibiotic resistance determinants. In V. cholerae, two
pathogencity genes, the heat-stable toxin gene, sto
(Ogawa and Takeda, 1993), and the mannose–fucose-
resistant haemagglutinin gene, mrhA (van Dongen et al.,
1987), have been found to be cassette encoded. In this
study we have identified the first functional antibiotic re-
sistance gene cassette, catB9, to be found in a SI. In addi-
tion, several potential progenitor cassettes with significant
homology to aminoglycoside, phosphinotricin, fosfomycin
and streptothricin resistance genes are present as well.
A variety of metabolic functions whose activities are not
related to antibiotic resistance or virulence have also been
determined for the gene cassettes of SIs in a number of
bacterial species (Barker and Manning, 1997; Rowe-
Magnus et al., 2001). As the majority of the cassettes
examined thus far appear to be unique to the host species
(Rowe-Magnus et al., 1999; Stokes et al., 2001; Vaisvila
et al., 2001), the reservoir of adaptive functions that can
be propagated by MRIs is potentially immense. Our
results confirm that integrons operate as a potent general
gene capture system, and this activity may have a re-
sounding impact on bacterial adaptation.
A striking difference between MRIs and SIs is the nature
of the attC sites that are associated with their accumu-
lated gene cassettes. The attC sites associated with the
gene cassettes of SIs display a strikingly high degree of
species-specific sequence relatedness, whereas the attC
sites of the cassettes found in MRIs share very little
sequence homology. Furthermore, nearly identical cas-
settes can be found in the SIs of remote species but each
is associated with the signature attC site of that species
(Rowe-Magnus et al., 1999). These observations led to
the proposal that gene cassettes are assembled within
individual SI-harbouring bacteria and are then dissemi-
nated by MRIs (Rowe-Magnus et al., 2001). In support of
this notion, to date 12 antibiotic resistance gene cassettes
have been identified with attC sites that are virtually
© 2002 Blackwell Science Ltd, Molecular Microbiology, 43, 1657–1669
Bacterial trafficking of integron gene cassettes 1665
identical in length and sequence to the signature attC
sites of various SIs (Rowe-Magnus et al., 2001). Nine of
these are associated with the signature attC sites of a
Xanthomonas species SI (also known as XSRs for Xan-
thomonas species repeats) and two are associated with
the signature attC sites of the V. cholerae SI (also known
as VCRs for V. cholerae repeats; Barker et al., 1994). We
found that a VCR abutted the catB9 gene. Interestingly,
the closest relatives to CATB9 (CATB6, CATB8 and
CAT accession no. X82455) are also structured as inte-
gron gene cassettes, but their associated attC sites
are strikingly different as they belong to the XSR or
SPR (for Shewanella putefaciens repeats) types. Our
identification of a third VCR-associated antibiotic re-
sistance gene cassette (CARB4 and dfrVI are also VCR-
associated; Mazel et al., 1998) suggests that SIs are an
important source of the antibiotic resistance gene cas-
settes found in MRIs.
The genetic organization of a MRI promotes coexpres-
sion of its gene cassettes from a single promoter, PC, and
selection for one resistance determinant often co-selects
for the maintenance of the entire array. Studies are cur-
rently under way in our laboratory to determine if similar
promoters reside within the attI regions of various SIs.
Even if this is the case, it is inconceivable that a single
promoter could drive expression of the potentially hun-
dreds of cassettes within a SI. As a case in point, the
catB9 gene cassette in the V. cholerae strain 63.32 SI was
phenotypically silent when located in its native seventh
position. Sequential repositioning of the cassette closer to
a strong promoter was paralleled by an increase in the
growth rate of E. coli transformants in the presence of Cm,
and integron-mediated repositioning of the catB9 cassette
into the first position of In3 produced CmR E. coli transcon-
jugants and V. cholerae 1275 inductants. The identifica-
tion of cassettes that contain their own promoters coupled
with transcriptional readthrough, the identification of cas-
settes that do not code for a discernible ORF but contain
identifiable –35 and –10 transcription initiation signals
and cassette repositioning are alternative ways in which
promoterless cassettes may be expressed.
Gene cassettes are probably exchanged in marine and
soil habitats and the forces governing their transfer are
probably highly variable. The antibiotic therapy regimes
of clinical environments typically select for resistance to
the antibiotics being administered. We observed that the
catB9 gene cassette was present in the SIs of Asian iso-
lates collected after 1961 but absent from isolates of the
same region collected prior to 1961. Several lines of evi-
dence, however, indicate that the intervention of a selec-
tive pressure for resistance to Cm around this point in time
to force V. cholerae to acquire the catB9 cassette or
favouring the expansion of a clone already harbouring the
catB9 cassette is unlikely. First, all of the V. cholerae
1666 D. A. Rowe-Magnus et al.
isolates that tested positive for the catB9 gene cassette
remained sensitive to Cm. Second, Cm has not been
widely used in the treatment of cholera (Kaper et al.,
1995) and the presence of the catB9 gene could not be
correlated with the specific use of Cm in the treatment of
epidemic and pandemic cholera. Furthermore, the catB9
cassette was variably present among the serovars of both
epidemic and non-epidemic strains of V. cholerae. Our
findings thus raise the intriguing question of why there is
a cat gene cassette in the V. cholerae SI. There is no
reason to assume that the original function of the resist-
ance gene homologues found in SIs was the defensive
inactivation of the antibiotics developed by human indus-
try or medicine. Indeed, more than 10 000 different natural
antibiotics produced by ascomycetes alone have been
catalogued (Berdy, 1995), any number of which may have
been the original substrates of the SI gene cassettes.
These genes may also have had a function aside from
conferring antibiotic resistance. In support of this notion,
we have found resistance gene homologues in the SI cas-
settes of a V. metschnikovii strain isolated in 1888 (Rowe-
Magnus et al., 1999), 40 years before the discovery of
penicillin. Thus, the gene cassettes of SIs, be they silent
genes or pseudogenes, may provide a genetic basis for
the evolution and subsequent retention of novel antimi-
crobial, virulence or metabolic functions.
The integron system allows bacteria to scavenge
foreign genes that may ultimately endow an adaptive
advantage. Likewise, as the cassettes are presumably
subject to episodic selection, unproductive gene cas-
settes may be eliminated through integrase-mediated
deletion events (Collis and Hall, 1992) to reduce the
genetic burden. With respect to the evolution of antibiotic
resistance, this may be particularly important. Several
studies have shown that, when acquiring antibiotic re-
sistance, a concomitant cost to fitness compared with
susceptible bacteria may arise in resistant strains in the
absence of selective pressure (Schrag and Perrot, 1996;
Levin et al., 1997; Schrag et al., 1997; Bjorkman et al.,
2000). The subsequent evolution of resistant bacteria
in the absence of selection commonly resulted in the
accumulation of secondary mutations that compensated
for these costs, rather than reversion to a drug-sensitive
state. Alternatively, as we have shown for V. cholerae,
‘positional regulation’ probably governs the expression
of many cassette-encoded functions as the majority
are promoterless and their transcription will be directly
dependent on their position relative to a sufficiently active
promoter (Levesque et al., 1994; Collis and Hall, 1995).
To this end, the evolution of promoterless antibiotic resis-
tance gene cassettes within integron environments may
permit bacteria to alternate between sensitive and resis-
tant states to offset the fitness constraints of resistance
under non-selective conditions. This could allow bacteria
© 2002 Blackwell Science Ltd, Molecular Microbiology, 43, 1657–1669
to circumvent the necessary ascent and selection of com-
pensatory mutations and still allow resistance to persist in
the absence of selection. Thus, otherwise phenotypically
sensitive strains may still be a genetic source for the evo-
lution of resistance to clinically relevant antibiotics through
integrase-mediated recombination events.
Experimental procedures
Bacterial strains and media
The following E. coli strains were used in this work:
DH5a (supE44DlacU169(F80lacZDM15) DArgFhsdR17recA1
endA1gyrA96thi-1relA1), b2150 (DdapA::(erm-pir) thrB1004,
pro, thi, strA, hsdS, lacZ DM15 (F¢ lacZ DM15 lacIq, traD36,
proA +, proB +) ), TOP10 (F- mcrA D(mrr-hdsRMS-mcrBC)
F80 lacZDM15 recA1 deoR araD139 D(ara-leu) 7697 galU
galK rpsL endA1 nupG) and UB5201 (Martinez and de la
Cruz, 1990). V. cholerae strains were provided by the Collec-
tion de l’Institut Pasteur (CIP). Antibiotics were used at the
following concentrations: nalidixic acid (Nal), 30 mg ml–1;
ampicillin (Ap), 100 mg ml–1; kanamycin (Kan), 25 mg ml–1; chlo-
ramphenicol (Cm), 25 mg ml–1; trimethoprim (Tp), 87 mg ml–1;
fosfomycin (Fm), 250 mg ml–1; phosphinothricine (Pt), 50 mg
ml–1 in Mueller–Hinton media. Susceptibility to virginiamycin
and pristinamycin was determined by a disc diffusion assay
with commercially available antibiotic discs (Sanofi Diagnos-
tics Pasteur, Marne-la-Coquette, France) containing 30 mg of
the antibiotic. Diaminopimelic acid (DAP) was supplemented
when necessary to a final concentration of 0.3 mM. PCR-
amplified genes were cloned using the TA TOPO cloning kit
(Invitrogen). Primers were obtained from Genset (France).
Sequencing was performed by MWG-Biotech AG (Germany)
or on a Pharmacia 4 ¥ 4 sequencing system.
Detection of resistance gene cassette homologues in
V. cholerae strain CIP 63.32
To verify that the resistance gene cassette homologues iden-
tified in V. cholerae strain N16961 were also present in the
SI of V. cholerae CIP 63.32, specific primers VC473 (TGCTG
TTTTGATGGTTGCTATGAT), VC387 (AAGAGGACATTTTG
TGGAAATTG), VC328 (AATGGAGTTTCTGATGAAAATTA)
and VC300 (TCCCTCTTGAGGCGTTTGTTATGT) were de-
signed from the sequence of strain N16961 and used in a
PCR reaction with VCR1 (GTCCCTCTTGAGGCGTTTGTTA)
and strain CIP 63.32 as matrix.
Conjugation assays
Plasmid R388 (Martinez and de la Cruz, 1990) was intro-
duced into V. cholerae 63.32 by conjugation from E. coli strain
b2150. The intI1 gene was amplified using the primers
IN1RBS (GAATTCGAGCTCTAACAAAGGAGCAAGCCATG
AAAACCGCCACTGCG) and IN1END (GGATCCATCAGG
CAACGACGGGCTGCTG), digested with EcoRI and BamHI
and cloned into pTRC99A (Pharmacia) digested with the
same enzymes to create plasmid p112. p112, which contains
the intI1 gene under the control of the isopropyl-beta-D-

thiogalactopyranoside (IPTG)-inducible synthetic Ptrc pro-
moter, was then introduced by electroporation (Chiang and
Mekalanos, 2000) into V. cholerae strain 63.32 (R388) to
create V. cholerae strain 1275. E. coli UB5201 served as the
recipient for conjugations with V. cholerae strain 1275. The
recombination and conjugation frequency (RCF) is the ratio
of CmR to TpR transconjugants (average of three independent
trials). The precise location of the recombination event was
established by PCR and sequencing of 10 random clones per
test set.
CAT cassette cloning and CAT assay
The CAT gene was amplified by PCR using the primers CAT1
(GCGGAATTCACCATGAACTTCTTTACGTCTCCATTTTC)
and CAT2 (AAACTGCAGTTGTTATGTGCGTGCTTTTGA
CT) from V. cholerae strain 63.32 genomic DNA. The product
was digested with EcoRI and PstI, cloned into pBAD24
(Guzman et al., 1995) digested with the same enzymes (to
create pBADVchCAT) and transformed into DH5a. CmR was
induced by addition of arabinose to a final concentration of
0.4%. Control strains were grown in 0.4% glucose to repress
expression of the CAT gene or contained pBAD24 without
insert. The CAT activity assay was conducted as previously
described (Gorman et al., 1982) on cell extracts from
the control E. coli strain UB5201 (R388) and from
UB5201a(R388::catB9)
Allele-specific fusion primer PCR and nested PCR
Following intI1 induction, primers DFRFUS (CACAACGGCT
GCCTAACGCCC) and attI1.1 (GAATTCGGCTTGTTATGACT
GTTTTTTTGTAC) were used in a PCR reaction with V.
cholerae strain 1275 as matrix. DFRFUS is a unique fusion
primer designed to hybridize specifically to the VCR–dfrB2
cassette junction that would arise from an IntI1-mediated
recombination event of any V. cholerae SI cassette at the attI1
site of plasmid R388. The PCR was conducted as follows:
94°C, 30 s; 60°C, 30 s; 72°C, 2 min, for 30 cycles with Pfu poly-
merase. The PCR products were cloned using a TA cloning kit
and transformed into TOP10 cells (invitrogen). Ten colonies
were chosen at random for sequence analysis. Homologues
of the following V. cholerae strain N16961 SI cassettes
were identified: accession numbers D82477 (catB9), F82451,
H82462, H82465, E82472, G82462, D82471, D82474,
E82471 and H82471. The primers VCR2 (Mazel et al., 1998)
and DFRFUS yielded no PCR product. None of the primers
alone produced a PCR product nor did either combination of
primers on uninduced cultures.
Nested PCR was conducted using primers cat2, 328-1
(ACAAATCTGCGCTCCCAACTC), 387-1 (GGGAACATGCA
AAACCTTGCAAC) or 473-1 (CAATGCTATAAGGCGGCAC
ATC) with the primer attI1.1 and the above bulk PCR product
as matrix. The PCR was conducted as described above. The
amplified fragments were sequenced directly with the corre-
sponding cassette-specific primers to confirm the attI site
location of each cassette.
Construction of sequential cassette deletion plasmids
The 3.8 kb region from the VchintIA gene to the end of the
© 2002 Blackwell Science Ltd, Molecular Microbiology, 43, 1657–1669
Bacterial trafficking of integron gene cassettes 1667
catB9 gene was amplified by PCR using Pfu polymerase
(Promega) with the primer pair I4 (CGCGGATCCGCGAA
CACTTAACAAAAACTGGG) and cat2, and V. cholerae strain
63.32 genomic DNA as matrix. The product was digested with
BamHI and PstI and cloned into pSU38 digested with the
same enzymes. Sequential deletions were cloned into the
SmaI and PstI sites of pSU38 following digestion of the Pfu
PCR products obtained with 14Cm1 (TGTGGTCGGAGAA
AAATTATGA), I4Cm2 (GTTGCCAACGAGAAAAAAGTA),
I4Cm3 (GTTTCTATCAAGAAAAATCATCAAA), I4Cm4 (GGT
GTTCATCAGCTATGTGAA), I4Cm5 (GAGCTTTATGGAGC
TAACTATG) or I4Cm6 (TGTGCTTTTAGGAAAATTTAAAAA
TG) and cat2 by PstI. Plasmids were transformed into E.
coli TOP10 (Invitrogen) and selected on Kan plates. To
test the Cm phenotype, the transformants were both plated
on 25 mg ml–1 Cm and inoculated into liquid culture con-
taining 25 mg ml–1 Cm. Growth following incubation at 37°C
was followed.
Nucleotide accession number
The catB9 and 1341 gene cassettes have been deposited in
GenBank under the accession numbers AF462019 and
AF462020 respectively.
Acknowledgements
We dedicate this paper to the memory of Professor Maurice
Hofnung, former Chef de l’Unité PMTG, for his encourage-
ment, guidance and friendship. We also thank Professor
Julian Davies (The University of British Columbia) for his
support, Brigitte Bourachot for technical help, Jean-Michel
Foumier for V. cholerae strains, and Jean-Marc Ghigo and
Martina Ochs for critical reading of the manuscript. D.R.-M.
is an EMBO and a Fondation de la Recherche Medicale
(FRM) Post-doctoral fellow. This work was supported by the
Institut Pasteur, the CNRS, the Programme de Recherche
Fondamentale en Microbiologie et Maladies Infectieuses et
Parasitaires from the MENRT and the DGA (contrat no.
0134020).
References
Arakawa, Y., Murakami, M., Suzuki, K., Ito, H.,
Wacharotayankun, R., Ohsuka, S., et al. (1995) A novel
integron-like element carrying the metallo-beta-lactamase
gene blaIMP. Antimicrobial Agents Chemotherapy 39:
1612–1615.
Avila, P., and de la Cruz, F. (1988) Physical and genetic map
of the IncW plasmid R388. Plasmid 20: 155–157.
Barker, A., Clark, C.A., and Manning, P.A. (1994) Identifica-
tion of VCR, a repeated sequence associated with a
locus encoding a hemagglutinin in Vibrio cholerae O1.
J Bacteriol 176: 5450–5458.
Barker, A., and Manning, P.A. (1997) VlpA of Vibrio
cholerae O1: the first bacterial member of the alpha
2-microglobulin lipocalin superfamily. Microbiology 143:
1805–1813.
Berdy, J. (1995) Are actynomycetes exhausted as a source
of secondary metabolites? In: Ninth Symposium on the
1668 D. A. Rowe-Magnus et al.
Actynomycetes, Moscow, Russia, 10–15 July 1994. New
York: Allerton Press, pp. 13–34.
Bjorkman, J., Nagaev, I., Berg, O.G., Hughes, D., and
Andersson, D.I. (2000) Effects of environment on com-
pensatory mutations to ameliorate costs of antibiotic resis-
tance. Science 287: 1479–1482.
Chiang, S.L., and Mekalanos, J.J. (2000) Construction of a
Vibrio cholerae vaccine candidate using transposon
delivery and FLP recombinase-mediated excision. Infect
Immun 68: 6391–6397.
Clark, C.A., Purins, L., Kaewrakon, P., Focareta, T., and
Manning, P.A. (2000) The Vibrio cholerae O1 chromoso-
mal integron. Microbiology 146: 2605–2612.
Collis, C.M., and Hall, R.M. (1992) Gene cassettes from the
insert region of integrons are excised as covalently closed
circles. Mol Microbiol 6: 2875–2885.
Collis, C.M., and Hall, R.M. (1995) Expression of antibiotic
resistance genes in the integrated cassettes of integrons.
Antimicrob Agents Chemother 39: 155–162.
Collis, C.M., Grammaticopoulos, G., Briton, J., Stokes, H.W.,
and Hall, R.M. (1993) Site-specific insertion of gene cas-
settes into integrons. Mol Microbiol 9: 41–52.
Dalsgaard, A., Forslund, A., Petersen, A., Brown, D.J., Dias,
F., Monteiro, S., et al. (2000a) Class 1 integron-borne,
multiple-antibiotic resistance encoded by a 150-kilobase
conjugative plasmid in epidemic Vibrio cholerae O1 strains
isolated in Guinea-Bissau. J Clin Microbiol 38: 3774–3779.
Dalsgaard, A., Forslund, A., Tam, N.V., Vinh, D.X., and Cam,
P.D. (1999) Cholera in Vietnam: changes in genotypes and
emergence of class I integrons containing aminoglycoside
resistance gene cassettes in Vibrio cholerae O1 strains
isolated from 1979 to 1996. J Clin Microbiol 37: 734–
741.
Dalsgaard, A., Forslund, A., Serichantalergs, O., and
Sandvang, D. (2000b) Distribution and content of class 1
integrons in different Vibrio cholerae O-serotype strains
isolated in Thailand. Antimicrob Agents Chemother 44:
1315–1321.
Davies, J.E. (1994) Inactivation of antibiotics and the dis-
semination of resistance genes. Science 264: 375–382.
Davies, J.E. (1997) Origins, acquisition and dissemination of
antibiotic resistance determinants. Ciba Foundation Symp
207: 15–27.
de la Cruz, F., and Davies, J. (2000). Horizontal gene trans-
fer and the origin of species: lessons from bacteria. Trends
in Microbiol 8: 128–133.
van Dongen, W.M.A.M., van Vlerken, M.M.A., and De Graaf,
F.K. (1987) Nucleotide sequence of a DNA fragment
encoding a Vibrio cholerae haemagglutinin. Mol General
(Life Sci Adv) 6: 85–91.
Falbo, V., Carattoli, A., Tosini, F., Pezzella, C., Dionisi, A.M.,
and Luzzi, I. (1999) Antibiotic resistance conferred by a
conjugative plasmid and a class I integron in Vibrio
cholerae O1 El Tor strains isolated in Albania and Italy.
Antimicrob Agents Chemother 43: 693–696.
Gorman, C.M., Moffat, L.F., and Howard, B.H. (1982) Recom-
binant genomes which express chloramphenicol acetyl-
transferase in mammalian cells. Mol Cell Biol 2:
1044–1051.
L.M., Belin, D., Carson, M.J., and Beckwith, J. (1995) Tight
regulation, modulation, and high-level expression by
© 2002 Blackwell Science Ltd, Molecular Microbiology, 43, 1657–1669
vectors containing the arabinose PBAD promoter. J
Bacteriol 177: 4121–4130.
Hall, R.M. (1997) Mobile gene cassettes and integrons:
moving antibiotic resistance genes in gram-negative bac-
teria. Ciba Foundation Symp 207: 192–202; discussion
202–5.
Hall, R.M., and Collis, C.M. (1995) Mobile gene cassettes
and integrons: capture and spread of genes by site-
specific recombination. Mol Microbiol 15: 593–600.
Hall, R.M., and Collis, C.M. (1998) Antibiotic resistance in
gram-negative bacteria: the role of gene cassettes and
integrons. Drug Resistance Updates 1: 109–119.
Hall, R.M., and Stokes, H.W. (1993) Integrons: novel DNA
elements which capture genes by site-specific recombina-
tion. Genetica 90: 115–132.
Heidelberg, J.F., Eisen, J.A., Nelson, W.C., Clayton, R.A.,
Gwinn, M.L., Dodson, R.J., et al. (2000) DNA sequence of
both chromosomes of the cholera pathogen Vibrio
cholerae. Nature 406: 477–483.
Hochhut, B., Lotfi, Y., Mazel, D., Faruque, S.M., Woodgate,
R., and Waldor, M.K. (2001) Molecular analysis of anti-
biotic resistance gene clusters in Vibrio cholerae O139
and O1 SXT constins. Antimicrob Agents Chemother 45:
2991–3000.
Jain, R., Rivera, M.C., and Lake, J.A. (1999) Horizontal gene
transfer among genomes: the complexity hypothesis. Proc
Natl Acad Sci USA 96: 3801–3806.
Jones, M.E., Peters, E., Weersink, A.M., Fluit, A., and
Verhoef, J. (1997) Widespread occurrence of integrons
causing multiple antibiotic resistance in bacteria. Lancet
349: 1742–1743.
Kaper, J.B., Morris, J.G. Jr and Levine, M.M. (1995) Cholera.
Clin Microbiol Rev 8: 48–86.
Levesque, C., Brassard, S., and Lapointe, J., and Roy, P.H.
(1994) Diversity and relative strength of tandem promoters
for the antibiotic-resistance genes of several integrons.
Gene 142: 49–54.
Levin, B.R., Lipsitch, M., Perrot, V., Schrag, S., Antia, R.,
Simonsen, L., Walker, N.M., and Stewart, F.M. (1997) The
population genetics of antibiotic resistance. Clin Infect Dis
24 (Suppl. 1): S9–S16.
Liebert, C.A., Hall, R.M., and Summers, A.O. (1999) Trans-
poson Tn21, flagship of the floating genome. Microbiol Mol
Biol Rev 63: 507–522.
Martin, C., Timm, J., Rauzier, J., Gomez-Lus, R., Davies,
J., and Gicquel, B. (1990) Transposition of an antibiotic
resistance element in mycobacteria. Nature 345: 739–743.
Martinez, E., and de la Cruz, F. (1988) Transposon Tn21
encodes a RecA-independant site-specific integration
system. Mol Gen Genet 211: 320–325.
Martinez, E., and de la Cruz, F. (1990) Genetic elements
involved in Tn21 site-specific integration, a novel mecha-
nism for the dissemination of antibiotic resistance genes.
EMBO J 9: 1275–1281.
Mazel, D., and Davies, J. (1999) Antibiotic resistance in
microbes. Cell Mol Life Sci 56: 742–754.
Mazel, D., Dychinco, B., Webb, V.A., and Davies, J. (1998)
A distinctive class of integron in the Vibrio cholerae
genome. Science 280: 605–608.
Naas, T., Poirel, L., Karim, A., and Nordmann, P. (1999)
Molecular characterization of In50, a class 1 integron

encoding the gene for the extended-spectrum beta-
lactamase VEB-1 in Pseudomonas aeruginosa. FEMS
Microbiol Lett 176: 411–419.
Naas, T., Mikami, Y., Imai, T., Poirel, L., and Nordmann, P.
(2001) Characterization of In53, a class 1 plasmid- and
composite transposon-located integron of Escherichia coli
which carries an unusual array of gene cassettes. J
Bacteriol 183: 235–249.
Nesvera, J., and Hochmannova, J., and Patek, M. (1998) An
integron of class 1 is present on the plasmid pCG4 from
gram-positive bacterium Corynebacterium glutamicum.
FEMS Microbiol Lett 169: 391–395.
Nield, B.S., Holmes, A.J., Gillings, M.R., Recchia, G.D.,
Mabbutt, B.C., Nevalainen, K.M., and Stokes, H.W. (2001)
Recovery of new integron classes from environmental
DNA. FEMS Microbiol Lett 195: 59–65.
Nunes-Duby, S.E., Kwon, H.J., Tirumalai, R.S., Ellenberger,
T., and Landy, A. (1998) Similarities and differences among
105 members of the Int family of site-specific recombi-
nases. Nucleic Acids Res 26: 391–406.
Ochman, H., Lawrence, J.G., and Groisman, E.A. (2000)
Lateral gene transfer and the nature of bacterial evolution.
Nature 405: 299–304.
Ogawa, A., and Takeda, T. (1993) The gene encoding the
heat-stable enterotoxin of Vibrio cholerae is flanked by
123-base pair direct repeats. Microbiol Immunol 37: 607–
616.
Ploy, M.C., Courvalin, P., and Lambert, T. (1998) Character-
ization of In40 of Enterobacter aerogenes BM2688, a class
1 integron with two new gene cassettes, cmlA2 and qacF.
Antimicrobial Agents Chemother 42: 2557–2563.
Poirel, L., Guibert, M., Bellais, S., Naas, T., and Nordmann,
P. (1999a) Integron- and carbenicillinase-mediated re-
duced susceptibility to amoxicillin–clavulanic acid in iso-
lates of multidrug-resistant Salmonella enterica serotype
typhimurium DT104 from French patients. Antimicrob
Agents Chemother 43: 1098–1104.
Poirel, L., Naas, T., Guibert, M., Chaibi, E.B., Labia, R., and
Nordmann, P. (1999b) Molecular and biochemical charac-
terization of VEB-1, a novel class A extended-spectrum
beta-lactamase encoded by an Escherichia coli integron
gene. Antimicrob Agents Chemother 43: 573–581.
© 2002 Blackwell Science Ltd, Molecular Microbiology, 43, 1657–1669
Bacterial trafficking of integron gene cassettes 1669
Poirel, L., Le Thomas, I., Naas, T., Karim, A., and Nordmann,
P. (2000) Biochemical sequence analyses of GES-1, a
novel class A extended-spectrum b-lactamase, and the
class 1 integron In52 from Klebsiella pneumoniae. Anti-
microbial Agents Chemother 44: 622–632.
Recchia, G.D., and Hall, R.M. (1995) Gene cassettes: a new
class of mobile element. Microbiology 141: 3015–3027.
Rowe-Magnus, D.A., and Mazel, D. (1999) Resistance gene
capture. Curr Opin Microbiol 2: 483–488.
Rowe-Magnus, D.A., Guerout, A.-M., and Mazel, D. (1999)
Super-integrons. Res Microbiol 150: 641–651.
Rowe-Magnus, D.A., Guerout, A.-M., Ploncard, P., Dychinco,
B., Davies, J., and Mazel, D. (2001) The evolutionary
history of chromosomal super-integrons provides an
ancestry for multi-resitant integrons. Proc Natl Acad Sci
USA 98: 652–657.
Schrag, S.J., and Perrot, V. (1996) Reducing antibiotic
resistance. Nature 381: 120–121.
Schrag, S.J., Perrot, V., and Levin, B.R. (1997) Adaptation to
the fitness costs of antibiotic resistance in Escherichia coli.
Proc R Soc Lond B Biol Sci 264: 1287–1291.
Stokes, H.W., O’Gorman, D.B., Recchia, G.D., Parsekhian,
M., and Hall, R.M. (1997) Structure and function of 59-base
element recombination sites associated with mobile gene
cassettes. Mol Microbiol 26: 731–745.
Stokes, H.W., Holmes, A.J., Nield, B.S., Holley, M.P.,
Nevalainen, K.M., Mabbutt, B.C., and Gillings, M.R. (2001)
Gene cassette PCR: sequence-independent recovery of
entire genes from environmental DNA. Appl Environ
Microbiol 67: 5240–5246.
Sundstrom, L. (1998) The potential of integrons and
connected programmed rearrangements for mediating
horizontal gene transfer. APMIS 84 (Suppl.): 37–42.
Sundstrom, L., Roy, P.H., and Skold, O. (1991) Site-specific
insertion of three structural gene cassettes in transposon
Tn7. J Bacteriol 173: 3025–3028.
Vaisvila, R., Morgan, R.D., Posfai, J., and Raleigh, E.A.
(2001) Discovery and distribution of super-integrons
among pseudomonads. Mol Microbiol 42: 587–601.
Vicente, A.C.P., Coehlo, A.M., and Salles, C.A. (1997) Detec-
tion of Vibrio cholerae and V. mimicus heat-stable toxin
gene sequence by PCR. J Med Microbiol 46: 398–402.
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

Annu. Rev. Microbiol. 2000. 54:641–79

Copyright c 2000 by Annual Reviews. All rights

PATHOGENICITY ISLANDS AND THE EVOLUTION

OF MICROBES

Jörg Hacker
Institut für Molekulare Infektionsbiologie, Universität Würzburg, D-97070 Würzburg,
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

Germany; e-mail: j.hacker@mail.uni-wuerzburg.de

James B. Kaper
Center for Vaccine Development and Department of Microbiology & Immunology,
University of Maryland School of Medicine, Baltimore, Maryland 21201;
e-mail: jkaper@umaryland.edu

Key Words gene transfer, virulence, genome structure, plasmid, bacteriophage

■ Abstract Virulence factors of pathogenic bacteria (adhesins, toxins, invasins,
protein secretion systems, iron uptake systems, and others) may be encoded by partic-
ular regions of the prokaryotic genome termed pathogenicity islands. Pathogenicity
islands were first described in human pathogens of the species Escherichia coli, but
have recently been found in the genomes of various pathogens of humans, animals, and
plants. Pathogenicity islands comprise large genomic regions [10–200 kilobases (kb)
in size] that are present on the genomes of pathogenic strains but absent from the
genomes of nonpathogenic members of the same or related species. The finding that
the G+C content of pathogenicity islands often differs from that of the rest of the
genome, the presence of direct repeats at their ends, the association of pathogenicity
islands with transfer RNA genes, the presence of integrase determinants and other mo-
bility loci, and their genetic instability argue for the generation of pathogenicity islands
by horizontal gene transfer, a process that is well known to contribute to microbial evo-
lution. In this article we review these and other aspects of pathogenicity islands and
discuss the concept that they represent a subclass of genomic islands. Genomic islands
are present in the majority of genomes of pathogenic as well as nonpathogenic bac-
teria and may encode accessory functions which have been previously spread among
bacterial populations.

CONTENTS
INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 642
COMMON FEATURES OF PATHOGENICITY ISLANDS . . . . . . . . . . . . . . . . . . 643
OCCURENCE OF PATHOGENICITY ISLANDS . . . . . . . . . . . . . . . . . . . . . . . . . 645
VIRULENCE-ASSOCIATED GENES ON PATHOGENICITY ISLANDS . . . . . . . . 649
Adhesins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 649
Secretion Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 650

0066-4227/00/1001-0641$14.00 641
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

642 HACKER ¥ KAPER

Invasins, Modulins, and Effectors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 654

Toxins . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 656
Iron Uptake Systems . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 658
Other Virulence-Associated Genes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 659
REGULATION OF PATHOGENICITY ISLANDS . . . . . . . . . . . . . . . . . . . . . . . . . 661
JUNCTION SITES OF PATHOGENICITY ISLANDS . . . . . . . . . . . . . . . . . . . . . . 664
THE ROLE OF tRNA LOCI . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 665
MOBILITY GENES ON PATHOGENICITY ISLANDS . . . . . . . . . . . . . . . . . . . . . 667
TRANSFER AND DELETION OF PATHOGENICITY ISLANDS . . . . . . . . . . . . . 668
CONCLUSION: Pathogenicity Islands, Genomic Islands,
and the Evolution of Microbes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 670
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

INTRODUCTION

Over the last few years, the genome sequences of >20 bacterial species have
been determined and analyzed. It became obvious that microbial genomes consist
primarily of core sequences, with a fairly homogeneous G+C content and codon
usage, which encode housekeeping functions and which carry gene clusters with
relatively low mutational capacity. These core sequences include ribosomal-RNA-
specific genes and genes that encode key metabolic proteins such as ATPases [for
a review see Hacker et al (37) and Morschhäuser et al (78)]. In addition to the
core genome, sequences were discovered that differ from the rest of the genome
in their G+C content and codon usage and that have been acquired via horizontal
gene transfer. For example, the overall amount of horizontally transferred DNA
in the Escherichia coli K-12 genome has been calculated to be ∼17% (63). These
regions, which have been designated genomic islands, encode accessory functions
such as additional metabolic activities, antibiotic resistance, or properties involved
in microbial fitness, symbiosis, or pathogenesis.
It has long been known that important pathogenicity factors can be encoded by
mobile genetic elements, which are capable of lateral gene transfer. The presence
of genes encoding diphtheria toxin of Corynebacterium diphtheriae on a lysogenic
bacteriophage was discovered nearly half a century ago, and in subsequent years
a variety of toxins have been shown to be encoded on bacteriophages, including
Shiga toxin of enterobacteria, cholera toxin of Vibrio cholerae, neurotoxins of
Clostridium botulinum, and the cytotoxin of Pseudomonas aeruginosa (11, 110).
Other examples of mobile virulence elements include the heat-stable enterotoxin
gene of E. coli, which is part of a transposon, and genes encoding a V. cholerae
hemagglutinin, which are located on an integron structure (68). In addition, impor-
tant virulence-associated genes of gram-negative pathogens (e.g. Shigella flexneri,
Salmonella enterica and Yersinia spp.) as well as of gram-positive pathogens
(e.g. Clostridium tetani and Enterococcus faecalis) are located on plasmids (33,
48, 82). Other virulence determinants are located on the chromosome, where
they are often associated in so-called virulence blocks or virulence cassettes. In
the early 1980s, it was discovered that the presence of chromosomally located
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

PATHOGENICITY ISLANDS 643

virulence-associated genes of uropathogenic E. coli (UPEC) differs among E. coli

strains (38, 66). This observation was also made for intestinal E. coli pathogens
and other pathogenic bacteria of different species and genera (69). In 1990, these
regions were termed pathogenicity islands (PAIs) (34). Although the presence of
PAIs was first described for human pathogens of the family Enterobacteriaceae, it
has become clear in recent years that animal and plant pathogens may also contain
PAIs. The majority of PAIs are located on the chromosome; however, they can
also be part of bacterial plasmids and phages (37).
This article describes the structures and functions of the various PAIs known to
date and discusses virulence functions encoded on PAIs and their impact on infec-
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

tious diseases. In addition, the mobility of PAI elements is summarized. Last, but
not least, we discuss PAIs as examples of genomic islands present in the majority of
bacterial genomes, where they act as landmarks for lateral gene transfer. The sig-
nificance of genomic islands, including PAIs, for microbial evolution is discussed.

COMMON FEATURES OF PATHOGENICITY ISLANDS

PAIs in pathogenic E. coli were the first described. It was soon discovered that
pathogenic bacteria of species other than E. coli, both gram positive and gram
negative, contain in their genomes DNA segments which share many of the below
mentioned features of PAIs. A simplified model of a bacterial PAI is shown in
Figure 1. PAIs possess most, if not all, of the characteristics listed below (36, 37).
PAIs carry genes encoding one or more virulence factors. They were first
described in human pathogens but are also present in plant pathogens such as
Pseudomonas syringae (50). One should keep in mind that in some pathogenic

Figure 1 Model of a bacterial pathogenicity island. The thin bold line represents regions of the
core genome; pathogenicity island-specific sequences are indicated. The box represent genes. The
arrows indicate the presence of direct repeats at the ends of the pathogenicity island. Abbreviations:
DR, direct repeats; int, integrase gene; vir, virulence-associated gene; mob, mobility gene; 1mob,
pseudo-mobility gene. mob genes encode integrases, transposases, or other proteins involved in
mobility of the prokaryotic genome.
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

644 HACKER ¥ KAPER

bacteria virulence factors contribute to pathogenic potential, whereas in other, non-

pathogenic bacteria the same factors may be important for survival and replication
in particular ecological niches, where they do not create any pathogenic potential
(e.g. the iron uptake system in pathogenic Yersinia species and nonpathogenic
E. coli) (88). In nonpathogenic hosts, these DNA segments may act as fitness is-
lands or ecological islands (see section on iron uptake systems below) rather than
as PAIs. A summary of important virulence features encoded by PAIs is given in
Table 1 and discussed below.
PAIs are present in the genomes of pathogenic organisms but absent from the
genomes of nonpathogenic organisms of the same or closely related species. PAIs
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

were first described as chromosomal DNA regions, but the increasing amount of
sequence data from extrachromosomal elements supports the view that PAIs may
be also part of plasmids or bacteriophage genomes.
PAIs occupy relatively large genomic regions. The majority of PAIs cover
DNA regions of ≥10–200 kb. Strains of various species, however, may also
carry insertions of small pieces of DNA which may encode virulence factors.
These DNA regions have been termed pathogenicity islets. From a heuristic point
of view, it is useful to distinguish between islands and islets to account for the

TABLE 1 Major virulence features encoded by pathogenicity islands

Virulence feature Examples

Adherence factors Diarrheagenic Escherichia coli

Uropathogenic Escherichia coli
Vibrio cholerae
Listeria spp.
Toxins Uropathogenic Escherichia coli
Staphylococcus aureus
Iron uptake systems Uropathogenic Escherichia coli
Shigella flexneri
Yersinia spp.
Invasions, modulins, effectors Diarrheagenic Escherichia coli
Salmonella spp.
Shigella spp.
Listeria spp.
Type III secretion systems Diarrheagenic Escherichia coli
Pseudomonas syringae
Erwinia spp.
Yersinia spp.
Salmonella spp.
Shigella spp.
Type IV secretion system Helicobacter pylori
Agrobacterium tumefaciens
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

PATHOGENICITY ISLANDS 645

apparent differences in size and structure. Some authors, however, prefer the term
PAI also for smaller DNA regions (28).
PAIs often consist of DNA regions that differ from the core genome in G+C
content and in different codon usage, which may reflect the generation of PAIs by
horizontal gene transfer. It should be mentioned here that differences in the G+C
contents of PAIs and the core genome will not be observed if the DNAs of the
donors and recipients have similar or identical G+C contents.
PAIs are often flanked by small directly repeated (DR) sequences. These se-
quences may be generated after integration of PAI-specific DNA regions into the
host genome via recombination.
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

PAIs are often associated with transfer RNA (tRNA) genes. tRNA loci often
act as integration sites for foreign DNA. The association of PAIs and tRNA loci
may therefore reflect the generation of PAIs by horizontal gene transfer.
PAIs often carry cryptic or functional genes encoding mobility factors such as
integrases, transposases, and insertion sequence (IS) elements or parts of these
elements. PAIs often do not represent homogeneous pieces of DNA but rather
are made up of mosaic-like structures which have been generated by a multistep
process.
PAIs often represent unstable DNA regions. Deletions of PAIs may occur via
the direct repeats (DRs) at their ends or via IS elements or other homologous
sequences located on PAIs. Additionally, a few PAIs [e.g. the high PAI (HPI)
of Yersinia pseudotuberculosis] have the capacity to move from one tRNA site to
another (7), and other PAIs may be mobilized and transmitted by bacteriophages
(e.g. those of Staphylococcus aureus and V. cholerae) (56, 64). Particular PAIs
represent integrated plasmids, conjugative transposons, or bacteriophages or parts
of such elements.

OCCURENCE OF PATHOGENICITY ISLANDS

PAIs occur in the genomes of various human, animal, and plant pathogens. A
list of PAIs described up to now (see Table 2) shows that many members of the
Enterobacteriaceae (e.g. E. coli, S. flexneri, S. enterica, and Yersinia spp.) cause
either intestinal or extraintestinal infections via virulence factors encoded on PAIs.
Enterobacteria, as well as V. cholerae, P. syringae, and others, show frequent gene
transfer via plasmids and bacteriophages. Such extrachromosomal elements in-
deed seem to represent one source of PAIs (37). Recently it was demonstrated
that the genome of the causative agent of Legionnaires’ disease, Legionella pneu-
mophila, also contains DNA segments which carry type IV secretion-specific genes
and which can be considered PAIs (96, 97).
In addition, bacteria that have the capacity to take up DNA from the environment
via natural transformation, such as Helicobacter pylori and Neisseria gonorrhoeae,
may carry PAIs (10; JP Dillard & HS Seifert, submitted for publication).
It seems, however, that naturally transformable organisms frequently introduce
 Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.
P1: FRD/FQU

646
August 11, 2000
P2: FQP

TABLE 2 Pathogenicity islands (PAIs) and PAI-associated genes of various pathogensa

11:13

HACKER

Associated
¥

Organism Description Functions Size (kb) Junction sequences

E. coli 536 (UPEC) PAI I536 Hemolysin 70 DR 16 bp selC

KAPER

E. coli 536 (UPEC) PAI II536 Hemolysin, P-fimbriae 190 DR 18 bp leuX

E. coli 536 (UPEC) PAI III536 S-Fimbriae 25 thrW
E. coli 536 (UPEC) PAI IV536 Yersiniabactin synthesis, uptake 45 asnT
Annual Reviews

E. coli J96 (UPEC) PAI IJ96 Hemolysin, P-fimbriae 170 pheV

E. coli J196 (UPEC) PAI IIJ96 Hemolysin, P-fimbriae, 110 DR 135 bp pheR
cytotoxic necrotizing factor 1 (CNF1)
E. coli CFT073 (UPEC) PAI ICFT073 Hemolysin, P-fimbriae 58 DR 9 bp metV
E. coli K1 kps PAI Capsule n.d. pheV
AR110-19

E. coli E2348/69 (EPEC) LEE Type III secretion, invasion 35 selC

E. coli O157:H7 (EHEC) LEE Type III secretion, invasion 43 selC
E. coli EPEC2 LEE Type III secretion, invasion 35 pheU
E. coli ETEC Tia-PAI Invasion 46 selC
Y. enterocolitica HPI Yersiniabactin synthesis, transport 43 DR 17 bp asnT
Y. pseudotuberculosis HPI Yersiniabactin synthesis, transport 36 DR 17 bp asnT, U, W
Y. pestis HPI ( pgm locus) Yersiniabactin synthesis, transport 102 IS100 asnT
Hemin uptake DR 17 bp
 Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.
P1: FRD/FQU
August 11, 2000

Yersinia spp. Yop virulon Type III secretion, effectors (YOPs) 47 Plasmid
Pathogenic E. coli HPI Yersiniabactin synthesis, transport 42 asnT
P2: FQP

S. flexneri SHI-1 Enterotoxin, protease 51 IS elements

11:13

S. flexneri SHI-2 Aerobactin synthesis, transport 23–30 selC

Shigella spp. Entry region Type III secretion, invasion 37 Plasmids
S. typhimurium SPI-1 Type III secretion, invasion into 40
epithelial cells, apoptosis
S. typhimurium SPI-2 Type III secretion, invasion into monocytes 40 valV
S. typhimurium SPI-3 Invasion, survival in monocytes 17 selC
S. typhimurium SPI-4 Invasion, survival in monocytes 25 Putative
Annual Reviews

tRNA gene
S. dublin SPI-5 Enteropathogenesis 7 serT
V. cholerae VPI TCP-adhesin, regulator 39.5 DR 30 bp ssrA
H. pylori cag Pai Type IV secretion, cag-antigen 40 DR 31 bp glr
AR110-19

D. nodosus vap region Vap-antigens 12 Dr 19 bp serV

D. nodosus vrl region Vrl-antigens 27
N. gonorrhoeae PAI Serum resistance, cytotoxin 60–70
L. pneumophila icm/dot region Type IV secretion ∼20
lvh region Type IV secretion 23
B. fragilis BfPAI Fragilysin, metalloprotease II 6
B. pertussis ptx-ptl locus Pertussis toxin tRNA (Asp)
P. syringae hrp cluster Type III secretion, effectors 35 Plasmid
PATHOGENICITY ISLANDS

X. campestis hrp cluster Type III secretion, effectors ∼20

(Continued )
647
 Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.
P1: FRD/FQU

648
August 11, 2000
P2: FQP
11:13

HACKER

TABLE 2 (Continued )
¥

Associated
Organism Description Functions Size (kb) Junction sequences
KAPER

Erwinia spp. hrp cluster Type III secretion, effectors

R. solanacearum hrp cluster Type III secretion, effectors
A. tumefaciens T-DNA Crown gall tumor induction 20 DR 25 bp Plasmid
Annual Reviews

Opine production
S. aureus SaPI1 Toxic shock syndrome toxin-1, 15.2 DR 17 bp
putative superantigen
L. monocytogenes prf vir gene cluster Phospholipases, listeriolysin 9
Metalloprotease, ActA protein regulator
AR110-19

L. ivanovii prf vir gene cluster Phospholipases, listeriolysin 9

(LIPI-1) Metalloprotease, ActA protein regulator
Internalins, Sphingomyelinase 18
LIPI-2 Internalin CD 4 tRNA Thr
inlCD-region
C. difficile Pathogenicity locus Tcd toxins 19
a
DR, direct repeat; T-DNA, transferred DNA; LEE, Locus of enterocyte effacement; Tia, toxigenic invasion locus A; HPI, high PAI; SHI, Shigella PAI; SPI, Salmonella PAI; VPI,
Vibrio PAI; BfPAI, Bacteriodes f ragilis PAI; SaPI, Staphylococcus aureus PAI; LIPI, Listeria ivanovii PAI; Tcd, Clostridium difficile toxin.
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

PATHOGENICITY ISLANDS 649

small pieces of DNA into their genomes and that the generation of PAIs repre-
sents an exceptional event. Gram-positive bacteria such as Listeria ivanovii and
S. aureus carry “classical” PAIs which exhibit the majority of the features dis-
cussed above. Other gram-positive pathogens, such as Listeria monocytogenes and
Clostridium difficile, carry on their genomes so-called virulence or pathogenicity
gene clusters, which exhibit a few features of PAIs but differ in many aspects
(23, 61, 64). It is obvious from Table 2 that the occurrence of PAIs is not restricted
to a particular group of pathogens; rather, PAIs are distributed among all groups
of pathogens. In the following sections, virulence factors encoded by PAI-specific
genes are described.
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

VIRULENCE-ASSOCIATED GENES
ON PATHOGENICITY ISLANDS
Virulence factors encoded on PAIs represent the entire spectrum of bacterial viru-
lence factors, from adhesins to toxins to host defense avoidance mechanisms. This
section briefly reviews the major classes of virulence factors found on PAIs but is
by no means an exhaustive summary.

