JOURNAL OF BACTERIOLOGY, May 1985, p. 633-640 Vol. 162, No.

2
0021-9193/85/050633-08$02.00/0
Copyright © 1985, American Society for Microbiology

Nucleotide Sequence of ermA, a

Macrolide-Lincosamide-Streptogramin B Determinant in
Staphylococcus aureus
ELLEN MURPHY
Department of Plasmid Biology, The Public Health Research Institute of the City of New York, Inc., New York,
New York 10016
Received 26 November 1984/Accepted 15 February 1985

The complete nucleotide sequence of ermA, the prototype macrolide-lincosamide-streptogramin B resistance

gene from Staphylococcus aureus, has been determined. The sequence predicts a 243-amino-acid protein that
is homologous to those specified by ermC, ermAM, and ermD, resistance determinants from Staphylococcus
aureus, Streptococcus sanguis, and Bacillus licheniformis, respectively. The ermA transcript, identified by
Northern analysis and S1 mapping, contains a 5' leader sequence of 211 bases which has the potential to encode
two short peptides of 15 and 19 amino acids; the second, longer peptide has 13 amino acids in common with
the putative regulatory leader peptide of ermC. The coding sequence for this peptide is deleted in several
mutants in which macrolide-lincosamide-streptogramin B resistance is constitutively expressed. Potential
secondary structures available to the leader sequence of the wild-type (inducible) transcript and to constitutive
deletion, insertion, and point mutations provide additional support for the translational attenuation model for
induction of macrolide-lincosamide-streptogramin B resistance.

Inducible resistance to macrolide-lincosamide-strepto-
gramin B (MLS) antibiotics was identified in gram-positive
organisms by Weaver and Pattee (47) shortly after the
introduction of erythromycin into clinical practice. Resist-
ance is associated with N6, N6-dimethylation of specific
adenine residues in 23S rRNA (21); methylated ribosomes
do not bind erythromycin. The enzymatic activity responsi-
ble is an S-adenosylmethionine-dependent methylase that is
active on 23S rRNA (39).
Resistance to the MLS spectrum of antibiotics is inducible
by erythromycin (Em) and the closely related macrolide
oleandomycin (Om). Induction occurs post-transcriptionally
(40) by a mechanism resembling transcriptional attenuation
(11, 16).
support to the translational attentuation model (11) for the
regulation of expression of erythromycin resistance.
MATERIALS AND METHODS
Strains and plasmids. Bacterial strains and plasmids are
shown in Table 1. The transposon Tn554 (28, 33), which
specifies resistance to erythromycin and spectinomycin, was
originally isolated in Staphylococcus aureus E1206 (20). Con-
stitutive mutants were selected by plating RN4689 on me-
dium containing 10 ,ug of tylosin, carbomycin, clindamycin,
or lincomycin per ml. As these antibiotics do not induce
resistance, colonies arising on plates containing them are
expected to contain mutations to constitutivity. The mutant
strains were purified by transductional outcross, and plasmid

Downloaded from https://journals.asm.org/journal/jb on 25 November 2021 by 37.238.174.7.

The sequences of the Emr determinants analyzed to date
indicate a strong evolutionary relationship among ermA,
ermC (Staphylococcus aureus) (11, 16), and ermAM (Strep-
tococcus spp.) (15); ermD (Bacillus licheniformis) (12) is less
closely related. Hybridization studies indicate that Strep-
tomyces erythreus, the producer of erythromycin, contains
nucleotide sequences homologous to ermD (18). This sup-
ports the hypothesis of Benveniste and Davies (1) that
antibiotic resistance genes evolved originally in the antibiot-
ic-producing strains themselves. Although the methylase
genes themselves are strikingly homologous and the overall
regulatory mechanism is conserved, there is no nucleotide
sequence homology in the regulatory leader regions of the
ermC, ermAM, and ermD transcripts.
This report describes the sequence of ermA, the determi-
nant for which the mechanism of erythromycin resistance
was first described (20). ErmA is found on the transposon
TnS54 (28, 33), which is derived from Staphylococcus au-
reus E1206. The ermA mRNA contains a 211-base leader
that potentially encodes two short peptides, one of which
has 13 of 19 amino acids in common with the ermC leader
peptide. The sequences of the wild type, inducible leader,
and of a number of constitutive mutants lend additional
633
DNA was prepared for sequence analysis.
Nucleotide sequencing. Sequences were determined by the
dideoxy chain termination method of Sanger et al. (37) by
using restriction fragments cloned into the M13 derivative
mpll (27). [ox-35S]-dATP (Amersham Corp.) and 8% gradient
gels were used (3). The constitutive mutants were sequenced
in situ (46); after digestion with BglII the DNA was heated
to 90°C with twofold excess of primer for 3 min and
quenched on ice. A synthetic oligonucleotide primer, 5'-
GTGCACCAGTATCGCAG-3' (provided by the Laboratory
of Applied Microbiology at the Public Health Research
Institute of the City of New York), that anneals to the ermA
leader was used.
Analysis of transcripts. In vivo RNA was prepared from 50
ml of late-logarithmic-phase cultures. Protoplasts (5, 30)
were suspended in 2.5 ml of 5 M guanidium isothiocyanate
(6), and the suspension was layered on 6 M CsCl and
centrifuged at 100,000 x g for 18 h (9). The RNA pellet was
precipitated twice with 8 M guanidine HCl and three times
with ethanol and was stored in 70% ethanol at -70°C. Just
before the gels were loaded, the suspension was centrifuged,
rinsed, dried, resuspended in gel loading buffer, and heated
to 65°C for 5 min. Ten micrograms of each RNA was
electrophoresed in a 1.5% agarose-formaldehyde gel (22) and
634 MURPHY J. BACTERIOL.

extracted twice with phenol-chloroform, ethanol precipi-

B.O
_-
tated, and loaded on an 8% gradient sequencing gel.
Predicted secondary structure of the ermA leader. Free
C .

