Downloaded from journal.pda.

org on August 23, 2010

Equivalence of Quality Control Strains of Microorganisms

Used in the Compendial Microbiological Tests: Are National
Culture Collection Strains Identical?
Anthony M. Cundell, Sonia Chatellier, Peter Schumann, et al.

PDA J Pharm Sci and Tech 2010, 64 137-155

 Downloaded from journal.pda.org on August 23, 2010

RESEARCH

Equivalence of Quality Control Strains of Microorganisms

Used in the Compendial Microbiological Tests: Are National
Culture Collection Strains Identical?
ANTHONY M. CUNDELL†*, SONIA CHATELLIER⫹, PETER SCHUMANN#, and RICHARD LILISCHKIS‡*
†
Merck Research Laboratories, Union, New Jersey, USA; ⫹bioMérieux, La Balme Les Grottes, France;
#
Deutsche Sammlung für Mikrobiologie und Zellkultur, Braunschweig, Germany; and ‡BTF bioMérieux,
Sydney, Australia ©PDA, Inc. 2010

ABSTRACT: The pharmacopoeias list a number of microorganisms to be used in the compendial microbiological tests for
confirming the growth-promoting, indicative, and inhibitory properties of the media and demonstrating the suitability of the
test for a specific test article. Major national culture collections are specified as the sources for these test strains based on
their history of deposition and maintenance and use in the compendial tests. Using these microorganisms, it has long been
assumed that these strains are interchangeable and that sourcing the strains from different culture collections has no impact
on the result of the media quality control and method qualification tests. In order to evaluate whether this assumption is
correct and to add more certainty to the procedures, we investigated whether there are detectable differences among isolates
of the same strain sourced from different culture collections. Using various phenotypic and genotypic identification and
strain typing methods, nine major pharmacopoeial species were analyzed. As expected, most of the species showed very
uniform patterns across the isolates, indicating that the strains were indeed identical. Surprisingly, the strains of Salmonella
enterica subsp. enterica serotype abony showed distinct differences at both the genotypic and the phenotypic level,
suggesting that the strains sourced from the different culture collections were not identical strains, or that they have
undergone detectable genetic shift from the time they were derived from the original depositor. Irrespective of the level of
genotypic or phenotypic homology identified here, there are no practical consequences on their performance in compendial
assays. It is concluded that the compendial strains investigated in this study are indeed equivalent and will perform
identically in compendial tests, making it safe to base pharmaceutical quality control procedures on the strains sourced from
any of the recognized national culture collections.

KEYWORDS: Microbial identification, Quality control, Strain typing, Pharmacopoeia, MALDI-TOF mass spectrom-
etry, Rep-PCR, Ribotyping, 16S rRNA gene sequencing, Serotyping

Introduction
The three compendia, the United States, European, and
Japanese Pharmacopeia (USP, Ph. Eur. and JP, respec-
tively), are most widely used as standard-setting organi-
zations to guide pharmaceutical quality control (QC).
Each of them is recognized by the respective authorities
as the official compendium, e.g., in the US by the Federal
Food, Drug, and Cosmetic (FDC) Act. The compendial
standards are used to determine the identity, strength,
quality, and purity of pharmaceutical articles. Although
* Corresponding authors: anthony.cundell@spcorp.com,
Tel:⫹1908-820-6966;richard.lilischkis@btf.biomerieux.
com, Tel: ⫹61 28877 9150
the compendial tests are strictly referee tests that would
be used to assess the microbiological quality of a phar-
maceutical article, they are typically referenced in regu-
latory filings, hence becoming product release tests. The
tests used in the pharmaceutical industry need to comply
with the compendia of the country the product is to be
sold in, often resulting in slightly different test needs.
After many years of effort, the chapters in the compendia
for microbial testing of nonsterile and sterile products
have been successfully harmonized and are in the pro-
cess of implementation (1–3). Only the USP official tests
are referenced as representative of the tripartite harmo-
nized tests.
In order to confirm method suitability, different
species of bacteria, yeasts, and molds as well as

Vol. 64, No. 2, March–April 2010 137

 Downloaded from journal.pda.org on August 23, 2010

TABLE I
USP-Listed Cultures for the Official Compendial Microbiological Test Methods

Compendial Test Species Compendial Species

具61典 Microbiological
Examination of Non-
sterile Products:
Microbial Enumeration
具62典 Microbiological
Examination of Non-
sterile Products: Tests
for Specified
Microorganisms
具71典 Sterility Tests
For the explanations of the acronyms see Table II.
a
A name change for this strain from A. niger to A. brasiliensis has been proposed (17).
Staphylococcus aureus
Pseudomonas aeruginosa
Bacillus subtilis
Escherichia coli
Aspergillus nigera
Candida albicans
Staphylococcus aureus
Pseudomonas aeruginosa
Escherichia coli
Salmonella enterica subsp.
enterica serotype typhimurium
Salmonella enterica subsp.
enterica serotype abony
Clostridium sporogenes (two
different strains listed)
Candida albicans
Staphylococcus aureus
Bacillus subtilis
Pseudomonas aeruginosa
Clostridium sporogenes
Candida albicans
Aspergillus nigera
ATCC 6538, NCIMB 9518, CIP 4.83,
NBRC 13276
ATCC 9027, NCIMB 8626, CIP
82.118, NBRC 13275
ATCC 6633, NCIMB 8054 CIP 52.62,
NBRC 3134
ATCC 8739, NCIMB 8545, CIP 53.126
NBRC 3972
ATCC 16404, IMI 149007, IP 1431.83
NBRC 9455
ATCC 10231, NCPF 3179, IP 48.72,
NBRC 1594
ATCC 6538, NCIMB 9518, CIP 4.83,
NBRC 13276
ATCC 9027, NCIMB 8626, CIP
82.118, NBRC 13275
ATCC 8739, NCIMB 8545, CIP 53.126
NBRC 3972
ATCC 14028
NBRC 100797, NCTC 6017, CIP 80.39
ATCC 11437, NBRC 14293, NCIMB
12343, CIP 100651 and ATCC
19404, NCTC 532, CIP 79.3
ATCC 10231,NCPF 3179, IP 48.72,
NBRC 1594
ATCC 6538, NCIMB 9518, CIP 4.83,
NCTC 10788
ATCC 6633, NCIMB 8054 CIP 52.62,
ATCC 9027, NCIMB 8626, CIP 82.118
ATCC 19404, NCTC 532 CIP 79.3
ATCC 10231, NCPF 3179 IP, 48.72
ATCC 16404, IMI 149007, IP 1431.83

suitable strains are specified. Moreover, the com-
pendia list a limited number of national culture
collections as sources for these strains as equivalent
(see Table I), with the assumption that the strains
are indeed identical (i.e., monoclonal) and could be
used interchangeably in the compendial tests. There
is, however, no formal information in the literature
to confirm this belief (4). The equivalency is solely
based on the culture history and the affirmation of
the culture collection that at some point in time the
strains were derived from each other or a common
source. There are other well recognized culture col-
lections that make the same strains available and
can be identified using the bioportal of the Univer-
sity of Gent (www.straininfo.ugent.be). Further-
more, these stains can be purchased as semi-
quantitative or quantitative prepared inocula from
commercial suppliers.