Adhesins
PAIs located in the genomes of various species and pathotypes encode adhesins,
which mediate the capacity of microbes to attach to specific eukaryotic receptor
molecules. Thus, P fimbriae, which represent important adherence factors of
UPEC, are encoded by UPEC-specific PAIs (35). These attachment factors have
the capacity to bind to galactose–α1-4–galactose-specific receptor molecules on
uroepithelial cells. P-fimbrial genes ( pap or prs) are often linked to gene clusters
hly and cfn, respectively encoding the UPEC-specific toxins α-hemolysin and
cytotoxic necrotizing factor 1, a linkage that argues for a strong coevolution of
these factors (4). In addition, UPEC, as well as sepsis- or meningitis-causing
E. coli, have the capacity to produce S fimbriae, which bind to sialic acid-specific
receptors on uroepithelial cells and on brain cells (79). S-fimbria-specific genes
(sfa) are part of a PAI that also carries genes for the iron uptake system iro, which
was initially found in Salmonella spp. (U Dobrindt, G Gottschalk, G Blum-Oehler,
J Hacker, unpublished data). Sequence data from PAIs of different UPEC isolates
show the presence of additional genes with significant homologies to adhesin gene
clusters such as those encoding the Proteus mirabilis fimbriae (Pmf), the heat-
resistant hemagglutinin of enterotoxigenic E. coli, and a saliva-binding protein of
Streptococcus sanguis (35).
The locus of enterocyte effacement (LEE) PAI of enteropathogenic and en-
terohemorrhagic E. coli (EPEC and EHEC, respectively) encodes an important
intestinal adherence factor called intimin [reviewed elsewhere (51, 52, 81)]. In-
timin is a 94- to 97-kDa outer-membrane protein encoded by the eae gene, which
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

650 HACKER ¥ KAPER

shows similarity to the Yersinia invasin adhesin. The role of intimin in disease has
been shown in volunteer and animal studies using isogenic eae mutants. EPEC
and EHEC interact with intestinal epithelial cells in a characteristic pattern, called
attaching and effacing, in which the epithelial cell brush border is effaced and the
bacteria intimately adhere to the epithelial cell membrane. Marked cytoskeletal
changes, including accumulation of polymerized actin, are seen in the epithelial
cell directly beneath the adherent bacteria. Intimin mediates the intimate adher-
ence of the bacteria to the host cell, whereas other factors on the LEE PAI (see
below) signal cytoskeletal changes. Intimin binds to the translocated intimin re-
ceptor (Tir) protein, which is translocated from the bacterium to the host cell via a
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

type III secretion system encoded on the LEE PAI (58). Tir is also encoded on the
LEE PAI and has effector functions in addition to serving as a receptor for intimin
(see below).
The Vibrio PAI (VPI) of V. cholerae (54, 55) encodes the toxin-coregulated
pilus (TCP), a type 4 pilus that is an essential intestinal adherence factor for this
species. The importance of the TCP as an intestinal colonization factor has been
shown in volunteer and animal studies [reviewed by Kaper et al (53)], although
the TCP has never been shown to actually mediate adherence to epithelial cells.
The TCP definitely mediates interbacterial adherence, and its role in intestinal
colonization may be to increase the colonizing mass of bacteria while some other
factor directly binds the bacteria to epithelial cells. Downstream of the tcp gene
cluster on the VPI is the acf (accessory colonization factor) gene cluster, whose
products play a role in chemotaxis, thereby assisting in intestinal colonization.
Another type 4 pilus of V. cholerae is the mannose-sensitive hemagglutinin,
which is encoded by a cluster of 16 genes (msh) flanked by 7-base-pair (-bp) DR
sequences (67). The 16.7-kb msh region is inserted between the yhdA and mreB
genes, which are adjacent to each other on the E. coli chromosome. Volunteer
studies have shown that the mannose-sensitive hemagglutinin is not involved in
intestinal colonization or any other aspect of cholera (102), thus indicating that,
although this region has features of a PAI, it is not involved in pathogenicity. How-
ever, finding that the mannose-sensitive hemagglutinin is essential for formation of
biofilms on abiotic surfaces (111) led Marsh & Taylor (67) to call this an environ-
mental persistence island. Another fitness or persistence island, in V. cholerae El
Tor, encodes an exopolysaccharide (EPSETr ) that is also involved in biofilm forma-
tion (117). EPSETr also mediates the rugose colony type and chlorine resistance.
Thus, the fitness/persistence islands encoding the mannose-sensitive hemagglu-
tinin and exopolysaccharide may be essential for maintenence of V. cholerae in an
environmental reservoir between epidemics.

Secretion Systems
Five distinct mechanisms for extracellular secretion of proteins, known as types I
through V, have been described in gram-negative bacteria. Such secretion mecha-
nisms are essential for the extracellular secretion of virulence factors to the surface
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

PATHOGENICITY ISLANDS 651

of the host cell or their direct translocation into the host cell. Extracellular protein
secretion via type II, IV, and V mechanisms requires the sec general secretion
pathway, whereas proteins secreted via type I and III mechanisms do not require
the sec system. Type III and type IV systems are the mechanisms that are most
closely associated with PAIs, although types I, II, and V can also be found on PAIs.

Type III Secretion Systems

Type III secretion systems have been found on PAIs in Salmonella, Yersinia,
Shigella, EPEC, and EHEC as well as in plant pathogens such as P. syringae,
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

Erwinia spp., Xanthomonas campestris, and Ralstonia solanacearum [reviewed

by Galán & Collmer (32) and by Hueck (46)]. Type III systems are not found on
nonpathogenic members of these species. Some pathogens have two distinct type
III systems; for example, Salmonella has complete type III systems on both the
SPI-1 and SPI-2 PAIs [reviewed by Groisman et al (33); see also Galan (31)], and
Yersinia enterocolitica has a chromosomally encoded type III system in addition
to the plasmid-encoded system (39). For most of these PAIs, genes encoding the
type III secretion system and genes encoding the proteins secreted by the type III
system comprise the great majority of the genes contained on the PAI. Although
the sequence and host cell activities of the secreted proteins differ according to the
individual pathogen, the majority of the >20 proteins that comprise the type III
secretion apparatus are highly conserved among the different pathogens [reviewed
by Hueck (46)]. Greater divergence is seen among the proteins secreted by the
type III systems, which are characterized by the lack of a cleavable amino-terminal
signal peptide.
The majority of the proteins comprising the type III secretion apparatus are
located in the inner membrane, but the functions of many of the type III secretion
apparatus proteins are not known. Members of the YscN family (e.g. InvC of S.
typhimurium, Spa47 of S. flexneri, and EscN of EPEC/EHEC) are ATPases that
are believed to provide energy for the secretion/assembly process (46). Proteins
of the LcrD family (e.g. InvA of S. typhimurium, MxiA of S. flexneri, and EscV
of EPEC/EHEC) and proteins of the YscC family (e.g. InvG of S. typhimurium,
MxiD of S. flexneri, and EscC of EPEC/EHEC) are believed to form channels in
the inner and outer membranes, respectively (112). Some type III systems, such as
those of S. typhimurium (62), P. syringae (91), and EPEC (59), have filamentous
surface structures that are distinct from flagella and are presumed to function as
channels for translocation of effector proteins into the host cell. Type III systems
have been proposed to act like a needle and syringe, injecting effector proteins into
the host cell.
Thus, PAIs of several mammalian and plant pathogens encode type III secretion
systems that are essential for virulence. In many cases, the PAIs are self-contained
units containing all genes encoding secreted proteins, the secretion apparatus, the
chaperones, etc. Many PAIs also contain genes encoding regulators of the type III
secretion system, but in most, if not all, PAIs the PAI-encoded regulators are in turn
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

652 HACKER ¥ KAPER

regulated by regulatory systems encoded outside the PAI (see below). In at least
one case, the LEE PAI of EPEC, the cloned PAI confers all known PAI-associated
phenotypes on E. coli K-12 (70), but in other cases, such as the closely related
LEE of EHEC O157:H7, the cloned PAI does not confer the expected phenotype
on K-12 (25). Some proteins secreted by a type III system are encoded outside
the PAI but still use the secretion system encoded in the PAI. For example, the
SopB and SopE proteins, first identified in S. enterica serovar Dublin, are secreted
via the type III secretion system encoded on the Salmonella SPI-1 PAI that is
located at 63 min on the S. enterica serovar Typhimurium chromosome. SopB is
encoded on the SPI-5 PAI, located at 20 min on the chromosome (114), and SopE
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

is encoded on a bacteriophage inserted at 61 min (76). It is unknown whether the

sopB and sopE genes were originally located on the SPI-1 island and subsequently
relocated or they were introduced into Salmonella spp. after SPI-1 was introduced
and somehow adapted to use the preexisting type III secretion system.

Type IV Secretion Systems

Like type III systems, type IV secretion systems have also been closely as-
sociated with PAIs. Type IV systems have been found to be essential for full
virulence of a variety of pathogens, including H. pylori, Bordetella pertussis,
L. pneumophila, Agrobacterium tumefaciens, and Brucella suis [reviewed by Burns
(9)]. In H. pylori, the type IV system encoded on the Cag PAI has recently been
shown to mediate the translocation of the CagA protein to host cells (95). CagA
plays an important role in the induction of host cellular growth changes induced
by H. pylori. In B. pertussis, the type IV secretion system (also called the Ptl
transporter) is essential for secretion of pertussis toxin to mammalian cells (9). A
type IV secretion system in B. suis was recently shown to be necessary for full
virulence of this organism in an in vitro infection model (81a).
The type IV secretion system (which has also been called a type V secretion
system, a term that we reserve for autotransporting proteins as discussed below)
was first described in the plant pathogen A. tumefaciens, wherein it mediates trans-
fer of DNA into plant cells. The type IV system in this species is encoded on the
∼200-kb Ti plasmid, which is essential for crown gall tumorigenesis of higher
plants [reviewed by Winans et al (113)]. Part of the Ti plasmid, called T-DNA,
is transferred into plant cells, where it is integrated into the host genome. The
transferred Agrobacterium genes are expressed in the plant cell, which causes the
infected cells to proliferate, ultimately resulting in crown gall tumors. Approxi-
mately 20 vir genes are responsible for this transfer of T-DNA, and many of the
Vir proteins show sequence similarity to Cag proteins of H. pylori and Ptl pro-
teins of B. pertussis (e.g. VirB4 is similar to Cag3 and PtlC, VirB7 is similar to
CagT and PtlI, etc.). In A. tumefaciens, the VirB4, VirB7, VirB9, VirB10, and
VirB11 proteins assemble to form the core of the transporter while VirB2 forms
a pilus. Little information is available on the functions of the type IV proteins of
the mammalian pathogens.
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

PATHOGENICITY ISLANDS 653

In L. pneumophila, a cluster of genes called dot (107) or icm (96, 97) was shown
to be essential for intracellular replication in host cells. The proteins encoded by
these genes show sequence similarity to the VirB proteins and are also essential for
transfer of plasmid DNA from one bacterial cell to another. The role of this type IV
secretion system in the pathogenesis of disease due to L. pneumophila is assumed to
involve transfer of protein virulence factors, rather than DNA, into eukaryotic cells,
but the translocated proteins have not yet been identified. Recently another gene
cluster, specific for a second type IV secretion system termed Lvh (for Legionella
vir homologs), was detected on a putative PAI of the L. pneumophila genome; the
function of this system, however, has yet to be elucidated (97).
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

Other Secretion Systems

Type III or type IV secretion systems are invariably associated with PAIs and are
nearly always essential for full virulence of a pathogen. Other protein secretion
systems of gram-negative bacteria have also been found to be encoded on PAIs,
but the association of these other systems with PAIs is not yet as strong as it is for
type III or type IV systems. As more genomic sequences become available and
the evolution of these secretion systems is further characterized, it is possible that
most, if not all, representatives of these specialized protein secretion systems will
be recognized as being part of PAIs or other genomic islands.
The type I secretion system was first described in the E. coli α-hemolysin, which
is encoded on PAIs in most UPEC strains and on the pO157 plasmid of EHEC
O157:H7. This secretion system requires three proteins: an inner-membrane
ATPase that provides energy for the system (HlyB for E. coli hemolysin), a mem-
brane fusion protein that spans the periplasm (HlyD), and the TolC outer-membrane
protein [for a review see Dobrindt & Hacker (22)]. Other examples of proteins se-
creted by type I systems include the B. pertussis adenylate cyclase, the Pasteurella
haemolytica leukotoxin, and proteases of P. aeruginosa and Erwinia chrysanthemi.
Type II secretion systems are exemplified by the pullulanase secretion system
of Klebsiella oxytoca, which requires 14 proteins, most of which are located in
the inner membrane [reviewed by Lory (65) and by Pugsley et al (85, 86)]. The
outer-membrane protein PulD forms a pore in the outer membrane, through which
the type II-secreted proteins are believed to pass. PulD is a homolog of the YscC
family of proteins, which are involved in type III secretion, and homologs of
other components of the pullulanase system are involved in secretion of type 4
pili of V. cholerae and other organisms. Type II systems are primarily involved
in extracellular secretion of degradative enzymes, although some toxins are also
secreted via this mechanism. Proteins secreted via type II systems include cel-
lulase and pectinase enzymes of Erwinia, elastase, phospholipase C, and toxin A
of P. aeruginosa, amylase and protease enzymes of Aeromonas hydrophila, and
chitinase, protease, and cholera toxin of V. cholerae. Although the genes encoding
the type II secretion apparatus are usually clustered, the genes encoding proteins
secreted by the apparatus are frequently unlinked to these clusters and in one case
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

654 HACKER ¥ KAPER

may not even be on the same chromosome. In V. cholerae, the Eps type II secre-
tion system is encoded on the large chromosome while a protein secreted by this
system, the Hap protease, is encoded on the small chromosome (105). Type II
systems can be found on mobile genetic elements as well as on the core genomes
of nonpathogenic members of the same species. For example, the pO157 viru-
lence plasmid of EHEC encodes an apparently intact type II system (8), but the
proteins secreted by this system and the contribution of this system to virulence
of EHEC are unknown. A cryptic yet apparently intact type II system is also
encoded in the E. coli K-12 chromosome (27), but the proteins secreted by this
system and the environmental conditions necessary for this system to function are
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

not known.
The type V secretion system encompasses a group of proteins known as auto-
transporters [reviewed by Henderson et al (42)]. This system has also been called
type IV in some publications. Examples of the autotransporters include the im-
munoglobulin A (IgA) protease of N. gonorrhoeae, pertactin of B. pertussis, the
vacuolating cytotoxin (Vac) of H. pylori, SepA of S. flexneri, and EspC of EPEC.
Rather than a cluster of genes encoding a multicomponent secretion apparatus, the
secreted protein and the secretion apparatus are encoded in a single open reading
frame. The autotransporters are exported from the cytoplasm via the sec pathway
with cleavage of an amino-terminal signal peptide. The C-terminal portion of the
protein then forms a β-barrel structure which inserts into the outer membrane and
serves as a pore for the passage of the mature protein. Some autotransporters, such
as pertactin, remain attached to the bacterial cell surface, while other members of
this class, such as IgA protease, are cleaved, thereby leaving the C-terminal por-
tion in the outer membrane and releasing the mature protein into the extracellular
milieu. Examples of autotransporters which are encoded on PAIs include EspC
of EPEC (52) and the IgA protease of N. gonorrhoeae (84). The SHI-1 (formerly
she) PAI of S. flexneri contains genes for two different yet related autotransporters,
SigA and Pic (formerly ShMu), within a 12-kb region (41, 87).

Invasins, Modulins, and Effectors

Some of the best characterized PAIs provide the host bacterium with the ability
to invade epithelial cells and/or modulate host cell activities. At least five PAIs
(named SPI-1 through SPI-5; Table 2), encoding various virulence phenotypes,
including epithelial cell invasion and macrophage apoptosis (SPI-1), intracellular
proliferation and systemic spread (SPI-2), intramacrophage survival (SPI-3 and
SPI-4), and intestinal fluid secretion and inflammation (SPI-5), have been de-
scribed for Salmonella [reviewed by Groisman et al (33)]. The best-characterized
Salmonella PAI, SPI-1, encodes a type III secretion system and several proteins
that are translocated into host cells via this type III system [reviewed by Galán
(31), Galán & Collmer (32), and Groisman et al (33)]. The SipB and SipC proteins
form part of the protein translocation system, allowing translocation of effector
proteins into epithelial cells. The SipA protein binds cellular actin and inhibits
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

PATHOGENICITY ISLANDS 655

depolymerization of actin filaments, thereby promoting localized actin cytoskele-

ton reorganization. The SopE protein, which is encoded outside SPI-I but secreted
via the SPI-1 type III secretion system (see above), initiates the invasion process
by activating small GTPases such as Rac and Cdc42, thereby leading to activation
of mitogen-activated protein kinase pathways and membrane ruffling. Another
translocated protein, SptP, inactivates Rac and Cdc42, thereby restoring normal
cell architecture after bacterial internalization (30). The SPI-2 island also encodes
a type III secretion system, but the proteins secreted by this system are not as well
defined as those secreted by the SPI-1 system. One protein translocated into host
cells by the SPI-2 secretion system, SpiC, was recently shown to inhibit cellular
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

trafficking in macrophages (106).

Two PAIs located on plasmids in Shigella and Yersinia spp. also encode proteins
that mediate internalization and other host cell responses. The type III secretion
system encoded on the Shigella virulence plasmid is essential for invasion of host
epithelial cells by this pathogen (73, 82, 104). Several Ipa proteins secreted by this
type III system are essential for invasion, although the exact functions of these
proteins are unknown. IpaB and IpaC are inserted into host cell membranes (3),
and IpaC has been specifically implicated in induction of actin polymerization and
filopodia formation (103). The IpaA protein binds to the focal adhesion protein
vinculin, resulting in F-actin depolymerization (5). The Yersinia Yop virulon pro-
tects the organism from macrophages by disrupting the phagocytic and signaling
functions of these cells and ultimately inducing apoptosis (15, 16, 48). This region
encodes a type III secretion system that translocates a number of effector proteins
into host cells. The YopH protein has a tyrosine phosphatase activity, and it in-
hibits phagocytosis and disrupts focal adhesions of HeLa cells. The YopE protein,
for which no enzymatic activity has been described, also inhibits phagocytosis and
disrupts the cytoskeleton of HeLa cells. The YpkA protein of Yersinia pestis and
Y. pseudotuberculosis, also known as YopO in Y. enterocolitica, is a serine/threonine
kinase whose target protein is as yet unidentified. The YopT protein induces a cy-
totoxic effect in cells, leading to disruption of the actin filaments and alteration
of the cell cytoskeleton, while mutants defective for production of YopM exhibit
reduced virulence in the mouse by an as-yet-unknown mechanism.
The LEE PAI of EPEC and EHEC encodes a type III secretion system that is es-
sential for formation of the attaching-and-effacing histopathology seen with these
pathogens [reviewed by Frankel et al (29) and by Kaper & colleagues (52, 81)].
The EspA, EspB, and EspD proteins are secreted via the type III system and are
believed to be part of a protein translocation complex with EspA, forming a fil-
amentous structure on the surface of the bacterium, and with EspB and EspD,
perhaps forming a pore in the host cell membrane (29, 109). The Tir protein is
translocated into host cell membranes via the type III system and serves as a recep-
tor for the LEE-encoded intimin adhesin (58). Tir may also serve to nucleate host
cell actin and to transmit signals to the host cell following intimin-Tir binding.
The human pathogen L. monocytogenes and the animal pathogen L. ivanovii
have the capacity to survive and replicate intracellularly in eukaryotic cells (61).
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

656 HACKER ¥ KAPER

The internalization of L. monocytogenes as well as of L. ivanovii is triggered by

membrane proteins termed internalins (InlA and InlB). The genes encoding the L.
monocytogenes-specific large internalins InlA and InlB are located in a particular
chromosomal region, but their expression depends on the presence of the PrfA
regulator, which is part of the PAI-like Prf virulence gene cluster present in the
genomes of both species. In addition to the large internalins InlA and InlB, Listeria
strains express so-called small internalins (InlC to InlI), which are secreted into
the medium and which also contribute to internalization and host cell specificity.
The genes encoding the small internalins of L. ivanovii are part of two PAIs termed
LIPI2 and the inlCD gene cluster. Whereas the latter is only 5 kb in size and is
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

linked to a tRNA cluster, LIPI2 is 18 kb and represents an unstable DNA region

exclusively located on the L. ivanovii genome.

Toxins
Toxin Genes on Mobile Genetic Elements
Many bacterial pathogens harbor plasmids or bacteriophages which encode impor-
tant toxins. Thus, the heat-labile and heat-stable enterotoxins of enterotoxigenic
E. coli, as well as pore-forming toxins of various enterobacteria, cytolysins of
enterococci, enterotoxins of S. aureus with superantigen activity, and neurotoxins
of pathogenic clostridia and Bacillus anthracis, are plasmid encoded [for a review
see Dobrindt & Hacker (22)]. In addition, the cholera toxin genes, the genes en-
coding diphtheria toxin, the Shiga toxin genes of pathogenic enterobacteria, the
C. botulinum neurotoxin, and various enterotoxins of S. aureus and Streptococcus
pyogenes with superantigen specificity are located on phages. In addition, unortho-
dox toxins, factors which modulate host signaling or which may act as effector
molecules in bacterial host-cell interactions, are encoded by plasmids and phages.
Because so many bacterial toxin genes are located on plasmids and bacteriophages,
it is not surprising that PAIs also often carry toxin-encoding genes.

Pore-Forming Toxins
The prototypes of pore-forming PAI-encoded toxins are α-hemolysins of UPEC
termed HlyA (22). These toxins are transported by a type I secretion system and
have the capacity to lyse erythrocytes and other eukaryotic cells following inser-
tion into the eukaryotic cell membrane. The genes encoding the type I transport
system, hlyB and hlyD, as well as a the hlyC gene encoding an HlyA-modifying
enzyme, are clustered on PAIs of UPEC and, to a minor extent, on bacterial plas-
mids. On PAIs, but not on plasmids, the hly gene cluster is often colocalized with
P-fimbrial genes and loci encoding cytotoxic necrotizing factor 1. The importance
of the hemolysin for UPEC is corroborated by the fact that particular strains often
carry two PAIs with two hly determinants on their genomes. Interestingly, the large
virulence plasmid of EHEC carries a hly gene cluster similar to the PAI-located hly
determinant of UPEC (8, 35). Another example of a pore-forming toxin encoded
by a PAI or PAI-like structures is listeriolysin O of pathogenic Listeria species.
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

PATHOGENICITY ISLANDS 657

Listeriolysin O is able to lyse erythrocytes and other cells in a cholesterol-dependent

manner but is also necessary for lysis of the vacuoles of eukaryotic cells before
L. monocytogenes spreads through the cytoplasm. The gene encoding listeriolysin
O is part of the so-called PrfA virulence gene cluster of L. monocytogenes and
L. ivanovii (also termed LIPI1), which carries virulence genes and the prf-specific
regulatory region (61). The prf locus exhibits some characteristics of PAIs but is
considered a chromosomal DNA segment rather than a typical PAI.

Proteases, Lipases, and Enterotoxins

Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org

Various bacteria of pathogenic and even nonpathogenic species and subtypes pro-
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

duce enzymes with proteolytic activities, many of which play a role in pathogenic-
ity. The corresponding genes are often located on PAIs or on PAI-like structures
of gram-positive bacteria. Thus, the prf virulence gene cluster of L. monocy-
togenes and L. ivanovii (LIPI1) carries the genes for two phospholipases ( plcA
and plcB) and a locus encoding a metalloprotease (mpl). In L. ivanovii, but not in
L. monocytogenes, another PAI (termed LIPI2) encodes a sphingomyelinase
(SmcL) (61). Furthermore, enterotoxigenic Bacteroides fragilis strains produce
another metalloprotease, termed fragilysin. This enterotoxin causes fluid accumu-
lation in lamb ligated ileal loops (115). The corresponding gene, bft, is inserted
into the B. fragilis chromosome on a small (6-kb) DNA segment, termed the BfPAI,
that appears to be of plasmid origin, suggesting a potential mechanism for intro-
duction into this species (28). In addition, several other proteolytic proteins with
enterotoxic activity have been found to be encoded on PAIs. The EspC protein of
EPEC, encoded within a 15-kb PAI that is present in some but not all lineages of
EPEC, is an autotransporter that increases the short-circuit current in rat jejunal
tissue (52). Related proteins with similar functions are the EspP protein, encoded
on the EHEC virulence plasmid, and the Pic protease/mucinase of S. flexneri SHI-
1, which degrades gelatin and mucin. It is interesting that another enterotoxin
of Shigella, ShET1, which causes fluid accumulation in rabbit ileal loops, is also
encoded in the SHI1 PAI of S. flexneri by two open reading frames that are on the
DNA strand opposite that encodes the Pic protease/mucinase (26, 87).

Toxins with Other Enzymatic Activities

Many toxins have the capacity to modify host cell components such as small G
proteins or elongation factors. While some of these toxins (e.g. cholera toxin
and diphtheria toxin) are encoded by bacteriophages, others are known to be
encoded by PAIs. Thus, the gene encoding cytotoxic necrotizing factor 1 is part
of PAIs of UPEC, where it is often located next to hly and pap gene clusters.
cytotoxic necrotizing factor 1 has the capacity to modify the RhoA protein, a small
GTPase, by deamidation. Interestingly, the cnf2 gene, encoding a toxin of similar
function, is part of a plasmid of enterotoxigenic E. coli (22, 35). Furthermore,
the genes encoding the enterotoxins of C. difficile, TcdA and TcdB, which exhibit
glycosyltransferase activity and also modify GTPases, are part of the pathogenicity
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

658 HACKER ¥ KAPER

locus, a DNA segment of 19 kb of the C. difficile genome which exhibits some, but
not all, features of PAIs (23). In addition, the pertussis toxin gene cluster ptx of
B. pertussis has features of PAIs. ptx encodes a very complex toxin which ADP-
ribosylates small G proteins. Another toxin with enzymatic activity is the Shiga
toxin (encoded by stx), which specifically cleaves the 28S eukaryotic ribosomal-
RNA molecule. Whereas in pathogenic E. coli and other enterobacterial species
the stx genes are part of phage genomes, in Shigella dysenteriae the stx locus is
located on the chromosome. The stx region exhibits features of integrated phages
which in turn could be considered as PAIs or PAI-like structures (71).
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

Superantigens
Superantigens have the capacity to bind to the T-cell receptor molecule and in turn
activate these cells in a nonspecific manner. Many of these toxins are produced
by S. pyogenes and S. aureus, in which the corresponding genes are often located
on bacteriophages or plasmids. The toxic shock syndrome toxin (TSST)-1 toxin
of S. aureus, however, is part of a 15-kb PAI (SaPI) which represents a defective
bacteriophage. The SaPI can be mobilized by helper phages and transferred to
other strains (64). Similar PAIs have been found in different S. aureus isolates,
indicating that the occurrence of SaPIs is a general phenomenon in S. aureus.

Iron Uptake Systems

Survival and multiplication of microbes in certain ecological niches depend on
the ability of these organisms to scavenge essential nutrients such as iron. For
pathogenic bacteria the acquisition of iron is a prerequisite for the infectious pro-
cess, but nonpathogenic organisms also need sufficient iron ions (74). At least two
different strategies are used by bacteria to meet their iron needs: the expression of
receptors for iron carriers (e.g. heme, hemoglobin, lactoferrin, and transferrin) and
the synthesis and secretion of siderophores, which are low-molecular-weight, high-
affinity iron-binding compounds. The gene clusters for two different siderophore
systems, yersiniabactin and aerobactin, are part of PAIs, while a hemin uptake
system of enterobacteria is part of an islet.
The siderophore system of the phenolate type termed yersiniabactin (Ybt) is
encoded by the so-called HPI first described for the pathogenic Yersinia spp.
Y. pestis, Y. pseudotuberculosis, and Y. enterocolitica (7). The Ybt system, how-
ever, is also encoded by nonpathogenic E. coli, by defined E. coli pathotypes
such as enteroaggregative E. coli (94), non-O157 EHEC (57), and extraintestinal
E. coli, as well as by other enterobacteria such as Citrobacter spp. and S. enterica
serotype IIIb (94; T Ölschläger, T Zhang, E Carniel, J Heesemann, W Rabsch, H
Tschaepe, H Karch, J Hacker, unpublished data). The HPI consists of a core ele-
ment, common to all HPI-positive species, comprising 12 yersiniabactin-encoding
genes and additional loci at the right end which are specific for particular species
and pathotypes. Y. pestis and Y. pseudotuberculosis also carry hemin storage locus
genes, which are located next to the HPI sequences but which are not part of the
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

PATHOGENICITY ISLANDS 659

island. It is interesting that the HPI is located next to one of three asparagine tRNA
loci. In Y. pseudotuberculosis the island has the capacity to move from one tRNA
locus to the other (7).
Aerobactin has long been known as an important hydroxamate iron uptake
system of pathogenic and nonpathogenic E. coli, in which the corresponding gene
cluster (termed aer or iut) may be part of large virulence plasmids which also
encode the colicin ColV. In S. flexneri, however, the aer (iut) locus is part of a 23-
to 30-kb PAI which is located next to the selC locus. The island, termed Shigella
PAI 2 (SHI-2), is unstable, and in species others than S. flexneri (e.g. S. boydii and
E. coli) the aerobactin PAI seems to be located at a chromosomal site different
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

from that of selC (80, 108). The finding that the aer determinant of the Shigella
PAI is linked to sequences with homology to ColV-specific genes confirms the
view that the PAI may be generated following integration of plasmids encoding
the aerobactin system as well as colicin biosynthesis. Whereas the two siderophore
systems yersiniabactin and aerobactin are encoded by PAI genes, the hemin uptake
system chu (for E. coli hemin uptake) or shu (for Shigella hemin uptake) is encoded
by a 5-kb DNA fragment located at position 78 on the E. coli linkage map. This
region is considered an islet rather an island, but nevertheless the chu (shu) gene
products seem to be important for the pathogenesis of EHEC and Shigella (116).
Iron uptake systems are encoded by pathogenic as well as nonpathogenic mem-
bers of the same species; however, they are more prevalent in pathogenic strains.
This supports the hypothesis that iron uptake systems, as part of genomic islands in
nonpathogenic strains, may contribute to the fitness of these organisms as well as
their adaptability to particular environments, whereas in pathogenic bacteria these
systems contribute to virulence. Accordingly, in nonpathogenic strains such DNA
fragments are considered fitness islands while in pathogenic strains they represent
PAIs.

Other Virulence-Associated Genes

O Antigen Synthesis
Although most novel pathogens arising by transfer of a PAI are thought to have
emerged hundreds, thousands, or millions of years ago, one novel pathogen with
a newly acquired (or at least newly recognized) PAI has emerged within the
past decade. Historically, V. cholerae strains associated with epidemic cholera
exclusively possessed the O1 lipopolysaccharide antigen. In 1992, epidemic
V. cholerae with a different O antigen, O139, emerged in India and Bangladesh.
The O139 clone arose from an O1 El Tor strain when a 22-kb region encoding the
O1 lipopolysaccharide antigen was replaced by a 35.8-kb region containing the wbf
genes, which are responsible for synthesis of the O139 polysaccharide [reviewed
by Karaolis & Kaper (55)]. Like PAIs, the O139 wbf region has a G+C content
(average, 42%) lower than that of the genome (47.5%), it contains an IS, and it
is inserted at a specific chromosomal site (99). The substitution of the O139 wbf
genes for the O1 genes had a dramatic effect on the virulence of V. cholerae and the
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

660 HACKER ¥ KAPER

epidemiology of cholera in India and Bangladesh. Before 1992, the majority of

cholera cases were in individuals <15 years of age, because substantial immunity
is acquired by adulthood in countries in which V. cholerae is endemic. When the
O139 strain emerged, the majority of cases were in individuals >15 years old whose
immunity against O1 strains apparently did not protect against O139 strains (53).

Serum Resistance
PAI-encoded proteins have also been implicated in mediating serum resistance.
The sac-4 gene, encoding serum resistance to N. gonorrhoeae, was recently re-
ported to be contained on a PAI that is found preferentially in isolates from persons
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

with disseminated gonococcal infections rather than those from cases of uncompli-
cated gonorrhea (JP Dillard & HS Seifert, submitted for publication). This island
is not present in strain FA1090, the strain for which the N. gonorrhoeae genomic
sequence was recently obtained. The Pic protein of S. flexneri and enteroaggrega-
tive E. coli, which is encoded on the SHI-1 PAI of S. flexneri, confers serum
resistance on E. coli K-12 strains expressing this protein (41).

Immunoglobulin A Proteases
Proteases of nonenterobacterial species with homology to the degradating pro-
teinase toxins of the Pic family (see section on toxins above) have the capacity
to cleave IgA1 molecules. For the enterobacterial IgA-like protease family (Tsh,
EspC, EspP, and Pic), an IgA-degrading activity has not been demonstrated so far
(41, 42). Other pathogenic organisms, such as Haemophilus influenzae, N. gon-
orrhoeae, Neisseria meningitidis, and Streptococcus pneumoniae, synthesize IgA
proteases, while commensal strains of these species do not show this activity. The
corresponding genes, however, have not formally been described as PAIs.

Apoptosis
Several PAI-encoded proteins that are secreted via type III secretion systems
have been shown to induce apoptosis in host cells. The SipB protein, encoded on
the Salmonella SPI-1 island, and the IpaB protein, encoded on the Shigella viru-
lence plasmid, induce apoptosis in macrophages by binding to caspase-1 (43, 44).
The Y. pseudotuberculosis YopJ and Y. enterocolitica YopP proteins, which are
homologous proteins encoded on the Yersinia virulence plasmids, also induce
apoptosis in macrophages (75, 77). These Yops exhibit a high degree of similarity
to AvrRxv from the plant pathogen X. campestris. AvrRxv, which mediates a
programmed cell death pathway in plant cells, is secreted by the type III secretion
system encoded in the hrp gene cluster (92), a PAI of plant pathogens.

Capsule Synthesis
Many pathogenic bacteria have the capacity to synthesize capsules; so far, how-
ever, the genomic locations of the corresponding genetic determinants have not
been carefully described. The capsule synthesis locus of extraintestinal E. coli
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

PATHOGENICITY ISLANDS 661

represents an exception, because in one particular E. coli K1 strain the 20-kb seg-
ment could be mapped next to the pheV tRNA locus. It is still uncertain whether
the kps gene cluster itself forms a PAI as suggested recently (13). A similar ob-
servation was made for the locus encoding the exopolysaccharide PAI of a biofilm-
positive subgroup of S. epidermidis. The corresponding genes, termed ica, are also
located on a particular DNA segment of >100 kb which is absent in biofilm-
negative strains. Further studies will be neccessary to answer the question of
whether the ica region can be considered a PAI (118; I Lösner, J Hacker,
W Ziebuhr, unpublished data).
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

Plant Tumorigenesis
The Ti virulence plasmid of A. tumefaciens encodes the ability to cause crown
gall tumors in higher plants. As noted above, the Ti plasmid encodes a type IV
secretion system (encoded by the virBDE regions) that transfers one, two, or three
fragments of DNA into host plant cells [reviewed by Winans et al (113)]. The
transferred DNA, which possesses characteristics more typical of plant DNA than
of bacterial DNA, encodes proteins that produce plant hormones such as auxin
that cause neoplastic growth of transformed plant cells. Another group of proteins
encoded by transferred DNA directs the synthesis by plant cells of opines, which
are in turn used by the bacterium for nutrition. The bacterial proteins required for
the uptake, degradation, and utilization of opines are encoded by other genes on
the Ti plasmid (e.g. occ, mop, and aga).

Genes and Gene Products of Unknown Function

Among the emerging DNA sequence data for partly or entirely sequenced PAIs
of various bacteria, many intact open reading frames which may encode putative
new virulence factors have been detected. It has been suggested that a functional
analysis of these genes and their products will certainly lead to the discovery of
new virulence strategies used by PAI-positive bacterial pathogens (35, 83). Ad-
ditionally, in the genomic islands of the gram-negative pathogen Dichelobacter
nodosus have been identified the vap and vpl regions, which exhibit many features
of PAIs. The contribution of the vap and vpl regions and their products to virulence
of the bacteria, however, is not yet clear. The DNA sequences of particular open
reading frames identified on the D. nodosus genomic islands show strong homolo-
gies to toxin, adhesin, and enzyme gene clusters of other pathogenic organisms,
and it has been suggested that the corresponding products may play a role in the
pathogenesis of this important veterinary pathogen (12).

REGULATION OF PATHOGENICITY ISLANDS

PAIs frequently contain genes encoding regulators of virulence factor genes located
on the same island. However, various arrangements of regulators and regulatory
factors are possible, including PAI-encoded regulators that are specific for PAI
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

662 HACKER ¥ KAPER

genes, PAI-encoded regulators that also regulate genes located outside the PAI,
and regulators encoded outside the PAI that regulate genes encoded on the PAI.
The last class includes regulators, present only in pathogens, which regulate
primarily virulence factors genes as well as regulators, present in both pathogenic
and nonpathogenic members of the species, that may also regulate housekeep-
ing genes in addition to virulence factors. The best-characterized regulatory sys-
tems of PAIs feature a regulatory cascade in which PAI-encoded regulators of
PAI-encoded virulence genes are themselves regulated by systems encoded out-
side the PAI.
Two major classes of regulators that are encoded on PAIs or that specifically
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

regulate PAI genes are AraC-like proteins and two-component response regulators.
Other classes of regulators include alternate sigma factors and histone-like proteins.
A few of the better-characterized PAI regulators are reviewed to illustrate the
spectrum of regulatory systems, but for many PAIs, little information concerning
regulation is currently available.
The VPI of V. cholerae contains at least three regulatory genes, toxT, tcpP, and
tcpH. ToxT is a member of the AraC family of transcriptional activators (21) and
activates transcription of the tcp genes, which code for the major virulence factor
encoded on the VPI, TCP. ToxT also regulates the ctx genes, encoding cholera
toxin, which are located on a filamentous phage inserted outside the VPI. ToxT
is itself regulated by TcpP and TcpH, which are cytoplasmic membrane proteins
encoded on the VPI (40). Transcription of toxT is also regulated by the ToxR and
ToxS proteins, which are encoded outside the VPI. ToxR and ToxS are present
in nonpathogenic strains of V. cholerae, as well as in other Vibrio spp., and they
regulate the OmpU and OmpT outer-membrane proteins of V. cholerae, which
are the major porins of this species. Thus, the ToxRS regulon, which controls
expression of housekeeping genes in V. cholerae, has been adapted by the toxT
regulator encoded on the VPI. [Expression of ctx and tcp is also regulated by other
factors encoded outside the VPI, such as cAMP receptor protein (CRP) (98)]. It is
interesting that TCP and ToxRS are encoded on the larger of the two V. cholerae
O1 chromosomes, whereas in the classical biotype, ctx genes are contained on both
the larger and the smaller chromosomes (105), an unusual example of prokaryotic
trans-chromosomal regulation.
Another well-characterized example of PAI genes regulated by proteins en-
coded within and outside of the PAI is the SPI-1 island of Salmonella [reviewed
by Cotter & Miller (17), Groisman et al (33), and Hueck (46)]. The HilA and
InvF proteins are encoded on SPI-1 and belong to the OmpR/ToxR and AraC
families of transcriptional activators, respectively. HilA regulates InvF, which in
turn regulates expression of the sip genes, encoding proteins secreted by the type
III secretion system, located on SPI-1 (19). HilA also directly regulates expres-
sion of the components of the type III system apparatus. The hilA gene is itself
regulated by proteins encoded both inside and outside SPI-1. The HilC and HilD
proteins, encoded on SPI-1, are members of the AraC family of regulators that
derepress expression of hilA (93). The PhoP and SirA proteins are encoded outside
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

PATHOGENICITY ISLANDS 663

SPI-1 and also modulate expression of hilA. PhoP is the response regulator of the
PhoP/PhoQ two-component regulatory system which responds to environmental
Mg2+ ion concentrations and acts as a negative regulator of HilA. SirA is the
response protein of another two-component system for which the cognate sensor
protein has not yet been identified. In contrast to the negative effect of PhoP, SirA
acts as a positive regulator of HilA expression. Environmental signals such as
pH, osmolarity, oxygen, and Mg2+ concentration are detected by such global reg-
ulatory systems, which then regulate expression of the SPI-1-encoded virulence
factors through HilA, supercoiling, or other mechanisms which have yet to be
characterized (17, 33).
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

The plasmid-encoded Yop virulon of pathogenic Yersinia spp. is regulated by

a global regulator called VirF (LcrF in Y. pestis and Y. pseudotuberculosis), which
belongs to the AraC family [reviewed by Cornelis et al (16)]. VirF directly binds
to the promoter regions of the several operons in the Yop virulon and is itself
strongly regulated by temperature, being expressed at 37◦ C but not at 26◦ C. This
temperature regulation involves a chromosomally encoded protein called YmoA
which resembles histone-like proteins and acts as a negative regulator of virF.
An additional level of control may involve feedback inhibition by a repressor(s)
(LrcQ in Y. pseudotuberculosis and YscM1 and YscM2 in Y. enterocolitica) that
is normally expelled via the type III secretion system and is similar to the secreted
anti-sigma factor involved in regulation of flagellum synthesis. In addition, another
regulator of pathogenic yersiniae, YbtA, is encoded by the HPI. YbtA belongs to
the AraC class of regulatory elements and seems to be involved in the expression
of both HPI-specific genes and determinants located outside the HPI (88).
The LEE PAIs of EPEC and EHEC encode a transcriptional activator of the
H-NS family called Ler which controls expression of the type III secretion system
and the Tir/intimin receptor/adhesin (72). Ler also controls expression of fimbriae
encoded outside the LEE (24) and is itself regulated by integration host factor (IHF)
and quorum sensing (100). The quorum-sensing regulatory system is encoded in
pathogenic as well as nonpathogenic E. coli strains, but the only E. coli genes so
far known to be regulated by quorum sensing are the LEE genes. EHEC is notable
for the low infectious dose required for disease (∼100 organisms), and it has been
proposed that the quorum-sensing system of EHEC allows the organism, in the
presence of high concentrations of coliforms in the colon, to induce expression of
the intimin and Tir proteins necessary for intestinal colonization (100). Regulation
of the EPEC LEE is not identical to that of the EHEC LEE, and variations include
the presence of an AraC homolog (PerA), encoded on the large virulence plasmid
of EPEC, which regulates expression of the EPEC, but not the EHEC, Ler and the
transcription of some LEE genes of EHEC, but not EPEC, by the RpoS alternate
sigma factor (100). UPEC strains carry fimbrial adhesin determinants that consist
of structural genes and two regulatory loci, which in the case of P fimbriae were
termed PapB and PapC. The Prf fimbrial operon located on PAI II of strain 536
also encodes such regulators which activate the S-fimbrial determinant (sfa) in
addition to the P-fimbrial genes. The sfa determinant is located on another PAI.
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

664 HACKER ¥ KAPER

The process in which one group of PAI-located genes is regulated by regulators

encoded by another PAI has been termed cross-talk activation (79).
An alternate sigma factor is also involved in regulation of the hrp gene cluster in
the plant pathogen P. syringae which appears to be the plant pathogen equivalent of
the PAIs found in mammalian pathogens [reviewed by Hutcheson (47)]. The hrp
cluster encodes a type III secretion system that is regulated by four proteins encoded
on the hrp PAI. HrpL is an alternate sigma factor that is essential for transcription
of all genes of the hrp regulon except hrpR, hrpS, and hrpL. The HrpL sigma factor
is similar to σ 28, which controls expression of genes involved in biosynthesis of
flagella. Transcription of hrpL itself is dependent on the σ 54 alternate sigma factor
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

and is positively regulated by the HrpR and HrpS proteins. The HrpR and HrpS
proteins both exhibit similarity to response regulators of two-component systems
such as NtrC but they do not appear to be typical two-component regulators,
because they lack the N-terminal domain that functions in typical two-component
regulators to modulate activity through phosphorylation. A fourth protein, HrpV,
negatively regulates genes in the hrp regulon by an unknown method, perhaps via
HrpR and HrpS. Homologs of the regulatory components HrpL, HrpS, and HrpV
have also been implicated in regulation of type III secretion systems on PAIs in
Erwinia strains that are pathogenic for plants.
The examples presented above represent only a sampling of the various regu-
lators and regulatory mechanisms found in PAIs. Similar regulatory mechanisms
are also seen with genes that are not encoded in PAIs. However, one recently
described regulatory mechanism, involving translation of rare tRNAs, may be
specific for regulation of genes encoded on PAIs. This mechanism, defined by the
so-called minor codon hypothesis, is described below in the section on the role of
tRNAs.