iI

r" 7 0
energy calculations for folded structures were based on the
z
N N K~ 'e
0))
W t-
0n X
0 40Dp;:X - OO
0,
rules of Tinoco et al. (44), by using the tabulated values of
I
oY c - Salser (36) as modified by Cech et al (4). The algorithm of
I aI >rr
Qc O c rr
Eo
Zucker and Stiegler (50) was useful in generating these
_ U.
I x al l 11 &-< /
cc a to 4 x
structures.
leader IormA I
* -> RESULTS
Nucleotide sequence of ermA. The sequencing strategy and
restriction map for erinA are depicted in Fig. 1. ErmA
4 -
roughly spans the region of 4.5 to 5.5 kilobases on the map of
4-< Tn554 (6,691 base pairs [bp]), the complete sequence of
< < <~~~~3 which will be published elsewhere (manuscript in prepara-
tion) and is transcribed from left to right as written, opposite
to the other known genes of Tn554. For this report the
transcriptional start site of ermA, bp 5493 of TnS54, has been
assigned bp 1. All but 23 bp of the ermA coding sequence,
1 597 bp and all of the 5' and 3' flanking sequences, were obtained
i 417bp from both strands of the constitutive mutant ermIl. A total
882 bp Probes of 64% of the coding region was verified on at least one
I- strand of the wild-type (inducible) strain, and the inducible
leader region was sequenced in its entirety in both direc-
FIG. 1. Sequencing strategy and restriction map of erinA. The tions.
map of ermA is presented in the 5' to 3' direction, opposite of The nucleotide sequence of ermA is given in Fig. 2. The
previous restriction maps (28). The first nucleotide of the erinA sequence contains an open reading frame of 243 amino acids,
transcript has been assigned bp 1. The map represents bp 5642 to starting with an AUG codon at position 212 which poten-
4420 of Tn554, from the Fnu4H site 5' to the ermnA transcript to the tially encodes a protein with a calculated molecular mass of
termination codon for spc, the adjacent gene, which is convergently
transcribed. The arrow marked with an asterisk represents the 28,380 daltons. The initiation codon is preceded by a strongly
sequence obtained with a synthetic oligonucleotide primer, 5'- complementary gram-positive Shine-Dalgarno sequence (26,
GTGCACCAGTATCGCAG-3'. provided by The Division of Ap- 38), AAGGAG, that has a free energy of pairing with the 3'
plied Microbiology of this Institute; the other sequences were end of Bacillis subtilis 16S rRNA of -13.4 kcal (-56.1 kJ).
obtained with the universal M13 primer. The open reading frame can be assigned to ermA based on
the following arguments. (i) ermA was previously mapped to
transferred to nitrocellulose in 20x SSC (1x SSC is 0.15 M the right half of Tn554, adjacent and to the right of spc (28).
NaCl plus 0.015 M sodium citrate) (42) without further (ii) In vitro manipulation at the SphI. ClaI, or AccI sites at
processing of the gel. Filters were baked, prehybridized, and bp 862, 783, or 716, respectively, results in a nonrevertable
hybridized in 40% formamide-5 x SSC-5 x Denhardt solu- Ems phenotype (M. Bastos and E. Murphy, unpublished
tion-50 mM sodium phosphate (pH 6.4)-10 mM EDTA-0. 1% data). (iii) The deduced protein sequence is homologous to
sodium dodecyl sufate-100 p.g of salmon sperm DNA per ml. those of ermC, ermAM, and ermD.
Identification of the ermA transcript. Total RNA prepared

Downloaded from https://journals.asm.org/journal/jb on 25 November 2021 by 37.238.174.7.

Dextran sulfate (10%) was included in the hybridization,
which was carried out at 51°C for 24 h; the filters were in vivo was subjected to Northern blot hybridization analy-
washed twice at room temperature in 2x SSC-0.1% sodium sis by using as a probe an 882-bp AccI-Fnu4H fragment
dodecyl sufate and twice at 55°C in 0.lx SSC-0.1% sodium containing 86% of the ermnA-coding region and the entire 5'
dodecyl sulfate. flanking sequences (Fig. 3A). A single transcript of ca. 1,070
SI mapping of transcripts. The 5' end of the ermA mRNA bases with homology to this probe is present in the wild-type
was determined by the method of Berk and Sharp (2). Total strain. Si mapping of the 5' ends of the transcript is shown
RNA (100 ,ug) was denatured at 85°C for 15 min with in Fig. 3B. Use of a 597-base Hinfl fragment from the
5'-end-labeled probes (25) and hybridized for 3 h at 42°C in wild-type strain as probe yielded a protected fragment of ca.
80% formamide-0.4 M NaCl-40 mM PIPES [piperazine-N, 280 bases (not shown). This indicated the approximate start
N'-bis(2-ethanesulfonic acid)] (pH6.5}-1 mM EDTA. The of the ermA transcript and established that the RsaI site at
samples were diluted 10-fold with S1 buffer containing 500 U bp 99, although 5' to the coding sequence, lies well within
of S1 per ml and incubated a further 30 min at 37°C, the transcript. A 417-bp RsaI-Hinfl subfragment was then

TABLE 1. Bacterial strains and plasmids

Strain Plasmid Description Derivation or reference
RN450 NCTC8325 cured of all known prophages (32)
RN451 RN450 relysogenized with 411 (32)
RN1389 Tn554 Emr Sp" (chromosomal) E1206-* RN450 (20, 33)
RN2864 Tn554 ernil *spc-l (chromosomal) (33)
RN4537 pEM9634 2.8-kilobase chromosomal insert in pT181 containing Tn554 insertion site (29)
RN4689 pEM9698 Tn554 Em' Sp" transposed into pEM9634 RN1389- RN4537
RN4491 pEM9631 Tn554 ermll spc-I transposed into pEM9634 RN2864- RN4537
VOL. 162, 1985 NUCLEOTIDE SEQUENCE OF ermA 635