138 PDA Journal of Pharmaceutical Science and Technology

 Downloaded from journal.pda.org on August 23, 2010
This study investigates whether there are detectable
differences of the same strains sourced from different
culture collections and if so whether this has an impact
on the outcome of the compendial tests. To achieve
this, multiple methods generally aimed at the identi-
fication at the species level (VITEK威 2, Biolog™,
API威, MALDI-TOF MS, and MicroSeq威, Applera
Corporation) and for typing at subspecies level (Di-
versiLab威, riboprinting, and serotyping) were em-
ployed. This study, however, did not attempt to compare
the merits and suitability of the different methods avail-
able for species identification and strain typing. In the
case of Salmonella enterica subsp. enterica serotype
abony where appreciable differences could be detected,
the impact on the suitability of the strains in compendial
tests such as the growth promotion test was evaluated.
Materials and Methods
Strains Evaluated
Bacterial and fungal strains (see Table II) were pur-
chased from the Australian Collection of Microorgan-
isms (ACM, Australia), the American Type Culture Col-
lection (ATCC威, USA), CAB International (CABI, UK),
the Collection de l’Institut Pasteur (CIP, France), the
Deutsche Sammlung von Mikroorganismen und Zellkul-
turen (DSMZ, Germany), the National Biological Re-
source Center (NBRC, Japan), the National Collections
of Industrial and Marine Bacteria (NCIMB, Scotland),
the National Collection of Pathogenic Fungi (NCPF,
UK), and the National Collection of Type Cultures
(NCTC™, Health Protection Agency, UK) and were
subcultured no more than five times prior to testing.
Please note that since the Japanese Institute for Fermen-
tation (IFO) joined the NBRC, all IFO numbers are
identical to their respective NBRC numbers.
VITEK 2 Microbial Identification
The frozen strains were subcultured twice on Colum-
bia agar with 5% sheep blood or Sabouraud dextrose
agar. All strains but Aspergillus isolates and Clostrid-
ium strains were identified by the colorimetric VITEK
2 cards (BCL, GN, GP, YST cards; bioMérieux,
Durham, NC) according to the instructions of the
manufacturer (McFarland standard of 0.5–2 depending
on the card used). A bacterial suspension made in
0.45% aqueous NaCl was adjusted to the recom-
mended McFarland standard with a VITEK 2 Den-
siChek™ instrument. The time period from prepara-
tion of the inoculum to loading of the card was less
Vol. 64, No. 2, March–April 2010
than 30 min. The card, placed on the tray and applied
to the VITEK 2 system, was automatically processed
in a vacuum chamber, incubated at 35.5 °C, and au-
tomatically subjected to a colorimetric measurement
by use of an optical reading head every 15 min for a
maximum incubation period of 8 –18 h depending on
the VITEK 2 card used. Data were analyzed using
VITEK 2 database version 4.01.
API Microbial Identification
Some of the isolates were also tested on API strips
(bioMérieux) according to the manufacturer’s instruc-
tions. ID32E was used for Gram-negative bacteria,
Candida albicans isolates were tested on ID32C strips,
Staphylococcus aureus was tested on ID32 Staph, and
Clostridium sporogenes was checked on ID32A.
Biolog Microbial Identification
A pure culture, isolated on soybean casein digest agar,
was subcultured onto Biolog BUG agar with 5% sheep
blood and incubated at 35 °C for 16 –18 h. From the
subculture, a uniform cell suspension was prepared to
a density of 61 ⫾ 2% light transmittance in Biolog
GN/GP-IF (Gram-negative/Gram-positive inoculating
fluid) containing 5 mM sodium thioglycolate. A 96-
well microplate was then inoculated with 150 ␮L/well
of the cell suspension. The microplate was incubated
at 35 °C for 16 –24 h with readings being taken during
this time interval.
Matrix-Assisted Laser-Desorption/Ionization
Time-of-Flight Mass Spectrometry (MALDI-TOF MS)
Sample preparation for MALDI-TOF MS protein anal-
ysis was carried out according to the ethanol/formic
acid extraction protocol recommended by Bruker Dal-
tonik (Bremen, Germany) as described previously (7).
First, about 10 mg biomass from agar cultures was
resuspended in 300 ␮L water by careful mixing. Then,
the suspension was mixed with 900 ␮L ethanol. The
biomass was collected by centrifugation and the pellet
was resuspended in 50 ␮L 70% formic acid. The
suspension was mixed carefully with 50 ␮L acetoni-
trile. After centrifugation, the supernatant was re-
moved immediately and aliquots of 1.5 ␮L were
placed on each spot of a stainless steel target plate.
After air-drying, 1.5 ␮L matrix solution (saturated
solution of alpha-cyanohydroxycinnaminic acid in
50% aqueous acetonitrile containing 2.5% trifluoro-
acetic acid) per spot was applied. MALDI-TOF MS
139
 TABLE II

140
History of Deposition of Compendial Cultures in Six Major Culture Collections
National Culture Collectionsa

Strain ATCC CIP DSMZ NBRCb NCIMB NCTC/NCPF

A. niger ATCC 16404 4 Warner IP 1431.83 no history record DSM 1988 4 IFO 9455 4 ATCC 16404 not available NCPF 2275 4
Lambert Res. Inst. ATCC 16404 ATCC 16404
Origin: SM Ringel; (S. M. Ringel)
Isolation: Blueberry, North
Carolina
B. subtilis ATCC 6633 4 N. R. CIP 52.62 4 R. E. Gordon NJ Agri. DSM 347 4 ATCC IFO 3134 4 Iowa State NCIMB 8054 4 ATCC 6633 NCTC 10400 4
Smith 231 4 K. F. Exper. Stat. 4 K. F. Kellerman 6633 Coll. (1A29) 4 ATCC ATCC 6633
Origin: K. F. Kellerman Kellerman 6633
Isolation: Unknown
C. albicans ATCC 10231 4 NIHA C. IP 48.72 4 ATCC 10231 DSM 1386 4 IFO 1594 4 ATCC 10231 not available NCPF 3179 4
W. Emmons 4 Wright ATCC 10231 ATCC 10231
Origin: Wright
Isolation: Man with
broncho-mycosis
C. sporogenes ATCC 114374 H.D CIP 100651 4 1983, ATCC DSM 1446 4 IFO 14293 4 Daigo NCIMB 12343 4 ATCC 11437 NCTC 12935 no
Vera NCIB 12343 11437 ATCC 11437 Nutritive Chem. (H. history record
Origin: HD Vera [L. S. McClung 2006] Yoshino) 4 ATCC
Isolation: Cotton plant 11437
C. sporogenes ATCC 19404 4 NCTC CIP 79.3 4 1979, ATCC 19404 DSM 1664 4 not available NCIMB 532 4 NCTC532 NCTC 532 4 M.
532 4 M. Robertson, ATCC 19404 Robertson, Lister
Origin: M Robertson Lister Inst.U.K Inst.U.K
Isolation: Gas gangrene
E. coli ATCC 8739 4 I. C. CIP 53.126 4 1953, NCIMB DSM 1576 4 ATCC IFO 3972 4 Takeda Chem. NCIMB 8545 4 ATCC 8739 NCTC 12923 no
Gunsalus strain Crooks 85454ATCC 8739 8739 Ind., Ltd. 4 Univ. Texas history record
Origin: G. C. Crooks 4 G. C. Crooks 4 ATCC 8739
Isolation: Feces
P. aeruginosa ATCC 9027 4 C. P. CIP 82.118 41982, Blanchard, DSM 1128 4 ATCC IFO 13275 4 ATCC 9027 NCIMB 8626 4 ATCC 9027 NCTC 12924 4
Hegarty, RH 813 Specia, France 4 ATCC 9027 9027 C.P Hegarty
Origin: CP Hegarty
Isolation: Outer ear infection
S. enterica subsp. enterica ATCC 14028 4 CDC CIP 104115 4 1994, ATCC 14028 DSM 19587 4 CIP not available NCIMB 13284 4 ATCC 14028 NCTC 12023 4
serotype typhimurium 6516-60 104115 4 ATCC ATCC 14028
14028
Downloaded from journal.pda.org on August 23, 2010

