Insights on Evolution of Virulence and Resistance from the Complete Genome Analysis of an Early Methicillin-Resistant Staphylococcus aureus Strain and a Biofilm-Producing Methicillin-Resistant Staphylococcus epidermidis Strain (original) (raw)

J Bacteriol. 2005 Apr; 187(7): 2426–2438.

Steven R. Gill,1,* Derrick E. Fouts,1 Gordon L. Archer,2 Emmanuel F. Mongodin,1 Robert T. DeBoy,1 Jacques Ravel,1 Ian T. Paulsen,1 James F. Kolonay,1 Lauren Brinkac,1 Mauren Beanan,1 Robert J. Dodson,1 Sean C. Daugherty,1 Ramana Madupu,1 Samuel V. Angiuoli,1 A. Scott Durkin,1 Daniel H. Haft,1 Jessica Vamathevan,1 Hoda Khouri,1 Terry Utterback,1,3 Chris Lee,1 George Dimitrov,1 Lingxia Jiang,1 Haiying Qin,1 Jan Weidman,1 Kevin Tran,1 Kathy Kang,1 Ioana R. Hance,1 Karen E. Nelson,1 and Claire M. Fraser1

Steven R. Gill

The Institute for Genomic Research,1 J. Craig Venter Science Foundation Joint Technology Center, Rockville, Maryland,3 Division of Infectious Diseases, Virginia Commonwealth University Health System, Richmond, Virginia2

Derrick E. Fouts

Gordon L. Archer

Emmanuel F. Mongodin

Robert T. DeBoy

Jacques Ravel

Ian T. Paulsen

James F. Kolonay

Lauren Brinkac

Mauren Beanan

Robert J. Dodson

Sean C. Daugherty

Ramana Madupu

Samuel V. Angiuoli

A. Scott Durkin

Daniel H. Haft

Jessica Vamathevan

Hoda Khouri

Terry Utterback

Chris Lee

George Dimitrov

Lingxia Jiang

Haiying Qin

Jan Weidman

Kevin Tran

Kathy Kang

Ioana R. Hance

Karen E. Nelson

Claire M. Fraser

*Corresponding author. Mailing address: Microbial Genomics, The Institute for Genomic Research, 9712 Medical Center Dr., Rockville, MD 20850. Phone: (301) 795-7572. Fax: (301) 838-0208. E-mail: gro.rgit@lligrs.

Received 2004 Jul 27; Accepted 2004 Dec 13.

Supplementary Materials

[Supplemental material]

GUID: 0CD97B74-3D2E-4C39-9473-F12509489DF5

GUID: 04DA3ED7-D1E7-42FA-8974-6361344E00D2

GUID: EB12D9E4-1CFB-40CD-B201-D8CF0F4377C7

Abstract

Staphylococcus aureus is an opportunistic pathogen and the major causative agent of numerous hospital- and community-acquired infections. Staphylococcus epidermidis has emerged as a causative agent of infections often associated with implanted medical devices. We have sequenced the ∼2.8-Mb genome of S. aureus COL, an early methicillin-resistant isolate, and the ∼2.6-Mb genome of S. epidermidis RP62a, a methicillin-resistant biofilm isolate. Comparative analysis of these and other staphylococcal genomes was used to explore the evolution of virulence and resistance between these two species. The S. aureus and S. epidermidis genomes are syntenic throughout their lengths and share a core set of 1,681 open reading frames. Genome islands in nonsyntenic regions are the primary source of variations in pathogenicity and resistance. Gene transfer between staphylococci and low-GC-content gram-positive bacteria appears to have shaped their virulence and resistance profiles. Integrated plasmids in S. epidermidis carry genes encoding resistance to cadmium and species-specific LPXTG surface proteins. A novel genome island encodes multiple phenol-soluble modulins, a potential S. epidermidis virulence factor. S. epidermidis contains the cap operon, encoding the polyglutamate capsule, a major virulence factor in Bacillus anthracis. Additional phenotypic differences are likely the result of single nucleotide polymorphisms, which are most numerous in cell envelope proteins. Overall differences in pathogenicity can be attributed to genome islands in S. aureus which encode enterotoxins, exotoxins, leukocidins, and leukotoxins not found in S. epidermidis.