CTTTACTTTTTAGTAAAGACAGTGGCTTCTCTTATCAAGTTTCAAAACATACTTATTTTGAAGAAAACGTCCATCTGAAGTCGTCAAGGT

+1 10 2S0 17-mer 30
GCAAAATTACATATAAAGGTTTATTCTAAAATGAAAAGATGATACAATCATATTCAGTTACATAAGGAGGTTTCAATTATGTGCACCAGT
TO_-35
-40 -10 t SDI MetCysThrSer
r,- Z1d- 6 otyc- 100 oermnIl, tyc- 101 _Zin- 70
-4 40 50 r60 170 90 1Q0 RsaI 110 120
ATCGCAGTAGTAGAAATTACTTTATCr CATTCATAATGAAAAAAATGGAAAGGAGATAAAAGT TGGGTACTTTTTCTATATTTGTTATT
lleAlaValValGIulleThrLeuSerHisSer * * SD2 MetGlyThrPheSeriloPh.Vaille
Zin-71--.~G cZd-6 15 tyc-101 ermIl, Zin-70, tyc-100
f 140 150 160 170 180 20 2 10
AATAAAGTTCGTTATCAACCAAATCAAAATTAATTGGTTATAATGAACGCTTAATGTCAGTTCATTATAACCAGTAAGGAGAAGGTTATA
AsnLysValArgTyrGlnProAsnGlnAsn * SD3
crb-12
220 230 240 250 260 270 SspI 28OHinfI 290 300
ATGAACCAGAAAAACCCTAAAGACACGCAAAATTTTATTACTTCTAAAAAGCATGTAAAAGAAATATTGAThCACACGAATATCAGTAAA
MetAsnGlnLysAsnProLysAspThrGlnAsnPhelleThrSerLysLysHlsValLysGtulIeLeuAsnHlsThrAsnileSerLys
310 320 MboI 330 340 350AZuI 360Hinf/T qj/MboI 380 390
CAAGACAACGTAATAGAAATCGGATCAGGAAAAGGACATTTTACCAAAGAGCTAGTCAAAATGAGTCGATGTTACTGCTATAGAAATT
GInAspAsnVaIlieGlulleGlySerGlyLysGlyHlsPheThrLysGIuLeuValLysMetSerArgSerValThrAlaIleGlulle
400 410 420 430 DdeI 440 45OHinfI 460 470 480
GATGGAGGCTTATGTCAAGTGACTAAAGAAGCGGTAAACCCCTCTGAGAATATAAAAGTGATTCAAACGGATATTCTAAAATTTTCCTTC
AspGlyGlyLeuCysGInValThrLysGIuAlaValAsnProSerGluAsnileLysValIleGlnThrAsptleLeuLysPheSerPhe
490 500 510 SspI 520 53ORsaI 540 550 560 570
CCAAAACATATAAACTATAAGATATATGGTAATATTCCTTATAACATCAGTKCGGATATTGTCAAAAGAATTACCTTTGAAAGTCAGGCT
ProLysHlsIleAsnTyrLystleTyrGlyAsnileProTyrAsnileSerThrAsplleValLysArgileThrPheGluSerGlnAla
AZuI 580 590 600 610 620 630AZuI 640 650 660
AAATATAGCTATCTTATCGTTGAGAAGGGATTTGCGAAAAGATTGCAAAATCTGCAACGAGCTTTGGGTTTACTATTAATGGTGGAGATG
LysTyrSerTyrLeulleValGluLysGlyPheAlaLysArgLeuGlnAsnLeuGlnArgAlaLeuGlyLeuLeuLeuMetValGluMet
670 680 ,sa-T 690 700 DdeI 7 1 0 AccI/Hinfl 730 740 750
GATATAAAAATGCTCAAAAAAGTACCACCACTATATTTTCATCCTAAGCCAAGTGTAGACTCTGTATTGATTGTTCTTGAACGACATCAA
AspiIeLysMetLeuLysLysValProProLeuTyrPheHlsProLysProSerValAspSerValLeulleValLeuGluArgHlsGIn
760 770 780 CiaI 790 800 810 820 830 840
CCATTGATTTCAAAGAAGGACTACAAAAAGTATCGATCTTTTGTTTATAAGTGGGTAAACCGTGAATATCGTGTTCTTTTCACTAAAAAC
ProLeulleSerLysLysAspTyrLysLysTyrArgSerPheValTyrLysTrpValAsnArgGluTyrArgValLeuPheThrLysAsn
850 860 SphI 870 Sspi880 890 TaqI 900XmnI 910 920 930
CAATTCCGACAGGCTTTGAAGCATGCAAATGTCACTAATATTAATAAACTATCGAAGGAACAATTTCTTTCTATTTTCAATAGTTACAAA
GInPheArgGInAlaLeuLysHlsAlaAsnValThrAsnileAsnLysLeuSerLysGluGlnPheLeuSerilePheAsnSerTyrLys

Downloaded from https://journals.asm.org/journal/jb on 25 November 2021 by 37.238.174.7.

940 950 960 970 980 990 1000 SspI 1010 1020
TTGTTTCACTAAATTAAAGTAATAAAGCGTTCTC1AATTTCACA4GAGGACGCTTTATTCTTCCCAAAAATTGTTCAATATTTATCAATA
LeuPheHls *
1030AhaIII1040 1050 1060
AATCAGTAGTTTTAAAAGTAAGCACCTGTTATTGCAATAAAATTAG
FIG. 2. Nucleotide sequence of ermA and the 5' regulatory region. The sequence shown corresponds to that shown in Fig. 1. Bases
complementary to the 3' end of 16S rRNA are underlined (SD1, SD2, and SD3). The major transcriptional start is bp 1. Extents of deletion
mutants are indicated. Inverted repeat sequences at the 3' end are indicated with arrows.

hybridized to wild-type RNA, yielding a major protected
fragment of 100 bases. This placed bp 1 of the transcript at a
position 211 bp 5' to the initiation codon, as indicated on Fig.
2 (bp 5493 of Tn554). The constitutive mutant ermIl
(RN4491), which contains a 116-bp deletion in the leader
region (see below), produced a correspondingly shorter
transcript (Fig. 3A, lane 3) than the wild type. RNA from
this mutant strain protected a 167-base fragment when a
481-bp Hinfl fragment (corresponding to the 597-bp wild-
type Hinfl fragment) was used as probe, placing the tran-
scriptional start site for this mutant at the same position as
for the wild-type (inducible) mRNA.
Analysis of constitutive mutants. Additional mutants con-
stitutive for MLS resistance were isolated by selection for
growth on noninducing antibiotics. All the mutants obtained
were fully constitutive, as judged by their cross-resistance to
all four of these antibiotics, as well as to erythromycin; none
exhibited an altered induction specificity similar to those
reported by Weisblum et al. (43, 48). The regulatory regions
of several mutants obtained by growth on each selective
drug were sequenced by using as a primer the oligonucleo-
tide shown in Fig. 2, which anneals to bp 22 to 38 of the
ermA mRNA. Two mutants failed to produce a sequencing
ladder and were found to contain gross deletions (>450 bp;
data not shown) involving the 597-bp Hinfl fragment that
spans the 5' flanking region (Fig. 1). Another mutant con-
tained no detectable alteration; since the method employed
would not identify any mutation mapping upstream from bp
 :.:|.g*
.w:. _
636 MURPHY J. BACTERIOL.