JUNCTION SITES OF PATHOGENICITY ISLANDS

PAIs represent distinct DNA segments which differ from the core genome by
several features (see above). Therefore, the boundary regions of the majority of
PAIs have a specific DNA composition, preferentially of DR DNA sequences.
As indicated in Figure 1, one of the two PAI junction sites is very often part of
the 30 end of tRNA genes or of similar gene loci encoding the small regulatory
RNAs (e.g. ssrA for D. nodosus) (12) or the gene glr, encoding glutamate racemase
(e.g. Cag PAI of H. pylori) (10). The length of the DRs ranges from 9 bp (in UPEC
CFT073) to 135 bp (PAI in UPEC J96 and LEE PAI in particular EHEC strains).
In most cases the DRs comprise 16 to 20 bp (35).
The DR DNA sequences were probably generated following insertion of pre-
PAI elements such as bacteriophages or plasmids into the core genomes of the host
organisms. It should be mentioned here that the 30 ends of tRNA loci very often
act as attachment sites for the integration of bacteriophages. Thus, the strong as-
sociation between tRNA genes and PAIs may argue for a development of PAIs
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

PATHOGENICITY ISLANDS 665

from former bacteriophages (89; see also section on mobility genes on PAIs
below). In addition, the DR segments may act as target sequences for the ac-
tion of integrases/excisionases in deletion processes, since many PAIs have the
tendency to delete (see section on transfer and deletion of PAIs below). In most
cases these deletion processes are RecA independent, and it seems that particular
integrases, preferentially those encoded by PAIs, may be involved in the deletion
processes. IS elements, in contrast, flank PAIs in very few cases (e.g. the PAI-like
element of the stx gene in S. dysenteriae) (71). IS elements or portions thereof
are, however, very often part of PAI segments. This is also true for Rhs sequences
and other repeated DNA segments which do not form junction sites for PAIs but
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

often act as targets for partial deletions of PAI-specific sequences (7).

THE ROLE OF tRNA LOCI

About 75% of the PAIs identified so far are associated with tRNA loci (see Table 2).
It has been known for many years that tRNA genes may act as landmarks for
the integration of foreign DNA, either of plasmids and phages or, in the case
of eukaryotes, of retroviruses (45, 89). It is interesting that extrachromosomal
genetic elements themselves often carry tRNA genes or parts of them. This is true
for plasmids of streptomycetes but also for bacteriophages of enterobacteria, such
as the T4 phage or the phage carrying Shiga toxin genes. In addition, in eukaryotes,
mitochondrial as well as plastidic DNA molecules harbor tRNA loci. It is therefore
suggested that the extrachromosomal elements integrate into the chromosomes of
their host organisms by homologous recombination between tRNA genes of the
chromosome and their extrachromosomal counterparts, thereby keeping the tRNA
loci intact. It seems that the 30 ends of the tRNA genes may preferentially act as
targets for integration. DNA sequence data for many PAIs and their associated
tRNA genes show that overlaps of 15–25 nucleotides exist between the 30 ends of
the tRNA loci and PAI-specific sequences (45).
Although in principle many tRNA genes can be used as integration sites for PAIs
and other extrachromosomal elements, in praxi particular preferences exist. The
tRNA genes frequently associated with PAIs (see Table 2) also quite often harbor
attachment sites for bacteriophages, indicative of the development of PAIs from
former bacteriophage genomes. One of the tRNA loci most frequently associated
with PAIs is the selC gene, encoding the selenocysteine-specific tRNAsec. In
enterobacteria the bacteriophage φR73 and five different PAI elements can be
located next to selC (35). Three different E. coli-specific PAIs, leading to different
E. coli pathotypes, as well as PAIs of S. enterica and S. flexneri are associated with
the selC locus. In addition, the two identical phenylalanine-specific tRNA genes
pheV and pheU (also termed pheR) often act as targets for PAI formation. The idea
that PAIs may represent a particular subtype of genomic island is corroborated by
the fact that not only three different PAIs but also two genomic islands, encoding
an alternative sucrose uptake system in S. enterica and the nitrogenase system in
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

666 HACKER ¥ KAPER

Mesorhizobium loti, are located next to the pheR and pheU loci (37; see Tables
2 and 3). In addition, tRNA-associated PAIs also have the capacity to be deleted
from the genomes of their respective pathogens, leading to a truncation of the
tRNA-specific genes. In the case of UPEC strain 536, the tRNA genes leuX and
selC, located next to PAI I and PAI II, became truncated at their 30 ends after
deletion of the two PAIs, which consequently leads to tRNA-specific mutations. It
therefore seems that only tRNA genes with another identical locus in the genome
(e.g. phenylalanine-specific tRNAs), genes which encode tRNAs with wobble
capacity (e.g. leuX ), or nonessential tRNAs (e.g. selC ) are frequently used as
integration sites for PAIs.
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

Many authors have suggested that the presence of particular tRNAs that are
less frequently used than others (so-called minor tRNAs) may have a modulatory
effect on the translational efficiency of particular target genes (the minor codon
hypothesis). The evolutionary advantage of extrachromosomal genetic elements
carrying additional tRNA genes may be a more effective expression of genes with
a larger amount of the corresponding codon located on the respective element.
Thus, the genes present on PAI II of UPEC strain 536 that are located next to
the leuX tRNA have a significantly larger number of leuX-specific codons than
genes located on the rest of the core genome. Therefore, it was not surprising
that leuX mutants were less virulent than their leuX-positive counterparts, because
many genes located on PAI II, such as that encoding α hemolysin, and also other
genes were less frequently expressed (90). For type I fimbriae, whose expression

TABLE 3 Examples of genomic islands

Organism Property Type of island Genetic feature Size (kb)

Mesorhizobium loti Nitrogen fixation Symbiosis Integrated plasmid 500

Pseudomonas putida Chlorocatechol Degradation Integrated plasmid 105
degradation
Salmonella senftenberg Sucrose uptake Metabolism Conjugative 100
transposon
Staphylococcus aureus MecA protein Resistance Location on 52
chromosome
Salmonella Multiresistance Resistance Location on 14
typhimurium DT 104 chromosome
Various bacteria Type III secretion Secretion Location on Variable
Type IV secretion chromosome
or plasmid
Various bacteria Iron uptake Fitness Location on Variable
chromosome
or plasmid
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

PATHOGENICITY ISLANDS 667

is also affected by the presence of leuX, it was shown that the translation of the
fimB-specific positive regulator gene, which harbors an unusually large number of
leuX codons, is strongly affected by the presence of the leuX-specific tRNA5Leu.
It is interesting that the leuX gene itself is regulated by global regulators such as
the alternative sigma factors RpoS and RpoH, providing additional support for
the minor codon hypothesis (U Dobrindt & J Hacker, unpublished data). Besides
leuX, other minor tRNA genes, such as argU and leuZ, influence the expression of
particular virulence-associated genes such as the fimbrial genes of S. enterica.
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

MOBILITY GENES ON PATHOGENICITY ISLANDS

A variety of mobility genes are found on PAIs, particularly genes encoding phage
integrases and ISs/transposons. However, the presence and functionality of these
mobility genes can differ greatly among PAIs, from genes that encode intact and
functional integrases to mutated genes that encode nonfunctional integrases to PAIs
that contain no apparent integrase genes. These mobility genes and the protein
products they encode can lead to the formation, integration, deletion, and mobility
of PAIs as described below in the section on transfer and deletion of PAIs.
Several PAIs contain sequences that show homology to genes encoding in-
tegrases of bacteriophage P4 and retronphage φ73. P4 integrase sequences are
present in the HPI of Yersinia, a PAI which is also found in enteroaggregative
E. coli (94), some Shiga toxin-producing E. coli (STEC) serotypes (57), and UPEC
(PAI IV536) (94). This integrase appears to be intact and functional in the majority
of strains, in contrast to the P4 integrases encoded in PAI II536 of UPEC and in
E. coli K-12, both of which appear to be defective due to the presence of premature
stop codons. Integrases that are more closely related to the integrase of φ73 are
found in PAI I536 of UPEC, the SHI-2 PAI of Shigella species (108), and in the vap
(for virulence-associated protein) regions of D. nodosus (12, 35). The integrase
present in the V. cholerae VPI most closely resembles that of the cryptic bacterio-
phage CP4-57 (53% amino acid identity) but also exhibits substantial homology
with the P4 integrase (32% identity) and the D. nodosus vap-encoded integrase
(34% identity). The substantial homology among integrases from different genera
suggests that these sequences could be transferred among a wide range of bacterial
species.
The integrases and other bacteriophage proteins encoded on PAIs are usually
assumed to have played a role in the original mobilization of the PAI, but there
are some phages that appear to have been inserted into the PAI after the origi-
nal mobilization of the PAI into the chromosome. One such phage is the 933L
prophage of the EHEC O157:H7 LEE, which is absent from the closely related
LEE of EPEC E2348/69 (83). Similar phages may have nevertheless played a role
in the mobilization of other PAIs, because an integrase gene nearly identical to
that of 933L is also present in the Shigella SHI-2 PAI (108).
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

668 HACKER ¥ KAPER

Numerous IS elements and transposon sequences can be found in PAIs. In

some cases, such elements may have played a role in the original mobilization
or subsequent rearrangement of the PAI, or they may have inserted afterward and
played little or no role in PAI formation. The insertion of IS605 into the cag PAI of
H. pylori appears to have played no role in the original mobilization of this PAI
but has played a major role in its subsequent evolution (18). The SHI-2 PAI of
S. flexneri, which encodes the aerobactin iron transport system, contains numer-
ous partial and complete IS elements, including IS3, IS629, IS600, IS2, and IS1
(80, 108). Comparison of SHI-2 islands from S. flexneri strains SA100 (serotype
2a) and M90T (serotype 5a) revealed essentially identical insertion sites, aerobactin
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

genes, and IS elements on the left side of the PAI but substantial differences in
IS elements and other sequences on the right side of the island. The SHI-2 PAIs
in Shigella sonnei and Shigella boydii have the same insertion site and aerobactin
genes but exhibit major differences in the IS elements on their left sides.
In addition to mediating rearrangements of PAIs over long time frames (macro-
evolution), IS elements can also mediate important changes over very short time
frames, as is seen in the phase variation of the polysaccharide intercellular adhesin
of Staphylococcus epidermidis, which is a microevolutionary process. Alternating
insertion and excision of IS256 into different sites of the S. epidermidis ica gene
cluster controls expression of the polysaccharide intercellular adhesin, which is a
major component of staphylococcal biofilms (118).

TRANSFER AND DELETION

OF PATHOGENICITY ISLANDS
The VPI of V. cholerae was recently shown to be mobilizable from one strain of
V. cholerae to another by transduction (56). This island encodes several proteins
whose predicted sequences are similar to those of proteins of bacteriophages or
eukaryotic viruses as well as a protein that is highly homologous to the E. coli
cryptic prophage (CP4-57) integrase (60). Furthermore, the VPI is flanked by
20-bp DR sequences that resemble phage att sites (56). Culture supernatants of
VPI-positive V. cholerae strains contained sequences present on the VPI, as shown
by polymerase chain reaction, but did not contain sequences of other chromosomal
genes located outside the VPI (56). Cell-free phage preparations from a strain
marked by insertion of a kanamycin resistance gene into the VPI were found to be
capable of transferring the marked VPI into a previously VPI-negative recipient
strain. These results suggest that transfer of the VPI via phage transduction could
play an important role in the evolution of pathogenic V. cholerae strains from
nonpathogenic strains.
The staphyloccocal PAIs (SaPIs) encode the TSST, a property which has been
previously shown to be transferable. It became clear from studies by Novick &
coworkers that the SaPI is part of a defective bacteriophage of 15 kb which has the
capacity to be excised and circulated by helper phages such as φ13 and φ80α (64).
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

PATHOGENICITY ISLANDS 669

After excision, the islands are transduced to other staphylococcal strains with high
frequencies. Thus, the SaPI belongs to the group of mobile PAIs that may have
been derived from bacteriophages.
Phage transfer has been implicated in the transfer of other PAIs and PAI-like
structures. The genes responsible for the serotype 1a O antigen of S. flexneri
are encoded on a 5.8-kb region that is flanked by 45-bp att sequences and IS
elements (1). The overall G+C content of this area (40%) is significantly below
the 50% Shigella genome content. On one side of this region is a gene whose
predicted protein product shows significant homology to integrases of P22 and
other bacteriophages. This region is inserted at the thrW tRNA gene, and in fact
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

the 30 end of this gene is identical to the attR sequence. One final example is the
SopE protein of Salmonella, which is secreted via the type III secretion system
encoded on the SPI-1 PAI. The SopE protein is not encoded on SPI-I but instead
is encoded within a temperate bacteriophage belonging to the P2 family (76).
Certain PAIs do not have the capacity to move from one bacterial strain to
another but are, however, able to jump from one site to another in the genome of
a bacterial pathogen. The HPIs of pathogenic enterobacteria, encoding the iron
uptake system yersiniabactin in Y. pseudotuberculosis and Y. pestis, carry intact
integrase genes and are flanked by DRs of 17 kb (88). The HPIs in Yersinia are
usually located in the asparagine-specific tRNA gene asnT. In particular strains of
Y. pseudotuberculosis, however, the HPI first detected in asnT, over a short time
period, has also been found to be located in two other asparagine-specific tRNA
genes, asnW and asnV, indicating a transfer of the element from one target site to
another (7). It has been suggested that such mobility events of PAIs may also occur
during infections and may represent examples of PAI-induced microevolutionary
processes.
In addition, certain PAIs have a tendency for deletion of particular PAI-specific
sequences or of the whole genetic element. On one hand, such deletions may occur
over longer time intervals in order to optimize the structure of the PAI elements and
to reduce the genetic burden by eliminating genes whose products are not further
used (see also below and Figure 2). On the other hand, in contrast to the processes
of macroevolution, in many organisms PAIs show deletions in a short time range.
Such processes of microevolution have been detected for the Cag PAI of H. pylori,
which has a tendency to decrease in size over short time intervals (10). In addition,
the HPI and also the hemin storage locus of Y. pestis carry many Rhs elements
which may act as landmarks for frequent deletion processes (7). Furthermore,
in other organisms, such as UPEC, whole PAI elements can be deleted. In these
cases recombinases/integrases located on the corresponding PAI may be involved
in the deletion processes (4, 35). For PAI I and PAI II of E. coli 536, however,
the PAI-encoded recombinases are not intact, thereby suggesting that integrases
of other PAIs or encoded by the rest of the genome are probably involved in the
deletion processes. It is suggested that the deletion processes may play a role in the
adaptation of pathogenic microbes during certain stages of infection. Therefore,
genetic flexibility of pathogenic microbes may create selective advantages over
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

670 HACKER ¥ KAPER

Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

Figure 2 Evolutional stages of pathogenicity island formation. The vertical line represents
the evolutionary progress; the horizontal line represents a time scale. The thin line represents
parts of the core genome; the double line represents pathogenicity island-specific DNA. The
boxes indicate genes, and the asterisks indicate mutations. Abbreviations: tRNA, transfer RNA
genes; int, integrase gene; vir, virulence-associated gene; mob, mobility gene (e.g. gene encoding
transposase).

other, less-flexible organisms and may finally result in proper replication in host
organisms or other ecological niches.

CONCLUSION: Pathogenicity Islands, Genomic Islands,

and the Evolution of Microbes
The key processes of Darwinian evolution can be described by the terms genetic
variability, phenotype formation, and natural selection. There is no doubt that the
permanent generation of new genetic variants represents the main prerequisite for
evolution. The term macroevolution describes evolutionary processes which oc-
cur within a longer period of time and which lead to the formation of new species
or pathotypes (78). In contrast, microevolution, which generates new variants of
given species or pathotypes, takes only days or weeks. Both processes, macro- and
microevolution, are also involved in the generation and adaptation of pathogenic
microbes.
The majority of pathogenic bacteria show evidence of horizontal gene transfer,
which together with the generation of point mutations and genetic rearrangements
represents one major source for the creation of new genetic variants. The analysis
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

PATHOGENICITY ISLANDS 671

of the structure of PAIs revealed that these genetic elements have been generated
after lateral gene transfer processes. Thus, PAIs contributed to the development of
pathogenic variants, preferentially in macroevolution. The ongoing processes of
rearrangements, deletions, and transfer of PAIs, however, also have a strong impact
on microevolution and adaptation of pathogenic microbes during acute infectious
processes.
As indicated in Figure 2, PAIs can be considered genetic fossils owing to their
particular stages in the evolutionary process of pathogens. Integrated newly ac-
quired DNAs seem to represent, from the evolutionary point of view, the youngest
types of PAIs. The complete or incomplete phages which are identical to S. aureus
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

PAIs and the V. cholerae VPI encoding the TCP may represent PAIs at the begin-
ning of their career as evolutionary chronometers (56, 64). If a PAI significantly
contributes to the fitness and probably also the pathogenic potency of its host
organism the mobility genes involved in transfer, deletion, or excision will be
inactivated or even deleted. In other words, selection for stability and homing of
the formerly acquired DNA takes place. The HPIs of Yersinia, the PAIs of UPEC,
and the Cag island of H. pylori may represent such “middle-aged,” adolescent-
type PAIs (18, 88). Over time, for example periods of thousands of years, the sizes
of successful PAIs tend to be reduced to include just the important genes required
for their key functions, and the PAIs become more and more condensed (e.g. the
LEE island of EPEC) (52). Thus, the Salmonella SPI-1, which was introduced
into the genome of the common ancestor of S. enterica and Salmonella bongori
100 million years ago, is already part of the core genome of both organisms (33).
It is our assumption that there exist genetic segments which are former PAI regions
but already have been adapted to the core genome of the host and which are no
longer detectable as PAIs. In these cases the evolutionary progress has terminated.
Over the past few years, research on PAIs has attracted much attention. How-
ever, important questions concerning the development and selection of PAIs remain
unanswered. One major question is: Where do the PAI-specific DNA segments
come from? From our point of view it is very difficult to answer this question.
The overall number of prokaryotes in the soils, fresh waters, and oceans has been
estimated to comprise 150,000–200,000 different species; however, only ∼4000
have been described so far. The majority of the bacterial species are not culturable
in the laboratory, and therefore analyses of their genomes have been largely im-
possible. It has been suggested that the enormous amount of genetic information
present in the genomes of species not yet described may be a source for many ge-
netic elements, including PAIs. In the case of the Yersinia HPI it is suggested that
members of the Pseudomonadae may be donors for this DNA; however, this has
not yet been verified. The further development of microbial ecology will certainly
open new avenues to approach the important question of the source of PAI-specific
DNA in the future.
Another question is: What are the selective forces for the development of PAIs?
Pathogenic organisms as well as nonpathogenic microbes have to overcome envi-
ronmental restrictions for survival and replication and have to compete successfully
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

672 HACKER ¥ KAPER

in their ecological niches. These ecological processes, more so than the processes
of disease formation, are of evolutionary relevance. The occurrence of infectious
diseases does not necessarily contribute to the evolutionary fitness of microbes.
It is therefore suggested that from an evolutionary view point, many PAIs rep-
resent fitness islands because they were selected in the natural environment of
microbes to stimulate their survival and replication. Thus, P fimbriae of UPEC,
whose genes are components of PAIs, contribute to adherence of these microbes
in the gut (14). Iron uptake systems, of course, are necessary for the survival
of microbes in various niches. As the ecological niches and reservoirs of many
pathogens (e.g. pathogenic Shigella) are not yet known, it is not possible now to
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

fully answer the question on the selective pressures leading to PAI generation.
The third question of interest is: What are the host equivalents of PAIs? The
acquisition of new genetic material by the microbes, of course, may generate new
phenotypes which may interact with host structures. If the acquisition and genera-
tion of PAIs were to generate more-aggressive microbes, they would destroy their
hosts and would therefore be unsuccessful in evolutionary terms. On the other
hand, the generation of new relevant microbial phenotypes for host interaction
should also select for new host structures such as receptor variants or alterations
in signal transduction pathways. Such changes in host genes will be important
for overcoming hyperpathogenicity and for development of new strategies for an
appeasement of hosts and microbes. These host strategies are largely unknown
to date, and the study of host-pathogen coevolution at the genetic level will be a
major topic in the years to come. The development of DNA chips which represent
whole host genomes, including the human genome, will certainly help to address
these important questions in the future.
The accumulating sequence data for genomes of pathogenic as well as non-
pathogenic bacteria have shown that genetic structures similar to PAIs are also
part of the genomes of many nonpathogenic bacteria. These regions were termed
genomic islands because they were found to possess many features of PAIs, often
differing from the rest of the core genome in G+C content and coding usage,
and they often carry mobility genes (integrases and transposases). As with PAIs,
genomic islands are often be linked to tRNA genes and may be flanked by re-
peated DNA at their ends. Last, but not least, genomic islands may represent
genome regions of high instability. Many genomic islands, of course, do not
contribute to pathogenicity, and their gene products are rather involved in symbio-
sis, in degradation of xenobiotic compounds, in the generation of new metabolic
properties, or in the expression of antibiotic resistance phenotypes (6, 37, 49; for
details see Table 3). As already mentioned, iron uptake systems and bacterial
secretion pathways can be expressed either by pathogenic or by nonpathogenic
bacteria. In both cases they seem to contribute to fitness and adaptation of the
strains. It therefore appears that PAIs represent a genomic-island subgroup whose
members are, by definition, restricted to pathogenic bacteria. The analysis of
PAIs, however, not only opens new insights into the evolution of pathogens but also
contributes to the formulation of evolutional principles of prokaryotes in general.
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews AR110-19

PATHOGENICITY ISLANDS 673

ACKNOWLEDGMENTS
We thank Ute Hentschel (Würzburg) for helpful comments and Claudia Borde
(Würzburg) for editorial assistance. We apologize to the many colleagues whose
work we could not specifically cite due to space limitations. Our own work related
to this topic was supported by the Deutsche Forschungsgemeinschaft (SFB479,
Projekt A1) and the Fond der Chemischen Industrie as well as by the National
Institutes of Health (grants AI21657, AI19716, and AI41325 to JB Kaper).

Visit the Annual Reviews home page at www.AnnualReviews.org

Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

LITERATURE CITED
1. Adhikari P, Allison G, Whittle B, Verma NK.
1999. Serotype 1a O-antigen modification:
molecular characterization of the genes in-
volved and their novel organization in the
Shigella flexneri chromosome. J. Bacteriol.
181:4711–18
2. Anderson DM, Schneewind O. 1997. A
mRNA signal for the type III secretion of
Yop proteins by Yersinia enterocolitica. Sci-
ence 278:1140–43
3. Blocker A, Gounon P, Larquet E, Niebuhr
K, Cabiaux V, et al. 1999. The tripartite type
III secreton of Shigella flexneri inserts IpaB
and IpaC into host membranes. J. Cell Biol.
147:683–93
4. Blum G, Ott M, Lischewski A, Ritter A, Im-
rich H, et al. 1994. Excision of large DNA
regions termed pathogenicity islands from
tRNA-specific loci in the chromosome of an
Escherichia coli wild-type pathogen. Infect.
Immun. 62:606–14
5. Bourdet-Sicard R, Rudiger M, Jockusch
BM, Gounon P, Sansonetti PJ, Tran VN.
1999. Binding of the shigella protein IpaA to
vinculin induces F-actin depolymerization.
EMBO J. 18:5853–62
6. Briggs CE, Fratamico PM. 1999. Molecular
characterization of an antibiotic resistance
gene cluster of Salmonella typhimurium
DT104. Antimicrob. Agents Chemother.
43:846–49
7. Buchrieser C, Brosch R, Bach S, Guiyoule
A, Carniel E. 1998. The high-pathogenicity
island of Yersinia pseudotuberculosis can be
inserted into any of the three chromosomal
asn tRNA genes. Mol. Microbiol. 30:965–
78
8. Burland V, Shao Y, Perna NT, Plunkett
G, Sofia HJ, Blattner FR. 1998. The com-
plete DNA sequence and analysis of the
large virulence plasmid of Escherichia coli
O157:H7. Nucleic Acids Res. 26:4196–204
9. Burns DL. 1999. Biochemistry of type
IV secretion. Curr. Opin. Microbiol. 2:25–
29
10. Cesini S, Lange C, Xiang Z, Crabtree JE,
Ghiara P, et al. 1996. cag, a pathogenic-
ity island of Helicobacter pylori, encodes
type-I specific and disease-associated vir-
ulence factors. Proc. Natl. Acad. Sci. USA
93:14648–53
11. Cheetham BF, Katz ME. 1995. A role for
bacteriophages in the evolution and trans-
fer of bacterial virulence determinants.
Mol. Microbiol. 18:201–8
12. Cheetham BF, Whittle G, Katz ME. 1999.
Are the vap regions of Dichelobacter no-
dosus pathogenicity islands? See Ref. 51a,
pp. 203–18
13. Cieslewicz M, Vimr E. 1997. Reduced
polysialic acid capsule expression in Es-
cherichia coli K1 mutants with chromo-
somal defects in kpsF. Mol. Microbiol.
26:237–49
14. Connell H, Hedlund M, Agace W, Svan-
borg C. 1997. Bacterial attachment to uro-
epithelial cells: mechanisms and conse-
quences. Adv. Dent. Res. 11:50–58
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews
674 HACKER ¥ KAPER
15. Cornelis GR, Boland A, Boyd AP, Geui-
jen C, Iriarte M, et al. 1998. The virulence
plasmid of Yersinia, an antihost genome.
Microbiol. Mol. Biol. Rev. 62:1315–52
16. Cornelis GR, Wolf-Watz H. 1997. The
Yersinia Yop virulon: a bacterial system
for subverting eukaryotic cells. Mol. Mi-
crobiol. 23:861–67
17. Cotter PA, Miller JF. 1998. In vivo and ex
vivo regulation of bacterial virulence gene
expression. Curr. Opin. Microbiol. 1:17–
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.
26
18. Covacci A, Rappuoli R. 1999. cag, the
pathogenicity island of Helicobacter py-
lori, triggers host response. See Ref. 51a,
pp. 189–202
19. Darwin KH, Miller VL. 1999. InvF is re-
quired for expression of genes encoding
proteins secreted by the SPI1 type III secre-
tion apparatus in Salmonella typhimurium.
J. Bacteriol. 181:4949–54
20. Deleted in proof
21. DiRita VJ. 1992. Co-ordinate expression
of virulence genes by ToxR in Vibrio
cholerae. Mol. Microbiol. 6:451–58
22. Dobrindt U, Hacker J. 1999. Plasmids,
phages and pathogenicity islands: lessons
on the evolution of bacterial toxins. In The
Comprehensive Sourcebook of Bacterial
Protein Toxins, ed. J Alouf, J Freer, pp.
3–23. New York: Academic
23. Eichel-Streiber von C, Braun V. 1999.
Virulence-associated mobile elements in
bacilli and clostridia. See Ref. 51a, pp.
233–64
24. Elliott SJ, Hutcheson SW, Dubois MS,
Mellies JL, Wainwright LA, et al. 1999.
Identification of CesT, a chaperone for
the type III secretion of Tir in en-
teropathogenic Escherichia coli. Mol. Mi-
crobiol. 33:1176–89
25. Elliott SJ, Yu J, Kaper JB. 1999. The
cloned locus of enterocyte effacement
from enterohemorrhagic Escherichia coli
O157:H7 is unable to confer the attaching
and effacing phenotype upon E. coli K-12.
Infect. Immun. 67:4260–63
AR110-19
26. Fasano A, Noriega FR, Maneval DR Jr,
Chanasongcram S, Russell R, et al. 1995.
Shigella enterotoxin 1: an enterotoxin of
Shigella flexneri 2a active in rabbit small
intestine in vivo and in vitro. J. Clin. In-
vest. 95:2853–61
27. Francetic O, Pugsley AP. 1996. The cryptic
general secretory pathway ( gsp) operon of
Escherichia coli K-12 encodes functional
proteins. J. Bacteriol. 178:3544–49
28. Franco AA, Cheng RK, Chung GT, Wu S,
Oh HB, Sears CL. 1999. Molecular evo-
lution of the pathogenicity island of en-
terotoxigenic Bacteroides fragilis strains.
J. Bacteriol. 181:6623–33
29. Frankel G, Phillips AD, Rosenshine I,
Dougan G, Kaper JB, Knutton S. 1998.
Enteropathogenic and enterohaemorrhagic
Escherichia coli: more subversive ele-
ments. Mol. Microbiol. 30:911–21
30. Fu Y, Galan JE. 1999. A salmonella protein
antagonizes Rac-1 and Cdc42 to mediate
host-cell recovery after bacterial invasion.
Nature 401:293–97
31. Galán JE. 1999. Interaction of Salmonella
with host cells through the centisome 63
type III secretion system. Curr. Opin. Mi-
crobiol. 2:46–50
32. Galán JE, Collmer A. 1999. Type III se-
cretion machines: bacterial devices for
protein delivery into host cells. Science
284:1322–28
33. Groisman EA, Blanc-Portard AB, Uchiya
K. 1999. Pathogenicity islands and the evo-
lution of Salmonella virulence. See Ref.
51a, pp. 127–50
34. Hacker J, Bender L, Ott M, Wingender J,
Lund B, et al. 1990. Deletions of chro-
mosomal regions coding for fimbriae and
hemolysins occur in vivo and in vitro in
various extraintestinal Escherichia coli iso-
lates. Microb. Pathog. 8:213–25
35. Hacker J, Blum-Oehler G, Janke B, Nagy
G, Goebel W. 1999. Pathogenicity islands
of extraintestinal Escherichia coli. See Ref.
51a, pp. 59–76
36. Hacker J, Blum-Oehler G, Mühldorfer I,
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews
Tschäpe H. 1997. Pathogenicity islands of
virulent bacteria: structure, function and
impact on microbial evolution. Mol. Mi-
crobiol. 23:1089–97
37. Hacker J, Kaper JB. 1999. The concept of
pathogenicity islands. See Ref. 51a, pp. 1–
11
38. Hacker J, Knapp S, Goebel W. 1983. Spon-
taneous deletions and flanking regions of
the chromosomal inherited hemolysin de-
terminant of an Escherichia coli O6 strain.
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.
J. Bacteriol. 154:1145–52
39. Haller JC, Carlson SJ, Pierson DE. 1999. A
chromosomally-encoded type III secretion
apparatus of Yersinia enterocolitica affects
virulence in the mouse model of infection.
Abstr. ASM Conf. Cell Biol. Approach Mi-
crob. Pathog., Portland, OR, p. 25. Wash-
ington, DC: Am. Soc. Microbiol.
40. Häse CC, Mekalanos JJ. 1998. TcpP pro-
tein is a positive regulator of virulence gene
expression in Vibrio cholerae. Proc. Natl.
Acad. Sci. USA 95:730–34
41. Henderson IR, Czeczulin J, Eslava C, Nor-
iega F, Nataro JP. 1999. Characterization of
Pic, a secreted protease of Shigella flexneri
and enteroaggregative Escherichia coli. In-
fect. Immun. 67:5587–96
42. Henderson IR, Navarro-Garcia F, Nataro
JP. 1998. The great escape: structure and
function of the autotransporter proteins.
Trends Microbiol. 6:370–78
43. Hersh D, Monack DM, Smith MR,
Ghori N, Falkow S, Zychlinsky A.
1999. The Salmonella invasin SipB in-
duces macrophage apoptosis by binding
to caspase-1. Proc. Natl. Acad. Sci. USA
96:2396–401
44. Hilbi H, Moss JE, Hersh D, Chen Y, Aron-
del J, et al. 1998. Shigella-induced apopto-
sis is dependent on caspase-1 which binds
to IpaB. J. Biol. Chem. 273:32895–900
45. Hou YM. 1999. Transfer RNAs and
pathogenicity islands. Trends Biochem.
Sci. 24:295–98
46. Hueck CJ. 1998. Type III protein secretion
systems in bacterial pathogens of ani-
AR110-19
PATHOGENICITY ISLANDS 675
mals and plants. Microbiol. Mol. Biol.
Rev. 62:379–433
47. Hutcheson SW. 1999. The hrp cluster of
Pseudomonas syringae: a pathogenicity
island encoding a type III protein translo-
cation complex? See Ref. 51a, pp. 309–
29
48. Iriarte M, Cornelis GR. 1999. The 70-
kilobase virulence plasmid of Yersiniae.
See Ref. 51a, pp. 91–126
49. Ito T, Katayama Y, Hiramatsu K. 1999.
Cloning and nucleotide sequence de-
termination of the entire mec DNA
of pre-Methicillin-resistant Staphylococ-
cus aureus N315. Antimicrob. Agents
Chemother. 43:1449–58
50. Jackson RW, Athanassopoulos GT, Tsi-
amis G, Mansfield JW, Sesma A, et al.
1999. Identification of a pathogenicity is-
land, which contains genes for virulence
and avirulence, on a large native plasmid
in the bean pathogen Pseudomonas sy-
ringae pathovar phaseolicola. Proc. Natl.
Acad. Sci. USA 96:10875–80
51. Kaper JB, Gansheroff LJ, Wachtel WR,
O’Brien AD. 1998. Intimin-mediated ad-
herence of Shiga toxin-producing Es-
cherichia coli and attaching-and-effacing
pathogens. In Escherichia coli O157:H7
and Other Shiga Toxin-Producing E. coli
Strains, ed. JB Kaper, AD O’Brien, pp.
148–56. Washington, DC: Am. Soc. Mi-
crobiol.
51a. Kaper JB, Hacker J, eds. 1999.
Pathogenicity Islands and Other Mobile
Virulence Elements. Washington, DC:
Am. Soc. Microbiol.
52. Kaper JB, Mellies JL, Nataro JP. 1999.
Pathogenicity islands and other mobile
genetic elements of diarrheagenic Es-
cherichia coli. See Ref. 51a, pp. 33–58
53. Kaper JB, Morris JG Jr, Levine MM.
1995. Cholera. Clin. Microbiol. Rev.
8:48–86
54. Karaolis DKR, Johnson JA, Bailey CC,
Boedeker EC, Kaper JB, Reeves PR.
1998. A Vibrio cholerae pathogenicity
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews
676 HACKER ¥ KAPER
island associated with epidemic and pan-
demic strains. Proc. Natl. Acad. Sci. USA
95:3134–39
55. Karaolis DKR, Kaper JB. 1999. Patho-
genicity islands and other mobile virulence
elements of Vibrio cholerae. See Ref. 51a,
pp. 167–87
56. Karaolis DKR, Somara S, Maneval DR
Jr, Johnson JA, Kaper JB. 1999. A bacte-
riophage encoding a pathogenicity island,
a type-IV pilus and a phage receptor in
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.
cholera bacteria. Nature 399:375–79
57. Karch H, Schubert S, Zhang D, Zhang W,
Schmidt H, et al. 1999. A genomic is-
land, termed high-pathogenicity island, is
present in certain non-O157 Shiga toxin-
producing Escherichia coli clonal lineages.
Infect. Immun. 67:5994–6001
58. Kenny B, DeVinney R, Stein M, Rein-
scheid DJ, Frey EA, Finlay BB. 1997. En-
teropathogenic E. coli (EPEC) transfers its
receptor for intimate adherence into mam-
malian cells. Cell 91:511–20
59. Knutton S, Rosenshine I, Pallen MJ,
Nisan L, Neves BC, et al. 1998. A novel
EspA-associated surface organelle of en-
teropathogenic Escherichia coli involved
in protein translocation into epithelial cells.
EMBO J. 17:2166–76
60. Kovach ME, Shaffer MD, Peterson KM.
1996. A putative integrase gene defines
the distal end of a large cluster of ToxR-
regulated colonization genes in Vibrio
cholerae. Microbiology 142:2165–74
61. Kreft J, Vazquez-Boland JA, Ng E, Goebel
W. 1999. Virulence gene clusters and puta-
tive pathogenicity islands in Listeriae. See
Ref. 51a, pp. 219–32
62. Kubori T, Matsushima Y, Nakamura
D, Uralil J, Lara-Tejero M, et al.
1998. Supramolecular structure of the
Salmonella typhimurium type III protein
secretion system. Science 280:602–5
63. Lawrence JG, Ochman H. 1998. Molec-
ular archaeology of the Escherichia coli
genome. Proc. Natl. Acad. Sci. USA
95:9413–17
AR110-19
64. Lindsay JA, Ruzin A, Ross HF, Kurepina
N, Novick RP. 1998. The gene for toxic
shock toxin is carried by a family of mo-
bile pathogenicity islands in Staphylococ-
cus aureus. Mol. Microbiol. 29:527–43
65. Lory S. 1998. Secretion of proteins and
assembly of bacterial surface organelles:
shared pathways of extracellular protein
targeting. Curr. Opin. Microbiol. 1:27–35
66. Low D, David V, Lark D, Schoolnik
G, Falkow S. 1984. Gene clusters gov-
erning the production of hemolysin and
mannose-resistant hemagglutination are
closely linked in Escherichia coli serotype
O4 and O6 isolates from urinary tract in-
fections. Infect. Immun. 43:353–58
67. Marsh JW, Taylor RK. 1999. Genetic
and transcriptional analyses of the Vib-
rio cholerae mannose-sensitive hemagglu-
tinin type 4 pilus gene locus. J. Bacteriol.
181:1110–17
68. Mazel D, Dychinco B, Webb VA, Davis J.
1998. A distinctive class of integron in the
Vibrio cholerae genome. Science 280:605–
8
69. McDaniel TK, Jarvis KG, Donnersberg
MS, Kaper JB. 1995. A genetic locus of
enterocyte effacement conserved among
diverse enterobacterial pathogens. Proc.
Natl. Acad. Sci. USA 92:1664–68
70. McDaniel TK, Kaper JB. 1997. A
cloned pathogenicity island from en-
teropathogenic Escherichia coli confers
the attaching and effacing phenotype on
E. coli K-12. Mol. Microbiol. 23:399–407
71. McDonough MA, Butterton JR. 1999.
Spontaneous tandem amplification and
deletion of the Shiga toxin operon in
Shigella dysenteriae 1. Mol. Microbiol.
34:1058–69
72. Mellies JL, Elliott SJ, Sperandio V, Don-
nenberg MS, Kaper JB. 1999. The Per reg-
ulon of enteropathogenic Escherichia coli:
identification of a regulatory cascade and a
novel transcriptional activator, the locus of
enterocyte effacement (LEE)-encoded reg-
ulator (Ler). Mol. Microbiol. 33:296–306
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews
73. Ménard R, Sansonetti P, Parsot C, Vasselon
T. 1994. Extracellular association and cy-
toplasmic partitioning of the IpaB and
IpaC invasins of S. flexneri. Cell 79:515–
25
74. Mietzner TA, Morse SA. 1994. The role
of iron-binding proteins in the survival
of pathogenic bacteria. Annu. Rev. Nutr.
14:471–93
75. Mills SD, Boland A, Sory MP, van der
Missen P, Kerbourch C, et al. 1997.
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.
Yersinia enterocolitica induces apoptosis
in macrophages by a process requiring
functional type III secretion and translo-
cation mechanisms and involving YopP,
presumably acting as an effector protein.
Proc. Natl. Acad. Sci. USA 94:12638–
43
76. Mirold S, Rabsch W, Rohde M, Stender
S, Tschäpe H, et al. 1999. Isolation of a
temperate bacteriophage encoding the type
III effector protein SopE from an epidemic
Salmonella typhimurium strain. Proc. Natl.
Acad. Sci. USA 96:9845–50
77. Monack DM, Mecsas J, Ghori N, Falkow
S. 1997. Yersinia signals macrophages to
undergo apoptosis and YopJ is necessary
for this cell death. Proc. Natl. Acad. Sci.
USA 94:10385–90
78. Morschhäuser J, Köhler G, Ziebuhr W,
Blum-Oehler G, Dobrindt U, Hacker J.
2000. Evolution of microbial pathogens.
Phil. Trans. R. Soc. London B 355:695–
704
79. Morschhäuser J, Vetter V, Emödy L,
Hacker J. 1994. Adhesin regulatory genes
within large, unstable DNA regions of
pathogenic Escherichia coli: cross-talk be-
tween different adhesin gene clusters. Mol.
Microbiol. 11:555–66
80. Moss JE, Cardozo TJ, Zychlinsky A, Gro-
isman EA. 1999. The selC-associated SHI-
2 pathogenicity island of Shigella flexneri.
Mol. Microbiol. 33:74–83
81. Nataro JP, Kaper JB. 1998. Diarrheagenic
Escherichia coli. Clin. Microbiol. Rev.
11:142–201
AR110-19
PATHOGENICITY ISLANDS 677
81a. O’Callaghan D, Cazevielle C, Allardet-
Servent A, Boschiroli ML, Bourg G, et al.
1999. A homologue of the Agrobacterium
tumefaciens VirB and Bordetella pertus-
sis Ptl type IV secretion systems is essen-
tial for intracellular survival of Brucella
suis. Mol. Microbiol. 33:1210–20
82. Parsot C, Sansonetti PJ. 1999. The
virulence plasmid of Shigellae: an
archipelago of pathogenicity islands? See
Ref. 51a, pp. 151–65
83. Perna NT, Mayhew GF, PósfaI G, Elliott
S, Donnenberg MS, et al. 1998. Molec-
ular evolution of a pathogenicity island
from enterohemorrhagic Escherichia coli
O157:H7. Infect. Immun. 66:3810–17
84. Perrin A, Nassif X, Tinsley C. 1999. Iden-
tification of regions of the chromosome
of Neisseria meningitidis and Neisseria
gonorrhoeae which are specific to the
pathogenic Neisseria species. Infect. Im-
mun. 67:6119–29
85. Pugsley AP. 1993. The complete general
secretory pathway in gram-negative bac-
teria. Microbiol. Rev. 57:50–108
86. Pugsley AP, Francetic O, Possot OM,
Sauvonnet N, Hardie KR. 1997. Recent
progress and future directions in studies
of the main terminal branch of the gen-
eral secretory pathway in gram-negative
bacteria: a review. Gene 192:13–19
87. Rajakumar K, Sasakawa C, Adler B.
1997. Use of a novel approach, termed
island probing, identifies the Shigella
flexneri she pathogenicity island which
encodes a homolog of the immunoglob-
ulin A protease-like family of proteins.
Infect. Immun. 65:4606–14
88. Rakin A, Schubert S, Pelludat C, Brem
D, Heesemann J. 1999. The high-
pathogenicity island of yersiniae. See Ref.
51a, pp. 77–90
89. Reiter W, Palm DP, Yeats S. 1989. Trans-
fer RNA genes frequently serve as inte-
gration sites for prokaryotic genetic ele-
ments. Nucleic Acids Res. 17:1907–14
90. Ritter A, Gally DL, Olsen PB, Dobrindt
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews
678 HACKER ¥ KAPER
U, Friedrich A, et al. 1997. The PAI-
associated leuX specific tRNA5Leu affects
type 1 fimbriation in pathogenic Esch-
erichia coli by control of FimB recombi-
nase expression. Mol. Microbiol. 25:871–
82
91. Roine E, Wei W, Yuan J, Nurmiaho-Lassila
E-L, Kalkkinen N, et al. 1997. Hrp pilus:
an hrp-dependent bacterial surface ap-
pendage produced by Pseudomonas sy-
ringae pv. tomato DC3000. Proc. Natl.
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.
Acad. Sci. USA 94:3459–64
92. Rossier O, Wengelnik K, Hahn K, Bonas U.
1999. The Xanthomonas Hrp type III sys-
tem secretes proteins from plant and mam-
malian bacterial pathogens. Proc. Natl.
Acad. Sci. USA 96:9368–73
93. Schechter LM, Damrauer SM, Lee CA.
1999. Two AraC/XylS family members can
independently counteract the effect of re-
pressing sequences upstream of the hilA
promoter. Mol. Microbiol. 32:629–42
94. Schubert S, Rakin A, Karch H, Carniel E,
Heeseman J. 1998. Prevalence of the high
pathogenicity island of Yersinia species
among Escherichia coli strains that are
pathogenic to humans. Infect. Immun.
66:480–85
95. Segal ED, Cha J, Lo J, Falkow S, Tomp-
kins LS. 1999. Altered states: involvement
of phosphorylated CagA in the induction
of host cellular growth changes by Heli-
cobacter pylori. Proc. Natl. Acad. Sci. USA
96:14559–64
96. Segal G, Purcell M, Shuman HA. 1998.
Host cell killing and bacterial conjugation
require overlapping sets of genes within
a 22-kb region of the Legionella pneu-
mophila genome. Proc. Natl. Acad. Sci.
USA 95:1669–74
97. Segal G, Russo JJ, Shuman HA. 1999.
Relationships between a new type IV se-
cretion system and the icm/dot virulence
system of Legionella pneumophila. Mol.
Microbiol. 34:799–809
98. Skorupski K, Taylor RK. 1997. Cyclic
AMP and its receptor protein negatively
AR110-19
regulate the coordinate expression of
cholera toxin and toxin-coregulated pilus
in Vibrio cholerae. Proc. Natl. Acad. Sci.
USA 94:265–70
99. Sozhamannan S, Deng YK, Li M, Su-
lakvelidze A, Kaper JB, et al. 1999.
Cloning and sequencing of the genes
downstream of the wbf gene cluster of Vib-
rio cholerae serogroup O139 and analysis
of the junction genes in other serogroups.
Infect. Immun. 67:5033–40
100. Sperandio V, Mellies JL, Nguyen W, Shin
S, Kaper JB. 1999. Quorum sensing con-
trols expression of the type III secretion
gene transcription and protein secretion in
enterohemorrhagic and enteropathogenic
Escherichia coli. Proc. Natl. Acad. Sci.
USA 96:15196–201
101. Stein MA, Leung KY, Zwick M, del Por-
tillo FG, Finlay BB. 1996. Identification
of a Salmonella virulence gene required
for formation of filamentous structures
containing lysosomal membrane glyco-
proteins within epithelial cells. Mol. Mi-
crobiol. 20:151–64
102. Tacket CO, Taylor RK, Losonsky G, Lim
Y, Nataro JP, et al. 1998. Investigation
of the roles of toxin-coregulated pili and
mannose-sensitive hemagglutinin pili in
the pathogenesis of Vibrio cholerae O139
infection. Infect. Immun. 66:692–95
103. Tran Van Nhieu G, Caron E, Hall A,
Sansonetti PJ. 1999. IpaC induces actin
polymerization and filopodia formation
during Shigella entry into epithelial cells.
EMBO J. 18:3249–62
104. Tran Van Nhieu G, Sansonetti PJ. 1999.
Mechanism of Shigella entry into epithe-
lial cells. Curr. Opin. Microbiol. 2:51–
55
105. Trucksis M, Michalski J, Deng YK,
Kaper JB. 1998. The Vibrio cholerae
genome contains two unique circular
chromosomes. Proc. Natl. Acad. Sci. USA
95:14464–69
106. Uchiya K, Barbieri MA, Funatao K, Shah
AH, Stahl PD, Groisman EA. 1999. A
P1: FRD/FQU P2: FQP
August 11, 2000 11:13 Annual Reviews
Salmonella virulence protein that inhibits
cellular trafficking. EMBO J. 18:3924–33
107. Vogel JP, Andrews HL, Wong SK, Is-
berg RR. 1998. Conjugative transfer by
the virulence system of Legionella pneu-
mophila. Science 279:873–76
108. Vokes SA, Torres AG, Reeves SA, Payne
SM. 1999. The aerobactin iron trans-
port system genes in Shigella flexneri
are present within a pathogenicity island.
Mol. Microbiol. 33:63–73
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.
109. Wachter C, Beinke C, Mattes M, Schmidt
MA. 1999. Insertion of EspD into epithe-
lial target cell membranes by infecting en-
teropathogenic Escherichia coli. Mol. Mi-
crobiol. 31:1695–707
110. Waldor MK, Mekalanos JJ. 1996. Lyso-
genic conversion by a filamentous
phage encoding cholera toxin. Science
272:1910–14
111. Watnick PI, Fullner KJ, Kolter R. 1999.
A role for the mannose-sensitive hemag-
glutinin in biofilm formation by Vibrio
cholerae El Tor. J. Bacteriol. 181:3606–9
112. Wattiau P, Woestyn S, Cornelis G.
1996. Customized secretion chaperones
in pathogenic bacteria. Mol. Microbiol.
20:255–62
113. Winans SC, Kalogeraki V, Jafri S,
Akakura R, Xia Q. 1999. Diverse roles
of Agrobacterium Ti plasmid-borne genes
in the formation and colonization of plant
tumors. See Ref. 51a, pp. 289–307
AR110-19
PATHOGENICITY ISLANDS 679
114. Wood MW, Jones MA, Watson PR,
Hedges S, Wallis TS, Galyov EE. 1998.
Identification of a pathogenicity is-
land required for Salmonella entero-
pathogenicity. Mol. Microbiol. 29:883–
91
115. Wu S, Lim KC, Huang J, Saidi RF, Sears
CL. 1998. Bacteroides fragilis entero-
toxin cleaves the zonula adherens protein,
E-cadherin. Proc. Natl. Acad. Sci. USA
95:14979–84
116. Wyckoff EE, Duncan D, Torres AG, Mills
M, Maase K, Payne SM. 1998. Structure
of the Shigella dysenteriae heaem trans-
port locus and its phylogenetic distribu-
tion in enteric bacteria. Mol. Microbiol.
28:1139–52
117. Yildiz FH, Schoolnik GK. 1999. Vib-
rio cholerae O1 El Tor: identification
of a gene cluster required for the ru-
gose colony type, exopolysaccharide pro-
duction, chlorine resistance, and biofilm
formation. Proc. Natl. Acad. Sci. USA
96:4028–33
118. Ziebuhr W, Krimmer V, Rachid S, Loss-
ner I, Götz F, Hacker J. 1999. A novel
mechanism of phase variation of vir-
ulence in Staphylococcus epidermidis:
evidence for control of the polysaccha-
ride intercellular adhesin synthesis by
alternating insertion and excision of the
insertion sequence element IS256. Mol.
Microbiol. 32:345–56
 Annual Review of Microbiology
Volume 54, 2000