The staphylococci are a diverse group of bacteria that cause diseases ranging from minor skin infections to life-threatening bacteremia. In spite of large-scale efforts to control their spread, they persist as a major cause of both hospital- and community-acquired infections worldwide. In the hospital setting alone, they are responsible for upwards of one million serious infections per year (41). The two major opportunistic pathogens of this genus, Staphylococcus aureus and Staphylococcus epidermidis, colonize a sizable portion of the human population. The predominant species, S. epidermidis, is fairly widespread throughout the cutaneous ecosystem, whereas S. aureus is carried primarily on mucosal surfaces. Within this context, staphylococci generally have a benign symbiotic relationship with their host. However, breach of the cutaneous organ system by trauma, inoculation needles, or direct implantation of medical devices enables the staphylococci to gain entry into the host and acquire the role of a pathogen. S. epidermidis is primarily associated with infections of implanted medical devices, such as prosthetic heart valves and joint prostheses (49). On the other hand, S. aureus is a more aggressive pathogen, causing a range of acute and pyogenic infections, including abscesses, bacteremia, central nervous system infections, endocarditis, osteomyelitis, pneumonia, urinary tract infections, chronic lung infections associated with cystic fibrosis, and several syndromes caused by exotoxins and enterotoxins, including food poisoning and scalded skin and toxic shock syndromes (32, 41).

Successive acquisition of resistance to most classes of antimicrobial agents, such as penicillins, macrolides, aminoglycosides, chloramphenicol, and tetracycline has made treatment and control of staphylococcal infections increasingly difficult. The widespread use of methicillin and other semisynthetic penicillins in the late 1960s led to the emergence of methicillin-resistant S. aureus (MRSA) and S. epidermidis (MRSE), which continue to persist in both the health care and community environments (45). Currently, greater than 60% of S. aureus isolates are resistant to methicillin and some strains have developed resistance to more than 20 different antimicrobial agents (40). The remaining effective therapy against most strains of multidrug-resistant staphylococci, including MRSA and MRSE, is the glycopeptide antibiotic vancomycin (51). However, the emergence in 1997 (6) of S. aureus with intermediate levels of resistance to vancomycin (vancomycin-intermediate S. aureus) and the most recent emergence of S. aureus with high levels of resistance to vancomycin (vancomycin-resistant S. aureus) (7) has limited its effectiveness. Finally, the increasing incidence of hypervirulent community-acquired S. aureus (45, 48) has become a major concern to the global health community and reinforced the critical need for new methods of control and treatment.

We have determined the complete genome sequences of S. aureus COL, an early MRSA isolate, and S. epidermidis RP62a, an MRSE biofilm-producing clinical isolate. Comparison of these genomes with other sequenced staphylococcal genomes provides insights into genome features that contribute to increasing pathogenicity in S. aureus and has led to the identification of novel genome islands in S. epidermidis that may contribute to the evolution of this species from a commensal pathogen to a more aggressive pathogen.

MATERIALS AND METHODS

Bacterial sources.

S. aureus COL was obtained from Brian Wilkinson (Illinois State University), who has maintained the culture as a frozen stock since 1976. The COL strain was reportedly isolated as a penicillinase-negative strain in the early 1960s from the operating theatre in a hospital in Colindale, England (17, 43). COL was one of the first MRSA isolates to be identified and has been used extensively in biochemical investigations of methicillin and vancomycin resistance (16).

S. epidermidis RP62a (ATCC 35984) is a slime-producing strain isolated during the 1979 to 1980 Memphis, Tennessee, outbreak of intravascular catheter-associated sepsis (9, 10). RP62a is capable of accumulated growth and subsequent biofilm formation, which contribute to its pathogenicity in foreign-body infections (22).

Other strains used for comparative genomic analyses are S. aureus Mu50 (28), N315 (28), and MW2 (3) and S. epidermidis ATCC 12228 (54). Mu50 is a clinical MRSA strain isolated in 1996 from a Japanese patient with infection of a surgical incision site which was resistance to vancomycin therapy (19, 20). N315 is a Japanese clinical MRSA isolate identified in 1982 (28, 37). MW2 is a highly virulent community MRSA strain isolated in 1998 from a 16-month-old girl in North Dakota and initially associated with four pediatric deaths in Minnesota and North Dakota (3, 5). S. epidermidis ATCC 12228 is a non-biofilm-forming reference strain also isolated in the United States (2, 54).