40, appropriate fragments were cloned into M13rnpll and A

sequenced. No mutation was detected between bp -70 and 123
+280. This mutation may affect the copy number of the -2772
vector plasmid; it was not studied further. All of the remain-
ing constitutive mutants analyzed map) within the 211-bp -1676
leader sequence (Fig. 2). Most are deletions of 90 to 130 bp -124 2
that remove SD2 and the entire coding sequence for the
19-amino-acid peptide 2. crb-12 contains a tandem duplica- ...665
tion of 25 bp from SD3 through the first 12 bp of the
ermA-coding sequence. Only one point mutation, lin-71, a
C G--G
G C transversion at nucleotide 131, was isolated. B1 23 4 56 7 8
Possible effects of these mutations on the potential second- _
_
ary structure of the ermA leader are discussed below. _S S _
._
_
.. .l
_:

DISCUSSION
Comparison of ermA and other methylases. The amino acid .s
sequences of ermA (Staphylococcus aureus), ermC (Staph- -
X
ylococcus aureus), ermAM (Streptococcus sanguis), and _.
ermD (B. licheniformis) are compared in Fig. 4 and Table 2. .e
.X:

The ermA protein is somew,hat more closely related to the w

.;L
other Staphylococcus aureus determinant, ermC, than to w
ermAM or ermD. Among the three closely related genes the *

nucleic acid homology is quite constant.

Codon usage. The codon usage for ermA (Fig. 5) reflects .... ..
the low G + C content of the ermA-coding sequence
(32.5%). With a few exceptions (UUG and UUA for Leu; ..=t
ACC, ACG versus ACA, and ACU for Thr; GAC and GAU :tf
for Asp), there is a clear preference for A or U bases in the :al'.:

third position. Thus, for the eight duet codons postulated by :w:

Grosjean and Fiers (10) to be preferred by highly expressed .jX *:

genes, ermA utilizes the A + U-rich codons in every case, |

:.....

four of which fall in the highly expressed category and four *..:

of which do not. This usage is similar to that of ermC and of

other staphylococcal proteins whose nucleotide sequences I * ?*:: .
are known, such as staphylokinase (35), protein A (45),
staphylococcal nuclease (41), and the toxic shock syndrome B}4i:

antigen (B. Kreiswirth and R. Novick, personal communi-

cation).
Transcription signals. The most probable -10 sequence for
the ermA promoter is CATATT (Fig. 2), an acceptable fit to
the canonical sequence TATAAT (34). A second possibility,

Downloaded from https://journals.asm.org/journal/jb on 25 November 2021 by 37.238.174.7.

TACAAT, is adjacent to the first. Based on the spacing to
the transcriptional start site, the former sequence is more
likely to be the functional one, although the latter is a better FIG. 3. Analysis of transcripts. (A) Northern analysis. RNA (10
fit. It is possible that both are used to some extent; this might ±Lg) from RM4565 (lane 1, negative control), RN4689 (lane 2,
account for some of the minor protected species in the Si wild-type ermA), and RN4491 (lane 3, ermII), were run on a 1.5%
experiments. A possible -35 sequence is indicated in Fig. 2; agarose-formaldehyde gel as described in the text and probed with
it is preceded by an A + G-rich sequence at about -40, an 882-bp Fnu4H-AccI fragment (see Fig. 1). Arrows mark the
positions of pT181 DNA restriction fragments (19). (B) Si analysis.
noted by Murray and Rabinowitz (31) in their study of RNA (100Lug) was hybridized with 5' and 3'-end-labeled ermA
probes as described in the text and incubated for 30 min with 500 U
of S1 per ml. Lanes 1 through 4, 417-bp Hinfl-RsaI fragment from
TABLE 2. Comparison of methylase determinants wild type (RN4689). Lane 1,'tRNA; lane 2, 100 ,ug of RN4689 RNA;
lane 3, no S1; lane 4, chemically derivatized sequencing ladder (25)
thnIdentical
Methylase residues onservative
replacements"
Nucleic acid
homology
(adenine plus guanine lane). Lanes 5 through 8, 481-bp Hinfl probe
determinants
from ermIl (RN4491). Lane 5, 100 jig of RN4689 RNA; lane 6,
No. Fraction No. Fraction (%) tRNA; lahe 7, no Si; lane 8, sequencing ladder. The 5'-end-labeled
probes were labeled with [-y-32PIATP and polynucleotide kinase and
ermA-ermC 141 0.58 36 0.15 61 strand separated. Due to incomplete strand separation of the 417-bp
ermA-ermAM 123 0.51 41 0.17 58 probe, two higher-molecular-weight bands appear in lane 2, the
ermAM-ermC 122 0.50 48 0.19 59 result of protection by an RNA which is transcribed from the
ermA-ermD 71 0.29 50 0.21 opposite strand of TnSS4 (leftward on Fig. 1 and 2).
ermC-ermD 76 0.31 45 0.18
ermAM-ermD 68 0.28 57 0.23 gram-positive promoters. Immediately 3' to the ermA stop
"Conservative replacements are according to Dayhoff et al. (7). They are codon is a 17-bp imperfect inverted repeat that may function
his, lys, arg; glu, gin, asp, asn; val, leu, ile, met; trp, phe, tyr; ala, ser, thr, gly, as a rho-independent transcription terminator for both ermA
pro; and cys. and spc, the adjacent, convergently transcribed gene encod-
VOL. 162, 1985 NUCLEOTIDE SEQUENCE OF ermA 637

1231131 42211 3232243431444422334122412131112313

ermD: MKKKNHKYRGKKLNRGESPNFSGQHLMHNKKLIEEIVDRANISIDDTVLELGAGKGALTT
ermAM: MK-KNIK---------YSQNF-----LTNEKVLNQIIKQLNLKETDTVYEIGTGKGHLTT
ermC: MNEKNIK---------HSQNF-----ITSKHNIDKIMTNIRLNEHDNIFEIGSGKGHFTL
ermA: MNQKNPK----- DTQNF-----ITSKKHVKEILNHTNISKQDNVIEIGSGKGHFTK

313243441222131332344343234444443323311222421242422322211132
ermD: VLSQKAGKVLAVENDSKFVDILTRKTAQHSNTKIIHQDIMKIH-LPKE-KFVVVSNIPYAI
ermAM: KLAKISKQVTSIELDSHLFNLSSEKLKLNIRVTLIHQDILQFQFPNKQRYKRVGSIPYHL
ermC: ELVKRCNFVTAIEIDHKLCKTTENKLVDHDNFQVLNKDILQFKFPKNQSYKIYGNIPYNI
ermA: ELVKMSRSVTAIEIDGGLCQVTKEAVNPSENIKVIQTDILKFSFPKHINYKIYGNIPYNI

2131323332324244322 11 13122113434 322231223433214324332423

ermD: TTPIMKMLLNNPASGFQKGIIVMEKGAAKRFTSKFIKNSYVLAWRMWFDIGIVREISKEH
ermAM: STQIIKKVVFESHASDIYL-IV-EEGFYKRTLDI--HRSLGLLLHTQVSIQQLLKLPAEC
ermC: STDIIRKIVFDSIANEIYL-IV-EYGFAKRLLNT--KRSLALLLMAEVDISILSMVPREY
ermA: STDIVKRITFESQAKYSYL-IV-EKGFAKRLQNL--QRALGLLLMVEMDIKMLKKVPPLY