Origin: CDC
Isolation: tissue animal
(pools of heart & liver
from 4 wk old chickens)
S. enterica subsp. enterica not available CIP 80.39 4 Inst. Pasteur 4 F. DSM 4224 4 CIP IFO100797 4 CIP 80.39 not available NCTC 6017 4 F.
serotype abony Kaufmann, K 103 80.39 Kaufmann, K 103
Origin: F. Kaufmann.
Denmark
Isolation: Unknown
S. aureus ATCC 6538 4 FDA 209P CIP 4.83 4 ATCC 6538 DSM 799 4 ATCC IFO 13276 4 ATCC 6538 NCIMB 9518 4 Diversey (UK) NCTC 10788 4
4 Walter Reed AMC 6538 Ltd 4 Diversey Corp. ATCC 6538
Origin: Walter Reed AMC Chicago 4 ATCC 6538
Isolation: human lesion
Strain numbers in bold were analyzed in this study.
a
ATCC ⫽ American Type Culture Collection (USA), NCIMB ⫽ National Collections of Industrial and Marine Bacteria (Scotland), CIP ⫽ Collection de l’Institut Pasteur (France), NBRC ⫽ National
Biological Resource Center (Japan), NCTC ⫽ National Collection of Type Cultures (UK), DSM ⫽ Deutsche Sammlung von Mikroorganismen (Germany), IFO ⫽ Institute for Fermentation (Japan).
b

PDA Journal of Pharmaceutical Science and Technology

Since IFO joined NBRC, all IFO strain numbers are identical to NBRC numbers, i.e., B. subtilis IFO 3134 ⫽ NBRC 3134.
 Downloaded from journal.pda.org on August 23, 2010
was conducted using a Microflex威 L20 mass spec-
trometer (Bruker Daltonik) equipped with an N2 laser.
All spectra were recorded in linear, positive ion mode.
The acceleration voltage was 20 kV. Spectra were
collected as a sum of 500 shots across a spot. A mass
range of 2000 –20,000 m/z was used for analysis. The
spectra were analyzed by using the BioTyper software,
Version 1.1 (Bruker Daltonik).
16S rRNA Base Sequencing
Sequencing of the 16S rRNA gene was carried out
either as a partial 5⬘ sequence of 500 bp or the full-
length 1500 bp using the MicroSeq 16S rDNA proto-
col (Applied Biosystems, Foster City, CA) and was
conducted at the Australian Genome Research Facility
in Melbourne, Australia.
Ribotyping
Automated ribotyping was performed according to the
instructions of the manufacturer of the Riboprinter
microbial characterization system (DuPont Qualicon,
Wilmington, DE) as described previously (5). The
automated process included bacterial cell lysis and
digestion of the DNA by using either the restriction
enzyme EcoRI or PvuII. DNA fragments were sepa-
rated by size using agarose gel electrophoresis. The
DNA fragments were hybridized with a labeled rRNA
operon probe derived from Escherichia coli, and the
bands were detected by a chemiluminescent substrate.
The image was recorded by using a customized charge-
coupled device camera. Data were normalized to a stan-
dard marker set. The band patterns were compared using
the BioNumerics威 software (Applied Maths NV, St-
Martens-Latem, Belgium) and clustering was carried out
by the unweighted-pair group method with the arithmetic
mean (UPGMA) based on Pearson’s correlation coeffi-
cient (optimization coefficient, 1.2%).
Repetitive Sequence-Based Polymerase Chain Reaction
(Rep-PCR)
The bacterial isolates were cultured on plates of Co-
lumbia 5% sheep blood agar with the exception of C.
sporogenes, which was cultured in brain heart infusion
broth. Yeasts and molds were subcultured on Sab-
ouraud dextrose agar plates. The DNA from each
culture was extracted from a 10-␮L loop of colony
mass using the UltraClean microbial DNA isolation kit
(Mo Bio Laboratories, Solana Beach, CA). DNA (30
ng/␮L) was amplified using the DiversiLab kit appro-
Vol. 64, No. 2, March–April 2010
priate for the microorganism (Bacterial Barcodes, Inc.,
Athens, GA) following the manufacturer’s instruc-
tions. The amplicons were then analyzed using the
DiversiLab system including a microfluidics chip
(Bacterial Barcodes). Data analysis was performed
with the Web-based DiversiLab software version 3.3
using the Pearson Correlation coefficient and UPGMA
to automatically compare the rep-PCR-based DNA
fingerprints of unknown isolates.
Inter Simple Sequence Repeats (ISSR) and Variable
Number Tandem Repeats (VNTR)
ISSR and VNTR analyses were performed by CABI,
as described by Grünig et al. (8) and Bridge et al. (9),
respectively, except that the annealing temperature in
the VNTR assays was 45 °C for Aspergillus niger and
48 °C for C. albicans.
Serotyping
Strains of S. enterica subsp. enterica serotype abony
were subjected to serological typing according to the
Kauffman-White scheme (6) at the Institute of Clinical
Pathology and Medical Research in Sydney, Australia.
Growth Promotion Testing
Freeze-dried inocula containing a precise number of
viable cells of S. enterica subsp. enterica serotype
abony strains ACM 5080 and CIP 80.39 were prepared
using the proprietary BioBall technology (BTF bio-
Mérieux, Sydney, Australia) (10). Briefly, cultures
were grown in a modified 2YT medium and counted
and sorted using flow cytometry. Droplets were snap-
frozen in liquid nitrogen and freeze-dried. The freeze-
dried BioBall was rehydrated directly on xylose lysine
deoxocholate agar (XLD), soybean-casein digest agar
(SCDA) ⫽ trypticase soy agar (TSA), or nutrient agar
(NA) in 100 ␮L 0.9% NaCl solution. For experiments
utilizing non-freeze-dried cultures, the strains were
cultured in 2YT into mid-log phase, counted and
sorted as above, and dropped directly onto the agar
plates. After spreading and drying, the plate was in-
cubated at 30 –35 °C for the time indicated and the
colonies were counted or their diameter measured.
Results and Discussion
Strain Culture History
In the past, culture collections would extensively ex-
change strain isolates deposited at only one culture
141
 Downloaded from journal.pda.org on August 23, 2010
collection with others in order to make the strains
widely accessible to the research and industry com-
munities. In every culture collection, the same strain
assumes a new identity code, and the dependencies
become less apparent. Table II lists the culture history
of the deposition of compendial cultures in six major
culture collections. For the purpose of this study we
follow the definition of Brenner and coworkers and
define a strain as the descendant of a single isolation in
pure culture stored in a culture (strain) collection,
typically from a single colony isolated on a plate, that
is, monoclonal (11). One strain of a species is desig-
nated as the type culture due to its early history of
deposition in the culture collection and its taxonomic
status, with all other strains sufficiently similar in both
phenotype and genotype to be considered within the
species. For example, Pseudomonas aeruginosa
ATTC 10145 is the type strain for both the genus and
species while P. aeruginosa ATCC 9027 isolated from
an inner ear infection was designated as a compendial
QC strain in USP 具61典, 具62典, and 具71典.
Strain Identification Confirmation at Species Level
In order to confirm the species identity, three commer-
cially available phenotypic identification methods
were employed. API, Biolog, and VITEK 2 identifi-
cation reagents are based on biochemical methods
mainly measuring carbon source utilization and enzy-
matic activities. While both API strips and VITEK 2
cards contain biochemical substrates, their nature,
concentration, and number in the manual API strip and
the automated VITEK 2 card are different, leading to
a large menu of claimed species. As expected, culture
collection strains were correctly identified and no dif-
ference based on the source of the isolate was noted.
The BIOLOG results were not always definitive, as
some strains were identified incorrectly as phenotyp-
ically related bacteria. Salmonella typhimurium ATCC
14028 was identified as Salmonella gp 1 (cholerae-
suis) or Citrobacter sedlakii while S. typhimurium
NCTC 12023 was identified as Salmonella gp 1 (chol-
eraesuis). Salmonella abony CIP 80.39, ACM 5080,
and NCTC 6017 were identified as Salmonella gp 1
(choleraesuis).
Genetic Typing Discriminates Strains
Despite confirming the species in each case, the phe-
notypic identification methods did not uncover possi-
ble subtle differences among strains of the same spe-
cies or changes that might have occurred during
142
storage and subcultivation. In order to gain more res-
olution power, two commercially available strain typ-
ing methods were employed, ribotyping (Riboprinter,
DuPont Qualicon) and rep-PCR (DiversiLab, Bacterial
Barcodes). Both methods rely on differences in the
genomic DNA. Although both methods return results
that look very similar, there are fundamental differ-
ences as to how the banding patterns are generated and
how much information they bear. Briefly, for ribotyp-
ing, the extracted genomic DNA is digested with a
specific restriction enzyme (typically EcoRI) and the
resulting fragments are separated by gel electrophore-
sis and transferred onto a membrane. On this mem-
brane, the DNA is subjected to Southern hybridization
using a probe derived from the rRNA operon of E.
coli. Hybridization highlights the fragments of
genomic DNA with sufficient homology to the probe
to create a specific banding pattern. The length of the
detected fragments depends on the location of the
target sites for the restriction endonuclease used.
Rep-PCR, on the other hand, is not limited to the
conserved rRNA coding sequences on the genomic
DNA but utilizes repetitive DNA sequences dispersed
throughout the genome. In a PCR reaction utilizing
primers complementary to those repetitive DNA ele-
ments, fragments of different lengths are generated.
The amplified fragments are separated and detected by
microfluidics-based electrophoresis. Both methods re-
sult in a pattern of DNA fragments, the origin of which
however is fundamentally different. Comprehensive
descriptions of both technologies can be found in the
literature (5, 8, 12, 13). As expected, ribotyping the
strains of Bacillus subtilis, E. coli, P. aeruginosa, and
S. aureus specified in the USP and EP Ph. Eur. re-
sulted in very similar banding patterns, regardless of
the culture collection from which they were sourced
(Figure 1). This set of ribotyping experiments was
performed with EcoRI, the restriction enzyme most
commonly used for ribotyping. For comparison, an
unrelated strain of the same species was chosen ran-
domly and subjected to the same ribotyping and
showed marked differences in the banding pattern.
Pearson correlation analysis documented for each spe-
cies a very close clustering of the isolates of the same
strain from different culture collections, whereas a
much lower correlation percentage of the unrelated
strain to that cluster was observed. In the case of P.
aeruginosa, and to a lesser extent S. aureus, two
unrelated strains appeared to be more similar to each
other than to the other strain cluster (Figure 1, C and
D). It should be noted here that this close clustering in
PDA Journal of Pharmaceutical Science and Technology
 Downloaded from journal.pda.org on August 23, 2010