CONTENTS
THE LIFE AND TIMES OF A CLINICAL MICROBIOLOGIST, Albert
Balows 1
ROLE OF CYTOTOXIC T LYMPHOCYTES IN EPSTEIN-BARR
VIRUS-ASSOCIATED DISEASES, Rajiv Khanna, Scott R. Burrows 19
BIOFILM FORMATION AS MICROBIAL DEVELOPMENT, George
O'Toole, Heidi B. Kaplan, Roberto Kolter 49
MICROBIOLOGICAL SAFETY OF DRINKING WATER, U. Szewzyk,
R. Szewzyk, W. Manz, K.-H. Schleifer 81
THE ADAPTATIVE MECHANISMS OF TRYPANOSOMA BRUCEI
FOR STEROL HOMEOSTASIS IN ITS DIFFERENT LIFE-CYCLE
ENVIRONMENTS, I. Coppens, P. J. Courtoy 129
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.

THE DEVELOPMENT OF GENETIC TOOLS FOR DISSECTING THE

BIOLOGY OF MALARIA PARASITES, Tania F. de Koning-Ward,
Chris J. Janse, Andrew P. Waters 157
NUCLEIC ACID TRANSPORT IN PLANT-MICROBE
INTERACTIONS: The Molecules That Walk Through the Walls, Tzvi
Tzfira, Yoon Rhee, Min-Huei Chen, Talya Kunik, Vitaly Citovsky
187
PHYTOPLASMA: Phytopathogenic Mollicutes, Ing-Ming Lee, Robert E.
Davis, Dawn E. Gundersen-Rindal 221
ROOT NODULATION AND INFECTION FACTORS PRODUCED BY
RHIZOBIAL BACTERIA, Herman P. Spaink 257
ALGINATE LYASE: Review of Major Sources and Enzyme
Characteristics, Structure-Function Analysis, Biological Roles, and
Applications, Thiang Yian Wong, Lori A. Preston, Neal L. Schiller 289
INTERIM REPORT ON GENOMICS OF ESCHERICHIA COLI, M.
Riley, M. H. Serres 341
ORAL MICROBIAL COMMUNITIES: Biofilms, Interactions, and
Genetic Systems, Paul E. Kolenbrander 413
ROLES OF THE GLUTATHIONE- AND THIOREDOXIN-
DEPENDENT REDUCTION SYSTEMS IN THE ESCHERICHIA COLI
AND SACCHAROMYCES CEREVISIAE RESPONSES TO OXIDATIVE
STRESS, Orna Carmel-Harel, Gisela Storz 439
RECENT DEVELOPMENTS IN MOLECULAR GENETICS OF
CANDIDA ALBICANS, Marianne D. De Backer, Paul T. Magee, Jesus
Pla 463
FUNCTIONAL MODULATION OF ESCHERICHIA COLI RNA
POLYMERASE, Akira Ishihama 499
BACTERIAL VIRULENCE GENE REGULATION: An Evolutionary
Perspective, Peggy A. Cotter, Victor J. DiRita 519
LEGIONELLA PNEUMOPHILA PATHOGENESIS: A Fateful Journey
from Amoebae to Macrophages, M. S. Swanson, B. K. Hammer
567
THE DISEASE SPECTRUM OF HELICOBACTER PYLORI : The
Immunopathogenesis of Gastroduodenal Ulcer and Gastric Cancer, Peter
B. Ernst, Benjamin D. Gold 615
PATHOGENICITY ISLANDS AND THE EVOLUTION OF
MICROBES, Jörg Hacker, James B. Kaper 641
DNA SEGREGATION IN BACTERIA, Gideon Scott Gordon, Andrew
Wright 681
 POLYPHOSPHATE AND PHOSPHATE PUMP, I. Kulaev, T.
Kulakovskaya 709
ASSEMBLY AND FUNCTION OF TYPE III SECRETORY SYSTEMS,
Guy R. Cornelis, Frédérique Van Gijsegem 735
PROTEINS SHARED BY THE TRANSCRIPTION AND
TRANSLATION MACHINES, Catherine L. Squires, Dmitry Zaporojets
775
HOLINS: The Protein Clocks of Bacteriophage Infections, Ing-Nang
Wang, David L. Smith, Ry Young 799
OXYGEN RESPIRATION BY DESULFOVIBRIO SPECIES, Heribert
Cypionka 827
REGULATION OF CARBON CATABOLISM IN BACILLUS
SPECIES, J. Stülke, W. Hillen 849
IRON METABOLISM IN PATHOGENIC BACTERIA, Colin Ratledge,
Lynn G Dover 881
Annu. Rev. Microbiol. 2000.54:641-679. Downloaded from arjournals.annualreviews.org
by Secretaria de Ciencia y Tecnologia de Argentina on 06/21/05. For personal use only.
Genetic context of ISCR1 elements found in association with antibiotic resistance genes as part of complex class 1 integrons.
ISCRs are represented by yellow boxes. The qac/sul 3′ ends of the class 1 complex integrons are denoted by gray and black boxes,
respectively. The blue box denoting the intgene (encoding the integrase) represents the 5′ end of the class 1 integron. Gene
cassettes forming parts of class 1 integrons are represented by white boxes with aqua vertical lines. Genes that are not part of
cassettes have been given their own individual color as indicated by the key in Fig. 4. Open reading frames are indicated with
open boxes and the direction of their transcription indicated with arrows. See the text for a detailed description of class 1
integrons. (A) Complex class 1 integrons In6 and In7, in which ISCR1 elements were first observed. (B to F) Genetic context of
ISCR1 associated with trimethoprim resistance dfr genes, ciprofloxacin resistance qnrA genes, aminoglycoside resistance genes,
class C β-lactamase genes, and class A β-lactamase genes, respectively. Accession numbers of the nucleotide sequences of the
various gene arrays are included on the right. Microbiol. Mol. Biol. Rev. June 2006 vol. 70 no. 2 296-316
 Table 1. Putative and proved integrative and conjugative elements.a,b

Size Integrase
Elements Genus or species (kb) Other carried functions type Integration sites References

TCrERLa Bacteroides fragilis 80 Tetr ND Few sites Salyers et al. (1995b)

CTnDOTa Bacteroides 65 Tetr Ermr Tyr Few sites (GTTNNTTTGG) Cheng et al. (2001)
Symbiosis islanda Mezorhizobium loti R7A 502 Symbiosis Tyr 3¢ of a gene encoding tRNAPhe gene Sullivan et al. (2002)
CTnscr94 Salmonella enterica 100 Sucrose utilization ND 3¢ of two genes encoding a tRNAPhe Hochhut et al. (1997)
SXTa Vibrio cholerae 99.5 Sufr Trmr Camr Strr Tyr 5¢ of a gene encoding RF3 Hochhut and Waldor (1999)
R391a Providencia rettgeri 94 Kanr Hgr Tyr 5¢ of a gene encoding RF3 Murphy and Pembroke (1999)
clc Pseudomonas sp. B13 105 Chlorocatechol degradation Tyr 3¢ of two genes encoding a tRNAGly Ravatn et al. (1998b)
bph–sal Pseudomonas putida 90 Biphenyl and salicylate degradation ND ND Nishi et al. (2000)
pSAM2a Streptomyces ambofaciens 10.9 None identified Tyr 3¢ of a gene encoding a tRNAPro Pernodet et al. (1984)
pSE101a Saccharopolyspora erythraea 11.3 None identified Tyr 3¢ of a gene encoding a tRNAThr Brown et al. (1994)
Streptomyces lividans c Numerous sites
Tn1549 a Enterococcus sp. 34 Vanr Tyr A+T rich Garnier et al. (2000)
Tn916 a Enterococcus faecalis 18 Tetr Tyr A+T rich Clewell et al. (1995)

© 2002 Blackwell Publishing Ltd, Molecular Microbiology, 46, 601–610

Clostridium difficile One preferential site Wang et al. (2000a)
Tn5397 a C. difficile 21 Tetr Ser Single site Wang et al. (2000b)
Bacillus subtilis c Numerous sites
Tn5252 a Streptococcus pneumoniae 47 Camr Tyr Single intergenic site Vijayakumar and Ayalew (1993)
Tn5276 a Lactococcus lactis 70 Sucrose utilization and nisin synthesis Tyr Few sites (TTTTTG) Rauch and de Vos (1992)
pRS01/sex factor L. lactis 48.4 Tellurium resistance ND Single site Gasson et al. (1995)
EfaC1a E. faecalis 25.3 ND Tyr 3¢ end of a gene encoding a ThrtRNA Burrus et al. (2002)
EfaC2a E. faecalis 32.7 ND Tyr 3¢ of a gene encoding a GMP-synthase Burrus et al. (2002)
EfaD2a E. faecalis ND ND Tyr ND Burrus et al. (2002)
Tn5801a Staphylococcus aureus 25.8 Tetr Tyr 3¢ of a gene encoding a GMP-synthase Kuroda et al. (2001)
CW459tet(M)a Clostridium perfringens ND Tetr Tyr 3¢ of a gene encoding a GMP-synthase Roberts et al. (2001)
ICELm1a Listeria monocytogenes 21.3 Putative Cd2+ resistance Tyr 3¢ of a gene encoding a GMP-synthase Burrus et al. (2002)
ICESt1a Streptococcus thermophilus 34.7 Restriction-modification Sth368I Tyr 3¢ end of a gene encoding a Burrus et al. (2002)
fructose-1,6-diphosphate aldolase

a. Other related integrative and conjugative elements have not been indicated in this table.
b. The names of the proven ICEs are not underlined. The names of the putative ICEs are underlined.
c. pSE101 and Tn5397 were not found in S. lividans and B. subtilis respectively.
Cam, chloramphenicol; Erm, erythromycin; Hg, mercury; Kan, kanamycin; ND, not determined; Ser, serine recombinase; Str, streptomycin; Suf, sulphamethoxazole; Tet, tetracycline; Trm,
trimethoprim; Tyr, tyrosine recombinase; Van, vancomycin;
Conjugative transposons 603
 Tetracycline
A 1 kb Conjugation resistance RegulationRecombination

Tn5397
Group II
intron

88% 88% 84%

Tn916

67%

Tn1549

Conjugation Vancomycin resistance Recombination

Regulation
B
1 kb
Recombination Conjugation

EfaC1

0%
20%
40%
70%
100%
% identity of proteins

> 90% nt identity with pAD1

EfaC2
606 V. Burrus, G. Pavlovic, B. Decaris and G. Guédon

Conjugation Recombination

C
1 kb Recombination Regulation Conjugation
EfaD2

0%
20%
40%
70%
100%
% identity of proteins

pAM373

Conjugation Replication

Fig. 1. Comparison of the gene organization of some putative or proven conjugative elements.
A. Conjugative transposons Tn5397, Tn916 and Tn1549.
B. Putative site-specific integrative and conjugative elements EfaC1 and EfaC2.
C. Putative integrative conjugative element/conjugative transposon EfaD2 and conjugative plasmid pAM373. Boxes represent the ORFs, with the proposed direction of transcription shown by the
arrow. ORF names beginning with ‘orf ’ (Tn5397, Tn916, Tn1549), ‘EF’ (EfaC1, EfaC2, EfaD2) or ‘EP’ (pAM373) in the GenBank annotation are abbreviated with the corresponding number. The
colour of the boxes depicts the putative function of each ORF deduced from functional analysis or from BLASTP comparison: Blue, conjugative transfer; grey, antibiotic resistance; green, putative
transcriptional regulator related to those of Tn916; orange, serine recombinase; red, excisionase and tyrosine recombinase; black, plasmid replication functions; white, other putative functions
or no putative function.
In (A), the ORFs are connected by grey shading when significant nucleotide identity (given between brackets) is detected; the ORFs are connected by light yellow shading when the predicted
proteins share significant sequence identity (<40% in BLASTP comparison).
In (B) and (C), the ORFs are connected by a yellow/orange shading when the predicted proteins share significant sequence identity; the colour of this shading is linked to the percentage amino
acid identity. The whole sequence of each element is presented, except for EfaD2, as its ends are not known. In Tn5397, the group II intron is depicted by a square pattern. In Tn5397, orf5 is

© 2002 Blackwell Publishing Ltd, Molecular Microbiology, 46, 601–610

not described by Roberts et al. (2001), but is found by sequence comparison with Tn916. In EfaC1, the black line indicates ORFs closely related to those of the conjugative plasmid pAD1 (>90%
identity in nucleotide sequence). Whereas the ORFs EF2546 from EfaC1 and EF0166 from EfaC2 encode distantly related putative integrases (29% amino acid identity), they are not connected
in (B).
© 1988 Nature Publishing Group
© 1988 Nature Publishing Group
© 1988 Nature Publishing Group
© 1988 Nature Publishing Group
 Blackwell Science, LtdOxford, UKMMIMolecular Microbiology0950-382XBlackwell Publishing Ltd, 2004? 2004523643650Review ArticleStress-directed mutationsB. E. Wright

Molecular Microbiology (2004) 52(3), 643–650 doi:10.1111/j.1365-2958.2004.04012.x

MicroReview

Stress-directed adaptive mutations and evolution

Barbara E. Wright* question concerns the origin of the variants from which
Division of Biological Sciences, University of Montana, the fittest survive. Dogma has it that these variants arise
Missoula, MT 59812, USA. from pre-existing ‘random’ mutations that are not directed
by the different conditions of stress that select them. In
other words, there would be no specific biochemical
Summary
mechanisms initiated by each stressor that can increase
Comparative biochemistry demonstrates that the mutation rates in those genes related to that stress.
metabolites, complex biochemical networks, However, specific, stress-directed mutagenesis (SDM)
enzymes and regulatory mechanisms essential to all would offer an enormous advantage for evolution and
living cells are conserved in amazing detail through- would be selected. Therefore, if SDM exists it should be
out evolution. Thus, in order to evolve, an organism present today in organisms coping with adverse condi-
must overcome new adverse conditions without cre- tions in their environment. Increased mutation rates would
ating different but equally dangerous alterations in its be directed to specific genes that must mutate to alleviate
ongoing successful metabolic relationship with its the stress, while avoiding random genome-wide damage.
environment. Evidence suggests that stable long- The search for directed mutations has been stymied by
term acquisitive evolution results from minor the problem posed by Delbrück in (1946): if a mutation
increases in mutation rates of genes related to a par- can only be observed under conditions that may cause it,
ticular stress, with minimal disturbance to the bal- as well as select it, how can we tell the difference? Does
anced and resilient metabolism critical for responding the stress cause the mutation before the mutant is
to an unpredictable environment. Microorganisms selected? This conundrum has understandably led to con-
have evolved specific biochemical feedback mecha- fusion, controversy, and, by default, acceptance of the
nisms that direct mutations to genes derepressed by Dogma first articulated more than 100 years ago by Weis-
starvation or other stressors in their environment. mann (1893). Recent advances in this field reveal mech-
Transcription of the activated genes creates localized anisms by which a specific stress does in fact cause
supercoiling and DNA secondary structures with mutations in related genes, thus resolving the issue raised
unpaired bases vulnerable to mutation. The resulting by Delbrück.
mutants provide appropriate variants for selection by This MicroReview attempts to critique and relate three
the stress involved, thus accelerating evolution with different areas of research concerned with mechanisms
minimal random damage to the genome. This model underlying adaptive mutations: (i) acquisitive evolution, in
has successfully predicted mutation frequencies in which gain-of-function mutations confer new and advan-
genes of E. coli and humans. Stressed cells observed tageous capabilities upon organisms; (ii) experimental
in the laboratory over hundreds of generations accu- evolution, observed over hundreds of generations using
mulate mutations that also arise by this mechanism. glucose limited/starved microbial populations, and (iii) bio-
When this occurs in repair-deficient mutator strains chemical mechanisms underlying SDM.
with high rates of random mutation, the specific
stress-directed mutations are also enhanced.
Acquisitive evolution

Investigators in this field have provided beautiful examples

Introduction
of evolving biosynthetic and catabolic pathways respond-
Evolution is defined as the result of environmental stress ing to environmental stress. These mutations modify exist-
that selects the fittest. The yet unresolved and fascinating ing gene-enzyme systems to initiate pathways using a
new carbon source as the previous source becomes lim-
Accepted 7 January, 2004. *For correspondence. E-mail
barbara.wright@mso.umt.edu; Tel. (+1) 406 243 6676; Fax (+1) 406 iting to survival (for reviews see Hegeman and Rosen-
243 4184. berg, 1970; Lin et al., 1976; Clarke, 1984; Mortlock,

© 2004 Blackwell Publishing Ltd

644 B. E. Wright
1984). Early investigations in this field demonstrated that
cells starved in the presence of a carbon source they
cannot use produced mutants that could use it. The most
common mechanism involved is gene derepression result-
ing in the constitutive production of a previously inducible
enzyme or an enzyme with altered substrate specificity.
Many examples exist; some of the earliest are the use of
altrose-galactoside via b-galactosidase (Lederberg,
1951); b-glycerolphosphate via alkaline phosphatase
(Torriani and Rothman, 1961) and xylitol via ribitol dehy-
drogenase (Lerner et al., 1964). As discussed below, such
derepressed genes are actively transcribed and thereby
become targets for SDM. The metabolic steps required to
metabolize a new related substrate are similar to those for
existing pathways. Therefore, relatively minor changes to
a duplicated copy of the existing gene may be required for
recruitment to serve a new function (reviewed in Wright,
2000). Other frequently observed mutations in response
to carbon source starvation confer increased permeability
for the limiting metabolite. Examples of this kind will be
discussed below in the section on experimental evolution.
The field of acquisitive evolution has been enhanced by
biotechnologists, who have published many papers on the
results of SDM. Genetic engineering is used to achieve
strain improvement by increasing the yield of a valuable
product for example, or allowing the metabolism of a
xenobiotic compound (van der Meer et al., 1992). Although
genetically engineered microbial strains are less compet-
itive in natural environments than those that originally
evolved and were selected under those conditions, they
have provided valuable insight into mechanisms by which
new catabolic and biosynthetic pathways arise during evo-
lution. Of particular relevance are studies that are similar
to investigations by evolutionary biologists analysing
mechanisms of acquisitive evolution. Single point muta-
tions have been shown to extend the substrate range for
the degradation of many organic compounds, for example,
xylene monooxygenase (Abril et al., 1989) and catachol
dioxygenase (Ramos et al., 1987). Gain-of-function muta-
tions for metabolizing a new substrate typically require
starvation conditions and long exposure times to that sub-
strate as the sole carbon source (e.g. Nochur et al., 1990).
Some examples of specific SDM in starving bacteria
are particularly well-documented and pertinent to this
review. Kasak et al. (1997) described the accumulation of
phenol-utilizing mutants (Phe+) in starving cultures of a
promoterless phenol degradation operon of P. putida in
the presence, but not the absence, of phenol (the possibility
that the absence of phenol eliminates Phe+ mutants from
starving populations was excluded). Sequence analysis of
these mutants revealed base substitutions, deletions and
insertions that in effect created a promoter. These muta-
tions differed in type and frequency in starving as com-
pared to growing populations. Because the frequency of
© 2004 Blackwell Publishing Ltd, Molecular Microbiology, 52, 643–650
Phe+ mutations found on selective plates from stationary-
phase cells was higher than from exponentially growing
cells, they concluded that specific factors in starving cells
must facilitate mutations conferring the ability to utilize
phenol. Other investigators have also observed that
mutants arising from SDM in starving bacteria differ from
those found during growth, implicating different molecular
mechanisms (Prival and Cebula, 1992; Rosenberg et al.,
1994; Foster and Trimarchi, 1995; Notley-McRobb and
Ferenci, 1999a; Sung and Yasbin, 2002). Although star-
vation and gene derepression are probably the predomi-
nant cause of SDM in nature, others include direct effects
on supercoiling by variables such as osmolarity, temper-
ature or anaerobiosis. For example, deletion mutations of
the stress-induced spvB gene in Salmonella (Massey
et al., 1999) only occur at a specific concentration of NaCl
that would presumably induce a particular level of super-
coiling, which is known to create DNA secondary structures
containing unpaired bases vulnerable to mutations (Ripley
and Glickman, 1983; Wright et al., 2002; Wright et al.,
2003). Supercoiling also plays a critical role in selective
gene transcription (Lefstin and Yamamoto, 1998), for exam-
ple, in sensitive genes such as proU (Higgins et al., 1988)
and Hg-MerR (Ansari et al., 1992).
Experimental evolution
A number of laboratories study mutations that arise over
many generations in microbial populations limited or
starved for a carbon source (for example, Dykhuizen and
Hartl, 1983; Death et al., 1993; Notley-McRobb and Fer-
enci, 1999a, b; Funchain et al., 2000; Giraud et al., 2001;
Notley-McRobb et al., 2003). The consequences of glu-
cose starvation (in contrast to less ‘global’ nutrients such
as phosphate or an amino acid) are extremely widespread
and difficult to sort out, as all areas of metabolism are
affected directly or, eventually, indirectly. A multitude of
genes are derepressed and activated as a result of the
stringent response (Cashel et al., 1996), for example, relA
and spotT which regulate the accumulation of ppGpp that
is required for the accumulation of rpoS-encoded ss. This
central regulator then initiates a cascade derepression of
starvation-responsive genes including those protecting
the cells from additional stressors such as heat, osmotic
shock, anaerobiosis, and oxidative damage (Gentry et al.,
1993; Hengge-Aronis, 1996). The cAMP-CRP complex
accumulates and activates about two-thirds of the carbon
starvation responsive genes. Derepressed genes include
those improving outer membrane permeability and glu-
cose transport, and those encoding enzymes using alter-
nate carbon sources, such as lactose, maltose, galactose
and ribose (Saier et al., 1996). A burst of supercoiling
accompanies this widespread derepression and a 30-fold
higher mutation rate is seen in severely glucose-limited

cultures with maximally derepressed genes than in mildly
limited cultures (Notley-McRobb and Ferenci, 1999a). As
discussed below, such derepressed genes are not only
activated and selected in starving cells, but also become
targets for SDM that may provide beneficial mutations and
increased fitness for the stressed cells. The majority of
mutations found in experimental evolution experiments do
not appear to be random, but occur repeatedly in these
newly derepressed genes and in replicate cultures.
The most frequently identified mutants in glucose-
starved cultures show improved uptake of the limiting nutri-
ent (Dykhuizen and Hartl, 1983). A literature search for
large data-bases of sequenced mutations from such cul-
tures was found in the work of Dardonville and Raibaud
(1990) and Notley-McRobb and Ferenci (1999a,b). The
majority of mutations (17, see Table 2 in Notley-McRobb
and Ferenci, 1999b) occurred in the mgl operator and
showed greater rates of glucose transport than the mglD
null mutations inactivating a repressor that binds to the
operator. These authors believe that the mglO mutations,
located within a 14 bp sequence, were non-random and
somehow directed to the operator. The regulatory muta-
tions increasing glucose transport were analysed using a
new computer program, mfg, to determine if SDM were
responsible for their appearance (Wright et al., 1999;
2002; 2003; 2004). This program (see below) has suc-
cessfully predicted mutation frequencies in E. coli and
humans, based on the extent to which a mutable base is
unpaired during transcription and upon the stability of the
DNA secondary structure in which it is unpaired. The eight
mglO point mutations were analysed and six were found
to be similar to those described in amino acid auxotrophs
of E. coli (Wright et al., 2003) and in human p53 cancers
(Wright et al., 2002). The mutable sites were located in
small loops very near or between stems of DNA stem-loop
structures (SLSs) with high stability, and their predicted
mutability was high, as they were unpaired and exposed
most of the time (averaging 94% of their folds) during
transcription (unpublished data). Of the nine mglD-
encoded null repressor mutants available for analysis most
were similar to the mglO mutable bases but were less
frequently unpaired and had lower predicted mutabilities.
Investigations with experimental evolution in glucose-
limited cultures are not designed to determine mutation
frequencies. However, we have reported reversion rates
of a derepressed lacZ auxotroph which serves as a model
for the analysis of ‘forward’ mutations in derepressed
genes. In this study relative reversion rates of each base
in a lacZ stop codon were determined, following glucose
starvation and derepression of this gene (including its
defective codon). There was a good correlation between
the relative number of mutations in the three bases and
mutation frequencies predicted by the mfg program
(Wright et al., 2003; and see below).
© 2004 Blackwell Publishing Ltd, Molecular Microbiology, 52, 643–650
Stress-directed mutations 645
The most fascinating results were seen in the analysis
of 14 point mutations of malT, encoding the specific MalT
activator protein (Dardonville and Raibaud, 1990; Notley-
McRobb and Ferenci, 1999a) which regulates at least five
‘maltose box’ promoters and LamB glycoporin expression
critical to outer membrane permeability (Death et al.,
1993). These mutations are observed only in response to
severe glucose limitation, and result in constitutive expres-
sion of the regulon. Quite unexpectedly, they were seen
to share the same characteristics found in the mutable
bases of a model system of the immune response (Wright
et al., 2004). Ninety per cent of these malT point muta-
tions occurred at G:C base pairs; they were rarely
unpaired and located at the ends of SLSs (unpublished
data). The most striking characteristic linking these muta-
ble bases to those in the immune system is that the rarely
unpaired bases are preferentially influenced by transcrip-
tion. There is an inverse correlation (P = 0.03) between
the extent to which mutable malT bases are unpaired and
the increase in stability of their SLSs as a function of
increased levels of transcription. This correlation suggests
a mechanism by which transcription could determine the
order and intensity of base mutability (Wright et al., 2004).
The gain-of-function malT mutations are clustered in two
small regions of the gene, and they all result in the same
discrete and well-defined change in MalT conformation.
The altered conformation confers constitutivity on this key
activator protein which binds to specific nucleotide
sequences in the promoters it activates (Dardonville and
Raibaud, 1990). Adaptations enhancing antibody affinity
for antigens during the immune response also involve
alterations in protein conformation. We speculate that
such adaptations required the evolution of localized
groups of unpaired mutable bases allowing coordinated
trials with simultaneous mutations at many sites to
achieve a best fit for binding to their target. In any event,
it appears that mutations occurring in derepressed genes
of glucose-starved cultures are not random, but share the
mechanisms of SDM elicited by many kinds of stressors
in the environment (see below).
Compared to wild type, the frequency of mismatch
repair defective mutator strains is higher in experimental
evolution experiments where strong selection is applied
(Cox and Gibson, 1974; Notley-McRobb et al., 2003).
Funchain et al. (2000) have monitored the function of
more than 700 genes (~15% of the genome) in repair-
deficient mutants after 1000 generations in rich media and
documented the presence of extensive damage not seen
in wild type under the same conditions. As discussed
above, the magnitude and variety of mutations expected
in nutritionally stressed mutators is expected to be much
greater than in cells grown in rich media. Distinguishing
the multitude of specific, transcription-induced mutations
expected in glucose-starved mutator cells from the
646 B. E. Wright
genome-wide random mutations resulting from deficien-
cies in repair would be impossible. Moreover, the magni-
tude of specific SDM will be amplified in strains deficient
in repair (see below).
Mechanisms of SDM
During cell growth, there are many DNA-destabilizing
events that will increase mutation rates: recombination,
transcription, error-prone DNA polymerases, proofreading
errors during DNA replication, and repair-deficient muta-
tors. In stressed cells, however, when growth is absent or
minimized, replication-associated events may not be major
sources of mutations compared with SDM. As mentioned
earlier, the most efficient process for overcoming stress
and accelerating evolution would be the initiation of spe-
cific feedback mechanisms by which each kind of stress
targets hypermutation to those genes that must mutate to
overcome the stress. Although there are a number of
examples of specifically directed mutations caused by
artificially induced transcription in growing cells (reviewed
in Wright, 2000), evolution requires mutations directed by
stress. To my knowledge, the first example of SDM was
reported by Lipschutz et al. (1965), who examined the
effect of phosphate starvation on reversion rates of an
alkaline phosphatase mutant in derepressed compared
with repressed genes. Reversion rates in the derepressed
gene (in the absence of orthophosphate) were 15–20-fold
higher than in the repressed gene indicating . . . ‘that the
susceptibility of . . . bases to chemical changes increases
during synthesis of mRNA . . . from an opening of the DNA
helix or from an uncovering of the gene during derepres-
sion’. With the advent of gene arrays we now know that
each kind of stress is limited to about 1% of the genome
and specifically derepresses and activates those genes
mRNA
RNAP
_
_
A-
G
C
-A
-G-
-
G-C
A-T
G C-G
mutation G-C
T -A
T C
A C
C T
C
Fig. 1. A model depicting the effect of transcription on mutations. Transcription drives supercoiling which creates and stabilizes SLSs containing
unpaired bases vulnerable to mutation.
related to the stress. Examples include phosphate starva-
tion (Antelmann et al., 2000; Ishige et al., 2003), amino
acid starvation (Rudner et al., 1999; Wright et al., 1999),
superoxide stress (Pomposiello et al., 2001), and osmotic
stress (Wolf et al., 2003).
Reversions of amino acid auxotrophs are convenient
models for analysing the specific consequences of
starvation on mutations in related genes (Wright et al.,
1999; 2003). For example, Fig. 1. depicts mechanisms by
which gene derepression can initiate a series of events
resulting in higher mutation rates in trpA of the trp
operon. As the absence of tryptophan removes gene
repression, transcription is activated in all the biosyn-
thetic trp genes, including the defective trpA– mutant
allele 23 (A-to-G) (Wright et al., 2003). Transcription
drives supercoiling (Liu and Wang, 1987), and the DNA
secondary stem-loop structures (SLSs) created by super-
coiling can be quantitatively measured (Dayn et al.,
1992). Stem-loop structures probably occur in the wake
of the transcription bubble, as the bubble itself contains
too little ssDNA to form predicted structures. Stem-loop
structures are sequence-dependent; that is, they form in
supercoiled DNA because of the proximity of inverted
complements such as the two 5 nt segments forming the
stem in Fig. 1. Templated mutations are perhaps the
most convincing evidence for the location of mutable
bases in SLSs (Ripley and Glickman, 1983). A recent
example is the simultaneous mutation of three contigu-
ous bases templated by three unpaired bases in the
opposing strand of a SLS in the lacZ gene (Wright et al.,
2003).
Supercoiling and SLSs are highly localized to the vicin-
ity of each RNA polymerase complex (Rahmouni and
Wells, 1992) and the secondary structures are character-
ized by unpaired bases located in a loop or at the base
+
+
Transcription
trpA allele 23
© 2004 Blackwell Publishing Ltd, Molecular Microbiology, 52, 643–650

of a stem. Increasing levels of transcription will increase
supercoiling, the lengths of ssDNA, the size and stability
of SLSs (Balke and Gralla, 1987; Figueroa and Bossi,
1988), and thereby increase mutation rates. Unpaired
bases in these secondary structures are vulnerable to
mutation because of their intrinsic thermodynamic insta-
bility (especially Gs and Cs) and their availability to
nucleotide-altering enzymes (Lindahl, 1993). The inci-
dence of mutations in such unpaired bases [including,
perhaps, ‘contingency’ loci (Moxon et al., 1994)] would of
course be amplified in mutator strains lacking mismatch
repair (see above).
Because extensive evidence indicates that mutable
bases in actively transcribed genes of stressed cells are
unpaired and located in SLSs (reviewed in Wright, 2000),
their mutation rates could be predicted knowing (i) how
frequently a base is unpaired in the successive SLSs that
contain it during transcription, and (ii) which of those SLSs
is the most stable, as that is the one in which the base
would be maximally exposed and most likely to mutate. A
new computer algorithm named mfg (http://biol-
ogy.dbs.umt.edu/wright/upload/mfg.html) was therefore
developed to provide this information and to calculate a
Mutability Index (MI) for each base, defined as the abso-
lute value of [(percentage of folds in which the base is
unpaired) (highest –DG of all folds in which the base is
unpaired)] (Wright et al., 2003). Thus, primary importance
is given to the most stable SLS in which a given base is
unpaired and exposed for the longest period of time; this
should be the structure in which the base has the highest
probability of mutation.
The mfg program was used to predict reversion fre-
quencies in derepressed genes of 15 auxotrophs, and the
correlation was good between MIs and relative mutation
frequencies determined experimentally (Wright et al.,
© 2004 Blackwell Publishing Ltd, Molecular Microbiology, 52, 643–650
Stress-directed mutations 647
Fig. 2. Mutation frequencies in genes of E. coli
and humans were predicted successfully by the
mfg program.
A. Predicted compared with experimentally
determined reversion rates in the derepressed
E. coli auxotroph, trpA and the non-induced lacI
repressor.
B. Predicted compared with observed mutation
frequencies in hypermutable bases of p53 and
in control bases in nearby codons.
2003). This correlation is shown for a series of trpA–
mutant alleles in Fig. 2A and compared with 35 back-
ground mutable bases in the constitutively expressed con-
trol gene, lacI.
As demonstrated by ‘The Unity in Biochemistry’ [first
recognized by Kluyver and Donker (1926)], the process of
evolution has bestowed a fundamental unity of biochemi-
cal behaviour on all forms of life: the same metabolites,
biosynthetic and catabolic pathways, enzymes and regu-
latory mechanisms. Mutagenesis in microbes and
humans, then, should have much in common. Moreover,
higher rates of transcription are associated with hypermu-
tation in both cancer (Ghosh and Mitchell, 1999) and the
immune response (Bachl et al., 2001), both of which have
been analysed with the mfg program (Wright et al., 2002;
2004). The mfg program was used to predict mutation
frequencies in an unprecedented database of 14 000
hypermutable bases of the human p53 tumor suppressor
gene. As seen in Fig. 2B, an excellent correlation was
found for the hypermutable bases as compared to control
bases located in nearby codons. Plotting the number of
mutations against –DG alone did not give a statistically
significant correlation, i.e. both variables in the above
equation were necessary to predict mutation frequencies
successfully (Wright et al., 2002). The p53 data are com-
pelling in demonstrating that the assumptions underlying
the mfg program are essentially correct. We conclude that
mutable bases in stressed cells are in fact unpaired and
located in SLSs caused by supercoiling in the wake of the
transcription bubble. The mutability of such a base will be
affected by the: (i) per cent of total folds in which the base
is unpaired during transcription; (ii) stability of the most
stable SLS in which the base is unpaired; (iii) intrinsic
thermodynamic instability of each base, and (iv) level of
transcription.
648 B. E. Wright
As DNA sequence and thermodynamics determine the
first three variables, the two extrinsic variables that can
influence mutability are transcription and supercoiling. As
discussed above, various environmental stressors can
enhance mutations directly via transcription and/or super-
coiling, and epigenetic factors increasing gene expression
are also clearly associated with several tumour models
(Ghosh and Mitchell, 1999). The data suggest that, if
stressors cause direct damage to the hypermutable bases
analysed in these studies, the resulting increase in muta-
tion frequencies must depend upon, and reflect, the intrin-
sic mutabilities (MIs) of these bases.
Conclusions
In nature, where cell division must often be negligible as
a result of multiple adverse conditions, beneficial muta-
tions for evolution can arise in specific response to stres-
sors that target related genes for derepression. Specific
transcription of these genes then results in localized DNA
secondary structures containing unpaired bases vulnera-
ble to mutation. Many environmental stressors can also
affect supercoiling and SDM directly.
Background mutation rates in nature remain within a
very narrow range: sufficiently high to adapt to environ-
mental stress, yet sufficiently low to avoid random damage
to the genome (Kimura, 1967; Drake, 1991). This sensi-
tivity is consistent with the low incidence of mutators in
wild-type populations and with the conclusion that optimal
conditions for stable, long-term evolutionary advances in
response to stress generally involve minor increases in
mutation frequency. The resulting advantage to the organ-
ism may necessarily be minor, in order to avoid rocking
the metabolic boat of complex and finely tuned interac-
tions that have been selected in the past and must be
maintained during further evolution.
Acknowledgements
I thank Arnold Demain, Scott Samuels, Frank Rosenzweig,
Jacqueline Reimers and Karen Schmidt for valuable discus-
sions and suggestions, and Maryann Hubbard for manuscript
preparation. This work was supported by National Institutes
of Health grants R15CA88893 and R55CA99242 and the
Stella Duncan Memorial Research Institute.
References
Abril, M.A., Michan, C., Timmis, K.N., and Ramos, J.L. (1989)
Regulator and enzyme specificities of the TOL plasmid-
encoded upper pathway for degradation of aromatic
hydrocarbons and expansion of the substrate range of the
pathway. J Bacteriol 171: 6782–6790.
Ansari, A.Z., Chael, M.L., and O’Halloran, T.V. (1992)
Allosteric underwinding of DNA is a critical step in posi-
© 2004 Blackwell Publishing Ltd, Molecular Microbiology, 52, 643–650
tive control of transcription by Hg-MerR. Nature 355: 87–
89.
Antelmann, H., Scharf, C., and Hecker, M. (2000) Phosphate
starvation-inducible proteins of Bacillus subtilis: proteomics
and transcriptional analysis. J Bacteriol 182: 4478–4490.
Bachl, J., Carlson, C., Gray-Schopfer, V., Dessing, M., and
Olsson, C. (2001) Increased transcription levels induce
higher mutation rates in a hypermutating cell line. J Immu-
nol 166: 5051–5057.
Balke, V.L., and Gralla, J.D. (1987) Changes in the linking
number of supercoiled DNA accompany growth transitions
in Escherichia coli. J Bacteriol 169: 4499–4506.
Cashel, M., Gentry, D.R., Hernandez, V.J., and Vinella, D.
(1996) The stringent response. In Escherichia Coli and
Salmonella: Cellular and Molecular Biology, Vol. 1.
Neidhardt, F.C., Curtis, R. III, Ingraham, J.L., Lin, E.C.C.,
Low, K.B., Magasanik, B., et al., (eds). Washington, DC:
American Society for Microbiology, pp. 1458–1496.
Clarke, P.H. (1984) The evolution of degradative pathways.
In Microbial Degradation of Organic Compounds. Gibson,
D.T, (ed.). New York: Marcel Dekker, pp. 11–27.
Cox, E.C., and Gibson, T.C. (1974) Selection for high muta-
tion rates in chemostats. Genetics 77: 169–184.
Dardonville, B., and Raibaud, O. (1990) Characterization of
malT mutants that constitutively activate the maltose reg-
ulon of Escherichia coli. J Bacteriol 172: 1846–1852.
Dayn, A., Malkhosyan, S., and Mirkin, S.M. (1992) Transcrip-
tionally driven cruciform formation in vivo. Nucleic Acids
Res 20: 5991–5997.
Death, A., Notley, L., and Ferenci, T. (1993) Derepression of
LamB protein facilitates outer membrane permeation of
carbohydrates into Escherichia coli under conditions of
nutrient stress. J Bacteriol 175: 1475–1483.
Delbruck, M., and Bailey, W.T. Jr (1946) Induced mutations
in bacterial viruses. Cold Spring Harb Symp Quant Biol 11:
33–37.
Drake, J.W. (1991) A constant rate of spontaneous mutation
in DNA-based microbes. Proc Natl Acad Sci USA 88:
7160–7164.
Dykhuizen, D.E., and Hartl, D.L. (1983) Selection in chemo-
stats. Microbiol Rev 47: 150–168.
Figueroa, N., and Bossi, L. (1988) Transcription induces gyra-
tion of the DNA template in Escherichia coli. Proc Natl
Acad Sci USA 85: 9416–9420.
Foster, P.L., and Trimarchi, J.M. (1995) Adaptive reversion of
an episomal frameshift mutation in Escherichia coli
requires conjugal functions but not actual conjugation. Proc
Natl Acad Sci USA 92: 5487–5490.
Funchain, P., Yeung, A., Stewart, J.L., Lin, R., Slupska, M.M.,
and Miller, J.H. (2000) The consequences of growth of a
mutator strain of Escherichia coli as measured by loss of
function among multiple gene targets and loss of fitness.
Genetics 154: 959–970.
Gentry, D.R., Hernandez, V.J., Nguyen, L.H., Jensen, D.B.,
and Cashel, M. (1993) Synthesis of the stationary-phase
sigma factor ss is positively regulated by ppGpp. J Bacteriol
175: 7982–7989.
Ghosh, R., and Mitchell, D.L. (1999) Effect of oxidative DNA
damage in promoter elements on transcription factor bind-
ing. Nucleic Acids Res 27: 3213–3218.
Giraud, A., Matic, I., Tenaillon, O., Clara, A., Radman, M.,