12121212132232412344 32313333233124222221234 211221

ermD: FSPPPKVDSAMVRITRKKDAPLSHKHYIAFRG-LAEYALKEPNIPLCVALRGIFTPRQMK
ermAM: FHPKPKVNSVLIKLTRHTTD-VPDKYWKLYTYFVSKWVNREYRQ--------LFTKNQ--
ermC: FHPKPKVNSSLIRLSRKKSR-ISHKDKQKYNYFVMKWVNKEYKK--------IFTKNQ--
ermA: FHPKPSVDSVLIVLERHQPL-ISKKDYKKYRSFVYKWVNREYRV--------LFTKNQ--

23322111323222343422312322223212
ermD: HLRKSLK------INNEKTVGTLTENQWAVIFNTMTQYVMHHKWPRANKRKPGEI
ermAM: -FHQAMKHAKVNNLSTVTYEQVLSIFNSYLLFNGRK
ermC: -FNNSLKHAGIDDLNNISFEQFLSLFNSYKLFNK

Downloaded from https://journals.asm.org/journal/jb on 25 November 2021 by 37.238.174.7.

ermA: -FRQALKHANVTNINKLSKEQFLSIFNSYKLFH
FIG. 4. Comparison of ermA, ermC (11, 16), ermAM (15), and ermD (12) methylases. The alignment of ermD is taken from Gryczan et
al. (12). The numbers across the top indicate the number of different amino acids at each position.
ing spectinomycin resistance (E. Murphy, Mol. Gen. Genet.,
in press).
Regulation of ermA. A detailed model for the regulation of
ermC by translational attenuation has been proposed inde-
pendently by two laboratories (11, 16). In this model, a
141-base leader at the 5' end of the ermC mRNA is proposed
to form a secondary structure that sequesters the translation
initiation signals. Within the leader is a short open reading
frame that potentially encodes a 19-amino-acid peptide.
Induction occurs when ribosomes that have bound erythro-
mycin stall during translation of this peptide; the stalled
ribosome remains bound to the mRNA, disrupting the sec-
ondary structure and allowing it to refold into a structure in
which the ermC initiation signals are unpaired. Erythromy-
cin-free ribosomes may then bind and initiate translation of
methylase. Detailed analysis of mutants constitutively ex-
pressing erythromycin resistance (14, 17) provided support
for this model, as did the demonstration that induction
requires translation of the leader peptide (8) by ribosomes
that are capable of binding erythromycin (13).
The regulation of ermA appears to be similar in its overall
mechanism. In contrast to the single leader peptide of ermC,
ermD, and ermAM, the ermA leader potentially encodes two
small peplides of 15 and 19 amino acids. Both are preceded
by strongly complementary ribosomal binding sites (AG =
-19.2 and -15.6 kcal [-80.3 and 65.3 kJ], respectively).
Two possible secondary structures for the 211-base leader of
the ermA mRNA are shown in Fig. 6. Structures A and B are
equally probable on thermodynamic grounds; structure A is
likely to be kinetically favored, as segments 3 and 4 will pair
before segment 5 has been completely transcribed. Neither
structure A nor B can actively translate methylase because
the initiation cbdon for ermA is base paired. In the presence
of inducing concentrations of erythromycin, translation of
peptide 1 by a ribosome that has bound erythromycin would
result in ribosome stall in segment 1, freeing segment 2 and
allowing translation of peptide 2. In turn, stalling in segment
3 by ribosomes translating peptide 2 will either directly free
segment 6, which contains the AUG codon for ermA (struc-
ture B), or, for structure A, will free segment 4 to pair with
638 MURPHY J. BACTERIOL.

UUU ( Phe) 9 UCU (Ser) 5 UAU (Tyr) 9 UGU (Cys) 1 possibly including one that gave rise to ermC and ermA from
UUC (Phe) 4 UCC (Sor) 1 UAC (Tyr) 2 UGC (Cys) 0 an ancestral gene.
UUA (Lou) 3 UCA (Ser) 3 UAA ( ... ) 1 UGA (...) o
The failure of antibiotics other than erythromycin and
UUG (Leu) 7 LCG (Ser) 1 UAG ( ... ) O UGG (Trp) 1 oleandomycin to act as inducers of ermC and of ermD has
CUU (Leu) 4 CCU (Pro) 3 CAU (His) 6 CGU (Arg) 2 been ascribed to subtle differences between the mechanism
CUC (Leu) 1 CCC (Pro) 1 CAC (His) 2 CGC (Arg) 0 of action of various MLS antibiotics. The transpeptidization
CUA (Lou) 5 CCA (Pro) 5 CAA (Gln) 9 CGA (Arg) 5 reaction is most effectively blocked by erythromycin and
CUG (Lou) 1 CCG (Pro) 0 CAG (Gin) 3 CGG (Arg) 0 oleandomycin when the nascent peptide is rather large;
AUU ( I le) 11 ACU (Thr) 5 AAll (Asn) 10 AGU (Sor) 6 amino acids with bulky, hydrophilic side chains are most
AUC (Ile) 4 ACC (Thr) 2 AAC (Asn) 8 AGC ( Ser) 1 sensitive to inhibition (24). Other macrolides and lincosa-
AUA (Ite) 7 ACA (Thr) 0 AM (Lys) 22 AGA (Arg) 2 mides inhibit transpeptidization when the peptidyl-tRNA
AUG (Met) 5 ACG (Thr) 4 AAG (Lys) 10 AGG (Arg) 0
donor is very small, preventing the formation of even the
GUU (Val) 5 GCU (Ala) 4 GAU (Asp) 4 GGU (Gly) 2 first peptide bond (23). Thus the sequence and location of
GUC (Val) 3 GCC (Ala) 0 GAC (Asp) 4 GGC (Gly) 1 amino acids in the regulatory peptide have been postulated
GUA (Val) 7 GCA (Ala) 1 GAA (Glu) 8 GGA (Gly) 5 to contribute to the position at which ribosomes stall (11);
GUG (Val) 3 GCG (Ala) 2 GAG (Glu) 4 GGG (Gly) 0 premature stalling, or release of ribosomes, would not free
FIG. 5. Codon usage for ermA. the mRNA to refold into a structure in which the translation