NCTC 10788 and ATCC 6538 vs DSM 799 to CIP

4.83 (Figure 1, C and D, respectively). A mutation in
the recognition site of the restriction enzyme on the
genomic DNA will result in the absence of a particular
fragment and possibly the reappearance of a fragment
of a different size. An additional or a missing band
was therefore interpreted as a much more dramatic
change than pattern migration differences.

Ribotyping can be performed using any restriction

endonuclease or even combinations of different re-
striction endonucleases in order to cover more target
sites and potentially to reveal more information. Ad-
ditionally, all E. coli strains were ribotyped using
PvuII as the restriction enzyme (Figure 2A). The many
more fragments detected indicated that there are more
target sites for PvuII present in the rRNA operon of E.
coli than for EcoRI. Despite the increased complexity

Figure 1

Ribotyping of the same strain sourced from different

culture collections. Unrelated strains of each species
were included for comparison. Bacterial cells were
lysed, the DNA extracted and digested with EcoR1.
The resulting fragments were separated using aga-
rose gel electrophoresis, continuously membrane-
blotted and hybridized with a labeled probe derived
from the rRNA operon of E. coli. The banding pat-
tern generated by chemiluminescent detection was
normalized using a standard marker set. The banding
Figure 2
patterns were analysed using BioNumerics software
and clustering was carried out based on Pearson’s
A. Ribotyping of E. coli with the restriction enzyme
correlation coefficient. (A–D) The same strain
PvuII. B. Repeat Ribotyping. (A) Ribotyping of the
sourced from 5– 6 different culture collections was
same set of E. coli strains as in Figure 1B with the
ribotyped. Unrelated strains of each species were in-
restriction enzyme PvuII. Note that the complexity
cluded for comparison.
of the banding pattern is much higher as compared
to the EcoR1-based analysis, however, there is no
the Pearson correlation was observed despite slight significant difference between the compendia strain
differences in the migration distance of the individual cluster and the unrelated E. coli strain. (B) Repeat
isolates (e.g., see the “drift” in P. aeruginosa fragment Ribotyping. P. aeruginosa NCTC 12924 was ri-
migration from NCIMB 8626 to NCTC 12924 or the botyped independently ten times to determine the
slight difference in migration of S. aureus patterns of method-inherent variability.