Genome sequencing, assembly, and ORF prediction.

S. aureus strain COL and S. epidermidis strain RP62a were sequenced to closure by the random shotgun method, with cloning, sequencing, and assembly completed as described previously for genomes sequenced at The Institute for Genomic Research (TIGR) (39). One small-insert plasmid library (2.0 to 3.0 kb) and one medium-insert plasmid library (10 to 12 kb) was constructed for each strain by random mechanical shearing and cloning of genomic DNA. In the initial random-sequencing phase, eightfold sequence coverage was achieved from the two libraries (one sequenced to fivefold coverage and the other sequenced to threefold coverage). The sequences from the respective strains were assembled separately with TIGR Assembler or Celera Assembler (www.tigr.org). All sequence and physical gaps were closed by editing the ends of sequence traces, primer walking on plasmid clones, and combinatorial PCR, followed by the sequencing of the PCR product.

An initial set of open reading frames (ORFs) that likely encode proteins was identified with GLIMMER (14), and those shorter than 90 bp as well as some of those with overlaps were eliminated. A region containing the likely origin of replication was identified, and bp 1 was designated adjacent to the dnaA gene, located in this region. All ORFs were searched against a nonredundant protein database as previously described (39). Frameshifts and point mutations were detected and corrected where appropriate. The remaining frameshifts and point mutations are considered authentic, and the corresponding regions were annotated as authentic frameshift or authentic point mutation, respectively. The ORF prediction and gene family identifications were completed by methodology described previously (39). Two sets of hidden Markov models (HMMs) were used to determine ORF membership in families and superfamilies. These included 721 HMMs from Pfam, version 2.0, and 631 HMMs from the TIGR ortholog resource. TMHMM (27) was used to identify membrane-spanning domains in proteins.

Comparative genomics.

For the identification of species-specific and strain-specific genes, all predicted ORFs from the TIGR-sequenced staphylococcal genomes (S. aureus COL and S. epidermidis RP62a) and published staphylococcal genomes (S. aureus N315 and Mu50 (28), S. aureus MW2 (3), and S. epidermidis ATCC 12228 (54) were searched against an in-house database composed of 195 prokaryotic, 8 eukaryotic, 175 phage, 63 virus, and 46 plasmid genomes with WU-BLASTP (http://BLAST.wustl.edu). Those genes that matched a nonself genomic sequence at a P value of ≤10−5, an identity of ≥35%, and match lengths of at least 75% of the length of both query and subject sequences were considered nonunique. These comparisons were used to generate match tables (see Supplemental Table 6 [http://www.tigr.org/tdb/staphylococcus; all supplemental tables are at this website]). Single nucleotide polymorphisms (SNPs) were identified by comparing the genome of S. aureus COL to those of S. aureus N315, Mu50, and MW2 and by comparing the genome of S. epidermidis RP62a to that of S. epidermidis ATCC 12228 with MUMer (15). Because we did not have access to underlying sequence data for published staphylococcal genomes, identification of SNPs was based on the final draft sequence. By mapping the position of the SNP to the annotation in the S. aureus COL and S. epidermidis RP62a genomes, it was possible to determine the location of the SNP (intergenic versus intragenic) and its effect on the deduced polypeptide (synonymous versus nonsynonymous). For each deduced polypeptide, the degree of relatedness across strains was calculated by using a BLAST score ratio. The BLASTP raw score was obtained for the alignment against itself (REF_SCORE) and the most similar protein in the query strains (QUE_SCORE). Scores were normalized by dividing the QUE_SCORE for each query genome by REF_SCORE. Normalized scores were plotted as xy coordinates.