5, displacing segment 6, allowing translation and also caus- ®8) AGG-65.2 Kcal
ing isomerization of structure A to structure B. In either 4000
tyclOO

case, translation of ermA would be transient and completely

dependent upon the continued presence of erythromycin as
I A
AA
AA cld6
inducer, as there is no stable structure in which segment 6 is UAA CA>150
unpaired. The ermC mRNA, on the other hand, can poten- A-U A U
tially form a metastable structure in which methylase can be UG
C-G
UA AA//U
C
translated; this structure, although considerably less stable
than the inactive structures, is the kinetically favored one cId6 AFS /

\C-U
A
tyclOl
ermIl

SD2
2
A-U
C 6U,
A-U

and may contribute to the basal level of ermC methylase t---V-

C-G
U-A
~~~~~~~A-U
3 cA-U 4 200
(11). A-U
U-A
U-A
A-U A
A
SD3
Analysis of the mutations that result in constitutive ex- 50 U-A U-A t C-u
AU ,U-A U-A tycG-O A

pression of ermA provides biological support for the model. A UAC

AU G
A
C-U
6U UN A'
CU-
A
A lt G erml
The point mutant lin-71 involves a base that pairs with the UAU-A -C--IU

ermA AUG codon in structure B; this substitution not only A-U G ~~I inMM U-C tYClOl A-U

rA C-6 A~ \6 5 A- 6
destabilizes the pairing of segments 3 and 6 but also allows A G ~~
~~~~A U-A

segment 3 to form a small hairpin loop between bp 129 and U-A

A-U UG
A-U- 100
A-U
A-U C-G
A-U m
143 that stabilizes the new structure. Segment 6 cannot form A-U
C-U1
U-A
S-C
U-A
U-A
crbI U-Al
U-A.
5 G
U U A-U UUCA6-CCASAAAAACCCU
the equivalent hairpin, leaving the initiation codon unpaired. A-UU C A C C A-U CUU
U U A U U U
Gln Lys Asn Pro
In structure A the effect of this mutation is to destabilize the
pairing of segments 3 and 4, permitting isomerization to the
mutant B-type structure. Iin-71 is located at the same posi-

Downloaded from https://journals.asm.org/journal/jb on 25 November 2021 by 37.238.174.7.