Vol. 64, No. 2, March–April 2010 143

 Downloaded from journal.pda.org on August 23, 2010
of the banding pattern, PvuII was unable to discrimi-
nate the unrelated strain DSM 1563 from the compen-
dial strain cluster, whereas EcoRI produced distinct
fragment patterns clearly distinguishing these strains
(compare Figures 1B and 2A). In order to gain insight
into the repeatability of the method, the same isolate
of P. aeruginosa (NCTC 12924) was ribotyped inde-
pendently 10 times (Figure 2B). The resulting patterns
were very similar with slight variations in the migra-
tion distances (Figure 2B).
When the same set of compendial strains was analyzed
using rep-PCR, again no appreciable differences could
144
be detected (Figure 3). For each strain cluster, the
overall similarity was always above 96%. As the band-
ing pattern is dependent on the location of repetitive
elements dispersed throughout the genome and is not
focused on a conserved region, the rep-PCR analysis
offers strain discrimination information distinct from
the ribotyping results. In the rep-PCR results there was
no evidence of a shift of the banding patterns as seen
in ribotyping. The authors contribute this observation
to the different methods used to separate the DNA
fragments in the different systems, as well as the
normalization of the data by the software. In each
rep-PCR analysis, at least one strain was included in
duplicate to facilitate the discrimination of true ge-
netic differences from method-inherent variability.
For example, the parallel analysis of two samples of B.
subtilis ATCC 6633 taken from the exact same culture
did not cluster next to each other, whereas the two
samples of B. subtilis NCTC 10400 did (Figure 3A).
The algorithm calculating the similarity takes both the
position and the intensity of the bands into account
and picks up minor differences in band intensities,
which are hard to quantify with the naked eye. This
repeat analysis was expanded using P. aeruginosa as
an example (Figure 4A). Ten independent repeat rep-
PCR analyses resulted in indistinguishable patterns
with only minor differences in band intensity. A com-
plete absence or reappearance of a prominent band
was never observed. While all samples clustered to-
gether, these small method-inherent variations in in-
tensity, which occur in both ribotyping and rep-PCR,
are likely to be the reason why the order of the
Figure 3
Rep-PCR of the same strain sourced from different
culture collections. Bacterial strains were cultured
and the DNA extracted from a 10 ␮L loop of colony
mass. Amplification of the repetitive genomic se-
quences was carried out using a specific reagent set
for each species. The resulting fragments were sep-
arated on a microfluidics chip and the banding
pattern recorded. The banding patterns were ana-
lyzed using the DiversiLab software. Clustering
was carried out based on Pearson’s correlation
coefficient and UPGMA. (A–D) The same strain
sourced from 4 –5 different culture collections was
analyzed. For every panel at least one strain was
included in duplicate to demonstrate the level of
reproducibility.
PDA Journal of Pharmaceutical Science and Technology
 Downloaded from journal.pda.org on August 23, 2010
Figure 4
A. Repeat rep-PCR experiments. B. and C. Rep-
PCR of yeast and mold. (A) Repeat rep-PCR ex-
periments. P. aeruginosa NCTC 12924 was ana-
lyzed by rep-PCR ten times in separate reactions to
determine the method-inherent variability. Rep-
PCR of (B) Aspergillus niger and (C) Candida albi-
cans. Note that a name change for this strain of A.
niger to A. brasiliensis has been proposed (17).
samples within each cluster may be different (compare
Figures 1 and 3).
It would be desirable to be able to set an independent,
universal cut-off limit for similarity or Pearson coef-
ficient above which different isolates are considered to
be identical across all taxa. The algorithms underlying
the output of commercial typing platforms are differ-
ent and often undisclosed. For example, the Ribo-
Vol. 64, No. 2, March–April 2010
printer defines 96% similarity as the limit above which
the patterns of this “Ribogroup” are indistinguishable
and isolates are therefore considered identical. In the
daily use in QC of a large national culture collection,
this fixed arbitrary number has proven to be of limited
value (P. Schumann, unpublished results). The con-
cept appears not to be uniformly applicable to patterns
with a large number of bands of varying intensity
(type 1) vs patterns with just a few bands of similar
intensity (type 2). It is also limited by the reproduc-
ibility of the method. For this reason we have not used
Ribogroups in the discussion of the results but calcu-
lated dendrograms instead. Type 1 patterns can pro-
duce a much wider bandwidth of results and are more
likely to be affected by reproducibility issues than
type 2. Accordingly, identical type 2 strains should
cluster with a much higher Pearson coefficient than
type 1 strains. Furthermore, there is no apparent cor-
relation between the typing pattern and the Gram type.
Finally, differences in band intensities will also be
influenced by the number of repeats of the rRNA
operon of different species. On the other hand, the
relative differences of the “out-group pattern” (pattern
of a different strain of the same species) will influence
the level at which identical strains appear “similar”.
Because the diversity of strains within each species is
quite different from species to species it is impossible
to define a standardized out-group across all species.
As with the above bacterial species, no significant
differences between the isolates of the molds and
yeasts were detected when tested by rep-PCR (Figure
4, B and C). In the analysis of C. albicans, an ex-
tremely low level of variation between the repeats and
within the cluster was observed, leading to an overall
similarity higher than 99%. The low sequence homol-
ogy between the rRNA operons of prokaryotes and
eukaryotes does not allow eukaryotes to be analyzed
with ribotyping. Therefore, in addition to rep-PCR, A.
niger and C. albicans were analyzed by ISSR PCR and
VNTR PCR, two specialized genetic fingerprinting
methods different from rep-PCR (8, 9). Distinct fin-
gerprints were obtained for all isolates tested. For A.
niger, four primer sets resulted in satisfactory ampli-
fication products. ATCC 16404, NCPF 3179, IP
1431.83, NBRC 9455, and IMI 149007 exhibited no
visible differences in the banding patterns, whereas
IMI 91855, an unrelated A. niger strain, displayed
quite distinct profiles (data not shown). Similar results
were obtained from the C. albicans isolates tested.
Again, all strains gave rise to informative fingerprints
with four primer sets. There were no detectable ge-
145
 Downloaded from journal.pda.org on August 23, 2010
Figure 5
Riboprinting of C. sporogenes with A. EcoRI and B.
PvuII. Riboprinting of C. sporogenes (A) using
EcoRI and (B) PvuII as restriction enzymes prior to
blotting and hybridization. A higher variability in
the banding patterns was observed for C. sporo-
genes when compared to the other bacterial species.
The unrelated DSM 767 could only be separated
from the compendia strains using PvuII, which
gave a distinctly different banding pattern (see ar-
row in B). Note that the resulting clustering even
for the compendial strain is different in the PvuII
and the EcoR1 digests, pointing toward a limit of
reproducibility rather than real differences.
netic differences among C. albicans ATCC 10231,
NCPF 3179, IP 48.72, and NBRC 1594. IMI 165713,
an unrelated C. albicans strain used as a control, again
exhibited distinct profiles under the same conditions.
The amplification profiles have not been included in
this publication but can be obtained from the authors
upon request.
At first glance, C. sporogenes appears to have a higher
variability in the ribotyping patterns than the afore-
mentioned microorganisms, leading to a lower per-
centage value in the Pearson correlation (Figure 5, A
and B). This variation appeared to be caused by dif-
ferences in the electrophoretic separation rather than a
difference in the banding pattern. Furthermore, neither
rep-PCR nor pulsed field gel electrophoresis (PFGE)
displayed significant differences among NCTC 12935,
ATCC 11437, and NBRC 14293 (data not shown).
Interestingly, even the unrelated C. sporogenes strain
DSM 767 could not be distinguished from the com-
146
pendial strain cluster using EcoRI-digested DNA.
When the same set of strains was ribotyped utilizing
PvuII as the restriction enzyme, DSM 767 displayed a
clear difference in banding pattern (Figure 5B, arrow).
The banding patterns of the compendial strains were
very similar, and again the electrophoretic mobility
may have caused the remaining variability. In the
interpretation of Southern hybridization experiments
such as ribotyping, one needs to be careful with con-
clusions drawn from bands around 50 kb. An incom-
plete restriction digest will result in partially or undi-
gested DNA. These large fragments tend to have an
electrophoretic mobility in standard agarose gels sim-
ilar to that of DNA fragments of 50 kb in length,
irrespective of their actual length. A hybridization
signal at 50 kb might, therefore, at least partly be
caused by undigested DNA and is therefore not infor-
mative.
Taken together, the analysis of the compendial strains
of A. niger, B. subtilis, C. albicans, C. sporogenes, E.
coli, P. aeruginosa, and S. aureus did not reveal
significant differences either on the phenotypic or
genotypic level.
Salmonella Enterica subsp. Enterica Serotype Abony:
Equivalent Strains Are Not Identical
The compendia specify two Salmonella strains as suit-
able for QC of the Microbial Limits Test (USP 62, Ph.
Eur.2.6.13). According to the new nomenclature, both
strains are of the same species and subspecies (S.
enterica subsp. enterica) (14) and differ only in the
serotype (typhimurium and abony). Initially, there
were concerns whether even genotyping methods
would be able to distinguish the two serotypes. How-
ever, both rep-PCR and ribotyping clearly distin-
guished S. typhimurium from S. abony (compare Fig-
ures 6 and 7). S. typhimurium clustered very closely in
the rep-PCR and in the ribotyping (Figures 6A and
7A).
Unexpectedly, the rep-PCR analysis divided the S.
abony strains into two clearly distinguishable groups
(Figure 6B), with ACM 5080 and NCTC 6017 dis-
playing a different banding pattern than CIP 80.39,
NBRC 100797, and DSM 4224. This distinction could
be confirmed by ribotyping (Figure 7B). This finding
was puzzling because the compendia, the culture col-
lections, and probably the overwhelming majority of
the pharmaceutical microbiology community have
long assumed that all of these strains were identical. In
PDA Journal of Pharmaceutical Science and Technology
 Downloaded from journal.pda.org on August 23, 2010
Figure 6
Rep-PCR of Salmonella strains. A. S. typhimurium
B. Two groups of S. abony can be differentiated. (A)
The rep-PCR patterns of S. typhimurium sourced
from ATCC and NCTC do not reveal significant
differences and appear to be highly reproducible.
(B) The same approach differentiated two groups
of S. abony. ACM 5080 and NCTC 6017 appear to
be distinctly different from the strains sourced
from CIP, NBRC, and DSMZ. One strain of each
group was analyzed in duplicate to judge the re-
producibility of this assay.
an attempt to resolve this paradoxical finding, the
culture history of these strains was investigated (Fig-
ure 8). The strain was first isolated by F. Kaufmann
from the Statens Seruminstitut, Denmark. On the April
4, 1940, it was deposited with the National Collection
of Type Cultures (NCTC), UK (B. Holmes, personal
communication), and it was in parallel sent to L.
LeMinor at the French Salmonella Reference Labora-
tory. In 1980, LeMinor deposited the strain with the
Collection of the Pasteur Institute in France. From
these two collections, the strains were distributed to
other national culture collections including those of
Germany (DSMZ), Japan (NBRC), and Australia
(ACM). The distribution pattern as outlined in Figure
8 matches exactly the distinction found by genetic
typing, with CIP 80.39, DSM 4224, and NBRC
100797 forming one group distinct from the group
consisting of NCTC 6017 and ACM 5080. It is there-
fore likely that the differences pre-existed at the two
Vol. 64, No. 2, March–April 2010
culture collections, NCTC and CIP, where the strain
was first deposited.
Confirming the Differences
In order to confirm the differences between the two
groups of strains, a number of further analyses were
conducted. For the genus Salmonella, serotyping ac-
cording to the Kauffmann-White scheme is the most
widely used method for typing and has a long history
in the field. Serotyping performed for S. enterica
subsp. enterica serotype abony revealed that all iso-
lates tested displayed the exact same antigenic struc-
ture of 1,4,5,12:b:e,n,x (data not shown). This absence
of differences in the antigenic structure of the isolates
was anticipated because, for the genus Salmonella,
culture collections and reference laboratories have tra-
ditionally used serotyping as a means of QC.
Next, sequencing of 500 bp of 5⬘ end of the DNA
coding for the 16S rRNA was conducted. CIP 80.39,
Figure 7
Riboprinting of Samonella strains. A. S. typhi-
murium B. Two groups of S. abony can be differ-
entiated. (A) Riboprinting patterns of S. typhi-
murium sourced from DSMZ, ATCC, and NCTC
do not reveal significant differences. (B) Riboprint-
ing differentiated the same two groups of S. abony
as found using rep-PCR. Note the presence of a
band around 15 kb in strains sourced from CIP,
NBRC, and DSMZ whereas this band is absent in
ACM 5080 and NCTC 6017. The reverse is true for
the band running close to 5 kb.
147
 Downloaded from journal.pda.org on August 23, 2010