Fons, M., and Taddei, F. (2001) Costs and benefits of high
mutation rates: adaptive evolution of bacteria in the mouse
gut. Science 291: 2606–2608.
Hegeman, G.D., and Rosenberg, S.L. (1970) The evolution
of bacterial enzyme systems. Annu Rev Microbiol 24: 429–
462.
Hengge-Aronis, R. (1996) Regulation of gene expression
during entry into stationary phase. In Escherichia Coli and
Salmonella: Cellular and Molecular Biology, Vol. 1.
Neidhardt, F.C., Curtis, R. III, Ingraham, J.L., Lin, E.C.C.,
Low, K.B., Magasanik, B., et al., (eds). Washington, DC:
American Society for Microbiology.
Higgins, C.F., Dorman, C.J., Stirling, D.A., Waddell, L.,
Booth, I.R., May, G., and Bremer, E. (1988) A physiological
role for DNA supercoiling in the osmotic regulation of gene
expression in S. typhimurium and E. coli. Cell 52: 569–584.
Ishige, T., Krause, M., Bott, M., Wendisch, V.F., and Sahm,
H. (2003) The phosphate starvation stimulon of Coryne-
bacterium glutamicum determined by DNA microarray
analyses. J Bacteriol 185: 4519–4529.
Kasak, L., Horak, R., and Kivisaar, M. (1997) Promoter-cre-
ating mutations in Pseudomonas putida: a model system
for the study of mutation in starving bacteria. Proc Natl
Acad Sci USA 94: 3134–3139.
Kimura, M. (1967) On the evolutionary adjustment of spon-
taneous mutation rates. Genet Res Camb 9: 23–43.
Kluyver, A.J., and Donker, H.J.L. (1926) Die Einheit in der
Biochemie. Chem d Zelle u Gew 13: 134–190.
Lederberg, J. (1951) In Genetics in the 20th Century. Dunn,
L.C., (ed.). New York: Macmillan, pp. 263.
Lefstin, J.A., and Yamamoto, K.R. (1998) Allosteric effects of
DNA on transcriptional regulators. Nature 392: 885–888.
Lerner, S.A., Wu, T.T., and Lin, E.C.C. (1964) Evolution of a
catabolic pathway in bacteria. Science 146: 1313–1315.
Lin, E.C.C., Hacking, A.J., and Aguilar, J. (1976) Experimen-
tal models of acquisitive evolution. Bioscience 26: 548–
555.
Lindahl, T. (1993) Instability and decay of the primary struc-
ture of DNA. Nature 362: 709–715.
Lipschutz, R., Falk, R., and Avigad, G. (1965) Mutation rates
of the induced and repressed locus of alkaline phos-
phatase in Escherichia coli. Israel J Med Sci 1: 323–324.
Liu, L.F., and Wang, J.C. (1987) Supercoiling of the DNA
template during transcription. Proc Natl Acad Sci USA 84:
7024–7027.
Massey, R.C., Rainey, P.B., Sheehan, B.J., Keane, O.M.,
and Dorman, C.J. (1999) Environmentally constrained
mutation and adaptive evolution in Salmonella. Curr Biol 9:
1477–1480.
van der Meer, J.R., de Vos, W.M., Harayama, S., and
Zehnder, A.J. (1992) Molecular mechanisms of genetic
adaptation to xenobiotic compounds. Microbiol Rev 56:
677–694.
Mortlock, R.P. (1984) Microorganisms as Model Systems for
Studying Evolution. New York, USA: Plenum Press.
Moxon, E.R., Rainey, P.B., Nowak, M.A., and Lenski, R.E.
(1994) Adaptive evolution of highly mutable loci in patho-
genic bacteria. Curr Biol 4: 24–33.
Nochur, S.V., Roberts, M.F., and Demain, A.L. (1990) Muta-
tion of Clostridium thermocellum in the presence of certain
carbon sources. FEMS Microbiol Lett 71: 199–204.
© 2004 Blackwell Publishing Ltd, Molecular Microbiology, 52, 643–650
Stress-directed mutations 649
Notley-McRobb, L., and Ferenci, T. (1999a) The generation
of multiple co-existing mal-regulatory mutations through
polygenic evolution in glucose-limited populations of
Escherichia coli. Environ Microbiol 1: 45–52.
Notley-McRobb, L., and Ferenci, T. (1999b) Adaptive mgl-
regulatory mutations and genetic diversity evolving in
glucose-limited Escherichia coli populations. Environ
Microbiol 1: 33–43.
Notley-McRobb, L., Seeto, S., and Ferenci, T. (2003) The
influence of cellular physiology on the initiation of muta-
tional pathways in Escherichia coli populations. Proc R Soc
Lond B Biol Sci 270: 843–848.
Pomposiello, P.J., Bennik, M.H., and Demple, B. (2001)
Genome-wide transcriptional profiling of the Escherichia
coli responses to superoxide stress and sodium salicylate.
J Bacteriol 183: 3890–3902.
Prival, M.J., and Cebula, T.A. (1992) Sequence analysis of
mutations arising during prolonged starvation of Salmo-
nella typhimurium. Genetics 132: 303–310.
Rahmouni, A.R., and Wells, R.D. (1992) Direct evidence for
the effect of transcription on local DNA supercoiling in vivo.
J Mol Biol 223: 131–144.
Ramos, J.L., Wasserfallen, A., Rose, K., and Timmis, K.N.
(1987) Redesigning metabolic routes: manipulation of TOL
plasmid pathway for catabolism of alkylbenzoates. Science
235: 593–596.
Ripley, L.S., and Glickman, B.W. (1983) Unique self-
complementarity of palindromic sequences provides DNA
structural intermediates for mutation. Cold Spring Harb
Symp Quant Biol 47 Part 2: 851–861.
Rosenberg, S.M., Longerich, S., Gee, P., and Harris, R.S.
(1994) Adaptive mutation by deletions in small mononucle-
otide repeats. Science 265: 405–407.
Rudner, R., Murray, A., and Huda, N. (1999) Is there a link
between mutation rates and the stringent response in
Bacillus subtilis? Ann N Y Acad Sci 870: 418–422.
Saier, M.H., Ramseier, T.M., and Reizer, J. (1996) Regula-
tion of carbon utilization. In Escherichia Coli and Salmo-
nella: Cellular and Molecular Biology, Vol. 1. Neidhardt,
F.C., Curtis, R. III, Ingraham, J.L., Lin, E.C.C., Low, K.B.,
Magasanik, B., et al. (eds). Washington, DC: American
Society for Microbiology, pp. 1325–1343.
Sung, H.M., and Yasbin, R.E. (2002) Adaptive, or
stationary-phase, mutagenesis, a component of bacterial
differentiation in Bacillus subtilis. J Bacteriol 184: 5641–
5653.
Torriani, A., and Rothman, F. (1961) Mutants of Escherichia
coli gene essential for survival in stationary phase. J Bac-
teriol 81: 835–840.
Weismann, A. (1893) Uber die Vererbung. Jena, Germany:
Gustav Fischer.
Wolf, A., Kramer, R., and Morbach, S. (2003) Three pathways
for trehalose metabolism in Corynebacterium glutamicum
ATCC13032 and their significance in response to osmotic
stress. Mol Microbiol 49: 1119–1134.
Wright, B.E. (2000) A biochemical mechanism for nonrandom
mutations and evolution. J Bacteriol 182: 2993–3001.
Wright, B.E., Longacre, A., and Reimers, J.M. (1999) Hyper-
mutation in derepressed operons of Escherichia coli K-12.
Proc Natl Acad Sci USA 96: 5089–5094.
Wright, B.E., Reimers, J.M., Schmidt, K.H., and Reschke,
650 B. E. Wright
D.K. (2002) Hypermutable bases in the p53 cancer gene
are at vulnerable positions in DNA secondary structures.
Cancer Res 62: 5641–5644.
Wright, B.E., Reschke, D.K., Schmidt, K.H., Reimers, J.M.,
and Knight, W. (2003) Predicting mutation frequencies in
stem-loop structures of derepressed genes: implications
for evolution. Mol Microbiol 48: 429–441.
Wright, B.E., Schmidt, K.H., and Minnick, M.F. (2004) Mech-
anisms by which transcription can regulate somatic hyper-
mutation. Genes Immun 5 (in press).

© 2004 Blackwell Publishing Ltd, Molecular Microbiology, 52, 643–650


References and Notes
1. N. D. Mathur et al., Nature 394, 39 (1998).
2. R. Movshovich et al., Phys. Rev. B 53, 8241 (1996).
3. H. Hegger et al., Phys. Rev. Lett. 84, 4986 (2000).
4. F. Steglich et al., Phys. Rev. Lett. 43, 1892 (1979).
5. P. Gegenwart et al., Phys. Rev. Lett. 81, 1501 (1998).
6. F. Thomas, J. Thomasson, C. Ayache, C. Geibel, F.
Steglich, Physica B 186 –188, 303 (1993).
7. D. Jaccard, H. Wilhelm, K. Alami-Yadri, E. Vargoz,
Physica B 259 –261, 1 (1999).
8. A. T. Holmes, A. Demuer, D. Jaccard, Acta Phys. Pol. B.
34, 567 (2003).
9. B. Bellarbi, A. Benoit, D. Jaccard, J. M. Mignot, H. F.
Braun, Phys. Rev. B 30, 1182 (1984).
10. H. Q. Yuan et al., Acta Phys. Pol. B 34, 533 (2003).
11. J. W. Allen, R. M. Martin, Phys. Rev. Lett. 49, 1106
(1982).
12. A. Onodera et al., Solid State Commun. 123, 113
(2002).
13. Materials and methods are available as supporting
material on Science Online.
14. H. Q. Yuan, unpublished data.
15. E. Vargoz, D. Jaccard, J. Y. Genoud, J.-P. Brison, J.
Flouquet, Solid State Commun. 106, 631 (1998).
16. J. Thomasson, Y. Okayama, I. Sheikin, J.-P. Brison,
D. Braithwaite, Solid State Commun. 106, 637
(1998).
17. Y. Onishi, K. Miyake, J. Phys. Soc. Jap. 69, 3955 (2001).
18. A. Holmes, D. Jaccard, K. Miyake, in preparation (avail-
able at http://xxx.lanl.gov/abs/cond-mat/0306054).
REPORTS
19. P. Monthoux, G. G. Lonzarich, in preparation (avail-
able at http://xxx.lanl.gov/abs/cond-mat/0309551).
20. O. Stockert et al., Acta Phys. Pol. B 34, 963 (2003).
21. We thank P. Gegenwart, G. G. Lonzarich, K. Miyake, P.
Monthoux, J. A. Mydosh, O. Stockert, and H. Wilhelm
for helpful discussions and S. Mederle for assistance
in some aspects of the high-pressure work.
Supporting Online Material
www.sciencemag.org/cgi/content/full/302/5653/2104/
DC1
Materials and Methods
References and Notes
18 September 2003; accepted 18 November 2003

Adaptation Limits Diversification niche 1 (I ⬎ 0.5) to diversify (i.e., for a

mutant specializing in niche 2 to be able to

of Experimental Bacterial invade and be maintained). For a variety of

different functional forms of Wi, we find that

Downloaded from http://science.sciencemag.org/ on April 12, 2017

the better adapted the population to niche 1
Populations (increasing I), the greater the phenotypic ef-
fect of a mutation required for invasion of
Angus Buckling,1* Matthew A. Wills,1 Nick Colegrave2 niche 2 (Fig. 1). If the probability of a muta-
tion is inversely proportional to its phenotyp-
Adaptation to a specific niche theoretically constrains a population’s ability to ic effect, then the specialization will make
subsequently diversify into other niches. We tested this theory using the diversification less likely.
bacterium Pseudomonas fluorescens, which diversifies into niche specialists We tested the hypothesis that adaptation
when propagated in laboratory microcosms. Numerically dominant genotypes within an existing niche in rugged fitness land-
were allowed to diversify in isolation. As predicted, populations increased in scapes constrains diversification using the com-
fitness through time but showed a greatly decreased ability to diversify. Sub- mon plant-colonizing bacterium Pseudomonas
sequent experiments demonstrated that niche generalists and reductions in fluorescens (8). Rejection of the hypothesis
intrinsic evolvability were not responsible for our data. These results show that would imply that the landscape is not very rug-
niche specialization may come with a cost of reduced potential to diversify. ged relative to the magnitude of mutational ef-
fect. When propagated in spatially structured
Despite much theoretical (1–6) and empirical fect, will be required for mutants to be com- environments (a static glass bottle containing
(7–14) work investigating the conditions pro- petitive when ascending an alternative peak, nutrient-rich medium), isogenic P. fluorescens
moting the evolution of sympatric (within- reducing the potential to diversify. populations rapidly diversify, generating numer-
population) diversity, little is known about Consider an environment with two avail- ous niche specialist genotypes that are readily
the constraints. Here, we demonstrate that able niches (1 and 2) with the population distinguished by their (heritable) colony mor-
adaptation itself is likely to limit a popula- being regulated at the level of the niche, such phologies on agar plates (8). The genotypes can
tion’s ability to diversify, when evolution that half the population derives from each be grouped into three distinct classes on the
occurs in “rugged fitness landscapes.” Fitness niche. The fitness of an individual in each basis of colony morphology and niche occupa-
landscapes are created by plotting fitness niche (Wi) depends on the value of a pheno- tion. Smooth (SM) morphs resemble the ances-
against all possible genotypes within the pop- typic trait (I), which can vary between 0 and tral genotype and largely inhabit the same, liq-
ulation (4–6, 15, 16). Assuming large popu- 1, and there is a trade-off between fitness in uid-phase, niche; wrinkly-spreader (WS)
lations, selection drives populations uphill each niche such that W1 ⫽ f (I ) and W2 ⫽ morphs form a mat at the air-broth interface, and
toward fitness peaks (adaptation). If several f (1 ⫺ I ). Using Levene’s classic model of the much less frequent fuzzy-spreader (FS)
ecological niches exist and organisms show species coexistence (1), we can consider the morphs colonize the harsher, less aerobic bot-
fitness trade-offs between niches (niche spe- size of the phenotypic (mutational) change tom of the vials. Fitness trade-offs between
cialization), or genetic interactions (epistasis) required for a population partially adapted to niches have been demonstrated (8, 17) by the
are a major determinant of fitness, the fitness
landscape may be rugged, with multiple Fig. 1. Magnitude of mutational ef-
peaks and valleys (4–6, 15, 16). When a fect required for adaptation to a
population is poorly adapted and in a fitness new niche, hence diversification, as
valley, multiple peaks can be ascended, hence a function of adaptation to niche 1
the potential for high rates of diversification. (population phenotype). Different
lines show different relationships
However, as the population ascends a peak, between phenotype (I) and fit-
mutations of increasingly large effect, or a ness in niche 1 (W1): W1 ⫽ I (F);
number of concurrent mutations of small ef- W1 ⫽ I 2 (‘); W1 ⫽ I 0.5 (E).
Fitness functions in niche 2 (W2) are
the same as in niche 1, but with I
1
Department of Biology and Biochemistry, University replaced with 1 ⫺ I.
of Bath, Bath BA2 7AY, UK. 2Institute of Cell, Animal
and Population Biology, University of Edinburgh,
King’s Buildings, Edinburgh EH9 3JT, UK.
*To whom correspondence should be addressed. E-
mail: bssagjb@bath.ac.uk

www.sciencemag.org SCIENCE VOL 302 19 DECEMBER 2003 2107

 REPORTS
operation of negative frequency-dependent se-
lection; genotypes have a fitness advantage
when rare because there is less intense compe-
tition within their niche (18, 19).
To investigate whether adaptation within
existing niches constrained diversifications, we
initiated six P. fluorescens populations with a
single clone [SBW25 (20)] and propagated
them for a week in spatially heterogeneous
environments. Diversity was determined from
colony morphologies on agar plates (8, 10, 17,
21), and the numerically most dominant SM-
like genotype in each population was isolated
and used to initiate a single subsequent culture,
which was propagated as above. This procedure
was repeated five times for each of the six
replicate lines, resulting in one genotype from
each of six transfers for each line (a total of 36
suites of novel genotypes when they were al-
lowed to diversify in replicate (22), suggesting
that the reduction in diversification is due, to
some extent, to a deterministic loss in accessi-
ble genotypes during the climbing of a partic-
ular adaptive peak.
Our proposed explanation for this de-
creasing ability to diversify (with increasing
transfer number) is niche-specific adaptation.
We tested this hypothesis by competing all
genotypes from all time points with a genet-
ically marked isogenic mutant of the ances-
tral genotype in the selective environment
(22). Fitness of genotypes did indeed increase
through time in all replicates (Fig. 2B)
(paired t test of ancestral versus transfer 6
selection rate constant: t ⫽ 15.27, P ⬍
0.0001; sign test of correlations between fit-
ferring 1% of the complete culture to fresh
media, rather than individual genotypes. We
found no evidence of diversity decreasing
through time, refuting generalist evolution as
the main explanation for our data (sign test
of correlations between number of genotypes
and transfer number for the six lines:
P ⬎ 0.2).
A second explanation is an intrinsic re-
duction in evolvability (for example, reduced
mutation rates) through time, possibly as a
result of genetic drift caused by our selection
regime. Such intrinsic reductions are likely to
be apparent in all environments; hence, the
same relationships between time spent evolv-
ing and diversification should be apparent in
a novel environment. Conversely, if ascend-
ing local fitness peaks are responsible for the

Downloaded from http://science.sciencemag.org/ on April 12, 2017

evolved clones) (22). Consistent with the
evolved genotypes occupying the SM niche, the
six clones from the final transfer were able to
competitively exclude ancestral SM (22). In all
lines, diversification after six transfers was
much less than ancestral diversification, and
there was an overall negative correlation be-
tween transfer number and ability to diversify
(Fig. 2A) (sign test of correlations between
number of genotypes and transfer number for
the six lines: P ⫽ 0.03). Different colony types
dominated in different populations, suggesting
the ascent of different adaptive peaks. Howev-
er, individual evolved genotypes (that showed
little diversification) produced almost identical
ness and transfer number for the six lines:
P ⫽ 0.03). Furthermore, there was an overall
negative correlation between fitness and abil-
ity to diversify (sign test of correlations be-
tween fitness and number of genotypes for
the six lines: P ⫽ 0.03).
It is possible that the inability to diversify
is the result of types evolving that are supe-
rior in all niches (23). If this were the case,
we would expect diversity to decrease (as a
result of the evolution of generalists) under
conditions in which the selection regime in-
volves propagation of a sample of the whole
population. We evolved six populations un-
der identical conditions as above, but trans-
inability to diversify (our hypothesis), diver-
sification in a novel environment should be
independent of adaptation to the selective
environment: All genotypes are equally un-
likely to be approaching fitness peaks in the
fitness landscape created by the novel envi-
ronment. To examine this hypothesis, we
measured the ability of each of the genotypes
to diversify in two novel environments: in
spatially homogeneous environments (shaken
microcosms) and in spatially heterogeneous
environments (static microcosms) with low
nutrients (22). Diversification of the ancestral
genotype was considerably less in both envi-
ronments than in the original environment

Fig. 2. Diversification and fit-

ness through time. (A) The
ability of genotypes to diver-
sify and (B) their fitness
(selection rate constant)
against a standard competi-
tor, where a value of zero in-
dicates equal performance of
the two competing geno-
types, after 1 week of growth
in static KB media micro-
cosms, as a function of trans-
fer number. Different lines
and symbols represent the six
replicate lines.

Fig. 3. Diversification in novel

environments. The ability of
genotypes to diversify in two
novel environments: (A) shak-
en KB media microcosms and
(B) static minimal media mi-
crocosms as a function of
transfer number. Different
lines and symbols represent
the six replicate lines.

2108 19 DECEMBER 2003 VOL 302 SCIENCE www.sciencemag.org


(compare Figs. 2A and 3). Nevertheless, in
both environments there was no relation be-
tween ability to diversify and time spent
evolving in the original high-nutrient media
(Fig. 3) (sign tests of correlations between
number of genotypes and transfer number for
the six lines: P ⬎ 0.2 in both cases).
Adaptation can limit the ability of bac-
terial genotypes to diversify genetically.
This was not the result of generalist evolu-
tion or the evolution of an intrinsic reduc-
tion in evolvability, but was caused by
environment-specific adaptation. Given the
strong empirical support for both the im-
portance of environmental heterogeneity in
diversification (7–14) and epistasis (6, 24),
it is likely that rugged fitness landscapes, a
requirement for the observed effects, are
common. These results are therefore likely
to be generally relevant and may help to
explain patterns of diversity over both mi-
cro- and macroevolutionary time scales.
Consistent with recent interpretations of
macroevolutionary adaptive radiations
(25), we predict that in environments that
can potentially support similar levels of
diversity, diversification is more likely to
occur immediately following colonization
of the environment than through expansion
into new niches within the environment
after an extinction event.
References and Notes
1. H. Levene, Am. Nat. 87, 331 (1953).
2. P. W. Hedrick, M. E. Ginevan, E. P. Ewing, Annu. Rev.
Ecol. Syst. 7, 1 (1976).
3. P. W. Hedrick, Annu. Rev. Ecol. Syst. 17, 535 (1986).
4. T. Dobzhansky, Genetics and the Origin of Species
(Columbia Univ. Press, New York, 1937).
5. M. C. Whitlock, P. C. Phillips, F. B. G. Moore, S. J.
Tonsor, Annu. Rev. Ecol. Syst. 26, 601 (1995).
6. M. J. Wade, in Epistasis and the Evolutionary Process,
J. B. Wolf, E. D. Brodie, M. J. Wade, Eds. (Oxford Univ.
Press, Oxford, 2000), pp. 213–231.
7. R. Korona, C. H. Nakatsu, L. J. Forney, R. E. Lenski,
Proc. Natl. Acad. Sci. U.S.A. 91, 9037 (1994).
8. P. B. Rainey, M. Travisano, Nature 394, 69 (1998).
9. D. E. Rozen, R. E. Lenski, Am. Nat. 155, 24 (2000).
10. A. Buckling, P. B. Rainey, Nature 420, 496 (2002).
11. D. Schluter, Am. Nat. 157 (suppl.), S4 (2000).
12. S. M. Vamosi, D. Schluter, Proc. R. Soc. London Ser. B
269, 923 (2002).
13. R. F. Rosenzweig, R. R. Sharp, D. S. Treves, J. Adams,
Genetics 137, 903 (1994).
14. R. Kassen, J. Evol. Biol. 15, 173 (2002).
15. S. Wright, Genetics 16, 97 (1931).
16. H. J. Muller, Biol. Rev. Camb. Philos. Soc. 14, 261
(1939).
17. A. Buckling, R. Kassen, G. Bell, P. B. Rainey, Nature
408, 961. (2000).
18. F. J. Ayala, C. A. Campbell, Annu. Rev. Ecol. Syst. 5,
115 (1974).
19. M. L. Rosenzweig, Species Diversity in Space and Time
(Cambridge Univ. Press, Cambridge, 1995).
20. P. B. Rainey, M. J. Bailey, Mol. Microbiol. 19, 521 (1996).
21. R. Kassen, A. Buckling, G. Bell, P. B. Rainey, Nature
406, 508 (2000).
22. Materials and methods are available as supporting
material on Science Online.
23. G. Bell, Selection: The Mechanism of Evolution (Chap-
man & Hall, London, 1997).
24. A. R. Templeton, in Epistasis and the Evolutionary
Process, J. B. Wolf, E. D. Brodie, M. J. Wade, Eds.
(Oxford Univ. Press, Oxford, 2000).
www.sciencemag.org SCIENCE VOL 302 19 DECEMBER 2003
REPORTS
25. D. Schluter, The Ecology of Adaptive Radiations (Ox- Supporting Online Material
ford Univ. Press, Oxford, 2000). www.sciencemag.org/cgi/content/full/302/5653/2107/DC1
26. We thank L. Hurst and three anonymous referees for Materials and Methods
comments on the manuscript. This work was funded References
by the Royal Society and Natural Environment Re-
search Council (UK). 7 July 2003; accepted 14 October 2003
Melt Inclusions in Veins: Linking
Magmas and Porphyry Cu Deposits
Anthony C. Harris,1* Vadim S. Kamenetsky,1 Noel C. White,1
Esmé van Achterbergh,2 Chris G. Ryan2
At a porphyry copper-gold deposit in Bajo de la Alumbrera, Argentina, silicate-
melt inclusions coexist with hypersaline liquid- and vapor-rich inclusions in the
earliest magmatic-hydrothermal quartz veins. Copper concentrations of the
Downloaded from http://science.sciencemag.org/ on April 12, 2017
hypersaline liquid and vapor inclusions reached maxima of 10.0 weight % (wt
%) and 4.5 wt %, respectively. These unusually copper-rich inclusions are
considered to be the most primitive ore fluid found thus far. Their preservation
with coexisting melt allows for the direct quantification of important ore-
forming processes, including determination of bulk partition coefficients of
metals from magma into ore-forming magmatic volatile phases.
In porphyry ore deposits, metals are con- Bajo de la Alumbrera is an Au-rich porphyry
centrated by large volumes of magmatic Cu deposit where potassic alteration assemblag-
volatiles exsolved from crystallizing upper es overprint several phases of porphyritic dacite
crustal magma bodies (1). Whether an ore intrusions and are associated with the bulk of the
deposit forms depends on the availability of disseminated Cu-Fe sulfides and Au. High tem-
metals in the magma, the partitioning of perature (maximum of 750°C) and saline fluid
those metals into the volatile phase, and the (⬎35 wt % equivalent NaCl) of magmatic ori-
history of the fluid after release from the gin (as inferred from the calculated ␦18O and ␦D
magma (2– 4 ). Ores are typically associated compositions) formed these alteration assem-
with hydrothermal mineral assemblages blages (8). Some of the earliest Cu-Fe sulfides
produced by the interaction of magmatic occur in diffuse quartz veins, which are textur-
fluids with wall rocks (5). The earliest ally similar to those described as A veins (9);
formed hydrothermal alteration of wall however, the presence of silicate-melt inclusions
rocks is potassic (biotite-K-feldspar-quartz warrants a new vein subclass. Hereafter we refer
with or without magnetite assemblage) and to them as P veins, reflecting their primitive role
is caused by high temperature (350° to in the evolution of the magmatic-hydrothermal
800°C) and saline (up to 70 weight % system. Typically, they consist of sugary quartz,
equivalent NaCl) fluids (6, 7 ). Discovery with lesser amounts of K-feldspar and with or
of silicate-melt inclusions coexisting with without hornblende-biotite-magnetite-chalcopy-
fluid inclusions in magmatic-hydrothermal rite (CuFeS2) and pyrite.
quartz veins unambiguously links devola- The P veins contain silicate-melt inclusions
tilization of the magma with the associated that consist of silicate crystals, vapor bubbles,
porphyry ore deposit. This occurrence pre- salt crystals, and opaque oxide and sulfide crys-
serves the most primitive magmatic vola- tals (Fig. 1). These inclusions are similar to
tiles and the melt from which these were those in magmatic quartz phenocrysts in the
derived and, with the use of advanced mi- mineralized intrusions (10). Heating experi-
croanalysis techniques, allows chemical ments (11) on silicate-melt inclusions revealed
changes to be traced through the evolution consistent phase transformations: The dissolu-
of the hydrothermal system. Moreover, we tion of the salt phases (e.g., halite and sylvite)
are able to use those data to quantify the occurs between 105° and 560°C; dissolution of
magmatic-hydrothermal processes that lead crystalline silicate phases occurs between 650°
to the formation of porphyry Cu deposits. and 765°C. After heating the host quartz crys-
tals to 800°C at 1 atm external pressure for
1
Centre for Ore Deposit Research, University of several hours, the sample was quenched to pro-
Tasmania, Private Bag 79, Hobart, Tasmania 7001,
duce a silicate glass, which is extremely rich in
Australia. 2Commonwealth Scientific and Industrial
Research Organization Exploration and Mining, Post
K2O compared to whole rock analyses and has
Office Box 136, North Ryde, New South Wales a composition similar to K-feldspar (table
1670, Australia. S2). This composition may represent a che-
*To whom correspondence should be addressed. E- mically modified melt trapped in a dominantly
mail: A.Harris@utas.edu.au aqueous environment or may contain ex-
2109
 Adaptation Limits Diversification of Experimental Bacterial
Populations
Angus Buckling, Matthew A. Wills and Nick Colegrave (December 18,
2003)
Science 302 (5653), 2107-2109. [doi: 10.1126/science.1088848]

Editor's Summary

This copy is for your personal, non-commercial use only.

Downloaded from http://science.sciencemag.org/ on April 12, 2017

Article Tools Visit the online version of this article to access the personalization and article
tools:
http://science.sciencemag.org/content/302/5653/2107

Permissions Obtain information about reproducing this article:

http://www.sciencemag.org/about/permissions.dtl

Science (print ISSN 0036-8075; online ISSN 1095-9203) is published weekly, except the last week in
December, by the American Association for the Advancement of Science, 1200 New York Avenue NW,
Washington, DC 20005. Copyright 2016 by the American Association for the Advancement of Science;
all rights reserved. The title Science is a registered trademark of AAAS.
 REVIEW ARTICLE

The role of variable DNA tandem repeats in bacterial

adaptation
Kai Zhou, Abram Aertsen & Chris W. Michiels
Department of Microbial and Molecular Systems (M²S), Faculty of Bioscience Engineering, Laboratory of Food Microbiology and Leuven Food
Science and Nutrition Research Centre (LFoRCe), KU Leuven, Leuven, Belgium

Correspondence: Chris W. Michiels, Abstract

Laboratory of Food Microbiology, Kasteelpark
Arenberg 23, B-3001 Leuven, Belgium.
Tel.: +32 16 321578;
fax: +32 16 321960;
e-mail: chris.michiels@biw.kuleuven.be
Received 17 April 2013; revised 13 July
2013; accepted 26 July 2013. Final version
published online 28 August 2013.
DOI: 10.1111/1574-6976.12036
Editor: Grzegorz Wegrzyn
Keywords
polymorphic DNA; contingency loci; host–
pathogen interaction; stress tolerance;
evolution; phase variation.
MICROBIOLOGY REVIEWS
DNA tandem repeats (TRs), also designated as satellite DNA, are inter- or
intragenic nucleotide sequences that are repeated two or more times in a head-
to-tail manner. Because TR tracts are prone to strand-slippage replication and
recombination events that cause the TR copy number to increase or decrease,
loci containing TRs are hypermutable. An increasing number of examples illus-
trate that bacteria can exploit this instability of TRs to reversibly shut down or
modulate the function of specific genes, allowing them to adapt to changing
environments on short evolutionary time scales without an increased overall
mutation rate. In this review, we discuss the prevalence and distribution of
inter- and intragenic TRs in bacteria and the mechanisms of their instability.
In addition, we review evidence demonstrating a role of TR variations in bacte-
rial adaptation strategies, ranging from immune evasion and tissue tropism to
the modulation of environmental stress tolerance. Nevertheless, while bioinfor-
matic analysis reveals that most bacterial genomes contain a few up to several
dozens of intra- and intergenic TRs, only a small fraction of these have been
functionally studied to date.

the latter strategy is that random mutations are more

Introduction
To cope with rapidly changing environmental conditions
and ensure their survival, unicellular organisms have
evolved a plethora of adaptation strategies (Aertsen &
Michiels, 2004, 2005). Most of these strategies are based
on transient alterations in gene expression in response to
stressful conditions, and well-known examples include the
SOS response (controlled by RecA and LexA), the general
stress response (regulated by the sigma factor RpoS), the
stringent response (mediated by pppGpp and ppGpp),
and the heat shock response (mainly controlled by RpoH;
Massey & Buckling, 2002; Foster, 2005; Saint-Ruf &
Matic, 2006; Foster, 2007; Jolivet-Gougeon et al., 2011).
In addition, adaptation can also stem from the acquisi-
tion of stochastic mutations that alter the genotype,
which become positively selected and fixed in a popula-
tion if they coincide with a beneficial phenotype (Rando
& Verstrepen, 2007). However, an important drawback of
often deleterious than beneficial.
Interestingly, neither the type nor the frequency of
mutational events is randomly distributed over the
genome, and some DNA sequences have evolved to be
mutational hotspots that drive the variability of genes
whose activity can impact the adaptive potential of their
host. One type of such special sequences that is very
abundant in prokaryotic and eukaryotic genomes is
known as tandem repeats (TRs), a major class of direct
DNA repeats. While at first TRs were considered to be
junk DNA without any biological function, studies of the
human genome have revealed some of these repetitive
sequences to be hypermutable and the cause of diseases
such as fragile X syndrome, spinobulbar muscular atro-
phy, and huntington disease (Hannan, 2010). In addition,
accumulating evidence points out the potential role of
TRs as engines of genetic variability and bacterial
adaptation. In this review, we therefore focus on the