tion as a C G-uA. T transversion mutant of ermC (17).
Another mutation, crb-12, a 25-bp tandem duplication from ®0) AG=-61.8Kcal
SD3 through the first 12 bp of ermA, is similar to tyc-16, a
A AU
109-bp insertion mutation in ermC (14). The most favorable tyclOO
structure for the crb-12 leader leaves the duplicated, 3' tyC100
t~~~U
~~~~~~~~~c
c
u u
U

Shine-Dalgarno site unpaired; the alternative structures in AA6

G
AA
A
C-G
U A A-U
which the duplicated region is involved in base pairing with \A A
\\A
A-U
G-C
segment 3 or 5 is unfavored on both thermodynamic and \\U-A U-A
kinetic grounds. The remaining mutations contain deletions
A-U

UC-U6
4 A-U
U-A
A-U
S
that remove most of the potential for secondary structure 1 U A-U

(Fig. 6).
A
AyA) ttcermllu-U-A 2 tyciol
Peptide 2 has 13 of 19 amino acids in common with the cIdG \C C-U a
SD2 G-C
U-
in70, tyciOO, .rmI
ermC peptide; the nucleic acid homology in this region (SD2 C-U
U-A
MIS
U
A-t7
A-U
to SD3) is much greater than that between the structural U-A A9f Sf\ D3'
genes themselves (Fig. 7). Those bases postulated to be AU
50
U-A
U-A
CU 15 A-C--AA
A.
G
A-U

involved in secondary structure are more highly conserved 6uu

A UAC
A AUU 6
A
AA
C

C-U
A-U
(86%) than those in non-base-paired regions (69%). This
implies a very recent evolutionary divergence between ermA 6 - CA-6A-U
CnU6-C4-SU
A AC-6
U-
CU-A
-U 6
n 70
3
A-u

A
I

and ermC; they are the only two determinants whose regu-
latory regions are detectably related. The two sequences UAl ioGA-U-100
6
lU
;n71 G -C-6Ull
U-A

diverge upstream of SD2; this region is highly A + T-rich,

and it is probably not coincidental that a number of the
CA-U-6
A-U O-CU-A UUUUCUAUAUUUGUUAUUAAUAAA-CCASA_AAA
C A C C A-U
UA Crb12ACC U-A
v
CU
Gin Lys Asn Pro
constitutive deletion mutations terminate at this point; the FIG. 6. Possible secondary structures for the ermA mRNA.
region seems to be a hot spot for recombinational events, Locations of mutations are indicated. V, Insertion.
VOL. 162, 1985 NUCLEOTIDE SEQUENCE OF ermA 639

SD2 Met Gly Thr Phe Ser lie Phe Val lie Asn Lys Val Arg Tyr Gln Pro Asn
ermA: AAG AtAAj* T ATG GGt AcT|TTT tcT ATa TTT GTt ATt Aat AaA|GTT CgT TATJ CjjA AAt
ermC: AAGGAGgA AAAA T ATG GGc tT TTT ATt TTTJGTa ATc Agc A&A
GTT CaT TAT| CE AAc
Met Gly lie Phe Ser lie Phe Val Ile Ser Thr Val His Tyr Gin Pro Asn

Gln Asn Stop SD3 Met Asn GIn

ermA: cAA A7t TkA tTGGTTATAATGAA cTTAATgt Cg TTCATtATAACCA_ta igAGA[GGTTATA ATG AAC cAG
ermC: aAA AAaFTA-A gGGTT C G TTAATaagCAaa,TTCAT ATAA E aAGA*GTTATA ATG AAC gAG
Lys Lys Stop Met Asn Glu
FIG. 7. Comparison of ermA and ermC 5' regulatory regions. The sequence shown extends from SD2 of ermA through the third amino acid
of methylase. Bases in upper case are identical; bases in lower case are divergent. Nucleotides postulated to participate in the secondary
structure are boxed. Data for ermC are for the most stable structure (C) shown in reference 11; data for ermA are from structure B (Fig. 6).

initiation signals are unpaired. In the model proposed above ACKNOWLEDGMENTS

for induction of ermA translation of peptide 1 is a prerequi-
site for translation of peptide 2; furthermore, erythromycin-
induced stall must occur near the C-terminal end of peptide
1 for induction to be effective. The hydrophilic regions
(where stall is most likely to occur) of peptides 1 and 2, and
of the ermC peptide, are located at the C-terminal ends,
beginning with residue 13 in all three cases. The model also
provides an explanation for the lack of any mutations
affecting only peptide 1: deletion of peptide 1 would render
the regulatory region essentially identical to that of ermC,
and the resistance would remain inducible, whereas muta-
tions that prevented the translation of peptide 1 without
deletion of segment 1 (by mutation of SD1, for example)
would result in an Ems phenotype, like those mutations
obtained by Hahn et al (14) for ermC. Perhaps peptide 1
plays a role in the specificity of induction; mutants of ermA
with altered induction specificity have been reported (43,
48), although no mutants of this type were isolated in this
study. In this regard it is interesting that the streptococcal
ermAM determinant is inducible by the entire range of MLS
antibiotics (49); it specifies a 155-base leader sequence
encoding a 36-amino-acid peptide that can potentially form a
I thank L. Huwyler for expert technical assistance, D. Dubnau, C.
Lusty, and H. Nyunoya for valuable discussions, and R. Smith for
assistance with computer analysis of the sequence data.
This work was supported by Public Health Service grant GM27253
from the National Institutes of Health.
LITERATURE CITED
1. Benveniste, R., and J. Davies. 1973. Aminoglycoside antibiotic-
inactivating enzymes in actinomycetes similar to those present
in clinical isolates of antibiotic-resistant bacteria. Proc. Natl.
Acad. Sci. U.S.A. 70:2276-2280.
2. Berk, A. J., and P. A. Sharp. 1977. Sizing and mapping of early
adeno-virus mRNAs by gel electrophoresis of S1 endonuclease-
digested hybrids. Cell 12:721-732.
3. Biggin, M. D., T. J. Gibson, and G. F. Hong. 1983. Buffer
gradient gels and 35S label as an aid to rapid DNA sequence
determination. Proc. Natl. Acad. Sci. U.S.A. 80:3963-3965.
4. Cech, T. R., N. Tanner, I. Tinoco, B. R. Weir, M. Zucker, and
P. S. Perlman. 1983. Secondary structure of the Tetrahymena
ribosomal RNA intervening sequence: structural homology with
fungal mitochondrial intervening sequences. Proc. Natl. Acad.
Sci. U.S.A. 80:3903-3907.
5. Chang, S., and S. N. Cohen. 1979. High frequency transforma-
tion of Bacillus subtilis protoplasts by plasmid DNA. Mol. Gen.

Downloaded from https://journals.asm.org/journal/jb on 25 November 2021 by 37.238.174.7.

large number of secondary structures (15). The leader region
of ermD is similarly complex, although it is inducible only by
erythromycin and oleandomycin (12). The sequences of
these regulatory regions are not detectably related to one
another or to peptide 1 of ermA.
Induction kinetics and basal levels: predictions. In the ermC
system, SD1 is always available to translate the leader
peptide. For ermA, SD1 is unpaired, but the AUG codon is
involved in a short stem loop. The free energy contribution
of this hairpin is only -4.8 kcal (-20.1 kJ); this may be
sufficiently unstable to allow some translation to occur. In
addition, since stall of peptides 1 and 2 must be sequential,
induction requires that both of the ribosomes that initiate
translation at SD1 and SD2 have bound erythromycin, a
condition which will not be easily met at very low drug
concentrations. This cascade of regulatory peptides, and the
possibility that peptide 1 is not constantly translated, might
affect the induction kinetics in such a way that the rate of
induction at low concentrations of erythromycin is slower
than that of ermC. This might allow the cell to avoid
responding to extremely low levels of antibiotic. In addition,
the absence of a metastable active structure, as noted above,
may contribute to a lower basal level of ermA methylase,
which would in turn predict poor induction with high con-
centrations of drug. These predictions remain to be tested.
Genet. 168:111-115.
6. Chirgwin, J. M., A. E. Przybyla, R. J. MacDonald, and W. J.
Rutter. 1979. Isolation of biologically active ribonucleic acid
from sources enriched in ribonuclease. Biochemistry 18:
5294-5299.
7. Dayhoff, M. O., R. M. Schwartz, and B. C. Orcutt. 1978. A
model of evolutionary change in proteins, p. 345-353. In Atlas