Figure 8

Deposition history of Salmonella abony strain

74K103. The strain was deposited in 1940 with the
National Collection of Type Cultures (NCTC) and
sent to L. LeMinor in the same year, who forty
years later deposited the strain with the Collection
of the Institute Pasteur. From these two culture
collections the strain reached the DSMZ, NBRC,
and ACM.

NCTC 6017, and the NCTC 6017-derived ACM 5080

showed identical sequences. Only one base at position
64 displayed a slight difference between the two
groups. In CIP 80.39, base 64 was determined clearly
as a cytosine, whereas the same position in NCTC
6017 and ACM 5080 resulted in an unclear base
determination, with either a cytosine or a thymidine at
this position (Figure 9). This result might be due to
different rRNA operons (rrn) present in Salmonella.
Most microorganisms capable of rapid growth have
multiple rrn copies to support high levels of ribosome
production. Salmonella, similar to E. coli, retained 7
rrn copies (13, 15). The ambiguous base seen in
NCTC 6017 and ACM 5080 could therefore be the
result of a single nucleotide change in at least one of Figure 9
the rrn copies.
Sequence analysis of the 5ⴕ end of the 16S rRNA of
In an attempt to further confirm the differences be- Salmonella abony strains. Sequencing 500 bp of the
tween the two strains, MALDI-TOF MS of total pro- 5ⴕ end of the 16S rRNA using the MicroSeq 500
tein extracts was employed. In the analysis of whole protocol revealed a single nucleotide difference. At
cell extracts of bacteria, MALDI-TOF MS produces position 64 in CIP 80.39 there is a clear cytosine
spectra based on highly abundant, mainly cytosolic (. . . TTCGC . . .). At this position both of the
proteins, at least half of them very basic ribosomal strains NCTC6017 and ACM 5080 detect a cytosine
proteins (16). In the NCTC 6017 extract, about 20 and a thymidine (. . . TTC/TGC . . .). This could be
peaks in the mass-to-charge (m/z) range of 2000 to interpreted as artifact or that at this position the 7
12,000 could be detected (Figure 10). Interestingly, copies of the rRNA operon are not identical.
two prominent peaks in the NCTC 6017 spectrum at an