FEMS Microbiol Rev 38 (2014) 119–141 ª 2013 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
120
identification and distribution of TRs in bacteria, the
mechanisms behind their variability, and their biological
significance in bacterial adaptation.
Identification and distribution of TRs in
bacterial genomes
Definition of TRs
TRs are nucleotide sequences that are directly repeated in
a head-to-tail manner. According to the conservation of
the repeated sequence, TRs are classified as identical/per-
fect TRs or degenerated/imperfect TRs, respectively
(Fig. 1). Furthermore, TRs are commonly classified into
three categories according to their repeat unit size,
although there is no consensus definition of these catego-
ries (Richard et al., 2008). Repeats with unit size varying
from 1 to 9, 10 to 100, and > 100 bp are termed micro-
satellites, minisatellites, and macrosatellites, respectively
(Lopes et al., 2006). The term ‘satellite DNA’ originally
refers to the very large arrays of tandemly repeated non-
coding DNA (often hundreds of copies) that are charac-
teristic of large eukaryotic genomes, but, in the context of
bacterial genomes, is also used to include small and intra-
genic TRs.
In silico identification of TRs
The increasing availability of genome sequences and
specialized bioinformatics software greatly facilitates the
search and identification of TR loci on a genomewide
scale, which obviously is a prerequisite for understanding
their distribution, predicting their function, and tracking
(a)
IdenƟcal / Perfect TRs AGCTG
(unit sequence = 100% conserved)
Degenerated / Imperfect TRs AGCTG
(unit sequence < 100% conserved)
(b) Microsatellite
(unit size 1-9 bp)
Minisatellite
(unit size 10-100 bp)
Macrosatellite
(unit size > 100 bp)
Fig. 1. Schematic representation of different types of TRs. (a) Different conservation of repeat unit sequence. (b) Different sizes of repeat unit.
Space between repeat units has only been introduced to improve visual clarity of the figure.
ª 2013 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
K. Zhou et al.
their evolution. A variety of algorithms have been devel-
oped for detecting TRs, but it is important to be aware
that these may differ in their ability to detect different
types of TRs (Merkel & Gemmell, 2008; Treangen et al.,
2009; Kajava, 2012). Hence, the choice of search tool
should be determined by the TR type of interest, or the
parallel use of several algorithms is advisable when a wide
screen for TRs is performed. Furthermore, parameter set-
tings (i.e. alignment weights, definition of repeats, and
threshold scores) can also strongly affect outcome in terms
of number and consensus motif of TRs (Lim et al., 2013).
In particular, problems are still commonly encountered in
detecting imperfect TRs (Leclercq et al., 2007; Schaper
et al., 2012). Several algorithms are freely available
online, such as Tandem Repeat Finder (Benson, 1999) and
IMEx (Mudunuri & Nagarajaram, 2007). In addition,
several databases of annotated TRs in prokaryotes have
been established, such as TRs DB (http://minisatellites.
u-psud.fr), PSSRDb (http://pssrdb.cdfd.org.in) and
MICAS (http://180.149.48.108/micas/index.php). In the
next section, we will review the major findings from some
recent in silico studies of the genomic distribution of TRs
in bacteria.
The distribution of TRs in bacterial genomes
The analysis of TRs in bacterial genomes so far has
mainly focused on microsatellites with unit size 1–6 bp,
also termed ‘simple sequence repeats (SSRs)’. A number
of general observations regarding the distribution of SSRs
can be stated. First of all, the abundance of SSRs in bacte-
ria is lower than that in eukaryotes (Schlotterer et al.,
2006). Nevertheless, the number of SSRs is orders of
AGCTG AGCTG AGCTG AGCTG
TGCTG AGGTG AGCTG AGCTC
FEMS Microbiol Rev 38 (2014) 119–141
Variable tandem repeats and bacterial adaptation
magnitude higher than that of other repeat types (i.e.
minisatellite and macrosatellite) in the genomes of most,
if not all bacteria. Of course, this is not unexpected
because this SSR count included even mononucleotide
trimers (e.g. AAA), which account for about 70% of the
total number of SSRs. While SSRs are generally believed
to contribute to genome polymorphism and adaptation
potential of bacteria (Kassai-J
ager et al., 2008), the contri-
bution of these very small SSRs like mononucleotide
trimers is probably limited. In fact, a rough threshold of
minimum TR unit number (4–9) has been noted, below
which a SSR is not likely to mutate or be variable (Lai &
Sun, 2003; Dettman & Taylor, 2004; Kelkar et al., 2010).
Intriguingly, heptameric repeats were found to be over-
represented among these SSRs in most prokaryotes, and
it was hypothesized that the seemingly preferred 7 bp
length of a repeat unit might relate to the DNA segment
size that interacts with the active site of the DNA poly-
merase, thus facilitating the occurrence of polymerase
slippage (Mr
azek et al., 2007).
A remarkable feature of SSRs is their widely diverse dis-
tribution across species, even closely related ones, and this
may indicate that they are subject to rapid evolutionary
change (Yang et al., 2003; Mr
azek, 2006; Kassai-J
ager et al.,
2008). Analysis of more than 300 prokaryotic genomes
showed that the distribution of SSRs varied with the bacte-
rial species, genome size, and G + C content (Mr
azek
et al., 2007). More specifically, SSRs with small motif
(1–4 bp) are more abundant in small genomes and particu-
larly in host-adapted pathogens with reduced genomes
(< 2 Mb) and low G + C content (< 40%), such as Myco-
plasma and Haemophilus spp. (Moxon et al., 2006; Trean-
gen et al., 2009). In contrast, SSRs with a larger motif
(5–11 bp) are more frequent in nonpathogens and oppor-
tunistic pathogens with large genomes (> 4 Mb) and high
G + C content (> 60%), such as Burkholderia and Anabae-
na spp. Based on this observation, it was hypothesized that
the differential representations of SSRs in bacteria may cor-
relate with pathogenicity, but more work is needed to cor-
roborate this. Another interesting observation is that some
relatively large bacterial genomes (e.g. Pseudomonas aeru-
ginosa, c. 5 Mb) have fewer SSRs than would be predicted
based on their genome properties, but harbor compara-
tively more two-component sensor transducers. In
contrast, some host-adapted pathogens with small genome
size (i.e. Haemophilus influenzae, Neisseria meningitidis,
and Helicobacter pylori) have comparatively more SSRs, but
less two-component sensor transducers (Moxon et al.,
2006). Thus, it seems that environmental adaptability in
host-adapted pathogens depends primarily on SSR varia-
tions, while in opportunistic pathogens with a more versa-
tile lifestyle it depends primarily on two-component sensor
transducers.
FEMS Microbiol Rev 38 (2014) 119–141
121
Closer examination of the SSR distribution across the
genome shows significant differences in coding and non-
coding regions. Because bacterial genomes are more com-
pact than those of eukaryotes, they have comparatively
more intragenic than intergenic SSRs. For example, in
Escherichia coli K-12, 79.5% of SSRs locate in coding
regions (Gur-Arie et al., 2000), whereas in the genome
of the Japanese pufferfish (Fugu rubripes), only 11.6% of
SSRs are intragenic (Edwards et al., 1998). Generally, long
mono- and dinucleotide SSRs are excluded from coding
regions, probably because they have a higher probability
to rearrange and cause frameshift mutations in genes (Co-
enye & Vandamme, 2005; Ackermann & Chao, 2006; Orsi
et al., 2010; Lin & Kussell, 2012). In contrast, SSRs whose
unit size is a multiple of three nucleotides (3, 6, 9 …) are
overrepresented in open reading frames (ORFs) because
their expansion or contraction does not disrupt the read-
ing frame (Mr
azek et al., 2007). However, exceptions have
been reported. For example, tetranucleotide SSRs of H. in-
fluenzae are exclusively found in ORFs, which is consistent
with their role in phase variation (Power et al., 2009). An
interesting situation exists in the mycoplasmas, where long
trinucleotide repeats are overrepresented in Mycoplasma
genitalium, Mycoplasma gallisepticum, and Mycoplasma
hyopneumoniae, but occur mainly in intergenic regions in
the former two species, but in coding regions in the latter
one (Mr
azek, 2006). This difference in distribution is also
reflected in different functional roles. In M. gallisepticum,
the most prominent trinucleotide TRs are the GAA
repeats in the 5′ untranslated region of the 42 up to 70
vlpA adhesin gene paralogs that exist in each strain, which
regulate vlpA gene expression (Glew et al., 1998, 2000;
Liu et al., 2000; Papazisi et al., 2003). In contrast,
M. hyopneumoniae trinucleotide repeats are found mostly
within hypothetical ORFs, but also in some adhesins, and
their contraction or expansion results in variability of
amino acid repeats that are believed to play a role in pro-
tein–protein interaction or adhesion (Mr
azek, 2006).
A more detailed study on the occurrence of intragenic
TRs in 44 bacteria and archaea revealed additional
features (Lin & Kussell, 2012). Intragenic SSRs were
found more frequently near the termini (5′ and 3′ ends)
of the ORF rather than in the middle, which most likely
stems from biophysical constraints of protein structure.
In addition, SSR-induced frameshifts at the 3′ end are less
harmful than at other parts, because most of the
upstream coding region will not be affected. Nevertheless,
an overrepresentation of SSRs was found in the 5′ end in
ORFs of pathogens, probably because this allows SSR-
induced frameshifts to function as an ON/OFF switch for
these ORFs, which can be advantageous for pathogens
because it facilitates rapid adaptation of populations.
Similar observations had already been made earlier for
ª 2013 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
122
some intragenic mononucleotide repeats at the 5′ end of
genes (van Passel & Ochman, 2007; Janulczyk et al., 2010;
Orsi et al., 2010). However, it remains unclear and often
difficult to prove whether this type of distribution bias of
intragenic SSRs is linked to selection pressures in bacteria.
An argument in favor of such a link is that intragenic
SSRs show a preference for certain categories of genes. In
both Gram-negative (e.g. Haemophilus and Helicobacter)
and Gram-positive (e.g. Streptococcus) pathogens, SSR-
associated genes frequently encode virulence factors, cell
surface components, and restriction–modification
enzymes (van Belkum, 1999; Moxon et al., 2006; Guo &
Mr
azek, 2008; Power et al., 2009; Janulczyk et al., 2010).
On the other hand, several intragenic SSRs with numer-
ous repeat copies and a unit size that is not a multiple of
three are also found in housekeeping genes whose prod-
ucts are essential for important cellular processes, such as
cell division, energy production, and DNA replication
and repair (Guo & Mr
azek, 2008). Obviously, corre-
sponding TR rearrangements leading to reading frame
disruption are anticipated to be detrimental or even lethal
for the cell, and it remains unclear why such TRs have
been maintained during evolution.
Intergenic SSRs also show a nonrandom distribution,
being found more frequently in the immediate vicinity of
genes than at distant positions. For example, intergenic
SSRs of E. coli K-12 concentrate in a region up to 200 bp
from the start codon, which contains proximal regulators
of gene expression (Gur-Arie et al., 2000). Another study
showed that in most cases, the intergenic SSRs with
numerous copies are located upstream of the first gene in
prokaryotic operons (Guo & Mr
azek, 2008). Together,
both studies reflect the potential role of intergenic SSRs
in the regulation of gene expression.
The variability of TRs
The variability of TRs is thought to be one of the drivers
of genomic plasticity. The regions containing TRs are
potentially hypermutable by contraction (deletion) or
expansion (insertion) of TR units, and mutation frequen-
cies up to 10 1 have been reported in bacteria (Rando &
Verstrepen, 2007). Not surprisingly, the polymorphisms
found in TR loci provide a foundation for DNA genotyp-
ing approaches such as variable number TR-based typing
or multilocus variable repeat analysis, which are com-
monly applied for pathogen typing (Lindstedt, 2005,
2011; reviewed in Chiou, 2010).
The molecular mechanisms of TR variation
Based on extensive studies in both plasmid-based and
chromosomal systems, two nonexclusive mechanisms,
ª 2013 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
K. Zhou et al.
replication slippage and recombination, are currently
widely accepted to explain TR variation (Pearson et al.,
2005; Bichara et al., 2006; Gemayel et al., 2010). With
regard to the slippage mechanism, several models have
been proposed, and the most common one is strand-
slippage mispairing (SSM), also called DNA slippage or
polymerase slippage (Kornberg et al., 1964; Streisinger
et al., 1966). This model proposes that TR rearrange-
ments result from strand slippage during DNA replica-
tion. The process is initiated by the formation of a bulge
structure of unpaired repeats either on the template or
on the nascent strand. If the bulge is present on the tem-
plate strand, it will result in a TR contraction (deletion)
in the newly synthesized DNA. In contrast, TR expansion
(insertion) will result when a bulge forms on the nascent
strand (Fig. 2). Currently, substantial evidence supports
the involvement of replication in TR instability in bacte-
ria. In E. coli, for example, triplet nucleotide tracts in a
plasmid or in the chromosome were dramatically destabi-
lized in a dnaQ49 mutant. The dnaQ gene encodes the
3′–5′ exonucleolytic e-subunit of DNA polymerase III,
which is involved in proofreading, and it was therefore
suggested that this mutant failed to correctly remove
slipped structures in the TR tracts during replication (Iyer
et al., 2000; Zahra et al., 2007). Moreover, mutations in
the a subunit (encoded by dnaE) of DNA polymerase III
holoenzyme, c and s subunits of the clamp-loading com-
plex (encoded by dnaX), and b clamp (encoded by dnaN)
have also been shown to increase instability of microsatel-
lites and tandemly repeated DNA sequences (reviewed in
Bichara et al., 2006). In general, the effect of DNA repli-
cation on TR rearrangements provides evidence for the
replication slippage mechanism, because a prerequisite of
this mechanism is that DNA replication is stalled by the
secondary loop structures formed by TRs. Notably, the
SSM mechanism is not only widely accepted to account
for the majority of SSR variations, it is also invoked to
explain the genesis of TRs from unrepeated DNA (Levin-
son & Gutman, 1987; Waite et al., 2003; Lindb€ack et al.,
2011).
Besides SSM, recombination is considered as another
mechanism of TR instability that could explain phenom-
ena that cannot be explained by the slippage mechanism.
Generally, recombination is more important for the rear-
rangement of TRs with large unit size, whereas SSM is
the dominant mechanism underlying variation of TRs
with small unit size (Bi & Liu, 1996; Richard & P^aques,
2000; Bzymek & Lovett, 2001; Rocha, 2003; Gemayel
et al., 2010). Both homologous (RecA-dependent) and
illegitimate (RecA-independent) recombination can be
involved. Several models have been proposed for the
recombination mechanism, such as unequal crossover and
intramolecular recombination (Fig. 3), and evidence for
FEMS Microbiol Rev 38 (2014) 119–141
Variable tandem repeats and bacterial adaptation
5’
3’
5’
3’
Expansion
5’ 3’
3’
Fig. 2. Diagram illustrating the replication slippage mechanism of TR rearrangement. Repeat units are shown as blocks on nascent (light green)
and template (dark green) DNA strands. Shown is a partially replicated TR region undergoing transient dissociation and mispairing, resulting in a
bulge on the template or the nascent strand and leading to insertion or deletion of TRs, respectively. Space between repeat units has only been
introduced to improve visual clarity of the figure.
the involvement of recombination in TR variations is
accumulating. For example, it has been suggested that
double-strand DNA break (DSB) repair can induce TR
instability via homologous recombination in prokaryotes
and eukaryotes (Hebert & Wells, 2005; reviewed in Rich-
ard et al., 2008; Malkova & Haber, 2012), and mechanis-
tic models explaining this process have been elaborated.
One such model is the DSB repair slippage model, which
combines the double Holliday junction intermediate
pathway with the strand-slippage model (Richard et al.,
2008). In an alternative explanation, the synthesis-depen-
dent strand annealing pathway also contributes to the TR
rearrangements mediated by DSB repair (P^aques et al.,
1998, 2001; Richard et al., 1999; Richard & P^aques,
2000). This model is proposed for explaining TR rear-
rangements that are rarely associated with crossover
events (P^aques et al., 1998, 2001; Richard et al., 1999).
Factors influencing TR rearrangement
frequencies
While TRs generally are intrinsically prone to incur con-
traction or expansion, the actual frequency of these events
FEMS Microbiol Rev 38 (2014) 119–141
123
3’ ReplicaƟon
5’
3’
Strand
dissociaƟon
5’
ContracƟon
5’ 3’
5’ 3’ 5’
can vary widely depending on both intrinsic (structural)
and extrinsic (environmental) factors. Regarding intrinsic
TR features, a positive correlation has been established
between TR copy number and rearrangement frequency
in several studies. Through in silico genome analysis of
multiple strains of 42 fully sequenced prokaryotic species,
Lin & Kussell (2012) observed that the variability of three
types of SSRs (monomeric, dimeric, and trimeric repeats)
increased dramatically with the number of repeat units. In
another study, an exponential relation between the num-
ber of repeat units and rearrangement frequency was
observed in a comparison of 30 artificially constructed
TRs (unit lengths of 2, 10, and 20 nucleotides; number of
units between 2 and 50; sequence conservation between
62.5% and 100%; Legendre et al., 2007). This also
accounts for the fact that long TRs are relatively uncom-
mon (Lai & Sun, 2003). Similar findings have been
reported in several other studies (Goldstein & Clark,
1995; Brinkmann et al., 1998; Lai & Sun, 2003; Vogler
et al., 2006). Likewise, a positive relationship is generally
found between TR mutation frequency and the size of the
repeat unit (Sia et al., 1997; Schug et al., 1998; Eckert &
Hile, 2009; Bayliss et al., 2012) as well as the degree of
ª 2013 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
124
5’ 3’ (a)
3’ 5’
5’
3’
Unequal
crossover
5’
3’
Expansion
3’ 3’
5’ 5’
ContracƟon
Fig. 3. Diagram illustrating the recombination mechanism of TR rearrangement. DNA molecules with repeat units represented as blocks are
shown in different colors. (a) Model of unequal crossover. When repeat units of two different DNA molecules misalign, unequal crossover can
occur, resulting in repeat expansion in one crossover product and repeat contraction in the other. (b) Model of intramolecular recombination.
When recombination occurs between repeat units within a DNA molecule, two products are generated that have undergone a repeat
contraction.
conservation between repeats (Legendre et al., 2007). In
addition, the GC content of the TR is also an important
factor determining stability, because repeated sequences
are prone to form diverse non-B DNA structures (i.e.
hairpins and triplexes), which may cause pausing of the
DNA polymerase and replication fork collapse, and in
turn necessitate intervention of the repair and recombina-
tion machinery to reinitiate replication (Wells et al., 2005;
Choudhary & Trivedi, 2010). Besides, the orientation of
the repeats with respect to the direction of replication can
affect the mutation ratio as well, because TRs are more
prone to form secondary structures in one orientation
than in the other (Hebert et al., 2004), and replication
fidelity is not equal in the leading strand compared with
the lagging strand (Gawel et al., 2002). Intriguingly, based
on a theoretical model, Lai & Sun (2003) suggested that
expansion occurs more frequently for short microsatel-
lites, while contraction occurs more frequently for long
ones, suggesting the rearrangement pattern might be
dependent on the repeat type. However, the lack of exper-
imental evidence so far cannot validate this observation.
Aside from intrinsic factors, extrinsic environmental
conditions may also affect TR rearrangement frequencies.
For instance, it was shown that several TRs used in a
multilocus TR typing scheme of E. coli showed enhanced
ª 2013 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
K. Zhou et al.
(b)
3’ 3’
5’
5’
5’
3’
Intramolecular
3’ recombinaƟon
5’
5’ 3’
3’ 5’
ContracƟon
variation at increased growth temperature and upon starva-
tion in E. coli O157:H7, but not upon irradiation (Cooley
et al., 2010). Likewise, in sclB of Streptococcus pyogenes,
encoding a collagen-like surface protein, TR variations
occurred during growth in fresh human blood, but not in
medium (Rasmussen & Bj€ ork, 2001). However, the
underlying mechanisms by which environmental stresses
affect the TR mutation frequency are poorly understood.
The phenotypical impact of intergenic
TR variations in bacteria
Accumulating evidence indicates that rearrangements of in-
tergenic TRs can confer transcriptional evolvability (Jansen
et al., 2012). More specifically, SSRs positioned as cis-regu-
latory elements around the promoter region can induce
phase variation (i.e. stochastic, high frequency, reversible
switching of genotype, and/or phenotype) by modulating
the transcription of the corresponding genes. Most studies
indicate that intergenic SSRs, except monomeric SSRs,
involved in phase variation tend to be A/T rich, which
makes them prone to melting and SSM. In this section, we
review examples of this mechanism of phase variation
according to the different positions of intergenic TRs rela-
tive to the transcriptional start site (Fig. 4; Table 1).
FEMS Microbiol Rev 38 (2014) 119–141
Variable tandem repeats and bacterial adaptation
ATG
UAS RNA pol
SD
ORF
–35 –10 +1
A B C a novel repressor (NadR) prevents transcription of nadA
Fig. 4. Scheme showing possible positions of intergenic TRs in a
standard promotor region that can cause phase variation. Indicated
are an ORF, a promoter with 10 and 35 signatures that are
recognized by RNA polymerase (RNA pol), an upstream regulatory
sequence to which activators or repressors can bind (UAS), the
transcription initiation site (+1), a Shine–Dalgarno sequence for
ribosome binding (SD), and a translation start codon (ATG). Repeats
in regions A and B (double-headed arrows) can modulate gene
expression by affecting transcription initiation (see text TRs upstream
of 35 site of promotor and TRs between 35 and 10 sites of
promotor), whereas repeats in region C operate via as yet
unidentified means (see text TRs between transcriptional start site and
ORF; adapted from van der Woude & Baumler, 2004).
TRs upstream of 35 site of promotor
When TRs are present upstream of the 35 site, the
repeat copy number is expected to affect the binding of
transcription factors and thus modulate gene expression.
This has been reported in the pathogen N. meningitidis,
where expression of the nadA gene (encoding a protein
that functions as invasin and adhesin) is regulated by the
number of tetrameric repeats (TAAA) within the
upstream region of the RNA polymerase binding site,
resulting in three significantly distinct transcriptional lev-
els (high, intermediate, and low). The frequency of phase
variation between these levels has been estimated at
c. 4.4 9 10 4. Remarkably, the levels of nadA transcrip-
tion show a periodic rather than a monotonous relation
to the number of repeat units. As such, the transcription
level varies in the order low–high–intermediate–low–high
for repeat copy numbers of 4–5–6–7–8 and again for
9–10–11–12–13 repeats. Further mechanistic studies indi-
cated that variation of the tetrameric repeats affects the
binding of the integration host factor transcriptional
Table 1. Overview of mechanisms by which intergenic TRs can modulate gene expression
Location of TRs Mechanisms
Upstream of 35 site Affects transcription initiation by modifying binding
affinity of regulatory proteins
Between 35 and 10 sites Affects transcription initiation by altering the
distance of promoter elements
Between transcriptional start and ORF May modify binding affinity of regulatory
proteins or mRNA stability
Between two separate transcription Unknown
start sites
ORF, open reading frame; TR, tandem repeat.
FEMS Microbiol Rev 38 (2014) 119–141
125
regulator protein to the nadA promoter (Martin et al.,
2003, 2005). More recently, Metruccio et al. (2009) dis-
covered that depending on the number of TAAA repeats,
through binding of two operators flanking the variable
tetrameric repeat tract on both sides. As a result, it was
proposed that alteration of the spacing between these two
operators by variation of the number of TAAA repeats
may affect the ability of NadR to repress nadA expression
(Metruccio et al., 2009).
TRs between 35 and 10 sites of promotor
The spacing between the 35 and 10 sites of a standard
promoter is critical for the binding efficiency of RNA
polymerase (Fig. 4), and the optimal distance is around
17 bp in bacteria. Consequently, when TRs locate in that
region, copy number changes are expected to modulate
the transcription level of the genes in the transcriptional
unit. An example of this is found in the host-adapted
pathogen H. influenzae, which adheres to human epithe-
lial cells with the help of LKP fimbriae (also called long,
thick pili). These fimbriae are encoded by the hif gene
cluster and important for H. influenzae infection at differ-
ent stages (van Ham et al., 1993). Phase variation in the
expression of these fimbriae is mediated by a string of
dinucleotide repeats between the 35 and 10 sites
within the overlapping, but divergent promoter regions of
hifA and hifB genes, with TR copy numbers of 9, 10, or
11, respectively, resulting in no, high, and low expression
of both genes.
Another intriguing example is the FetA protein of Nei-
sseria gonorrhoeae, an iron-repressible protein functioning
as ferric enterobactin receptor. The expression of FetA
exhibits extremely rapid phase variation (switching
frequency up to 1.3% per generation) correlating with
polymorphism of a poly-C tract between the 10 and
35 regions of the fetA promoter. The various lengths of
the poly-C tract result in either high or low expression.
References
Miller et al. (1987), Martin et al. (2003, 2005),
Metruccio et al. (2009)
Willems et al. (1990), Yogev et al. (1991), van
Ham et al. (1993), Sarkari et al. (1994),
Carson et al. (2000), van der Ende et al. (2000)
Lafontaine et al. (2001), Attia & Hansen (2006)
Dawid et al. (1999)
ª 2013 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
126
It was suggested that phase variation of FetA reflects a
balance between the advantages of iron scavenging, on
the one hand, and evasion of the host immune response
(based on FetA immunogenicity), on the other (Carson
et al., 2000).
A more complex situation is found in the PorA outer
membrane protein in N. meningitidis. Not only is expres-
sion of the porA gene stochastically modulated by two
variable homopolymeric tracts (poly-G and poly-T)
between the 35 and 10 sites of the promoter region,
there is also a variable poly-A tract within the porA
coding region. Both sites are believed to serve evasion of
the host immune response to this protein and explain the
poor efficacy observed for PorA-based vaccines (van der
Ende et al., 2000).
Further examples include the genes encoding the lipo-
protein Vlps of Mycoplasma hyorhinis (Yogev et al.,
1991), the fimbrial subunits of Bordetella pertussis (Wil-
lems et al., 1990), and the outer membrane protein Opc
of N. meningitidis (Sarkari et al., 1994), and it can be
concluded that variable SSRs within the 35 and 10
region represent a widespread strategy of phase variation
by modulated gene expression in pathogens.
TRs between transcriptional start site and ORF
While intergenic TR-dependent phase variations in bacte-
ria mostly belong to the two classes described above,
some cases are mediated by TRs located between the tran-
scriptional start site and the ORF. In the Gram-negative
pathogen Moraxella catarrhalis, the UspA1 protein func-
tions as an adhesin to mediate binding to human epithe-
lial cells (Lafontaine et al., 2000). Its expression was
shown to be phase variable and correlated with adherence
capacity. Sequence analysis revealed a variable homopoly-
meric poly-G tract between the transcriptional start site
and the start codon to be responsible for the variability
in uspA1 expression. Stable mRNA and strong expression
of UspA1 was detected with a 10-bp G repeat tract, while
truncated mRNA and weak expression occurred with a 9-
bp G tract. Based on these observations, it was proposed
that alterations in the poly-G tract affect the binding effi-
ciency of transcriptional regulators and/or the stability of
the uspA1 mRNA (Lafontaine et al., 2001).
Interestingly, a similar case was recently uncovered in
another outer membrane protein of M. catarrhalis,
UspA2, which is involved in serum resistance and vitro-
nectin binding (Attia & Hansen, 2006). It was observed
that a tetrameric (AGAT) repeat tract, located between
the uspA2 transcription start site and the start codon, is
highly variable in M. catarrhalis isolates. Moreover, the
expression level of UspA2 and, as a result, the serum
resistance and vitronectin binding capacity of the cells
ª 2013 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
K. Zhou et al.
were increased as the repeat copy number increased, pos-
sibly because the variable number of AGAT repeats may
affect the secondary structure of the UspA2 mRNA
transcript.
TRs between two separated transcription start
sites
This special example of phase switching induced by
variable intergenic TRs has so far only been described in
H. influenzae (Dawid et al., 1999). HMW1 and HMW2
are two adhesins of H. influenzae that exhibit different
cellular binding specificities and are encoded by two sepa-
rate chromosomal loci hmw1AC and hmw2AC, respec-
tively. Both genes show 80% similarity, suggesting that
they can be considered as alleles (Barenkamp & Leininger,
1992; Ecevit et al., 2004). Promoter analysis revealed two
transcription start sites (P1 and P2) within the upstream
region of each gene and a heptameric (ATCTTTC) TR
array between P1 and P2 of both genes. The occurrence of
variations in the number of TR copies has been confirmed
in H. influenzae isolates from patients with chronic
obstructive pulmonary disease (Dawid et al., 1999; Cholon
et al., 2008). These variations can affect mRNA synthesis
and thereby influence the expression of corresponding pro-
teins. For example, an increasing number of TR units
resulted in a gradual decrease in specific mRNA synthesis
and protein expression and vice versa. However, the
underlying molecular mechanism remains obscure.
The phenotypical impact of intragenic
TRs in bacteria
There is abundant evidence that besides intergenic TRs,
also intragenic TRs can trigger phase variation (Table 2).
However, the underlying mechanisms are dependent on
the nature of the TR. More specifically, if the TR unit size
is not a multiple of three, rearrangements are able to
induce frameshift mutations as the cause of ON–OFF
phase variation. In comparison, phase variation induced
by TRs whose unit size is a multiple of three is more
complex and is probably related to specific structural and
functional alterations of the corresponding proteins
(Gemayel et al., 2010). In this section, we review studies
on the phenotypical impact of intragenic TRs, grouped
according to their location in different functional classes
of genes.
Cell surface structural genes with TRs
It has been noted that TRs are most abundant in genes
whose products are either exposed on the cell surface or
involved in the biogenesis of cellular surface structures,
FEMS Microbiol Rev 38 (2014) 119–141
Variable tandem repeats and bacterial adaptation 127

Table 2. Examples of intragenic TRs causing phase switching in bacteria

Affected moiety Bacterial species Gene(s) or operon Repeat motif* (5′–3′) References
Adhesin Haemophilus influenzae cha 56 bp Sheets & St Geme (2011)
Helicobacter pylori sabA CT Goodwin et al. (2008)
Neisseria gonorrhoeae opa CTCTT Murphy et al. (1989)
Mycoplasma hominis vaa A Zhang & Wise (1997)
Legionella pneumophila lcl 45 bp† Vandersmissen et al. (2010)
Capsule Streptococcus pneumoniae cps15bM TA van Selm et al. (2003)
cap8E 223 bp Waite et al. (2003)
tts 22 bp Waite et al. (2003)
Neisseria meningitidis siaD C Hammerschmidt et al. (1996b)
Escherichia coli neuO AAGACTC Deszo et al. (2005)
Effector Xanthomonas campestris avrBs3 102 bp Herbers et al. (1992), Kay et al. (2007)
Fe binding Haemophilus influenzae hgp CCAA Jin et al. (1999), Ren et al. (1999)
Neisseria meningitidis hpuA, hmbR G Lewis et al. (1999), Richardson &
Stojiljkovic (1999)
Flagellin Campylobacter coli flhA T Park et al. (2000)
Lipoprotein Mycoplasma hyorhinis vlp 24/26 bp† Rosengarten & Wise (1991)
Mycoplasma bovis vspA 18/24 bp† Lysnyansky et al. (1996)
Mycoplasma pulmonis vsa 34 bp Bhugra et al. (1995)
Lipopolysaccharides Haemophilus influenzae losA CGAGCATA Erwin et al. (2006)
Haemophilus influenzae lic1, lic2 A, lic3 A CAAT Hosking et al. (1999), Dixon et al. (2007)
Haemophilus influenzae lex2, oafA GCAA Griffin et al. (2003), Fox et al. (2005)
Neisseria meningitidis lgtA, C, D G Yang & Gotschlich (1996), Jennings
et al. (1999)
Helicobacter pylori futA, futB C & 21 bp‡ Appelmelk et al. (1999), Nilsson
et al. (2008)
Metabolism Escherichia coli xylB G Funchain et al. (2000)
Mismatch repair Escherichia coli mutL CTGGCG Shaver & Sniegowski (2003)
Salmonella Typhimurium mutL GCTGGC Chen et al. (2010)
Outer membrane protein Neisseria meningitidis porA G van der Ende et al. (2000)
Group B streptococci bca 246 bp Gravekamp et al. (1996, 1997)
Mycoplasma fermentans p78 A Theiss & Wise (1997)
Inner membrane protein Escherichia coli tolA 15-18 bp† Zhou et al. (2012a, b)
Peroxiredoxin Escherichia coli ahpC TCT Ritz et al. (2001)
Pilus Neisseria gonorrhoeae pilC G Jonsson et al. (1991, 1992)
Legionella pneumophila fimV 18 bp† Coil & Anne (2010)
Regulator Listeria monocytogenes ctsR GGT Karatzas et al. (2003, 2005)
prfA CAGGAGT Lindb€ ack et al. (2011)
R-M§ system I Haemophilus influenzae hsdM GACGA Zaleski et al. (2005)
Neisseria gonorrhoeae hsdS G Adamczyk-Poplawska et al. (2011)
R-M§ system III Haemophilus influenzae modA AGTC Srikhanta et al. (2005)
Neisseria spp. modA AGCC Srikhanta et al. (2009)
Neisseria spp. modB CCCAA Srikhanta et al. (2009)
Neisseria meningitidis modD ACCGA Seib et al. (2011)
Helicobacter pylori res C de Vries et al. (2002)
Helicobacter pylori modH G Srikhanta et al. (2011)
Pasteurella haemolytica mod CACAG Ryan & Lo (1999)
R-M§ system IV Escherichia coli mrr G Tesfazgi Mebrhatu et al. (2011)
TR, tandem repeat.
*The motif sequences of microsatellites (≤ 9 bp) are listed; for longer TRs, only the length is given.
†
Different repeat unit lengths or sequences have been reported for this TR locus.
‡
Two TR loci exist at different positions in the same gene.
§
Restriction–modification system.

such as lipopolysaccharides (LPS), adhesins, pili, fimbriae, et al., 2010; Jerome et al., 2011). Extensive studies in dif-
and capsules (Moxon et al., 1994; Jordan et al., 2003; ferent organisms and with different cell surface genes
Verstrepen et al., 2004; Gibbons & Rokas, 2009; Janulczyk support the notion that stochastic TR-based switching

FEMS Microbiol Rev 38 (2014) 119–141 ª 2013 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
128
contributes to the rapid generation of diversity in surface
structures, which in pathogens can serve as a mechanism
for escaping the immune response and/or for determining
tissue tropism. Some examples are reviewed more elabo-
rately below.
TRs within LPS biosynthesis genes
LPS is a complex macromolecule in Gram-negative bacte-
ria that is composed of three distinct parts (i.e. lipid A,
core sugar, and O-antigen side chains), and a large num-
ber of genes are involved in the synthesis and export of
LPS. As one of the major cell surface antigens, LPS is
often implicated in cell adhesion and virulence. Further-
more, because LPS molecules are an essential structural
component of the outer membrane that forms the outer
shell of the Gram-negative cell, it also determines bacte-
rial resistance to a variety of toxic chemicals, including
some antibiotics and xenobiotics. A notable feature of
LPS of some pathogens is the extensive intra- and inter-
strain heterogeneity of the glycoform structure (i.e. the
moieties comprising the core and O side chain sugars),
which is mainly due to incomplete biosynthesis during
the stepwise assembly of the sugar residues resulting from
the phase-variable expression of the corresponding
biosynthesis genes (Schweda et al., 2007). Typically, the
phase variability of these genes derives from the fact that
they contain nontrimeric TR tracts that exhibit stochastic
variation. In some pathogens, this type of stochastic vari-
ation occurs in more than one gene for LPS synthesis,
turning phase switching into a combinatorial process.
Unlike that of most Gram-negative bacteria, the core
LPS of H. influenzae lacks the homopolymeric sugar units
comprising the O-antigen side chains. Therefore, phase
variation in LPS biosynthesis of H. influenzae occurs
mainly through reversible expression switching of a subset
of TR-containing genes, involved in the addition of core
sugars (i.e. glucose and sialic acid) to the conserved
tri-heptose backbone and in the addition of phospho-
rylcholine or acyl groups to these core sugars (Moxon
et al., 2006; Fig. 5a). These genes include lic1A, lic1B,
lic1C, and lic1D (collectively designated as lic1 locus), and
lic2A, lic3A, lgtC, lex2, and oafA, each of which contains a
variable tetrameric repeat tract (reviewed in Schweda
et al., 2007). Rearrangements in the tetrameric repeat
tracts of these genes can individually turn their expression
on or off, resulting in a modified LPS molecule at the
single cell level, and a repertoire of different LPS epitopes
throughout the population (Fig. 5b; Moxon et al., 1994;
Bayliss et al., 2001; Moxon et al., 2006). Remarkably, all
the phase-variable genes involved in LPS biosynthesis of
H. influenzae account for structural elements directly rele-
vant to virulence as well (Schweda et al., 2007).
ª 2013 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
K. Zhou et al.
Interestingly, searching for genes harboring tetrameric
TRs has proven to be a successful strategy to identify
novel LPS biosynthesis genes in H. influenzae genomes
(Hood et al., 1996; Fox et al., 2005). Similar mechanisms
generating LPS diversity have also been described in
N. meningitidis, where the ON/OFF variation in expres-
sion of genes of the lgt gene family (encoding LPS bio-
synthesis functions) is mediated by stochastic expansions
or contractions in their intragenic homopolymeric tracts
and where the corresponding LPS structures have so far
been classified into 12 immunotypes (Jennings et al.,
1999; Berrington et al., 2002; Bayliss et al., 2008).
The O-chain of H. pylori LPS can be fucosylated, thereby
generating Lewis antigens that mimic human blood group
antigens and mediate immune evasion (reviewed in Kusters
et al., 2006). Lewis antigens of H. pylori are subject to
reversible, high-frequency phase variation, and one of the
mechanisms is the slipped-strand mispairing of poly-C
tracts within three fucosyltransferase genes (futA, futB, and
futC; Appelmelk et al., 1999; Wang et al., 1999). In addi-
tion, a unique TR region with a 21-bp unit size was recently
uncovered in the 3′ end of futA and futB, but not in futC.
Strikingly, although copy number variations of the 21-bp
TR tract in both futA and futB did not alter the reading
frame, they could affect the fucosyltransferase activity. In
fact, a correlation was observed between the copy number
of the 21-bp TRs and the number of O-antigen units being
fucosylated, and the addition of one repeat unit led to the
addition of an N-acetyl-b-lactosamine (LacNAc) unit in
the O-antigen polysaccharide (Nilsson et al., 2006, 2008).
These studies show that the variability of TRs in futA, futB,
and futC increases the antigen diversity and population
heterogeneity and thereby supports adaptability of
H. pylori to fluctuating conditions in the gastric mucosa.
TRs within capsule biosynthesis genes
As one of the most external structures on the bacterial
surface, capsules may completely conceal other antigenic
surface molecules or may be co-exposed with other anti-
gens, which are thought to be important for pathogenicity
of bacteria. Production of a capsule provides some patho-
genic bacteria with resistance to phagocytic and comple-
ment-mediated killing and, at the same time, affects
bacterial attachment to host cells. Consequently, the ability
to regulate capsule expression might confer a selective
advantage for pathogens to cope with host immune
responses during different stages of the infection process.
For example, it was found that acapsulate variants of
N. meningitidis serogroup B show much higher adherence
and invasion of epithelial cells than their capsulated pro-
genitors. Such variants can be generated at high frequency
due to SSM of a poly-C tract in the polysialyltransferase
FEMS Microbiol Rev 38 (2014) 119–141
Variable tandem repeats and bacterial adaptation 129

(a) PEtn

lic1A 4
PC 6Glc Hep Kdo Lipid A
β-1,4 α-1,5

α-1,3

Hep6 PEtn

α-1,2
lic3A lic2A
NeuAc Gal Glc Hep
α-2,3 β-1,4 β-1,2

(b)
NeuAc
lic3A (ON) (OFF) (ON) (ON)
Gal
lic2A (ON) (OFF) (ON) (OFF)
PC
lic1 (ON) (ON) (OFF) (OFF)

Type I Type II Type III Type IV

Fig. 5. Structural variations of LPS molecules induced by TR rearrangements in LPS biosynthesis genes. (a) Representation of one possible
structure of the LPS from Haemophilus influenzae strain Rd. The conserved tri-heptose backbone is highlighted in bold. Three phase-variable lic
genes involved in the addition of specific components to the backbone are shown. Kdo, 2-keto-3-deoxyoctulosonic acid; Hep, L-glycero-D-manno-
heptose; Glc, D-glucose; Gal, D-galactose; NeuAc, N-acetylneuraminic acid; PEtn, phosphoethanolamine; PC, phosphorylcholine. (b) Scheme
showing some of the possible LPS types generated on the cell surface of H. influenzae Rd by the combinatorial effect of three LPS synthesis
genes with phase-variable expression. lic2A is responsible for adding a galactose (Gal); lic3A, for adding sialic acid (NeuAc); and lic1, for adding
phosphorylcholine (PC). The repeat tracts are shown by hatched boxes in the gray arrow representation of each gene. The TRs in each of these
genes are subject to variation, which can cause a frameshift and thus block expression (indicated by ON or OFF). Types I–IV represent microbial
cells with different LPS antigens depending on lic1, lic3A, and lic2A gene expression (adapted from Moxon et al., 2006).

gene siaD (Hammerschmidt et al., 1996a, b; Spinosa et al.,
2007). On the other hand, meningococcus isolates from
the blood of meningitis patients are almost always capsu-
lated, while both capsulated and noncapsulated strains
typically coexist in the nose or throat of healthy carriers.
Because phase variation of siaD is reversible, it was there-
fore proposed that infection is initiated by acapsulate
strains, but that capsule biosynthesis is reactivated at a
later stage during infection, allowing N. meningitidis to
resist the host immune system and to cause disease (i.e.
sepsis and meningitis). However, other findings indicate
that the presence of a capsule does not necessarily preclude
invasiveness, and the role of the capsule and capsule phase
switching in different stages of meningococcal infection
may therefore be more subtle and strain dependent
(Spinosa et al., 2007; Bartley et al., 2013).
Besides binary ON/OFF switching, some bacteria can
also modulate the composition of their capsule by a
mechanism of TR-dependent phase switching. This is
exemplified by the polysialic acid capsule (K1 antigen) of
E. coli K1, which can be modified through phase-variable
expression of the sialyl O-acetylating activity, resulting in
an altered immunogenicity and susceptibility to glycosid-
ases. The phase-variable acetylation is driven by a hep-
tanucleotide TR tract (AAGACTC; copy number typically
14–39) within the O-acetyltransferase gene neuO (Deszo
et al., 2005). Loss or gain of a number of repeat units
that is not a multiple of three results in a disruption of
the neuO reading frame and subsequent NeuO expression
(Deszo et al., 2005). Functional analysis furthermore
revealed enhanced desiccation resistance, but reduced bio-
film formation in E. coli K1 with active NeuO, suggesting
not only a role in host interaction, but also a more subtle
ecological impact of phase-variable neuO expression
(Mordhorst et al., 2009). Interestingly, each set of three
repeat units encodes a protein structure designated the
poly(w) motif, and NeuO enzymatic activity was found
to increase with the number of poly(w) motifs, support-
ing maintenance of high repeat copy numbers in the
population (Bergfeld et al., 2007).

FEMS Microbiol Rev 38 (2014) 119–141 ª 2013 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
130
Another intriguing aspect of this capsular acetylation is
that the neuO gene resides in a lambdoid prophage
termed ‘CUS-3’, and mitomycin treatment of E. coli K1
can induce release of this phage, which specifically infects
K1 antigen-expressing bacteria (Deszo et al., 2005; King
et al., 2007). This suggests that variant neuO alleles can
be redistributed among K1-encapsulated bacteria via hori-
zontal transfer. Indeed, CUS-3/neuO has been found in
several serotypes, such as O18 and O45 (King et al.,
2007). On the other hand, although the receptor for
CUS-3 is polysialic acid, superinfection was not prevented
by NeuO-mediated acetylation (Vimr & Steenbergen,
2006; King et al., 2007). Notably, sialic acid is also known
as a modification of LPS in numerous pathogens, under-
scoring the vital role of O-acetylation in causing structure
variation of polysaccharide epitopes (also see TRs within
LPS biosynthesis genes).
Also the mini- and macrosatellites within capsule
biosynthesis genes of Gram-positive pathogens can medi-
ate capsule diversity. As such, noncapsulated serotype 3
Streptococcus pneumoniae strains were shown to carry an
out-of-frame perfect tandem duplication in one of the
capsule biosynthesis genes (i.e. cap3A). Interestingly,
based on the sequence and length of the duplication, at
least seven different cap3A alleles were found in different
nonencapsulated strains. Analysis of the phase reversion
frequency (OFF to ON) induced by TR contractions
revealed a positive correlation between the frequency of
reversion and the length of the duplication (Waite et al.,
2001). A similar mechanism of capsular phase variation
correlating with a 223-bp and a 22-bp perfect tandem
duplication in cap8E and tts was also demonstrated in
serotype 8 and serotype 37, respectively, of S. pneumoniae
(Waite et al., 2003).
TRs within adhesin-associated genes
Adhesins mediate bacterial attachment to and further
colonization of host tissues, but they also act as surface
antigens (Bayliss et al., 2001). The opacity proteins Opa
of Neisseria spp. constitute a family of closely related, but
size-variable outer membrane proteins, which enhance
the adherence to epithelial, leukocyte, and phagocytic cells
(reviewed in Sadarangani et al., 2011). They do not only
determine host and tissue specificity, but also facilitate
efficient cellular invasion (Carbonnelle et al., 2009).
Neisseria spp. strains have 3–11 opa genes whose phase-
variable expression is modulated by a pentanucleotide TR
tract (CTCTT) in their coding regions or by intergenic
recombination. As a result, a vast array of Opa variants
can be generated to confer differential molecular specifici-
ties, allowing Neisseria both to alter its tissue tropism and
to escape the host immune system. In fact, it has been
ª 2013 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
K. Zhou et al.
argued that the occurrence of different arrays of Opa
variants in clinical (disease) and carriage (nondisease) iso-
lates of N. meningitidis may be the result of immune
selection pressure (Callaghan et al., 2006, 2008; Sadaran-
gani et al., 2011).
In the same vein, adhesins of H. pylori are categorized
into two main subgroups, Hop (Helicobacter outer mem-
brane porins) and Hor (Hop-related). For some, but not
all of these adhesins, the corresponding host cell receptors
have been identified (reviewed in Backert et al., 2011).
For example, BabA binds to the Lewis B (Leb) antigen
expressed on the gastric mucosa (Ilver et al., 1998), and
SabA binds to glycosphingolipids of host cells that display
a sialyl-dimeric Lewis X (sialyl-Lex) antigen (Mahdavi
et al., 2002). Interestingly, expression of some of the
adhesins (i.e. BabB, HopZ, SabA, OipA) is subject to ON/
OFF phase variation due to the presence of variable dinu-
cleotide (CT) tracts within the corresponding genes. As a
consequence, the expression patterns of adhesins gener-
ated by stochastic phase variation can not only affect the
adhesion efficiency, but also alter the tissue tropism
(Yamaoka et al., 2002; Backert et al., 2011). Additionally,
the resulting repertoire of phase-variable adhesins can be
advantageous for H. pylori to escape the host immune
system.
A final example in this category is the Eap protein of
the Gram-positive pathogen Staphylococcus aureus. Eap is
a multifunctional adhesin with a variable TR tract, in
which the size of each repeat unit is not identical (93–
110 aa). A minimum of two TRs in the eap gene are
required for Eap to cause agglutination, adherence and
cellular invasion by S. aureus. Furthermore, as the repeat
copy number increases from 2 to 5, those capacities are
significantly enhanced, suggesting that TR copy number
expansion in eap supports host adaptation of S. aureus
(Hussain et al., 2008).
TRs within iron (heme) acquisition genes
Free iron is typically limited in the host and is usually
sequestered by iron-binding proteins (e.g. hemoglobin,
transferrin, and lactoferrin). Iron acquisition mechanisms
are therefore considered indispensable for bacterial patho-
genicity, and pathogens have evolved many different
strategies for adaptation to fluctuating iron concentra-
tions. As such, pathogens are able to extract the iron
from heme groups of iron-binding proteins via surface
receptors. In H. influenzae, a family of hemoglobin-bind-
ing and hemoglobin–haptoglobin-binding proteins (Hgp)
is known to mediate heme scavenging (Ren et al., 1998;
Jin et al., 1999; Morton et al., 1999). Individual strains of
H. influenzae have 1–4 hgp genes, of which knockout
analysis has confirmed that they are indispensable
FEMS Microbiol Rev 38 (2014) 119–141
Variable tandem repeats and bacterial adaptation
virulence determinants in invasive disease (Seale et al.,
2006). Interestingly, a CCAA tetranucleotide repeat tract
exists in each hgp gene, which is reminiscent of the LPS
biogenesis genes of H. influenzae harboring a CAAT
repeat tract (Moxon et al., 2006; see TRs within LPS
biosynthesis genes). Phase variation induced by repeat
rearrangements within the hgp genes has been observed;
however, its biological significance is still obscure.
A similar case exists in N. meningitidis where iron
acquisition is modulated by the variable poly-G tract in
the hpuA and hmbR genes that are involved in the bio-
synthesis of two hemoglobin receptors (Lewis et al., 1999;
Richardson & Stojiljkovic, 1999). More recently, it was
revealed that 91% of N. meningitidis pathogenic isolates,
but only 71% of commensal isolates have at least one
receptor in an ON state, suggesting expression of hemo-
globin receptor(s) to be important for the systemic spread
of meningococci (Tauseef et al., 2011).
TRs within genes involved in restriction–
modification systems
In addition to their role as drivers of genetic variation as
reviewed above, TRs can also drive epigenetic variation
when they affect restriction–modification (R-M) systems.
Currently, TR-dependent phase variations have been doc-
umented only in type I and III R-M systems. Type I R-M
systems are generally comprised of three subunits (S, M,
and R), which together form a holoenzyme that has both
methylation and restriction activity. In H. influenzae, the
type I R-M system HindI functions as the main defense
system against the entry of foreign DNA (Glover & Pie-
karowicz, 1972; Piekarowicz et al., 1974). Phase variation
of HindI is driven by a GACGA pentanucleotide repeat
tract in the methyltransferase subunit gene (hsdM), as
supported by the finding that an hsdM allele with four
repeat units encoding an HsdM protein of normal length
was associated with resistance to phage HP1c1 infection,
while gain or loss of one repeat unit resulted in phage
sensitivity (Zaleski et al., 2005). Interestingly, phase
switching of phage susceptibility could also independently
be conferred by LPS alterations induced by the variable
tetranucleotide repeat tract of the lic2A gene, which con-
firmed the role of Lic2A-modified LPS as the receptor of
HP1c1 phage (Zaleski et al., 2005; see also TRs within
LPS biosynthesis genes). Moreover, the frequencies of the
phase variations of phage susceptibility described above
are affected by Dam (deoxyadenosine methyltransferase)
activity, although the mechanism is not clear yet (Zaleski
et al., 2005).
Another example was found in the NgoAV type I R-M
system of N. gonorrhoeae, which is encoded by four genes:
hsdMNgoAV, hsdRNgoAV, hsdSNgoAV1, and hsdSNgoAV2. The
FEMS Microbiol Rev 38 (2014) 119–141
131
product of the hsdSNgoAV1 is responsible for the specific
recognition of the target site, whereas hsdSNgoAV2 is
nonfunctional. It was postulated that hsdSNgoAV1 and
hsdSNgoAV2 are actually truncated proteins derived from
an integral hsdS locus that has become interrupted by a
frameshift mutation resulting in the formation of a stop
codon between hsdSNgoAV1 and hsdSNgoAV2 (Piekarowicz
et al., 2001). Interestingly, a recent study uncovered a
variable poly-G tract within the 3′ end of hsdSNgoAV1 in
N. gonorrhoeae, and loss of a guanine in that tract
restores the fusion of the hsdSNgoAV1 and hsdSNgoAV2
genes, resulting in the generation of a new HsdSNgoAVD
protein, responsible for a novel NgoAV R-M system,
termed ‘NgoAVD’. The NgoAVD system has a modified
DNA recognition specificity, thereby conferring an altered
susceptibility to various phages (Adamczyk-Poplawska
et al., 2011).
Type III R-M systems only consist of two subunits, the
methyltransferase (encoded by a mod gene) and
restriction endonuclease (encoded by a res gene). In
host-adapted pathogens, ON/OFF phase variation of this
system has been reported due to variable TRs (with
repeat unit not being a multiple of three bases) within
either the mod or the res gene (Ryan & Lo, 1999; De Bolle
et al., 2000; de Vries et al., 2002; Srikhanta et al., 2005,
2009, 2011; also see review Srikhanta et al., 2010). Inter-
estingly, TR-induced phase variation in the mod gene has
been shown to affect the expression of a number of genes,
referred to as a phase-variable regulon or phasevarion
(Srikhanta et al., 2005, 2010). In H. influenzae strain Rd,
when mod expression was switched off by a TR rearrange-
ment in a tetrameric repeat tract, nine other genes were
down-regulated, and seven, up-regulated (Srikhanta et al.,
2005). In the same vein, N. gonorrhoeae formed signifi-
cantly thicker biofilms and thus may indirectly benefit
from an increased resistance to external stresses, when its
mod was in the OFF state. Additionally, a mod-ON
phenotype resulted in an increased ability to associate
with human cervical epithelial (pex) cells, whereas the
mod-OFF configuration enhanced the ability to invade
and survive within pex cells following invasion (Srikhanta
et al., 2009). Further studies will be needed to clarify
whether these phenotypes directly relate to mod deficiency
itself or to altered expression of one or more members of
the mod phasevarion.
Interestingly, most N. meningitidis and N. gonorrhoeae
strains have a second phase-variable methyltransferase
gene (Srikhanta et al., 2009), and some, even a third
(Seib et al., 2011), and it has been anticipated that the
combinatorial use of different phasevarions may contrib-
ute to further phenotypic variability (Seib et al., 2011).
Furthermore, Fox et al. (2007) suggested that the
apparent evolution of this type III R-M system into an
ª 2013 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
132 K. Zhou et al.

epigenetic mechanism for controlling gene expression has
resulted in loss of the DNA restriction function in some
strains. In the latter context, it is also interesting to note
the existence of solitary type IV restriction endonucleases
that have no cognate methyltransferase. Because these
peculiar enzymes display a specificity for methylated
DNA, it has been anticipated that they might function to
ward off the lateral acquisition of methyltransferases that
might affect the cell’s epigenetic regulation (Fukuda et al.,
2008; Tesfazgi Mebrhatu et al., 2011). As such, the phase
variability of methyltransferase functions might serve to
prevent detection of their activity during their establish-
ment in a new host (Tesfazgi Mebrhatu et al., 2011).
TRs within transcription activator-like effectors
Gram-negative plant-pathogenic bacteria of the genus
Xanthomonas can infect a broad spectrum of plants. A
common feature of their infection process is the injection
of virulence proteins (termed effectors) into host cells,
mostly by means of a type III secretion system. The type
III effectors are currently classified into nearly 40 groups
based on sequence similarity and biochemical activity,
and the largest group is the AvrBs3/PthA family, also
known as transcription activator-like effector (TALE)
family (reviewed in White et al., 2009). The most con-
spicuous feature of TALE genes is the variation of their
central domain, mostly consisting of 15.5–19.5 nearly
identical TRs with a 102-bp unit (G€ urlebeck et al., 2006;
Mak et al., 2013; Fig. 6).
A recent study revealed that the AvrBs3 can determine
bacterial fitness in plants during infection (Kay et al.,
2007). A set of upa (up-regulated by AvrBs3) genes among
which upa20, the key regulator of the plant cell hypertro-
phy phenotype, has been identified as AvrBs3 targets in
pepper plants. AvrBs3 acts as a transcription factor by
binding to a conserved promoter element (UPA box) of
upa20, resulting in up-regulation. Binding was shown to be
mediated by the TR region of AvrBs3, suggesting the
repeats to act as a DNA-binding motif (Kay et al., 2007).
More specifically, it was found that the number of base
pairs of the UPA box closely matches the TR copy number

(a)

TTSS DNA-binding TR domain TranscripƟonal

secreƟon acƟvaƟon
signal 0 1 17.5 domain

N HD NG NS NG NI NI NI HD HD NG NS NS HD HD HD NG HD NG C
Nuclear
localizaƟon
signal
LT P E Q V VA I A S H D G G KQ A L E T V Q R L L P V L C Q A H G
1 12/13 34

(b)

UPA box T A T A T AA A C C T NN C CC T C T
UPA gene

Fig. 6. Domain organization and molecular function of Xanthomonas TALE AvrBs3. (a) The AvrBs3 functional domains shown include a type
three secretion system (TTSS) signal sequence (dark blue), a central DNA-binding TR domain, a nuclear localization signal (purple), and a
transcriptional activation domain (olive green). The TR domain in this case comprises 17.5 imperfect repeat units, each of which consists of 34
amino acids (aa). The unit numbered zero (red, dashed rectangle) is not a true repeat because it has a different aa sequence, but it also
contributes to DNA binding. Each repeat binds to a base in the target sequence, and the binding specificity of the repeats is determined by aa
12 and 13 (known as RVD) and displayed with repeat-specific colors. The complete aa sequence of one repeat (no. 9) is shown with its RVD
highlighted in green. (b) After injection into the plant cell by the bacterial type three secretion system, AvrBs3 is targeted to the nucleus and will
bind with its TR domain to a specific DNA sequence known as UPA box. The consensus UPA box matches closely the binding specificity of the TR
region of AvrBs3 as determined by the different TR units and their RVD (aa 12 and 13). AvrBs3 binding activates transcription of several UPA
genes (adapted from Mak et al., 2013).