of protein sequence and structure, vol. 5, suppl. 3. National
Biomedical Research Foundation, Silver Spring, Md.
8. Dubnau, D. 1985. Induction of ermC requires translation of the
leader peptide. EMBO J. 4:533-537.
9. Glisin, V., R. Crkvenjakov, and C. Byus. 1974. Ribonucleic acid
isolated by cesium chloride centrifugation. Biochemistry
13:2633-2637.
10. Grosjean, H., and W. Fiers. 1982. Preferential codon usage in
prokaryotic genes: the optimal codon-anticodon interaction
energy and the selective codon usage in efficiently expressed
genes. Gene 18:199-209.
11. Gryczan, T. J., G. Grandi, J. Hahn, R. Grandi, and D. Dubnau.
1980. Conformational alteration of mRNA structure and the
posttranscriptional regulation of erythromycin-induced drug re-
sistance. Nucleic Acids Res. 8:6081-6097.
12. Gryczan, T. J., M. Israeli-Reches, M. Del Bue, and D. Dubnau.
1984. DNA sequence and regulation of ermD, a macrolide-lin-
cosamide-streptogramin B resistance element from Bacillus
licheniformis. Mol. Gen. Genet. 194:349-356.
13. Gryczan, T. J., M. Israeli-Reches, and D. Dubnau. 1984. Induc-
640 MURPHY
tion of macrolide-lincosamide-streptogramin B resistance re-
quires ribosomes able to bind inducer. Mol. Gen. Genet.
194:357-361.
14. Hahn, J., G. Grandi, T. J. Gryczan, and D. Dubnau. 1982.
Translational attenuation of ermC: a deletion analysis. Mol.
Gen. Genet. 186:204-216.
15. Horinouchi, S., W.-H. Byeon, and B. Weisblum. 1983. A com-
plex attenuator regulates inducible resistance to macrolides,
lincosamides, and streptogramin type B antibiotics in Strepto-
coccus sanguis. J. Bacteriol. 154:1252-1262.
16. Horinouchi, S., and B. Weisblum. 1980. Posttranscriptional
modification of mRNA conformation: mechanism that regulates
erythromycin-induced resistance. Proc. Natl. Acad. Sci. U.S.A.
77:7079-7083.
17. Horinouchi, S., and B. Weisblum. 1981. The control region for
erythromycin resistance: free energy changes related to induc-
tion and mutation to constitutive expression. Mol. Gen. Genet.
182:341-348.
18. Israeli-Reches, M., Y. Weinrauch, and D. Dubnau. 1984. Evo-
lutionary relationships of the Bacillus licheniformis macrolide-
lincosamide-streptogramin B resistance elements. Mol. Gen.
Genet. 194:362-367.
19. Khan, S. A., and R. P. Novick. 1983. Complete nucleotide
sequence ofpT181, a tetracycline-resistance plasmid from Staph-
ylococcus aureus. Plasmid 10:251-259.
20. Lai, C., and B. Weisblum. 1971. Altered methylation of ribosomal
RNA in an erythromycin-resistant strain of Staphylococcus
aureus. Proc. Natl. Acad. Sci. U.S.A. 68:856-860.
21. Lai, C., B. Weisbium, S. R. Fahnestock, and M. Nomura. 1973.
Alteration of 23S ribosomal RNA and erythromycin-induced
resistance to lincomycin and spiramycin in Staphylococcus
aureus. J. Mol. Biol. 74:67-72.
22. Lehrach, H., D. Diamond, J. M. Wozney, and H. Boedtker.
1977. RNA molecular weight determinations by gel electropho-
resis under denaturing conditions, a critical reexamination.
Biochemistry 16:4743-4751.
23. Mao, J. C., and E. E. Robishaw. 1971. Effects of macrolides on
peptide-bond formation and translocation. Biochemistry
10:2054-2061.
24. Mao, J. C., and E. E. Robishaw. 1972. Erythromycin, a pep-
tidyltransferase effector. Biochemistry 11:48644872.
25. Maxam, A., and W. Gilbert. 1980. Sequencing end-labeled DNA
with base-specific chemical cleavages. Methods Enzymol.
65:499-560.
26. McLaughlin, J. R., C. L. Murray, and J. C. Rabinowitz. 1981.
Unique features in the ribosome binding site sequence of the
gram-positive Staphylococcus aureus ,B-lactamase gene. J. Biol.
Chem. 256:11283-11291.
27. Messing, J. 1983. New M13 vectors for cloning. Methods
Enzymol. 101:10-89.
28. Murphy, E. 1983. Inhibition of Tn554 transposition: deletion
analysis. Plasmid 10:260-269.
29. Murphy, E., and S. Lofdahl. 1984. Transposition of TnS54 does
not generate a target duplication. Nature (London) 307:292-294.
30. Murphy, E., S. Phillips, I. Edelman, and R. P. Novick. 1981.
Tn554: isolation and characterization of plasmid insertions.
Plasmid 5:292-305.
31. Murray, C. L., and J. C. Rabinowitz. 1982. Nucleotide se-
quences of transcription and translation initiation regions in
Bacillus phage 4)29 early genes. J. Biol. Chem. 257:1053-1062.
32. Novick, R. P. 1967. Properties of a cryptic high-frequency
J. BACTERIOL.
transducing phage in Staphylococcus aureus. Virology
33:155-166.
33. Phillips, S., and R. P. Novick. 1979. A site-specific repressor-
controlled transposon in Staphylococcus aureus. Nature (Lon-
don) 278:476-478.
34. Rosenberg, M., and D. Court. 1979. Regulatory sequences
involved in the promotion and termination of RNA transcrip-
tion. Annu. Rev. Genet. 13:319-353.
35. Sako, T., and N. Tsuchida. 1983. Sequence of Staphylococcus
aureus staphylokinase. Nucleic Acids Res. 11:7679-7693.
36. Salser, W. 1977. Globin mRNA sequences: analysis of base
pairing and evolutionary implications. Cold Spring Harbor
Symp. Quant. Biol. 42:985-1002.
37. Sanger, F., S. Nicklen, and A. R. Coulson. 1977. DNA sequenc-
ing with chain-terminating inhibitors. Proc. Natl. Acad. Sci.
U.S.A. 74:5463-5467.
38. Shine, J., and L. Dalgarno. 1974. The 3' terminal sequence of
Escherichia coli 16S ribosomal RNA: complementarity to non-
sense triplets and ribosome binding sites. Proc. Natl. Acad. Sci.
U.S.A. 71:1342-1346.
39. Shivakumar, A. G., and D. Dubnau. 1981. Characterization of a
plasmid-specified ribosome methylase associated with macro-
lide resistance. Nucleic Acids. Res. 9:2549-2562.
40. Shivakumar, A. G., J. Hahn, G. Grandi, Y. Kozlov, and D.
Dubnau. 1979. Posttranscriptional regulation of an erythromy-
cin resistance protein specified by plasmid pE194. Proc. Natl.
Acad. Sci. U.S.A. 77:3903-3907.
41. Shortle, D. 1983. A genetic system for analysis of staphylococ-
cal nuclease. Gene 22:181-189.
42.
Southern, E. M. 1975. Detection of specific sequences among
DNA fragments separated by gel electrophoresis. J. Mol. Biol.
98:503-517.
43. Tanaka, T., and B. Weisblum. 1974. Mutant of Staphylococcus
aureus with lincomycin- and carbomycin-inducible resistance to
erythromycin. Antimicrob. Agents Chemother. 5:538-540.
44. Tinoco, I., P. N. Borer, B. Dengler, M. D. Levine, 0. C.
Uhlenbeck, D. M. Crothers, and J. Gralla. 1973. Improved
estimation of secondary structure in ribonucleic acids. Nature
(London) New Biol. 246:40-41.
45. Uhlen, M., B. Guss, B. Nilsson, S. Gatenbeck, L. Philipson, and
M. Lindberg. 1984. Complete sequence of the staphylococcal
gene encoding protein A. A gene evolved through multiple
duplications. J. Biol. Chem. 259:1695-1702.
46. Viera, J., and J. Messing. 1982. The pUC plasmids, an ml3mp7-
derived system for insertion mutagenesis and sequencing with
Downloaded from https://journals.asm.org/journal/jb on 25 November 2021 by 37.238.174.7.
synthetic universal primers. Gene 19:259-268.
47. Weaver, J. R., and P. A. Pattee. 1964. Inducible resistance to
erythromycin in Staphylococcus aureus. J. Bacteriol.
88:574-580.
48. Weisblum, B., C. Siddhikol, C. J. Lai, and V. Demohn. 1971.
Erythromycin-inducible resistance in Staphylococcus aureus:
requirements for induction. J. Bacteriol. 106:835-847.
49. Yagi, Y., T. S. McLellan, W. A. Frez, and D. B. Clewell. 1978.
Characterization of a small plasmid determining resistance to
erythromycin, lincomycin, and vernamycin B,, in a strain of
Streptococcus sanguis isolated from dental plaque. Antimicrob.
Agents Chemother. 13:884-887.
50. Zucker, M., and P. Stiegler. 1981. Optimal computer folding of
large RNA sequences using thermodynamics and auxiliary
information. Nucleic Acids Res. 9:133-148.