148 PDA Journal of Pharmaceutical Science and Technology

 Downloaded from journal.pda.org on August 23, 2010
Figure 10
MALDI-TOF MS analysis of protein extracts from
Salmonella abony strains CIP 80.39 and NCTC
6017. Total protein of the strains CIP 80.39 and
NCTC 6017 was extracted and subjected to matrix-
assisted laser desorption ionization time-of-flight
mass spectrometry (MALDI-TOF MS). Of the
peaks detected with a mass to charge (m/z) ratio
between 2000 and 12,000, a double peak at m/z of
6000 – 6100 was prominent in NCTC 6017, whereas
in CIP 80.39 these peaks were only minor.
m/z of 6008 and 6094 were absent in the spectrum of
CIP 80.39 (Figure 10, arrows). Whereas the adjacent
peaks at 6255/6256 could represent the homologue of
an E. coli large subunit ribosomal protein as identified
by Ryzhov and Fenselau (16), the identity of the
protein(s) at an m/z of 6008 and 6094 remains un-
known. It is remarkable, however, that not only ge-
netic differences between NCTC 6017 and CIP 80.39
could be detected but also that clear phenotypic dif-
ferences exist.
Since sequencing 500 bp of the 5⬘ end of the 16S
rRNA gene revealed only a minor and possibly am-
biguous difference (see above), the whole coding re-
gion of the 16S rRNA gene of CIP 80.39, NCTC 6017,
and ACM 5080 was sequenced (Figure 11). Full gene
sequencing confirmed the single nucleotide divergence
at base 40 (same position as base 64 in the MicroSeq
500 protocol) as seen before (Figure 11, marked in
black). PCR amplification prior to sequencing re-
vealed that CIP 80.39 yielded a smaller fragment than
the other two samples (data not shown). Sequencing of
Vol. 64, No. 2, March–April 2010
the PCR products and alignment of the sequences
uncovered a 310-bp deletion in the middle of the gene
in the CIP 80.39-derived PCR product. In addition,
multiple single-nucleotide changes could be detected
(Figure 11, black circles). Interestingly, transversions
from thymidine to adenine at position 972 and from
guanine to thymidine at position 973 resulted in the
creation of an additional recognition site for the re-
striction endonuclease EcoRI (Figure 11, open box).
As the 16S rRNA gene contributes less than one-third
to the length of the whole operon, it is conceivable that
similar changes might have occurred in other parts of
the operon not sequenced in this study. The observed
additional EcoRI site might be the reason for the
disappearance of the band around 5 kb that was de-
tected in the ribotyping pattern of ACM 5080 and
NCTC 6017 but not in the other three strains analyzed
(Figure 7B). Similar changes in EcoRI recognition
sites in other parts of the operon could contribute to
the band at 14 kb in CIP 80.39, NBRC 100797, and
DSM 4224, which is absent in ACM 5080 and NCTC
6017 (Figure 7B).
Will the Non-identical “Twins” Perform Differently in
Compendial Tests?
The finding that the equivalent compendial strains of
S. enterica subsp. enterica serotype abony are not
identical leads to the question of whether compendial
assays based on one of the strains should be revali-
dated and, if so, which of the strains should be used as
reference. Firstly, the compendia require the use of
designated species but not a specific strain. For exam-
ple, the harmonized Chapter 62 of the USP lists “Sal-
monella enterica subsp. enterica serotype abony such
as NBRC 100797, NCTC 6017 or CIP 80.39”. Accord-
ingly, any S. enterica subsp. enterica isolate that bears
the serotype abony would satisfy the criteria set out in
the USP and the other compendia. Secondly, and pos-
sibly more importantly, one could ask whether the
differences seen are influencing the outcome of a
particular compendial test. For S. enterica subsp. en-
terica serotype abony, this would be a growth promo-
tion assay. In order to address this question, spikes
containing a precise number of colony-forming units
(cfu) were generated using the BioBall technology
(10). Briefly, liquid cultures of CIP 80.39 and ACM
5080 were subjected to flow-cytometric counting and
sorting. Populations of viable single cells were se-
lected and sorted into droplets, snap-frozen in liquid
nitrogen, and freeze-dried.
149
 Downloaded from journal.pda.org on August 23, 2010

Figure 11

Full sequence alignment of the 16S rRNA of Salmonella abony strains. Most prominently, a fragment of 310 bp in length
is deleted in CIP 80.39 when aligned with NCTC 6017 and ACM 5080. Furthermore, multiple single nucleotide changes
were detected. Two changes at positions 972 and 973 resulted in the appearance of an additional Eco R1 restriction site,
which could be the reason for a specific banding pattern difference in the Riboprinting.
150 PDA Journal of Pharmaceutical Science and Technology
 Downloaded from journal.pda.org on August 23, 2010

Figure 11 (Continued)

Vol. 64, No. 2, March–April 2010 151

 Downloaded from journal.pda.org on August 23, 2010

Figure 11. (Continued)

152 PDA Journal of Pharmaceutical Science and Technology

 Downloaded from journal.pda.org on August 23, 2010

Figure 12

Growth promotion testing of Salmonella abony strains. Colony numbers of (A) freeze-dried and (B) fresh
cultures. (C) Growth kinetics on different media. The graphs show colony number mean and standard
deviation of triplicate plates. For the colony size, images of the plates were taken at the indicated time points.
The size of all colonies of one representative plate each was measured on calibrated printouts. The graph
represents mean and standard deviation of about 30 colonies each.

For ACM 5080 and CIP 80.39, the batch mean was
29.0 and 29.2 cfu per BioBall, respectively, measured
by plating on nutrient agar. Similar counts were ob-
served when plated on SCDA (Figure 12A). Direct
spreading of the BioBall on XLD agar resulted in a
reduction of colonies for both strains, with CIP 80.39
recovering slightly fewer colonies than ACM 5080.
The difference in cfu between ACM 5080 and CIP
80.39, however, was not significant when tested by
paired Student’s t-test. In order to investigate whether
differences in growth properties might have been ob-
scured by freezing and freeze-drying, the growth pro-
motion assay was repeated, omitting the freezing in
liquid nitrogen and freeze-drying (Figure 12B). This
time, the droplets containing the flow-cytometric
counted and sorted bacteria were captured directly
onto the agar plates, spread, and the plates incubated
as before. Again, there was no significant difference
between the colony numbers recovered from NA,
SCDA, or XLD plates. Finally, the growth kinetics on