ª 2013 Federation of European Microbiological Societies. FEMS Microbiol Rev 38 (2014) 119–141
Published by John Wiley & Sons Ltd. All rights reserved
Variable tandem repeats and bacterial adaptation
in AvrBs3. An elegant study demonstrated that each repeat
unit of AvrBs3 binds to one base pair of the UPA box and
that the base recognition specificity of each repeat is deter-
mined by two hypervariable amino acids (the 12th and
13th amino acids of each repeat unit), known as repeat var-
iable di-residues (RVDs; Boch et al., 2009). Moreover, a
minimum of 6.5 TRs in AvrBs3 are necessary for activating
upa gene expression (Boch et al., 2009; Moscou & Bogda-
nove, 2009; Scholze & Boch, 2011; Fig. 6). Most recently,
the structural basis for this kind of sequence-specific recog-
nition has been uncovered by crystallographic analysis of
an artificially engineered TAL effector hybridized with its
target DNA (Deng et al., 2012). AvrBs3-dependent modu-
lation of plant signaling pathways causes enlargement of
mesophyll cells and hypertrophy of the infected tissue,
which might help the bacteria to proliferate and escape
from infection sites to facilitate bacterial spreading (Kay
et al., 2007; Boch & Bonas, 2010). Consequently, TR varia-
tions in the avrBs3 gene preclude activation of the UPA
box causing a failure in inducing the hypersensitive
response in plants (Kay et al., 2007).
While the above studies highlight the biological impor-
tance of TALEs as major virulence determinants, TALEs
also provide interesting avenues for biotechnological appli-
cations. As such, it will be possible to engineer broader
pathogen resistance in crops by combining several UPA
boxes into the promoter of plant resistance genes like Bs3,
which will render these transgenic plants resistant to infec-
tion by bacteria delivering matching TAL effectors (Boch
& Bonas, 2010; Scholze & Boch, 2011). Alternatively, the
TR region of TAL effectors can be tailored as to recognize
and bind to a predefined DNA sequence, resulting in acti-
vating of the expression of downstream target genes (Mor-
bitzer et al., 2010). Moreover, engineered TAL effectors
can be fused with endonuclease domains to generate TAL
effector nucleases that can introduce cuts or double-strand
breaks in or near specific sequences on the chromosome,
targeting these loci for mutagenesis or recombinational
repair and gene therapy (Bogdanove & Voytas, 2011;
Mu~ noz Bodnar et al., 2013).
TRs within stress response genes
Regulation of stress response genes is one of the most
common strategies employed by bacteria to cope with
stresses. TRs have been identified in a number of stress
response genes (Rocha et al., 2002), and some studies
have addressed the role of these repeats in the modula-
tion of a stress response. For example, the gene encoding
the CtsR regulator (class III stress gene repressor) of
Listeria monocytogenes carries a triplet repeat (GGT) tract
with three copies. Stochastic deletion of one triplet in the
ctsR gene results in an inactive CtsR repressor, leading to
FEMS Microbiol Rev 38 (2014) 119–141
133
expression of the clp genes (Karatzas et al., 2003, 2005).
This alteration confers increased resistance to high hydro-
static pressure, heat, acid, and H2O2, but attenuates viru-
lence in L. monocytogenes (Karatzas & Bennik, 2002).
Another example is mutL, which is involved in mis-
match repair in Salmonella Typhimurium and E. coli. The
functional allele of mutL carries a trimeric hexanucleotide
repeat in the region encoding the ATP-binding pocket of
the protein. Spontaneous loss of one repeat unit resulting
in a mutator phenotype due to MutL deficiency has been
observed in long-term cultures of both E. coli and
S.Typhimurium (Shaver & Sniegowski, 2003; Chen et al.,
2010). Expansion of this TR region from 3 to 4 units and
from 2 to 3 units has also been observed and, in the
latter case, caused reversion of the mutator phenotype
(Shaver & Sniegowski, 2003; Chen et al., 2010; Le Bars
et al., 2013). This genetic switching of MutL may serve as
a strategy to control the balance between genetic stability
and mutability and thus serve as an element controlling
bacterial evolution. Interestingly, not only mutL, but also
many other genes (i.e. mutT, mutY, mutS, dinJ, and ruvC)
involved in DNA repair of E. coli harbor SSRs, further
underscoring the putative regulatory role of TRs in stress
response (Rocha et al., 2002).
Some membrane proteins are essential for membrane
integrity and as such for tolerance to a variety of toxic
chemicals and other stresses. One such protein, that is pres-
ent in many Gram-negative bacteria, is TolA, which har-
bors a variable TR tract composed of 8–16 imperfect copies
of a 15- to 18-bp repeat unit in E. coli (Levengood et al.,
1991; Zhou et al., 2012b). TR variations in TolA occurred
at a frequency of at least 6.9 9 10 5 in a clonal wild-type
population of E. coli MG1655 and were shown to modulate
stress tolerance, with the most outspoken TR-dependent
phenotype being deoxycholic acid tolerance (Zhou et al.,
2012a). However, the precise molecular mechanism under-
lying this phenotypic variation remains unclear.
A peculiar case is the triplet repeat (TCT) in E. coli
ahpC, where expansion of the TR tract with 1 unit con-
verts the AhpC protein from a peroxidase into a disulfide
reductase, as demonstrated by the ability of this newly
acquired enzyme activity to restore normal growth of a
mutant lacking thioredoxin and glutathione reductase
(Ritz et al., 2001). To our knowledge, this is the only
example of an intragenic TR variation generating a truly
novel function in bacteria.
Conclusions and outlook: TRs and
bacterial evolution
The generation of mutations is the basis of evolution in
bacteria and all other living organisms. Under changing
environmental conditions, the evolution of better adapted
ª 2013 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
134
descendants may be essential for survival of the species,
and an increased mutation rate obviously could confer
higher survival chances. On the other hand, the genera-
tion of random mutations all over the genome also causes
an important burden because most mutations are delete-
rious. Therefore, to be successful, bacteria will have to
maintain a balance between genome plasticity and stabil-
ity. One way to do this is to control the spatial distribu-
tion of mutations over the genome by evolving highly
mutable sequences that are associated with loci (also
known as contingency loci) that are needed for a flexible
response to environmental conditions or stresses, espe-
cially those that cannot be detected by conventional
bacterial sensors (e.g. phages or host receptors; Fonville
et al., 2011). TRs and their variability provide a paradigm
for this regulatory strategy of adaptability.
The formation of TRs is suggested to be a random pro-
cess based on replication slippage (Levinson & Gutman,
1987); however, only some of the formed TRs are
believed to be maintained during evolution. Generally,
variable TRs tend to localize in flexible genes involved in
the biosynthesis of surface structures and with a function
in adhesion and (for pathogens) invasion, although they
are occasionally also found in genes encoding critical cel-
lular functions like DNA replication (Moxon et al., 2006;
Guo & Mr
azek, 2008). The association between TRs and
cell surface structures is suggested to allow populations to
anticipate changes in the environment in order to
enhance their survival rate (Moxon et al., 1994). This is
particularly common and critical for pathogens such as
H. influenzae and H. pylori with limited genetic informa-
tion (i.e. reduced genome) to cope with complex environ-
ments (Razin et al., 1998).
As such, TR-dependent phase variation can be regarded
as a strategy for bacterial adaptation that is complemen-
tary to conventional mutations (i.e. single nucleotide
polymorphism), but has some distinct features. First, the
hypermutability (typical mutation rates of 10 2–10 5 per
generation) can be advantageous for adaption on a short
time scale, and it has been demonstrated using a theoreti-
cal model that TRs mediating stochastic switching can
evolve and be maintained under a wide range of alternat-
ing selection regimens (Palmer et al., 2013). Second, TR
rearrangements, both local and combinatorial, can effec-
tuate alterations at both the transcriptional and the
translational levels, resulting in either binary switching
(‘ON’ and ‘OFF’) or gradual control, and this facilitates
subtle adaptation of bacterial fitness under stress. Further,
TR-dependent mechanisms operate not only in classic
genetic, but also in epigenetic pathways and from the sin-
gle locus to the more global level of phasevarions, sug-
gesting the power of TR intermediate regulation in
bacterial adaptation.
ª 2013 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
K. Zhou et al.
Although rapid TR-dependent phase switching facili-
tates bacterial adaption on a short time scale, it is proba-
bly not necessary for bacteria to maintain hypermutable
regions in the absence of selection pressures, because of
the associated cost of DNA replication fidelity and meta-
bolic energy. An interesting, but mostly unanswered ques-
tion is how bacteria optimize the mutation rate of TRs in
phase-variable genes in a fluctuating environment. Some
intrinsic features of the TR sequence (i.e. conservation
and copy number of repeat unit) and cellular function
(i.e. DNA replication, recombination, and repair) are
known to affect the frequency of TR-dependent phase
variation and may be used to this purpose (Bayliss,
2009). On the other hand, some studies have shown a
modulation of the TR rearrangement frequency by envi-
ronmental stress (Kanbashi et al., 1997; Jackson & Loeb,
2000; Rasmussen & Bj€ ork, 2001; Srikhanta et al., 2009;
Cooley et al., 2010). However, in general, it remains
unclear how environmental signals are transduced to
modulate the switching rate of TR-containing genes.
In conclusion, TRs confer local sequence flexibility in
bacterial genomes, thereby allowing targeted mutation
and evolution. The genotypic and phenotypic variations
modulated by TRs are rapid and coordinative and sup-
port the generation of substantial biological diversity on a
short time scale. In spite of a certain metabolic cost, ‘pre-
pared genomes’ (i.e. with TR-based contingency loci;
Caporale, 1999) have a higher adaptability and thus a fit-
ness advantage in frequently fluctuating environments.
Acknowledgements
This work was supported by the Research Foundation –
Flanders (FWO – Vlaanderen; Research project G.0289.06)
and by the KU Leuven Research Fund (project METH/07/
03).
References
Ackermann M & Chao L (2006) DNA sequences shaped by
selection for stability. PLoS Genet 2: e22.
Adamczyk-Poplawska M, Lower M & Piekarowicz A (2011)
Deletion of one nucleotide within the homonucleotide tract
present in the hsdS gene alters the DNA sequence specificity
of type I restriction-modification system NgoAV. J Bacteriol
193: 6750–9675.
Aertsen A & Michiels CW (2004) Stress and how bacteria cope
with death and survival. Crit Rev Microbiol 30: 263–273.
Aertsen A & Michiels CW (2005) Diversify or die: generation
of diversity in response to stress. Crit Rev Microbiol 31:
69–78.
Appelmelk BJ, Martin SL, Monteiro MA et al. (1999) Phase
variation in Helicobacter pylori lipopolysaccharide due to
FEMS Microbiol Rev 38 (2014) 119–141
Variable tandem repeats and bacterial adaptation
changes in the lengths of poly(C) tracts in
alpha3-fucosyltransferase genes. Infect Immun 67:
5361–5366.
Attia AS & Hansen EJ (2006) A conserved tetranucleotide
repeat is necessary for wild-type expression of the Moraxella
catarrhalis UspA2 protein. J Bacteriol 188: 7840–7852.
Backert S, Clyne M & Tegtmeyer N (2011) Molecular
mechanisms of gastric epithelial cell adhesion and injection
of CagA by Helicobacter pylori. Cell Commun Signal 9: 28.
Barenkamp SJ & Leininger E (1992) Cloning, expression,
and DNA sequence analysis of genes encoding
nontypeable Haemophilus influenzae high-molecular-weight
surface-exposed proteins related to filamentous
hemagglutinin of Bordetella pertussis. Infect Immun 60:
1302–1313.
Bartley SN, Tzeng YL, Heel K et al. (2013) Attachment and
invasion of Neisseria meningitidis to host cells is related to
surface hydrophobicity, bacterial cell size and capsule. PLoS
ONE 8: e55798.
Bayliss CD (2009) Determinants of phase variation rate and
the fitness implications of differing rates for bacterial
pathogens and commensals. FEMS Microbiol Rev 33:
504–520.
Bayliss CD, Field D & Moxon ER (2001) The simple sequence
contingency loci of Haemophilus influenzae and Neisseria
meningitidis. J Clin Invest 107: 657–662.
Bayliss CD, Hoe JC, Makepeace K, Martin P, Hood DW &
Moxon ER (2008) Neisseria meningitidis escape from the
bactericidal activity of a monoclonal antibody is mediated
by phase variation of lgtG and enhanced by a mutator
phenotype. Infect Immun 76: 5038–5048.
Bayliss CD, Bidmos FA, Anjum A et al. (2012) Phase variable
genes of Campylobacter jejuni exhibit high mutation rates
and specific mutational patterns but mutability is not the
major determinant of population structure during host
colonization. Nucleic Acids Res 40: 5876–5889.
Benson G (1999) Tandem repeats finder: a program to analyze
DNA sequences. Nucleic Acids Res 27: 573–580.
̈
Bergfeld AK, Claus H, Vogel U & Muhlenhoff M (2007)
Biochemical characterization of the polysialic acid specific
O-acetyltransferase NeuO of Escherichia coli K1. J Biol Chem
282: 22217–22227.
Berrington AW, Tan YC, Srikhanta Y, Kuipers B, van der Ley
P, Peak IR & Jennings MP (2002) Phase variation in
meningococcal lipooligosaccharide biosynthesis genes. FEMS
Immunol Med Microbiol 34: 267–275.
Bhugra B, Voelker LL, Zou N, Yu H & Dybvig K (1995)
Mechanism of antigenic variation in Mycoplasma pulmonis:
interwoven, site-specific DNA inversions. Mol Microbiol 18:
703–714.
Bi X & Liu LF (1996) A replicational model for DNA
recombination between direct repeats. J Mol Biol 256:
849–858.
Bichara M, Wagner J & Lambert IB (2006) Mechanisms of
tandem repeat instability in bacteria. Mutat Res 598:
144–163.
FEMS Microbiol Rev 38 (2014) 119–141
135
Boch J & Bonas U (2010) Xanthomonas AvrBs3 family-type III
effectors: discovery and function. Annu Rev Phytopathol 48:
419–436.
Boch J, Scholze H, Schornack S, Landgraf A, Hahn S, Kay S,
Lahaye T, Nickstadt A & Bonas U (2009) Breaking the code
of DNA binding specificity of TAL-type III effectors. Science
326: 1509–1512.
Bogdanove AJ & Voytas DF (2011) TAL effectors:
customizable proteins for DNA targeting. Science 333:
1843–1846.
Brinkmann B, Klintschar M, Neuhuber F, H€ uhne J & Rolf B
(1998) Mutation rate in human microsatellites: influence of
the structure and length of the tandem repeat. Am J Hum
Genet 62: 1408–1415.
Bzymek M & Lovett ST (2001) Instability of repetitive DNA
sequences: the role of replication in multiple mechanisms. P
Natl Acad Sci USA 98: 8319–8325.
Callaghan MJ, Jolley KA & Maiden MC (2006)
Opacity-associated adhesin repertoire in hyperinvasive
Neisseria meningitidis. Infect Immun 74: 5085–5094.
Callaghan MJ, Buckee CO, Jolley KA, Kriz P, Maiden MC &
Gupta S (2008) The effect of immune selection on the
structure of the meningococcal Opa protein repertoire. PLoS
Pathog 4: e1000020.
Caporale LH (1999) Chance favors the prepared genome. Ann
NY Acad Sci 870: 1–21.
Carbonnelle E, Hill DJ, Morand P, Griffiths NJ, Bourdoulous
S, Murillo I, Nassif X & Virji M (2009) Meningococcal
interactions with the host. Vaccine 27(suppl 2): B78–B89.
Carson SD, Stone B, Beucher M, Fu J & Sparling PF (2000)
Phase variation of the gonococcal siderophore receptor
FetA. Mol Microbiol 36: 585–593.
Chen F, Liu WQ, Eisenstark A, Johnston RN, Liu GR & Liu
SL (2010) Multiple genetic switches spontaneously
modulating bacterial mutability. BMC Evol Biol 10: 277.
Chiou CS (2010) Multilocus variable-number tandem repeat
analysis as a molecular tool for subtyping and phylogenetic
analysis of bacterial pathogens. Expert Rev Mol Diagn 10:
5–7.
Cholon DM, Cutter D, Richardson SK, Sethi S, Murphy TF,
Look DC & St Geme III JW (2008) Serial isolates of
persistent Haemophilus influenzae in patients with chronic
obstructive pulmonary disease express diminishing
quantities of the HMW1 and HMW2 adhesins. Infect
Immun 76: 4463–4468.
Choudhary OP & Trivedi S (2010) Microsatellite or simple
sequence repeat (SSR) instability depends on repeat
characteristics during replication and repair. J Cell Mol Biol
8: 21–34.
Coenye T & Vandamme P (2005) Characterization of
mononucleotide repeats in sequenced prokaryotic genomes.
DNA Res 12: 221–233.
Coil DA & Anne J (2010) The role of fimV and the
importance of its tandem repeat copy number in twitching
motility, pigment production, and morphology in Legionella
pneumophila. Arch Microbiol 192: 625–631.
ª 2013 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
136
Cooley MB, Carychao D, Nguyen K, Whitehand L & Mandrell
R (2010) Effects of environmental stress on stability of
tandem repeats in Escherichia coli O157:H7. Appl Environ
Microbiol 76: 3398–3400.
Dawid S, Barenkamp SJ & St Geme III JW (1999) Variation in
expression of the Haemophilus influenzae HMW adhesins: a
prokaryotic system reminiscent of eukaryotes. P Natl Acad
Sci USA 96: 1077–1082.
De Bolle X, Bayliss CD, Field D, van de Ven T, Saunders NJ,
Hood DW & Moxon ER (2000) The length of a
tetranucleotide repeat tract in Haemophilus influenzae
determines the phase variation rate of a gene with
homology to type III DNA methyltransferases. Mol
Microbiol 35: 211–222.
de Vries N, Duinsbergen D, Kuipers EJ, Pot RG, Wiesenekker
P, Penn CW, van Vliet AH, Vandenbroucke-Grauls CM &
Kusters JG (2002) Transcriptional phase variation of a type
III restriction-modification system in Helicobacter pylori.
J Bacteriol 184: 6615–6623.
Deng D, Yan C, Pan X, Mahfouz M, Wang J, Zhu JK, Shi Y
& Yan N (2012) Structural basis for sequence-specific
recognition of DNA by TAL effectors. Science 335:
720–723.
Deszo EL, Steenbergen SM, Freedberg DI & Vimr ER (2005)
Escherichia coli K1 polysialic acid O-acetyltransferase gene,
neuO, and the mechanism of capsule form variation
involving a mobile contingency locus. P Natl Acad Sci USA
102: 5564–5569.
Dettman JR & Taylor JW (2004) Mutation and evolution of
microsatellite loci in Neurospora. Genetics 168: 1231–1248.
Dixon K, Bayliss CD, Makepeace K, Moxon ER & Hood DW
(2007) Identification of the functional initiation codons of a
phase-variable gene of Haemophilus influenzae, lic2A, with
the potential for differential expression. J Bacteriol 189:
511–521.
Ecevit IZ, McCrea KW, Pettigrew MM, Sen A, Marrs CF &
Gilsdorf JR (2004) Prevalence of the hifBC, hmw1A, hmw2A,
hmwC, and hia genes in Haemophilus influenzae isolates.
J Clin Microbiol 42: 3065–3072.
Eckert KA & Hile SE (2009) Every microsatellite is different:
intrinsic DNA features dictate mutagenesis of common
microsatellites present in the human genome. Mol Carcinog
48: 379–388.
Edwards YJ, Elgar G, Clark MS & Bishop MJ (1998) The
identification and characterization of microsatellites in the
compact genome of the Japanese pufferfish, Fugu rubripes:
perspectives in functional and comparative genomic
analyses. J Mol Biol 278: 843–854.
Erwin AL, Bonthuis PJ, Geelhood JL, Nelson KL, McCrea KW,
Gilsdorf JR & Smith AL (2006) Heterogeneity in tandem
octanucleotides within Haemophilus influenzae
lipopolysaccharide biosynthetic gene losA affects serum
resistance. Infect Immun 74: 3408–3414.
Fonville NC, Ward RM & Mittelman D (2011) Stress-induced
modulators of repeat instability and genome evolution.
J Mol Microbiol Biotechnol 21: 36–44.
ª 2013 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
K. Zhou et al.
Foster PL (2005) Stress responses and genetic variation in
bacteria. Mutat Res 569: 3–11.
Foster PL (2007) Stress-induced mutagenesis in bacteria. Crit
Rev Biochem Mol Biol 42: 373–397.
Fox KL, Yildirim HH, Deadman ME, Schweda EK, Moxon ER
& Hood DW (2005) Novel lipopolysaccharide biosynthetic
genes containing tetranucleotide repeats in Haemophilus
influenzae, identification of a gene for adding O-acetyl
groups. Mol Microbiol 58: 207–216.
Fox KL, Srikhanta YN & Jennings MP (2007) Phase variable
type III restriction-modification systems of host-adapted
bacterial pathogens. Mol Microbiol 65: 1375–1379.
Fukuda E, Kaminska KH, Bujnicki JM & Kobayashi I (2008)
Cell death upon epigenetic genome methylation: a novel
function of methyl-specific deoxyribonucleases. Genome Biol
9: R163.
Funchain P, Yeung A, Stewart JL, Lin R, Slupska MM & Miller
JH (2000) The consequences of growth of a mutator strain of
Escherichia coli as measured by loss of function among
multiple gene targets and loss of fitness. Genetics 154: 959–970.
Gawel D, Jonczyk P, Bialoskorska M, Schaaper RM &
Fijalkowska IJ (2002) Asymmetry of frameshift mutagenesis
during leading and lagging-strand replication in Escherichia
coli. Mutat Res 501: 129–136.
Gemayel R, Vinces MD, Legendre M & Verstrepen KJ (2010)
Variable tandem repeats accelerate evolution of coding and
regulatory sequences. Annu Rev Genet 44: 445–477.
Gibbons JG & Rokas A (2009) Comparative and functional
characterization of intragenic tandem repeats in 10
Aspergillus genomes. Mol Biol Evol 26: 591–602.
Glew MD, Baseggio N, Markham PF, Browning GF & Walker
ID (1998) Expression of the pMGA genes of Mycoplasma
gallisepticum is controlled by variation in the GAA
trinucleotide repeat lengths within the 5′ noncoding regions.
Infect Immun 66: 5833–5841.
Glew MD, Browning GF, Markham PF & Walker ID (2000)
pMGA phenotypic variation in Mycoplasma gallisepticum
occurs in vivo and is mediated by trinucleotide repeat length
variation. Infect Immun 68: 6027–6033.
Glover SW & Piekarowicz A (1972) Host specificity of DNA in
Haemophilus influenzae: restriction and modification in
strain Rd. Biochem Biophys Res Commun 46: 1610–1617.
Goldstein DB & Clark AG (1995) Microsatellite variation in
North American populations of Drosophila melanogaster.
Nucleic Acids Res 23: 3882–3886.
Goodwin AC, Weinberger DM, Ford CB, Nelson JC, Snider
JD, Hall JD, Paules CI, Peek Jr RM & Forsyth MH (2008)
Expression of the Helicobacter pylori adhesin SabA is
controlled via phase variation and the ArsRS signal
transduction system. Microbiology 154: 2231–2240.
Gravekamp C, Horensky DS, Michel JL & Madoff LC (1996)
Variation in repeat number within the alpha C protein of
group B streptococci alters antigenicity and protective
epitopes. Infect Immun 64: 3576–3583.
Gravekamp C, Kasper DL, Michel JL, Kling DE, Carey V &
Madoff LC (1997) Immunogenicity and protective efficacy
FEMS Microbiol Rev 38 (2014) 119–141
Variable tandem repeats and bacterial adaptation
of the alpha C protein of group B streptococci are inversely
related to the number of repeats. Infect Immun 65: 5216–
5221.
Griffin R, Cox AD, Makepeace K, Richards JC, Moxon ER &
Hood DW (2003) The role of lex2 in lipopolysaccharide
biosynthesis in Haemophilus influenzae strains RM7004 and
RM153. Microbiology 149: 3165–3175.
Guo X & Mr
azek J (2008) Long simple sequence repeats in
host-adapted pathogens localize near genes encoding
antigens, housekeeping genes, and pseudogenes. J Mol Evol
67: 497–509.
Gur-Arie R, Cohen CJ, Eitan Y, Shelef L, Hallerman EM &
Kashi Y (2000) Simple sequence repeats in Escherichia coli:
abundance, distribution, composition, and polymorphism.
Genome Res 10: 62–71.
G€
urlebeck D, Thieme F & Bonas U (2006) Type III effector
proteins from the plant pathogen Xanthomonas and their
role in the interaction with the host plant. J Plant Physiol
163: 233–255.
Hammerschmidt S, Hilse R, van Putten JP, Gerardy-Schahn R,
Unkmeir A & Frosch M (1996a) Modulation of cell surface
sialic acid expression in Neisseria meningitidis via a
transposable genetic element. EMBO J 15: 192–198.
Hammerschmidt S, M€ uller A, Sillmann H, M€ uhlenhoff M,
Borrow R, Fox A, van Putten J, Zollinger WD,
Gerardy-Schahn R & Frosch M (1996b) Capsule phase
variation in Neisseria meningitidis serogroup B by
slipped-strand mispairing in the polysialyltransferase gene
(siaD): correlation with bacterial invasion and the
outbreak of meningococcal disease. Mol Microbiol 20:
1211–1220.
Hannan A (2010) TRPing up the genome: tandem repeat
polymorphisms as dynamic sources of genetic variability in
health and disease. Discov Med 10: 314–321.
Hebert ML & Wells RD (2005) Roles of double-strand breaks,
nicks, and gaps in stimulating deletions of CTG.CAG
repeats by intramolecular DNA repair. J Mol Biol 353:
961–979.
Hebert ML, Spitz LA & Wells RD (2004) DNA double-strand
breaks induce deletion of CTG.CAG repeats in an
orientation-dependent manner in Escherichia coli. J Mol Biol
336: 655–672.
Herbers K, Conrads-Strauch J & Bonas U (1992)
Race-specificity of plant resistance to bacterial spot disease
determined by repetitive motifs in a bacterial avirulence
protein. Nature 356: 172–174.
Hood DW, Deadman ME, Jennings MP, Bisercic M,
Fleischmann RD, Venter JC & Moxon ER (1996) DNA
repeats identify novel virulence genes in Haemophilus
influenzae. Proc Natl Acad Sci USA 93: 11121–11125.
Hosking SL, Craig JE & High NJ (1999) Phase variation of
lic1A, lic2A and lic3A in colonization of the nasopharynx,
bloodstream and cerebrospinal fluid by Haemophilus
influenzae type b. Microbiology 145: 3005–3011.
Hussain M, Haggar A, Peters G, Chhatwal GS, Herrmann M,
Flock JI & Sinha B (2008) More than one tandem repeat
FEMS Microbiol Rev 38 (2014) 119–141
137
domain of the extracellular adherence protein of
Staphylococcus aureus is required for aggregation, adherence,
and host cell invasion but not for leukocyte activation.
Infect Immun 76: 5615–5623.
Ilver D, Arnqvist A, Ogren J et al. (1998) Helicobacter pylori
adhesin binding fucosylated histo-blood group antigens
revealed by retagging. Science 279: 373–377.
Iyer RR, Pluciennik A, Rosche WA, Sinden RR & Wells RD
(2000) DNA polymerase III proofreading mutants enhance
the expansion and deletion of triplet repeat sequences in
Escherichia coli. J Biol Chem 275: 2174–2184.
Jackson AL & Loeb LA (2000) Microsatellite instability
induced by hydrogen peroxide in Escherichia coli. Mutat Res
447: 187–198.
Jansen A, Gemayel R & Verstrepen KJ (2012) Unstable
microsatellite repeats facilitate rapid evolution of coding
and regulatory sequences. Genome Dyn 7: 108–125.
Janulczyk R, Masignani V, Maione D, Tettelin H, Grandi G &
Telford JL (2010) Simple sequence repeats and genome
plasticity in Streptococcus agalactiae. J Bacteriol 192:
3990–4000.
Jennings MP, Srikhanta YN, Moxon ER, Kramer M, Poolman
JT, Kuipers B & van der Ley P (1999) The genetic basis of
the phase variation repertoire of lipopolysaccharide
immunotypes in Neisseria meningitidis. Microbiology 145:
3013–3021.
Jerome JP, Bell JA, Plovanich-Jones AE, Barrick JE, Brown CT
& Mansfield LS (2011) Standing genetic variation in
contingency loci drives the rapid adaptation of
Campylobacter jejuni to a novel host. PLoS ONE 6: e16399.
Jin H, Ren Z, Whitby PW, Morton DJ & Stull TL (1999)
Characterization of hgpA, a gene encoding a haemoglobin/
haemoglobin-haptoglobin-binding protein of Haemophilus
influenzae. Microbiology 145: 905–914.
Jolivet-Gougeon A, Kovacs B, Le Gall-David S, Le Bars H,
Bousarghin L, Bonnaure-Mallet M, Lobel B, Guill
e F, Soussy
CJ & Tenke P (2011) Bacterial hypermutation: clinical
implications. J Med Microbiol 60: 563–573.
Jonsson AB, Nyberg G & Normark S (1991) Phase variation of
gonococcal pili by frameshift mutation in pilC, a novel gene
for pilus assembly. EMBO J 10: 477–488.
Jonsson AB, Pfeifer J & Normark S (1992) Neisseria
gonorrhoeae PilC expression provides a selective mechanism
for structural diversity of pili. P Natl Acad Sci USA 89:
3204–3208.
Jordan P, Snyder LA & Saunders NJ (2003) Diversity in
coding tandem repeats in related Neisseria spp. BMC
Microbiol 3: 23.
Kajava AV (2012) Tandem repeats in proteins: from sequence
to structure. J Struct Biol 179: 279–288.
Kanbashi K, Wang X, Komura J, Ono T & Yamamoto K
(1997) Frameshifts, base substitutions and minute deletions
constitute X ray-induced mutations in the endogenous tonB
gene of Escherichia coli K12. Mutat Res 385: 259–267.
Karatzas KA & Bennik MH (2002) Characterization of a
Listeria monocytogenes Scott A isolate with high tolerance
ª 2013 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
138
towards high hydrostatic pressure. Appl Environ Microbiol
68: 3183–3189.
Karatzas KA, Wouters JA, Gahan CG, Hill C, Abee T &
Bennik MH (2003) The CtsR regulator of Listeria
monocytogenes contains a variant glycine repeat region that
affects piezotolerance, stress resistance, motility and
virulence. Mol Microbiol 49: 1227–1238.
Karatzas KA, Valdramidis VP & Wells-Bennik MH (2005)
Contingency locus in ctsR of Listeria monocytogenes Scott A:
a strategy for occurrence of abundant piezotolerant isolates
within clonal populations. Appl Environ Microbiol 71:
8390–8396.
Kassai-J
ager E, Ortutay C, T
oth G, Vellai T & G
asp
ari Z
(2008) Distribution and evolution of short tandem repeats
in closely related bacterial genomes. Gene 410: 18–25.
Kay S, Hahn S, Marois E, Hause G & Bonas U (2007) A
bacterial effector acts as a plant transcription factor and
induces a cell size regulator. Science 318: 648–651.
Kelkar YD, Strubczewski N, Hile SE, Chiaromonte F, Eckert
KA & Makova KD (2010) What is a microsatellite: a
computational and experimental definition based upon
repeat mutational behavior at A/T and GT/AC repeats.
Genome Biol Evol 2: 620–635.
King MR, Vimr RP, Steenbergen SM, Spanjaard L, Plunkett G
III, Blattner FR & Vimr ER (2007) Escherichia coli
K1-specific bacteriophage CUS-3 distribution and function
in phase-variable capsular polysialic acid O acetylation. J
Bacteriol 189: 6447–6456.
Kornberg A, Bertsch LL, Jackson JF & Khorana HG (1964)
Enzymatic synthesis of deoxyribonucleic acid, XVI.
oligonucleotides as templates and the mechanism of their
replication. Proc Natl Acad Sci USA 51: 315–323.
Kusters JG, van Vliet AHM & Kuipers EJ (2006) Pathogenesis of
Helicobacter pylori infection. Clin Microbiol Rev 19: 449–490.
Lafontaine ER, Cope LD, Aebi C, Latimer JL, McCracken GH &
Hansen EJ (2000) The UspA1 protein and a second type of
UspA2 protein mediate adherence of Moraxella catarrhalis to
human epithelial cells in vitro. J Bacteriol 182: 1364–1373.
Lafontaine ER, Wagner NJ & Hansen EJ (2001) Expression of
the Moraxella catarrhalis UspA1 protein undergoes phase
variation and is regulated at the transcriptional level. J
Bacteriol 183: 1540–1551.
Lai Y & Sun F (2003) The relationship between microsatellite
slippage mutation rate and the number of repeat units. Mol
Biol Evol 20: 2123–2131.
Le Bars H, Bousarghin L, Bonnaure-Mallet M &
Jolivet-Gougeon A (2013) Role of a short tandem leucine/
arginine repeat in strong mutator phenotype acquisition in
a clinical isolate of Salmonella Typhimurium. FEMS
Microbiol Lett 338: 101–106.
Leclercq S, Rivals E & Jarne P (2007) Detecting microsatellites
within genomes: significant variation among algorithms.
BMC Bioinformatics 8: 125.
Legendre M, Pochet N, Pak T & Verstrepen KJ (2007)
Sequence-based estimation of minisatellite and microsatellite
repeat variability. Genome Res 17: 1787–1796.
ª 2013 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
K. Zhou et al.
Levengood SK, Beyer WF Jr & Webster RE (1991) TolA: a
membrane protein involved in colicin uptake contains an
extended helical region. P Natl Acad Sci USA 88: 5939–5943.
Levinson G & Gutman GA (1987) Slipped-strand mispairing: a
major mechanism for DNA sequence evolution. Mol Biol
Evol 4: 203–221.
Lewis LA, Gipson M, Hartman K, Ownbey T, Vaughn J &
Dyer DW (1999) Phase variation of HpuAB and HmbR,
two distinct haemoglobin receptors of Neisseria meningitidis
DNM2. Mol Microbiol 32: 977–989.
Lim KG, Kwoh CK, Hsu LY & Wirawan A (2013) Review of
tandem repeat search tools: a systematic approach to
evaluating algorithmic performance. Brief Bioinform 14:
67–81.
Lin WH & Kussell E (2012) Evolutionary pressures on simple
sequence repeats in prokaryotic coding regions. Nucleic
Acids Res 40: 2399–2413.
Lindb€ack T, Secic I & Rørvik LM (2011) A contingency locus
in prfA in a Listeria monocytogenes subgroup allows
reactivation of the PrfA virulence regulator during infection
in mice. Appl Environ Microbiol 77: 3478–3483.
Lindstedt BA (2005) Multiple-locus variable number tandem
repeats analysis for genetic fingerprinting of pathogenic
bacteria. Electrophoresis 26: 2567–2582.
Lindstedt BA (2011) Genotyping of selected bacterial
enteropathogens in Norway. Int J Med Microbiol 301: 648–653.
Liu L, Dybvig K, Panangala VS, van Santen VL & French CT
(2000) GAA trinucleotide repeat region regulates M9/pMGA
gene expression in Mycoplasma gallisepticum. Infect Immun
68: 871–876.
Lopes J, Ribeyre C & Nicolas A (2006) Complex
minisatellite rearrangements generated in the total or
partial absence of Rad27/hFEN1 activity occur in a single
generation and are Rad51 and Rad52 dependent. Mol Cell
Biol 26: 6675–6689.
Lysnyansky I, Rosengarten R & Yogev D (1996) Phenotypic
switching of variable surface lipoproteins in Mycoplasma
bovis involves high-frequency chromosomal rearrangements.
J Bacteriol 178: 5395–5401.
Mahdavi J, Sonden B, Hurtig M et al. (2002) Helicobacter
pylori SabA adhesin in persistent infection and chronic
inflammation. Science 297: 573–578.
Mak ANS, Bradley P, Bogdanove AJ & Stoddard BL (2013) Tal
effectors: function, structure, engineering and applications.
Curr Opin Struct Biol 23: 93–99.
Malkova A & Haber JE (2012) Mutations arising during repair
of chromosome breaks. Annu Rev Genet 46: 455–473.
Martin P, van de Ven T, Mouchel N, Jeffries AC, Hood DW &
Moxon ER (2003) Experimentally revised repertoire of
putative contingency loci in Neisseria meningitidis strain
MC58: evidence for a novel mechanism of phase variation.
Mol Microbiol 50: 245–257.
Martin P, Makepeace K, Hill SA, Hood DW & Moxon ER
(2005) Microsatellite instability regulates transcription factor
binding and gene expression. P Natl Acad Sci USA 102:
3800–3804.
FEMS Microbiol Rev 38 (2014) 119–141
Variable tandem repeats and bacterial adaptation
Massey RC & Buckling A (2002) Environmental regulation of
mutation rates at specific sites. Trends Microbiol 10: 580–
584.
Merkel A & Gemmell N (2008) Detecting short tandem
repeats from genome data: opening the software black box.
Brief Bioinform 9: 355–366.
Metruccio MM, Pigozzi E, Roncarati D, Berlanda Scorza F,
Norais N, Hill SA, Scarlato V & Delany I (2009) A novel
phase variation mechanism in the meningococcus driven by
a ligand-responsive repressor and differential spacing of
distal promoter elements. PLoS Pathog 5: e1000710.
Miller VL, Taylor RK & Mekalanos JJ (1987) Cholera toxin
transcriptional activator toxR is a transmembrane DNA
binding protein. Cell 48: 271–279.
Morbitzer R, R€ omer P, Boch J & Lahaye T (2010) Regulation
of selected genome loci using de novo-engineered
transcription activator-like effector (TALE)-type
transcription factors. P Natl Acad Sci USA 107:
21617–21622.
Mordhorst IL, Claus H, Ewers C et al. (2009)
O-acetyltransferase gene neuO is segregated according to
phylogenetic background and contributes to environmental
desiccation resistance in Escherichia coli K1. Environ
Microbiol 11: 3154–3165.
Morton DJ, Whitby PW, Jin H, Ren Z & Stull TL (1999)
Effect of multiple mutations in the hemoglobin- and
hemoglobin-haptoglobin-binding proteins, HgpA, HgpB,
and HgpC, of Haemophilus influenzae type b. Infect Immun
67: 2729–2739.
Moscou MJ & Bogdanove AJ (2009) A simple cipher governs
DNA recognition by TAL effectors. Science 326: 1501.
Moxon ER, Rainey PB, Nowak MA & Lenski RE (1994)
Adaptive evolution of highly mutable loci in pathogenic
bacteria. Curr Biol 4: 24–33.
Moxon R, Bayliss C & Hood D (2006) Bacterial contingency
loci: the role of simple sequence DNA repeats in bacterial
adaptation. Annu Rev Genet 40: 307–333.
Mr
azek J (2006) Analysis of distribution indicates diverse
functions of simple sequence repeats in Mycoplasma
genomes. Mol Biol Evol 23: 1370–1385.
Mr
azek J, Guo X & Shah A (2007) Simple sequence repeats in
prokaryotic genomes. Proc Natl Acad Sci USA 104: 8472–
8477.
Mudunuri SB & Nagarajaram HA (2007) IMEx: imperfect
microsatellite extractor. Bioinformatics 23: 1181–1187.
Mu~noz Bodnar A, Bernal A, Szurek B & L
opez CE (2013) Tell
me a tale of TALEs. Mol Biotechnol 53: 228–235.
Murphy GL, Connell TD, Barritt DS, Koomey M & Cannon
JG (1989) Phase variation of gonococcal protein II:
regulation of gene expression by slipped-strand mispairing
of a repetitive DNA sequence. Cell 56: 539–547.
Nilsson C, Skoglund A, Moran AP, Annuk H, Engstrand L &
Normark S (2006) An enzymatic ruler modulates Lewis
antigen glycosylation of Helicobacter pylori LPS during
persistent infection. P Natl Acad Sci USA 103: 2863–2868.
FEMS Microbiol Rev 38 (2014) 119–141
139
Nilsson C, Skoglund A, Moran AP, Annuk H, Engstrand L &
Normark S (2008) Lipopolysaccharide diversity evolving in
Helicobacter pylori communities through genetic
modifications in fucosyltransferases. PLoS ONE 3: e3811.
Orsi RH, Bowen BM & Wiedmann M (2010) Homopolymeric
tracts represent a general regulatory mechanism in
prokaryotes. Genomics 11: 102.
Palmer ME, Lipsitch M, Moxon ER & Bayliss CD (2013)
Broad conditions favor the evolution of phase-variable loci.
mBio 4: e00430–12.
Papazisi L, Gorton TS, Kutish G, Markham PF, Browning GF,
Nguyen DK, Swartzell S, Madan A, Mahairas G & Geary SJ
(2003) The complete genome sequence of the avian
pathogen Mycoplasma gallisepticum strain R(low).
Microbiology 149: 2307–2316.
P^aques F, Leung WY & Haber JE (1998) Expansions and
contractions in a tandem repeat induced by double-strand
break repair. Mol Cell Biol 18: 2045–2054.
P^aques F, Richard GF & Haber JE (2001) Expansions and
contractions in 36-bp minisatellites by gene conversion in
yeast. Genetics 158: 155–166.
Park SF, Purdy D & Leach S (2000) Localized reversible
frameshift mutation in the flhA gene confers phase
variability to flagellin gene expression in Campylobacter coli.
J Bacteriol 182: 207–210.
Pearson CE, Nichol Edamura K & Cleary JD (2005) Repeat
instability: mechanisms of dynamic mutations. Nat Rev
Genet 6: 729–742.
Piekarowicz A, Brzezi
nski R & Kauc L (1974) Host specificity
of DNA in Haemophilus influenzae: the in vivo action of the
restriction endonucleases on phage and bacterial DNA. Acta
Microbiol Pol A 27: 51–65.
Piekarowicz A, Klyz A, Kwiatek A & Stein DC (2001) Analysis
of type I restriction modification systems in the
Neisseriaceae: genetic organization and properties of the
gene products. Mol Microbiol 41: 1199–1210.
Power PM, Sweetman WA, Gallacher NJ, Woodhall MR,
Kumar GA, Moxon ER & Hood DW (2009) Simple
sequence repeats in Haemophilus influenzae. Infect Genet
Evol 9: 216–228.
Rando OJ & Verstrepen KJ (2007) Timescales of genetic and
epigenetic inheritance. Cell 128: 655–668.
Rasmussen M & Bj€ ork L (2001) Unique regulation of SclB – a
novel collagen-like surface protein of Streptococcus pyogenes.
Mol Microbiol 40: 1427–1438.
Razin S, Yogev D & Naot Y (1998) Molecular biology and
pathogenicity of mycoplasmas. Microbiol Mol Biol Rev 62:
1094–1156.
Ren Z, Jin H, Morton DJ & Stull TL (1998) hgpB, a gene
encoding a second Haemophilus influenzae hemoglobin- and
hemoglobin-haptoglobin-binding protein. Infect Immun 66:
4733–4741.
Ren Z, Jin H, Whitby PW, Morton DJ & Stull TL (1999)
Role of CCAA nucleotide repeats in regulation of
hemoglobin and hemoglobin-haptoglobin binding protein
ª 2013 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
140
genes of Haemophilus influenzae. J Bacteriol 181:
5865–5870.
Richard GF & P^aques F (2000) Mini- and microsatellite
expansions: the recombination connection. EMBO Rep 1:
122–126.
Richard GF, Dujon B & Haber JE (1999) Double-strand break
repair can lead to high frequencies of deletions within short
CAG/CTG trinucleotide repeats. Mol Gen Genet 261: 871–882.
Richard GF, Kerrest A & Dujon B (2008) Comparative
genomics and molecular dynamics of DNA repeats in
eukaryotes. Microbiol Mol Biol Rev 72: 686–727.
Richardson AR & Stojiljkovic I (1999) HmbR, a
hemoglobin-binding outer membrane protein of Neisseria
meningitidis, undergoes phase variation. J Bacteriol 181:
2067–2074.
Ritz D, Lim J, Reynolds CM, Poole LB & Beckwith J (2001)
Conversion of a peroxiredoxin into a disulfide reductase by
a triplet repeat expansion. Science 294: 158–160.
Rocha EP (2003) An appraisal of the potential for illegitimate
recombination in bacterial genomes and its consequences:
from duplications to genome reduction. Genome Res 13:
1123–1132.
Rocha EP, Matic I & Taddei F (2002) Over-representation of
repeats in stress response genes: a strategy to increase
versatility under stressful conditions? Nucleic Acids Res 30:
1886–1894.
Rosengarten R & Wise KS (1991) The Vlp system of
Mycoplasma hyorhinis: combinatorial expression of distinct
size variant lipoproteins generating high-frequency surface
antigenic variation. J Bacteriol 173: 4782–4793.
Ryan KA & Lo RY (1999) Characterization of a CACAG
pentanucleotide repeat in Pasteurella haemolytica and its
possible role in modulation of a novel type III
restriction-modification system. Nucleic Acids Res 27:
1505–1511.
Sadarangani M, Pollard AJ & Gray-Owen SD (2011) Opa
proteins and CEACAMs: pathways of immune engagement
for pathogenic Neisseria. FEMS Microbiol Rev 35: 498–514.
Saint-Ruf C & Matic I (2006) Environmental tuning of
mutation rates. Environ Microbiol 8: 193–199.
Sarkari J, Pandit N, Moxon ER & Achtman M (1994) Variable
expression of the Opc outer membrane protein in Neisseria
meningitidis is caused by size variation of a promoter
containing poly-cytidine. Mol Microbiol 13: 207–217.
Schaper E, Kajava AV, Hauser A & Anisimova M (2012)
Repeat or not repeat? Statistical validation of tandem repeat
prediction in genomic sequences. Nucleic Acids Res 40:
10005–10017.
Schlotterer C, Imhof M, Wang H, Nolte V & Harr B (2006)
Low abundance of Escherichia coli microsatellites is
associated with an extremely low mutation rate. J Evol Biol
19: 1671–1676.
Scholze H & Boch J (2011) TAL effectors are remote controls
for gene activation. Curr Opin Microbiol 14: 47–53.
Schug MD, Hutter CM, Wetterstrand KA, Gaudette MS,
Mackay TF & Aquadro CF (1998) The mutation rates of
ª 2013 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
K. Zhou et al.
di-, tri- and tetranucleotide repeats in Drosophila
melanogaster. Mol Biol Evol 15: 1751–1760.
Schweda EKH, Richards JC, Hood DW & Moxon ER (2007)
Expression and structural diversity of the lipopolysaccharide
of Haemophilus influenzae: implication in virulence. Int J
Med Microbiol 297: 297–306.
Seale TW, Morton DJ, Whitby PW, Wolf R, Kosanke SD,
VanWagoner TM & Stull TL (2006) Complex role of
hemoglobin and hemoglobin-haptoglobin binding proteins
in Haemophilus influenzae virulence in the infant rat model
of invasive infection. Infect Immun 74: 6213–6225.
Seib KL, Pigozzi E, Muzzi A, Gawthorne JA, Delany I,
Jennings MP & Rappuoli R (2011) A novel epigenetic
regulator associated with the hypervirulent Neisseria
meningitidis clonal complex 41/44. FASEB J 25: 3622–3633.
Shaver AC & Sniegowski PD (2003) Spontaneously arising
mutL mutators in evolving Escherichia coli populations are
the result of changes in repeat length. J Bacteriol 185:
6076–6082.
Sheets AJ & St Geme III JW (2011) Adhesive activity of the
Haemophilus cryptic genospecies Cha autotransporter is
modulated by variation in tandem peptide repeats. J
Bacteriol 193: 329–339.
Sia EA, Kokoska RJ, Dominska M, Greenwell P & Petes TD
(1997) Microsatellite instability in yeast: dependence on
repeat unit size and DNA mismatch repair genes. Mol Cell
Biol 17: 2851–2858.
Spinosa MR, Progida C, Tala A, Cogli L, Alifano P & Bucci C
(2007) The Neisseria meningitidis capsule is important for
intracellular survival in human cells. Infect Immun 75:
3594–3603.
Srikhanta YN, Maguire TL, Stacey KJ, Grimmond SM &
Jennings MP (2005) The phasevarion: a genetic system
controlling coordinated, random switching of expression of
multiple genes. P Natl Acad Sci USA 102: 5547–5551.
Srikhanta YN, Dowideit SJ, Edwards JL et al. (2009)
Phasevarions mediate random switching of gene expression
in pathogenic Neisseria. PLoS Pathog 5: e1000400.
Srikhanta YN, Fox KL & Jennings MP (2010) The phasevarion:
phase variation of type III DNA methyltransferases controls
coordinated switching in multiple genes. Nat Rev Microbiol
8: 196–206.
Srikhanta YN, Gorrell RJ, Steen JA, Gawthorne JA, Kwok T,
Grimmond SM, Robins-Browne RM & Jennings MP (2011)
Phasevarion mediated epigenetic gene regulation in
Helicobacter pylori. PLoS ONE 6: e27569.
Streisinger G, Okada Y, Emrich J, Newton J, Tsugita A,
Terzaghi E & Inouye M (1966) Frameshift mutations and
the genetic code. Cold Spring Harb Symp Quant Biol 31:
77–84.
Tauseef I, Harrison OB, Wooldridge KG et al. (2011)
Influence of the combination and phase variation status of
the haemoglobin receptors HmbR and HpuAB on
meningococcal virulence. Microbiology 157: 1446–1456.
Tesfazgi Mebrhatu M, Wywial E, Ghosh A, Michiels CW,
Lindner AB, Taddei F, Bujnicki JM, Van Melderen L &
FEMS Microbiol Rev 38 (2014) 119–141
Variable tandem repeats and bacterial adaptation
Aertsen A (2011) Evidence for an evolutionary antagonism
between Mrr and Type III modification systems. Nucleic
Acids Res 39: 5991–6001.
Theiss P & Wise KS (1997) Localized frameshift mutation
generates selective, high-frequency phase variation of a
surface lipoprotein encoded by a mycoplasma ABC
transporter operon. J Bacteriol 179: 4013–4022.
Treangen TJ, Abraham AL, Touchon M & Rocha EP (2009)
Genesis, effects and fates of repeats in prokaryotic genomes.
FEMS Microbiol Rev 33: 539–571.
van Belkum A (1999) Short sequence repeats in microbial
pathogenesis and evolution. Cell Mol Life Sci 56: 729–734.
van der Ende A, Hopman CT & Dankert J (2000) Multiple
mechanisms of phase variation of PorA in Neisseria
meningitidis. Infect Immun 68: 6685–6690.
van der Woude MW & Baumler AJ (2004) Phase and antigenic
variation in bacteria. Clin Microbiol Rev 17: 581–611.
van Ham SM, van Alphen L, Mooi FR & van Putten JP (1993)
Phase variation of H. influenzae fimbriae: transcriptional
control of two divergent genes through a variable combined
promoter region. Cell 73: 1187–1196.
van Passel MW & Ochman H (2007) Selection on the genic
location of disruptive elements. Trends Genet 23: 601–604.
van Selm S, van Cann LM, Kolkman MA, van der Zeijst BA &
van Putten JP (2003) Genetic basis for the structural
difference between Streptococcus pneumoniae serotype 15B
and 15C capsular polysaccharides. Infect Immun 71: 6192–
6198.
Vandersmissen L, De Buck E, Saels V, Coil DA & Anne J
(2010) A Legionella pneumophila collagen-like protein
encoded by a gene with a variable number of tandem
repeats is involved in the adherence and invasion of host
cells. FEMS Microbiol Lett 306: 168–176.
Verstrepen KJ, Reynolds TB & Fink GR (2004) Origins of
variation in the fungal cell surface. Nat Rev Microbiol 2:
533–540.
Vimr E & Steenbergen SM (2006) Mobile contingency locus
controlling Escherichia coli K1 polysialic acid capsule
acetylation. Mol Microbiol 60: 828–837.
Vogler AJ, Keys C, Nemoto Y, Colman RE, Jay Z & Keim P
(2006) Effect of repeat copy number on variable-number
tandem repeat mutations in Escherichia coli O157:H7. J
Bacteriol 188: 4253–4263.
Waite RD, Struthers JK & Dowson CG (2001) Spontaneous
sequence duplication within an open reading frame of the
pneumococcal type 3 capsule locus causes high-frequency
phase variation. Mol Microbiol 42: 1223–1232.
Waite RD, Penfold DW, Struthers JK & Dowson CG (2003)
Spontaneous sequence duplications within capsule genes
cap8E and tts control phase variation in Streptococcus
pneumoniae serotypes 8 and 37. Microbiology 149: 497–504.
FEMS Microbiol Rev 38 (2014) 119–141
141
Wang G, Rasko DA, Sherburne R & Taylor DE (1999)
Molecular genetic basis for the variable expression of
Lewis Y antigen in Helicobacter pylori: analysis of the
alpha (1,2) fucosyltransferase gene. Mol Microbiol 31:
1265–1274.
Wells RD, Dere R, Hebert ML, Napierala M & Son LS (2005)
Advances in mechanisms of genetic instability related to
hereditary neurological diseases. Nucleic Acids Res 33:
3785–3798.
White FF, Potnis N, Jones JB & Koebnik R (2009) The type III
effectors of Xanthomonas. Mol Plant Pathol 10: 749–766.
Willems R, Paul A, van der Heide HG, ter Avest AR & Mooi
FR (1990) Fimbrial phase variation in Bordetella pertussis: a
novel mechanism for transcriptional regulation. EMBO J 9:
2803–2809.
Yamaoka Y, Kita M, Kodama T, Imamura S, Ohno T, Sawai
N, Ishimaru A, Imanishi J & Graham DY (2002)
Helicobacter pylori infection in mice: role of outer
membrane proteins in colonization and inflammation.
Gastroenterology 123: 1992–2004.
Yang QL & Gotschlich EC (1996) Variation of gonococcal
lipooligosaccharide structure is due to alterations in poly-G
tracts in lgt genes encoding glycosyl transferases. J Exp Med
183: 323–327.
Yang J, Wang J, Chen L, Yu J, Dong J, Yao ZJ, Shen Y, Jin Q
& Chen R (2003) Identification and characterization of
simple sequence repeats in the genomes of Shigella species.
Gene 322: 85–92.
Yogev D, Rosengarten R, Watson-McKown R & Wise KS
(1991) Molecular basis of Mycoplasma surface antigenic
variation: a novel set of divergent genes undergo
spontaneous mutation of periodic coding regions and 5′
regulatory sequences. EMBO J 10: 4069–4079.
Zahra R, Blackwood JK, Sales J & Leach DR (2007)
Proofreading and secondary structure processing determine
the orientation dependence of CAG 9 CTG trinucleotide
repeat instability in Escherichia coli. Genetics 176: 27–41.
Zaleski P, Wojciechowski M & Piekarowicz A (2005) The role
of Dam methylation in phase variation of Haemophilus
influenzae genes involved in defence against phage infection.
Microbiology 151: 3361–3369.
Zhang Q & Wise KS (1997) Localized reversible frameshift
mutation in an adhesin gene confers a phase-variable
adherence phenotype in mycoplasma. Mol Microbiol 25:
859–869.
Zhou K, Michiels CW & Aertsen A (2012a) Variation of
intragenic tandem repeat tract of tolA modulates Escherichia
coli stress tolerance. PLoS ONE 7: e47766.
Zhou K, Vanoirbeek K, Aertsen A & Michiels CW (2012b)
Variability of the tandem repeat region of the Escherichia
coli tolA gene. Res Microbiol 163: 316–322.
ª 2013 Federation of European Microbiological Societies.
Published by John Wiley & Sons Ltd. All rights reserved
 Europe PMC Funders Group
Author Manuscript
J Mol Evol. Author manuscript; available in PMC 2014 April 14.
Published in final edited form as:
J Mol Evol. 2012 June ; 74(0): 273–280. doi:10.1007/s00239-012-9505-4.
Europe PMC Funders Author Manuscripts

Estimating the Relative Roles of Recombination and Point

Mutation in the Generation of Single Locus Variants in
Campylobacter jejuni and Campylobacter coli
Shoukai Yu,
Infectious Disease Research Centre, Institute of Veterinary, Animal, and Biomedical Sciences,
Massey University, Palmerston North, New Zealand; S.Yu1@massey.ac.nz
Paul Fearnhead,
Department of Mathematics and Statistics, Lancaster University, Lancaster, UK
Barbara R. Holland,
School of Mathematics and Physics, University of Tasmania, Hobart, Australia
Patrick Biggs,
Infectious Disease Research Centre, Institute of Veterinary, Animal, and Biomedical Sciences,
Massey University, Palmerston North, New Zealand
Martin Maiden, and
Department of Zoology, University of Oxford, Oxford, UK
Nigel French
Infectious Disease Research Centre, Institute of Veterinary, Animal, and Biomedical Sciences,
Massey University, Palmerston North, New Zealand; Allan Wilson Centre for Molecular Ecology
and Evolution, Massey University, Palmerston North, New Zealand; n.p.french@massey.ac.nz
Europe PMC Funders Author Manuscripts

Abstract
Single locus variants (SLVs) are bacterial sequence types that differ at only one of the seven
canonical multilocus sequence typing (MLST) loci. Estimating the relative roles of recombination
and point mutation in the generation of new alleles that lead to SLVs is helpful in understanding
how organisms evolve. The relative rates of recombination and mutation for Campylobacter jejuni
and Campylobacter coli were estimated at seven different housekeeping loci from publically
available MLST data. The probability of recombination generating a new allele that leads to an
SLV is estimated to be roughly seven times more than that of mutation for C. jejuni, but for C. coli
recombination and mutation were estimated to have a similar contribution to the generation of
SLVs. The majority of nucleotide differences (98 % for C. jejuni and 85 % for C. coli) between
strains that make up an SLV are attributable to recombination. These estimates are much larger
than estimates of the relative rate of recombination to mutation calculated from more distantly
related isolates using MLST data. One explanation for this is that purifying selection plays an
important role in the evolution of Campylobacter. A simulation study was performed to test the
performance of our method under a range of biologically realistic parameters. We found that our
method performed well when the recombination tract length was longer than 3 kb. For situations
in which recombination may occur with shorter tract lengths, our estimates are likely to be an
underestimate of the ratio of recombination to mutation, and of the importance of recombination
for creating diversity in closely related isolates. A parametric bootstrap method was applied to
calculate the uncertainty of these estimates.

© Springer Science+Business Media, LLC 2012

Correspondence to: Nigel French.
 Yu et al. Page 2

Keywords
Single locus variant; MLST; Purifying; selection; Evolution of Campylobacter; Mutation;
Recombination

Introduction
Europe PMC Funders Author Manuscripts

The genus Campylobacter is the major cause of gastroenteritis in many industrialized

countries (Tauxe et al. 1992), with approximately 1 % of the population throughout the
western world being affected by campylobacteriosis every year (The World Health
Organization, cited in Humphrey et al. 2007). The species Campylobacter jejuni (C. jejuni)
and Campylobacter coli (C. coli) are the main causes of bacterial food-borne disease in
developed countries, compared to other members of the family Campylobacteriaceae
(Konkel et al. 1999).

Substantial evidence for the presence of recombination at specific genes has been found in
several studies (Suerbaum et al. 2001; Fearnhead et al. 2005). The relative contributions of
recombination and point mutation to genetic diversity have also been investigated (Feil et al.
1999, 2000,2001; Sarkar and Guttman 2004). Although most research indicates that
recombination contributes more to genetic diversity than mutation, there is considerable
uncertainty about the relative number of events and the number of nucleotide differences
that may be attributable to these two processes (Schouls et al. 2003; Richman et al. 2003;
Fearnhead et al. 2005). This paper is focused on estimating the relative contributions of
recombination and point mutation to the generation of new alleles that lead to single locus
variants (SLVs), based on C. jejuni and C. coli from the seven gene multilocus sequence
typing (MLST) scheme.

An SLV is a pair of sequence types (STs) that differ at exactly one of the seven alleles that
make up the MLST profile (Feil et al. 2004). SLVs are pairs of STs that most likely share a
very recent common ancestor and the analysis of SLVs can be helpful in understanding the
Europe PMC Funders Author Manuscripts

evolution and molecular epidemiology of pathogens. The large collections of isolates that
have been characterized by MLST provide a good opportunity to study SLVs in detail.

This research is based on distinct STs of C. jejuni and C. coli in the PubMLST database
(http://pubmlst.org/campylobacter). In order to understand whether there are differences in
the mechanisms that produce SLVs across the genome, SLVs were divided into groups
depending on the locus at which the STs differ. The distribution of nucleotide differences
within SLVs was explored. The nucleotide differences between two STs that form an SLV
can be generated by two different kinds of events: recombination or mutation. Intuitively,
SLVs that comprise two STs which differ at many nucleotide positions are more likely to be
due to recombination, whereas those that differ at only a few nucleotide positions may be
the result of point mutations. In this study, an EM algorithm was applied to allocate SLVs
into either a point mutation only model or a recombination model. Two key parameters were
estimated: the probability that an SLV arose due to point mutation(s) only, and the relative
rate of recombination to mutation. In order to test the performance of our method, a
simulation study was performed under a range of biologically realistic parameters. When the
recombination tract length was longer than 3 kb our method performed well. Three kilobase
pairs is the average of the estimated mean tract lengths suggested by previous research on
Campylobacter (Schouls et al. 2003; Fearnhead et al. 2005; Wilson et al.2009; Biggs et al.
2011). When recombination occurs with shorter tract lengths, our estimates may
underestimate the ratio of recombination to mutation.

J Mol Evol. Author manuscript; available in PMC 2014 April 14.

 Yu et al. Page 3

Materials and Methods

Campylobacter Data
The data were taken from the PubMLST database (September 27, 2010); at that time the
PubMLST database contained 4676 distinct C. jejuni and C. coli STs. MLST is a way of
typing strains that is based on nucleotide sequences (Maiden et al. 1998). Using the MLST
technique (Dingle et al. 2001), these isolates are sequenced at seven housekeeping loci
Europe PMC Funders Author Manuscripts

(aspA, glnA, gltA, glyA, pgm, tkt, and uncA). These seven loci are widely dispersed around
the genome, which means there is a very low chance for one recombination to change two or
more loci.

We separated ST datasets for C. jejuni and C. coli, and excluded the 22 STs found in both
species. Furthermore, we separated C. coli by clades according to previous research
(Sheppard et al. 2008, 2011), and we chose C. coli clade 1 to investigate in detail because C.
coli clade 1 contains more STs, and is more diverse, compared to the other two clades
(Sheppard et al. 2008). We selected clade 1 from C. coli by extracting all STs that are
members of ST-828 clonal complex and ST-1150 clonal complex (Sheppard et al. 2010).
There are 3654 STs for C. jejuni, and 606 distinct STs for C. coli clade1.

Methods Overview
Either mutation(s) or recombination(s) can generate SLVs. In this paper, mutation is defined
as a single nucleotide change (a point mutation), whereas recombination represents the
transfer of several adjacent nucleotides from one DNA source to another. An event is either
a mutation or a recombination. An SLV can be generated by one or more events; however,
recombination will tend to mask mutation. We model separately the mutation and
recombination process to derive a probability model for the number of nucleotide
differences between STs, under both the assumption that the SLV has been created solely by
mutation, and that it has not. This then enables us to estimate the proportion of SLVs that
have been caused solely by mutation, and also estimate the relative rate of recombination to
mutation. More details of the analysis are given in the Supplementary Material.
Europe PMC Funders Author Manuscripts

Modeling SLV Evolution

The data consists of, for each SLV, the locus at which the pair of STs differ, and the number
of nucleotide differences at that locus. From this, we aim to infer how likely it is that the
differences observed at this locus arise from point mutation only, as opposed to being
produced by recombination.

To do this, we first model the distribution of nucleotide differences we would expect at an

SLV at a given locus if these differences are solely due to mutation. This can be done by
first calculating the probability of an SLV given the number of point of mutations that have
occurred in one locus as the likelihood function, introducing a prior distribution for the
number of mutations to occur between two STs in that locus. The former probability is
based on the need for all mutation events to occur at the same locus. Under the coalescent
theory, a geometric distribution is chosen to use as the prior distribution (Hein et al. 2005).
Under Bayesian theory, we can obtain the required conditional distribution (Equation 1 in
Supplementary Material). The resulting conditional distribution of the number of nucleotide
differences is concentrated on small numbers of nucleotide differences, and is robust to the
choice of prior.

Second, the probability of observing h (h = 1, 2, 3…) nucleotide differences introduced by

recombination was estimated using Bayesian methods. It was calculated by sampling the
alleles based on their frequencies in the current database. Two (simplifying) assumptions for

J Mol Evol. Author manuscript; available in PMC 2014 April 14.

 Yu et al. Page 4

the recombination model were made: (1) if recombination occurs between two alleles it
affects an entire locus rather than just part of a locus; and (2) we ignore the effect of any
additional mutation events. Under these assumptions, our model suggests that in most cases
recombination will introduce many more nucleotide differences than expected under the
mutation only model. Note that our results are robust to the assumption in (1) unless
recombination affects only small fragments of a locus, in these cases our assumption will
tend to lead to overestimates of the proportion of SLVs due to mutation only. Hence, it will
Europe PMC Funders Author Manuscripts

tend to underestimate the ratio of recombination to mutation.

Given these two models, we can then estimate the proportion of our SLVs at each locus that
are due to mutation only. In practice, we use an expectation-maximization (EM) algorithm
(Dempster et al. 1977) to infer this proportion. Finally, based on the estimated proportion of
SLVs at a given locus that is due to mutation only we estimate the probability that the single
event that led to the generation of a new allele was a mutation. The above analysis was
carried out by an R script (available by request from the first author).

To test the accuracy of our method for estimating the ratio of recombination to mutation,
MLST data were simulated under different known ratios of recombination to mutation with
different recombination tract lengths using SimMLST software (Didelot et al. 2009).

We used a parametric bootstrap to assess uncertainty in estimates. We simulated 100

datasets for both C. coli and C. jejuni. These datasets matched the true data in terms of
number of STs, relative rate of mutation to recombination, and overall mutation rate across
the seven gene loci. Within the simulations we assumed that the mutation rate and
recombination rate were the same across loci. For our simulated data we estimated the
probability of an event being a mutation, and calculated the variability of estimates of this
quantity across the simulations: both for estimates for a single locus, and for the estimate
obtained by averaging across loci. We consider estimates of this quantity as the variance of
the estimates changed little when we varied the true value of the relative rate of
recombination to mutation. Confidence intervals were then calculated using a normal
approximation, and transformed to confidence intervals for the relative rate of
recombination to mutation.
Europe PMC Funders Author Manuscripts

Results
SLV Analysis on the Campylobacter MLST Databases
From our downloaded dataset, there were 7417 SLVs (aspA: 992; glnA: 1045; gltA: 1250;
glyA: 773; pgm: 1580; tkt: 1060; and uncA 717) for C. jejuni, and 1842 SLVs (aspA: 110;
glnA: 179; gltA: 128; glyA: 292; pgm: 325; tkt: 647; and uncA: 161) for C. coli clade1. The
difference in the number of SLVs at each locus suggests that it is worthwhile estimating the
relative mutation and recombination rates separately for each locus.

The Distribution of Nucleotide Differences Between Each SLV for Each Locus
Each SLV relates to one pair of STs, and the plots (Figs. 1, 2) show the nucleotide
differences that occurred within those pairs of STs at each MLST locus for C. jejuni and C.
coli clade 1. These plots show that SLVs with a large number of nucleotide differences
(>45) occurred in every locus. The pairs of STs with a large number of nucleotide
differences (50–80) are almost certainly due to recombination, as it is highly unlikely that
more than 50 independent point mutations would occur at a single locus while the other six
loci remained the same. These large differences are likely to be due to recombination
between C. jejuni and C. coli (Sheppard et al. 2008; Wilson et al.2009). Species were
designated according to the PubMLST data, and only those SLVs that comprised STs that

J Mol Evol. Author manuscript; available in PMC 2014 April 14.

 Yu et al. Page 5

were assigned 100 % C. jejuni or C. coli were plotted. Even with this strict species
designation, there were still large nucleotide differences visible between SLVs within
species. There were second peaks in the range of 15-20 differences at the loci glyA, pgm,
and tkt for both C. jejuni and C. coli clade 1. These peaks are likely to be due to
recombination as well. The first peak of most loci (except for pgm for C. jejuni and tkt for C.
coli clade 1) represented approximately 100-200 SLVs for C. jejuni and around 100 SLVs
for C. coli clade 1 with only one nucleotide difference; most of these are more likely to be
Europe PMC Funders Author Manuscripts

due to mutation.

Relative Contributions of Recombination and Mutation Separately for C. jejuni and C. coli
Clade 1
Tables 1 and 2 demonstrate that recombination contributed more to the generation of SLVs
than did mutation for both the groups (C. jejuni and C. coli clade 1), but the range of
estimates vary for the two groups. The average ratio of recombination events to mutation
events from the seven loci is 6.96 (95 % CI 6.08, 8.09) for C. jejuni (Table 1), and 1.01 (95
% CI 0.78, 1.30) for C. coli clade 1 (Table 2).

For each locus, we also estimated the proportion of nucleotide differences introduced by
recombination as opposed to mutation, and this ranged from 97 % (gltA and glyA) to 99 %
(aspA, tkt, and uncA) for C. jejuni, and from 60 % (glnA) to 98 % (aspA) for C. coli clade 1.

We also investigated the robustness of the mutation model to different prior distributions of
the probability of events caused by mutations only. These suggest that the results in
Supplementary Table 1 and 2 are conservative regarding the importance of recombination in
producing new variation for C. jejuni and C. coli clade 1.

We see evidence for differences in the relative role of recombination to mutation across the
genes (Tables 1, 2). In particular, the parametric bootstrap results show that there is evidence
for a lower rate of recombination in glnA for C. coli and for glyA in C. jejuni, and for a
higher rate in aspA in C. coli. To assess the strength of this evidence, we looked at the
lowest (and the highest) estimated value of the relative rate of recombination to mutation
Europe PMC Funders Author Manuscripts

across the seven genes in our simulated data divided by the average of estimated rate across
the seven genes. For both C. coli and C. jejuni, we never observed an estimate as low as that
for glnA and glyA, respectively, across the 100 simulations in each case (the lowest
estimates were 0.36 and 0.62 for C. coli and C. jejuni, respectively, compared to observed
values of 0.23 and 0.43) or as high as aspA for C. coli (highest estimate was 2.04, compared
to an observed value of 2.21).

Discussion
We have analyzed SLVs to infer the relative importance of recombination and mutation to
generate differences between closely related C. jejuni and C. coli clade 1 isolates. The
higher average estimates for C. jejuni compared to C. coli demonstrates higher
recombination in C. jejuni, compared to C. coli. This is consistent with the existing
population structure (three clades) of C. coli, but not with apparent subclade structure in C.
jejuni (Sheppard et al.2008). We estimate that recombination contributes between 2.97 and
8.91 times more than mutation to events that generate new alleles for C. jejuni, depending
on the MLST locus, and between 0.23 and 2.23 for C. coli clade 1. The variations between
housekeeping genes within species also show the different evolution pressure on different
genes. For C. jejuni, glyA has less recombination contribution, compared to the other six
genes. For C. coli, glnA has less recombination contribution, compared to the other six
genes.

J Mol Evol. Author manuscript; available in PMC 2014 April 14.

 Yu et al. Page 6

Our analysis has similarities to that of Schouls et al. (2003), who used the approach
described by Feil et al. (2000) to estimate the relative rate of recombination and mutation for
C. jejuni. The original idea of Feil et al.’s method (2000) is put forward by Guttman and
Dykhuizen (1994). However, their method overestimates the ratio of recombination to
mutation, compared to ours. They also analyzed SLVs, though restricted to pairs of SLVs
within the same clonal complex. Furthermore, rather than the model-based approach we
consider, they used a simple rule to classify which SLVs had been caused by mutation as
Europe PMC Funders Author Manuscripts

opposed to recombination. The rule was that if a pair of SLVs varies by a single nucleotide
difference and one of the MLST alleles at the locus was unique, it is due to a mutation,
whereas all other pairs of SLVs are caused by recombination. This means that, under this
algorithm, SLVs that differ by two nucleotide differences could not have arisen by two
independent mutation events, and recombination events that mask mutation events are not
considered. Both assumptions may lead to an underestimate of mutation. The analysis of
Schouls et al. (2003) estimates that recombination is approximately eight times more likely
to change an allele than mutation. This is larger than our estimate, which is likely to be due
to these biases in the method used by Schouls et al. (2003). According to Feil et al.’s method
(2000), Schouls et al. (2003) estimated a recombination size of about 3.3 kb. We
implemented a simplified version of Feil et al.’s method (2000) (details in the
Supplementary Material), and the results show that under the 3 kb recombination size, the
ratio is overestimated.

Our estimates suggest a more important role for recombination in producing new diversity
into C. jejuni than more recent studies which have analyzed samples of C. jejuni isolates
from different source populations. Fearnhead et al. (2005) estimate that recombination rates
are, if anything, less than the mutation rates. While Vos and Didelot (2008), and Wilson et
al. (2009) give estimates of the proportion of nucleotide differences introduced by
recombination as opposed to mutation which are much smaller than the ones we obtained.
Both the studies concluded that the number of nucleotide differences introduced by
recombination are only approximately twice as many as those introduced by point mutation:
2.2 for Vos and Didelot (2008) and 2.67 (95 % CI 1.39, 4.95) for Wilson et al. (2009).
Europe PMC Funders Author Manuscripts

The difference between our study and these is that we analyze only SLVs, which means we
are looking at closely related STs for which there has been less time for selection to act.
Intuitively, selection is likely to be the strongest against recombination events that introduce
large differences, although it is possible that some recombination events may introduce a
section of DNA from an organism that is highly adapted and “successful” in the given
environment. Therefore, although we estimate that recombination is introducing more
differences than previously thought in our closely related and recently evolved STs, many of
these differences may be subsequently purged from the population due to weak purifying
selection. This is consistent with the effects of purifying selection described in Wilson et
al.’s paper (2009).

Whole genome analysis may provide a greater insight into the genome-wide evolution of
Campylobacter and provide further explanations for the apparent differences between
previous estimates of recombination and mutation. Recently, Biggs et al. (2011) analyzed
the genomes of two closely related Campylobacter ST-474 isolates that also had identical
flaA SVR regions and compared them to available C. jejuni reference strains. They
estimated that around 97 % of the nucleotide differences between these two closely related
isolates were caused by recombination. This estimate is similar to ours, and suggests that the
importance of recombination for driving changes in C. jejuni is not just confined to the
MLST housekeeping genes we have studied.

J Mol Evol. Author manuscript; available in PMC 2014 April 14.

 Yu et al. Page 7

The aim of this study was to increase our understanding of the evolution of C. jejuni and C.
coli by investigating the generation of SLVs. The availability of the large database of C.
jejuni and C. coli isolates provides a good opportunity to investigate the evolution of C.
jejuni and C. coli using SLVs. Using seven independent housekeeping loci we used the
method proposed in this paper to estimate that recombination contributes roughly seven
times as much as mutation to the generation of SLVs for C. jejuni, and equally for C. coli,
which provides further evidence that recombination plays a more important role in the
Europe PMC Funders Author Manuscripts

evolution of C. jejuni and C. coli than mutation.

Our results also point to important differences in terms of the forces driving evolution for C.
jejuni and C. coli; and suggest that the relative role of recombination to mutation may differ
between genes, and these differences themselves may be different for C. jejuni and C. coli.
Understanding what is causing these differences will be important for fully understanding
how these bacteria may evolve in the future. However, the fact that we observed differences
in recombination between C. jejuni and C. coli is consistent with the introgression
hypothesis of Sheppard et al.’s paper (2011), which implies that patterns of genetic
exchange have changed over time. The research on SLVs described in this paper could be
extended either by considering more genes, such as flagellin genes (flaA and flaB)
(Meinersmann and Hiett 2000), and porA, the gene encoding the major outer membrane
proteins (MOMPs) (Zhang et al. 2000; Clark et al. 2005), or by considering other species of
Campylobacter.

Supplementary Material
Refer to Web version on PubMed Central for supplementary material.

Acknowledgments
The authors acknowledge the Marsden Fund project 08-MAU-099 (Cows, starlings and Campylobacter in New
Zealand: unifying phylogeny, genealogy, and epidemiology to gain insight into pathogen evolution) for funding this
project. This publication made use of the Campylobacter Multi Locus Sequence Typing website (http://
pubmlst.org/campylobacter/) developed by Keith Jolley and sited at the University of Oxford (Jolley and Maiden
Europe PMC Funders Author Manuscripts

2010, BMC Bioinformatics, 11:595). The development of this site has been funded by the Wellcome Trust. BRH
acknowledges the Australian Research Council (Grant FT100100031).

References
Biggs PJ, Fearnhead P, Hotter G, Mohan V, Collins-Emerson J, Kwan E, Besser TE, Cookson A,
Carter PE, French NP. Whole-genome comparison of two Campylobacter jejuni isolates of the same
sequence type reveals multiple loci of different ancestral lineage. PLoS One. 2011; 6(11):e27121.
[PubMed: 22096527]
Clark CG, Bryden L, Cuff WR, Johnson PL, Jamieson F, Ciebin B, Wang G. Use ofthe Oxford
multilocus sequence typing protocol and sequencing of the flagellin short variable region to
characterize isolates from a large outbreak of waterborne Campylobacter sp. strains in Walkerton,
Ontario, Canada. J Clin Microbiol. 2005; 43:2080. [PubMed: 15872226]
Dempster AP, Laird NM, Rubin DB. Maximum likelihood from incomplete data via the EM
algorithm. J Roy Statist Soc Ser B. 1977; 39(1):1–38.
Didelot X, Lawson D, Falush D. Simmlst: simulation of multilocus sequence typing data under a
neutral model. Bioinformatics. 2009; 25:1442. [PubMed: 19286834]
Dingle KE, Colles FM, Wareing DRA, Ure R, Fox AJ, Bolton FE, Bootsma HJ, Willems RJL, Urwin
R, Maiden MCJ. Multilocus sequence typing system for Campylobacter jejuni. J Clin Microbiol.
2001; 39:14. [PubMed: 11136741]
Fearnhead P, Smith NGC, Barrigas M, Fox A, French N. Analysis of recombination in Campylobacter
jejuni from MLST population data. J Mol Evol. 2005; 61:333–340. [PubMed: 16044246]

J Mol Evol. Author manuscript; available in PMC 2014 April 14.

 Yu et al. Page 8

Feil EJ, Maiden MC, Achtman M, Spratt BG. The relative contributions of recombination and
mutation to the divergence of clones of Neisseria meningitidis. Mol Biol Evol. 1999; 16:1496.
[PubMed: 10555280]
Feil EJ, Smith JM, Enright MC, Spratt BG. Estimating recombinational parameters in Streptococcus
pneumoniae from multilocus sequence typing data. Genetics. 2000; 154:1439. [PubMed: 10747043]
Feil EJ, Holmes EC, Bessen DE, Chan MS, Day NPJ, Enright MC, Goldstein R, Hood DW, Kalia A,
Moore CE, et al. Recombination within natural populations of pathogenic bacteria: short-term
Europe PMC Funders Author Manuscripts

empirical estimates and long-term phylogenetic consequences. Proc Natl Acad Sci. 2001; 98:182.
[PubMed: 11136255]
Feil EJ, Li BC, Aanensen DM, Hanage WP, Spratt BG. eBURST: inferring patterns of evolutionary
descent among clusters of related bacterial genotypes from multilocus sequence typing data. J
Bacteriol. 2004; 186:1518. [PubMed: 14973027]
Guttman DS, Dykhuizen DE. Clonal divergence in Escherichia coli as a result of recombination, not
mutation. Science. 1994; 266:1380–1383. [PubMed: 7973728]
Hein, J.; Schierup, MH.; Wiuf, C. Gene genealogies, variation and evolution: a primer in coalescent
theory. Oxford University Press; Oxford: 2005.
Humphrey T, O’Brien S, Madsen M. Campylobacters as zoonotic pathogens: a food production
perspective. Int J Food Microbiol. 2007; 117:237–257. [PubMed: 17368847]
Jolley KA, Maiden MCJ. BIGSdb: scalable analysis of bacterial genome variationat thepopulation
level. BMC Bioinformatics. 2010; 11:595. [PubMed: 21143983]
Konkel ME, Gray SA, Kim BJ, Garvis SG, Yoon J. Identification of the enteropathogens
Campylobacter jejuni and Campylobacter coli based on the cadF virulence gene and its product. J
Clin Microbiol. 1999; 37:510. [PubMed: 9986804]
Maiden MCJ, Bygraves JA, Feil EJ, Morelli G, Russell JE, Urwin R, Zhang Q, Zhou J, Zurth K,
Caugant DA. Multilocus sequence typing: a portable approach to the identification of clones
within populations of pathogenic microorganisms. Proc Nat Acad Sci USA. 1998; 95:3140.
[PubMed: 9501229]
Meinersmann RJ, Hiett KL. Concerted evolution of duplicate fla genes in Campylobacter.
Microbiology. 2000; 146:2283. [PubMed: 10974116]
Richman AD, Herrera LG, Nash D, Schierup MH. Relative roles of mutation and recombination in
generating allelic polymorphism at an MHC class II locus in Peromyscus maniculatus. Genet Res.
2003; 82:89–99. [PubMed: 14768893]
Europe PMC Funders Author Manuscripts

Sarkar SF, Guttman DS. Evolution of the core genome of Pseudomonas syringae, a highly clonal,
endemic plant pathogen. Appl Environ Microbiol. 2004; 70:1999. [PubMed: 15066790]
Schouls LM, Reulen S, Duim B, Wagenaar JA, Willems RJL, Dingle KE, Colles FM, Van Embden
JDA. Comparative genotyping of Campylobacter jejuni by amplified fragment length
polymorphism, multilocus sequence typing, and short repeat sequencing: strain diversity, host
range, and recombination. J Clin Microbiol. 2003; 41:15. [PubMed: 12517820]
Sheppard SK, McCarthy ND, Falush D, Maiden MCJ. Convergence of Campylobacter species:
implications for bacterial evolution. Science. 2008; 320:237–239. [PubMed: 18403712]
Sheppard SK, Dallas JF, Wilson DJ, Strachan NJC, McCarthy ND, et al. Evolution of an agriculture-
associated disease causing Campylobacter coli clade: evidence from National Surveillance Data in
Scotland. PLoS ONE. 2010; 5(12)
Sheppard SK, McCarthy ND, Jolley KA, Maiden MCJ. Introgression in the genus Campylobacter:
generation and spread of mosaic alleles. Microbiology. 2011; 157:1066–1074. [PubMed:
21212120]
Suerbaum S, Lohrengel M, Sonnevend A, Ruberg F, Kist M. Allelic diversity and recombination in
Campylobacter jejuni. J Bacteriol. 2001; 183:2553. [PubMed: 11274115]
Tauxe, RV.; Nachamkin, I.; Blaser, MJ.; Tompkins, LS. Epidemiology of Campylobacter jejuni
infections in the United States and other industrialized nations. MBio; Washington, DC: 1992.
Vos M, Didelot X. A comparison of homologous recombination rates in bacteria and archaea. ISME J.
2008; 3:199–208. [PubMed: 18830278]

J Mol Evol. Author manuscript; available in PMC 2014 April 14.

 Yu et al. Page 9

Wilson DJ, Gabriel E, Leatherbarrow AJH, Cheesbrough J, Gee S, Bolton E, Fox A, Hart CA, Diggle
PJ, Fearnhead P. Rapid evolution and the importance of recombination to the gastroenteric
pathogen Campylobacter jejuni. Mol Biol Evol. 2009; 26:385. [PubMed: 19008526]
Zhang Q, Meitzler JC, Huang S, Morishita T. Sequence polymorphism, predicted secondary structures,
and surface-exposed conformational epitopes of Campylobacter major outer membrane protein.
Infect Immun. 2000; 68:5679. [PubMed: 10992471]
Europe PMC Funders Author Manuscripts
Europe PMC Funders Author Manuscripts

J Mol Evol. Author manuscript; available in PMC 2014 April 14.

 Yu et al. Page 10
Europe PMC Funders Author Manuscripts
Europe PMC Funders Author Manuscripts

Fig. 1.
SLVs of PubMLST data. The x axes represent the number of nucleotide differences between
STs that make up an SLV; y axes represent the number of recorded events. A represents the
nucleotide differences for SLVs in the PubMLST database for C. jejuni; others are the
nucleotide differences for SLVs by loci

J Mol Evol. Author manuscript; available in PMC 2014 April 14.

 Yu et al. Page 11
Europe PMC Funders Author Manuscripts
Europe PMC Funders Author Manuscripts

Fig. 2.
SLVs of PubMLST data. The x axes represent the number of nucleotide differences between
STs that make up an SLV; y axes represent the number of recorded events. A represents the
nucleotide differences for SLVs in the PubMLST database for C. coli clade 1; others are the
nucleotide differences for SLVs by loci

J Mol Evol. Author manuscript; available in PMC 2014 April 14.

 Yu et al. Page 12

Table 1
Allele lengths for each locus; estimates for C. jejuni for each housekeeping locus of the probability of an SLV
being caused by mutation only (p); the expected number of mutations for an SLV; the relative rate of
recombination to mutation; 95 % CI for the estimated relative rate of recombination to mutation; and the % of
Europe PMC Funders Author Manuscripts

nucleotide differences of an SLV that were introduced by recombination

Locus Allele lengths (bp) p Expected number mut. Relative rate of rec 95 % CI % Diff due to rec
aspA 477 0.09 0.11 8.91 (5.98, 16.06) 99
glnA 477 0.09 0.11 8.86 (5.96, 15.91) 98
gltA 402 0.10 0.12 7.84 (5.43, 13.12) 97
glyA 507 0.21 0.32 2.97 (2.40, 3.77) 97
pgm 498 0.10 0.15 7.14 (5.06, 11.41) 98
tkt 459 0.11 0.15 6.81 (4.87, 10.66) 99
uncA 489 0.12 0.16 6.21 (4.53, 9.37) 99
Average 472.71 0.12 0.16 6.96 (6.08, 8.09) 98
Europe PMC Funders Author Manuscripts

J Mol Evol. Author manuscript; available in PMC 2014 April 14.

 Yu et al. Page 13

Table 2
Allele lengths for each locus; estimates for C. coli clade 1 for each housekeeping locus of the probability of an
SLV being caused by mutation only (p); the expected number of mutations for an SLV; the relative rate of
recombination to mutation; 95 % CI for the estimated relative rate of recombination to mutation; and the % of
Europe PMC Funders Author Manuscripts

nucleotide differences of an SLV that were introduced by recombination

Locus Allele lengths (bp) p Expected number mut. Relative rate of rec 95 % CI % Diff due to rec
aspA 477 0.18 0.65 2.23 (1.13, 5.69) 98
glnA 477 0.80 0.84 0.23 (0.03, 0.53) 60
gltA 402 0.56 0.61 0.72 (0.35, 1.37) 80
glyA 507 0.63 0.72 0.54 (0.24, 1.04) 86
pgm 498 0.47 0.54 1.04 (0.54, 2.03) 90
tkt 459 0.51 0.62 0.85 (0.43, 1.63) 91
uncA 489 0.39 0.44 1.44 (0.75, 3.01) 89
Average 472.71 0.51 0.63 1.01 (0.78, 1.30) 85
Europe PMC Funders Author Manuscripts

J Mol Evol. Author manuscript; available in PMC 2014 April 14.