Vol. 64, No. 2, March–April 2010 153

 Downloaded from journal.pda.org on August 23, 2010
the different media determined by measuring colony
sizes did not differ significantly over a time period of
72 h on any of the media tested (Figure 12C). In
summary, growth promotion assays of strains of the
two genetically differentiable S. enterica subsp. en-
terica serotype abony “lineages” did not reveal signif-
icant differences in growth characteristics. The au-
thors suspect that the differences will not affect
suitability testing with specific test articles.
Conclusions
The compendia require standardized stable suspen-
sions of test strains to be used for growth promotion
and inhibitory properties testing as well as for method
validation. For each test, the compendia suggest a
suitable strain for each species and name a number of
culture collections as sources. Additionally, the same
strains can be sourced from a number of national
culture collections. The compendia require that the
cultures should be removed no more than five passages
from the original master seed lots. This allows for
inocula to be prepared in your laboratory or purchased
as titrated preprepared inocula without a significant
drift in the genotypic and phenotypic properties of the
cultures.
This study aimed to investigate whether a particular
strain held at different culture collections is indeed the
same strain or whether there are detectable differ-
ences, which could affect the outcome of a compendial
test. Phenotypic and genotypic identification and typ-
ing techniques employed in this study did not reveal
any reproducible differences. There was just one no-
table exception: The same original strain of S. enterica
subsp. enterica serotype abony deposited by F. Kauf-
mann 69 years ago appeared to have split into two
distinct lineages. These differences could have devel-
oped during the course of subcultivation previous to or
during the storage at the culture collections. They
might also be due to authentically different strains,
undistinguishable by the QC methods employed at the
time. It is beyond the scope of this study to determine
the likely cause for this divergence. More importantly,
however, at least from a practical view in the pharma-
ceutical QC laboratory, these strains did not behave
grossly different in growth promotion assays and
hence are both considered equivalently suitable for the
intended compendial tests. Taken together, the data
presented strongly suggest that all strains are equiva-
lent, regardless of which culture collection they were
sourced from.
154
This study was limited to the analysis of strains of the
major national culture collections. Subsequent work
should include strains sourced from other culture col-
lections as well as other microorganisms used in com-
pendial tests, for example, microbial assay of antibi-
otics and vitamins. Moreover, the managers of the
culture collections might be inspired to compare type
strains held in their collections, which are taxonomi-
cally more important than the QC strains used in the
pharmaceutical industry.
Acknowledgments
We are grateful to Barry Holmes for detailed informa-
tion on strain culture history, Scott Sutton for valuable
comments throughout the study, and Mimi Healy,
Mary Jane Hilbert, Danielle Dennis, and Nick Herman
for excellent technical work and review of the manu-
script.
References
1. USP 具61典. Microbiological Examination of Non-
Sterile Products: Microbial Enumeration Tests.
US Pharmacopeial Convention: Rockville, MD.
2. USP 具62典. Microbiological Examination of Non-
Sterile Products: Test for Specified Microorgan-
isms. US Pharmacopeial Convention: Rockville,
MD.
3. USP 具71典. Sterility Tests. US Pharmacopeial Con-
vention: Rockville, MD.
4. Sutton, S. Are all identical culture collection
strains the same? Pharmaceutical Microbiology
Forum Newsletter 2008, 14 (2), 6 –11.
5. Bruce, J. Automated system rapidly identifies and
characterizes microorganisms in food. Food Tech-
nology 1996, 50 (1), 77– 81.
6. Poppoff, Y. Antigenic formulas of the Salmonella
serovars. 8th ed. WHO Collaborating Centre for
Reference and Research on Salmonella, Institut
Pasteur, Paris, 2001.
7. Toth, E. M.; Schumann, P.; Borsodi, A. K.; Kéki,
Z.; Kovács, A. L.; Márialigeti, K. Wohlfahrtiimo-
nas chitiniclastica gen. nov., sp. nov., a new gam-
maproteobacterium isolated from Wohlfahrtia
PDA Journal of Pharmaceutical Science and Technology
 Downloaded from journal.pda.org on August 23, 2010
magnifica (Diptera: Sarcophagidae). Int. J. Syst.
Evol. Microbiol. 2008, 58 (Pt. 4), 976 –981.
8. Grünig, C. R.; Sieber, T. N.; Holdenrieder, O.
Characterisation of dark septate fungi (DSE) us-
ing inter-simple-sequence-repeat-anchored poly-
merase chain reaction (ISSR-PCR) amplification.
Mycological Research, 2001, 105 (1), 24 –32.
9. Bridge, P. D.; Pearce, D. A.; Rivera, A.; Ruther-
ford, M. A. VNTR derived oligonucleotides as
PCR primers for population studies in filamentous
fungi. Letters in Applied Microbiology 1997, 24
(5), 426 – 430.
10. Morgan, C. A.; Bigeni, P.; Herman, N.; Gauci,
M.; White, P. A.; Vesey, G. Production of precise
microbiology standards using flow cytometry and
freeze drying. Cytometry 2004, 62 (2), 162–168.
11. Brenner, D. J.; Staley, J. T.; Krieg, N. R. Classi-
fication of procaryotic organisms and the concept
of bacterial speciation: The Proteobacteria. Part
A. Introductory Essays. In Bergey’s Manual of Sys-
tematic Bacteriology; Springer, 2005; pp 27–32.
12. Healy, M.; Huong, J.; Bittner, T.; Lising, M.;
Frye, S.; Raza, S.; Schrock, R.; Manry, J.; Ren-
wick, A.; Nieto, R.; Woods, C.; Versalovic, J.;
Lupski, J. R. Microbial DNA typing by automated
Vol. 64, No. 2, March–April 2010
repetitive-sequence-based PCR. J. Clin. Micro-
biol. 2005, 43 (1), 199 –207.
13. Lehner, A. F.; Harvey, S.; Hill, C. W. Mapping
and spacer identification of rRNA operons of Sal-
monella typhimurium. J. Bacteriol. 1984, 160 (2),
682– 686.
14. Tindall, B. J.; Grimont, P. A.; Garrity, G. M.;
Euzéby, J. P. Nomenclature and taxonomy of the
genus Salmonella. Int. J. Syst. Evol. Microbiol.
2005, 55 (1), 521–524.
15. Condon, C.; Liveris, D.; Squires, C.; Schwartz, I.;
Squires, C. L. rRNA operon multiplicity in Esch-
erichia coli and the physiological implications of
rrn inactivation. J. Bacteriol. 1995, 177 (14),
4152– 4156.
16. Ryzhov, V.; Fenselau, C. Characterization of the
protein subset desorbed by MALDI from whole
bacterial cells. Anal. Chem. 2001, 73 (4), 746 –
750.
17. Varga, J.; Kocsubé, S.; Tóth, B.; Frisvad, J. C.;
Perrone, G.; Susca, A.; Meijer, M.; Samson, R. A.
Aspergillus brasiliensis sp. nov., a biseriate black
Aspergillus species with world-wide distribution.
Int. J. Syst. Evol. Microbiol. 2007, 57 (8), 1925–
1932.
155
 Downloaded from journal.pda.org on August 23, 2010

An Authorized User of the electronic PDA Journal of Pharmaceutical Science and

Technology (the PDA Journal) is a PDA Member in good standing. Authorized Users are
permitted to do the following:

·Search and view the content of the PDA Journal

·Download a single article for the individual use of an Authorized User
·Assemble and distribute links that point to the PDA Journal
·Print individual articles from the PDA Journal for the individual use of an Authorized User
·Make a reasonable number of photocopies of a printed article for the individual use of an
Authorized User or for the use by or distribution to other Authorized Users

Authorized Users are not permitted to do the following:

·Except as mentioned above, allow anyone other than an Authorized User to use or access the
PDA Journal
· Display or otherwise make any information from the PDA Journal available to anyone other
than an Authorized User
·Post articles from the PDA Journal on Web sites, either available on the Internet or an Intranet,
or in any form of online publications
·Transmit electronically, via e-mail or any other file transfer protocols, any portion of the PDA
Journal
·Create a searchable archive of any portion of the PDA Journal
·Use robots or intelligent agents to access, search and/or systematically download any portion
of the PDA Journal
·Sell, re-sell, rent, lease, license, sublicense, assign or otherwise transfer the use of the PDA
Journal or its content
·Use or copy the PDA Journal for document delivery, fee-for-service use, or bulk reproduction or
distribution of materials in any form, or any substantially similar commercial purpose
·Alter, modify, repackage or adapt any portion of the PDA Journal
·Make any edits or derivative works with respect to any portion of the PDA Journal including any
text or graphics
·Delete or remove in any form or format, including on a printed article or photocopy, any
copyright information or notice contained in the PDA Journal