A comparative database of all staphylococcal ORFs was generated for position effect determination by identifying all matches among the six sequenced genomes by a BLAST-Extend-Repraze (BER) search (P < 0.1; bit score > 50). These BER matches were then run through Position Effect software (TIGR) to determine conservation of gene order. The query and hit genes from each match were defined as anchor points in gene sets composed of adjacent genes, with up to 10 genes upstream and downstream from each anchor gene used in creating the gene sets. An optimal alignment between the ordered gene sets was calculated by using percent similarity from BER and applying a linear gap penalty of 100. Positive-scoring optimal alignments containing gene sets of four or more matching genes were stored in the database.

Nucleotide sequence accession numbers.

Nucleotide sequences for S. aureus COL (accession numbers CP000046 for the chromosome sequence and CP000045 for the plasmid sequence) and S. epidermidis (accession numbers CP000029 for the chromosome sequence and CP000028 for the plasmid sequence) have been deposited at GenBank. The genome sequences and the annotation of the TIGR-sequenced strains are available in the TIGR Comprehensive Microbial Resource at www.tigr.org. The S. aureus COL SAXXXX and S. epidermidis RP62a SEXXXX locus numbers are listed as SACOLXXXX and SERPXXXX, respectively, in GenBank.

RESULTS AND DISCUSSION

S. aureus is one of the leading causes of infectious disease in hospital settings and, recently, is an increasing cause of disease in the community (45, 48). Since the emergence of MRSA in the 1970s, S. aureus has continued to acquire additional antimicrobial resistance factors to the point where some isolates are resistant to more than 20 different antimicrobial agents (40). Development of antimicrobial resistance factors along with additional virulence factors and their movement through this species have likely occurred through gene transfer mediated by mobile genome islands, bacteriophage, plasmids, transposons, and insertion sequences (IS). The most recent example of such gene movement is the acquisition of the Enterococcus faecalis Tn_1546_ vancomycin resistance element by plasmid-mediated transfer into S. aureus (52). S. epidermidis, the less-virulent member of this genus frequently associated with hospital-acquired and biomedical device infections, has also acquired multiple resistance factors through similar processes. It is likely that gene transfer among multiple members of the staphylococcal species is a frequent event, allowing for adaptation to shifting host environments.

Genome features and islands.

General genome features of S. aureus COL and S. epidermidis RP62a, along with those of S. aureus N315 (28), Mu50 (28), and MW2 (3) and S. epidermidis ATCC 12228 (54) are presented in Supplemental Table 1 (http://www.tigr.org/tdb/staphylococcus). The genome sequences of two clinical S. aureus isolates, MSSA476 and MRSA252, were published (21) just prior to submission of the manuscript and were not included in our whole-genome comparisons. Significant aspects of MSSA476 and MRSA252 are, however, included within the following results and discussion. Whole-genome analysis indicated that the genomes of S. aureus and S. epidermidis are syntenic throughout a well-conserved core region (data not shown), with differences the result of genomic elements including genome islands (νSa, νSe, SSC_mec_, and staphylococcus cassette chromosome [SSC]-like elements), integrated prophage, IS elements, composite transposons, and integrated plasmids (Table 1) which are associated with disease and virulence. These genomic elements make up approximately 7% of the S. aureus COL genome and 9% of the S. epidermidis RP62a genome, percentages that are similar to those for other gram-positive pathogens, such as group A streptococcus (∼10%) (4) but lower than that for Enterococcus faecalis (25%) (39).

TABLE 1.

Genomic islands in six sequenced staphylococcal genomes

Inland	Genes found on island_b_	Type, location of island (predicted integration sequence_a_) for:	Allele family member(s)
S. aureus	S. epidermidis
COL	Mu50	N315	MW2	RP62A	ATCC 12228
vSa1	Enterotoxin genes (seb, tsst, ear)	903332-919283 (A)	NP	NP	NP	NP	NP	SaPI1, SaPI3
vSa2	sec, tsst	NP	NP	NP	NP	NP	NP	SaPIbov
vSa3	Type I, fhuD; type II, sel2, sec4, ear	NP	I, 868373-882872 (B)	NP	II, 839358-853808 (B)	NP	NP	SaPI3
vSa4	Type I, sel, sec3, tsst; type II, four unknown ORFs	II, 2072899-2076041 (C)	I, 2148912-2133235 (D)	I, 2056679-2072358 (D)	II, 2097809-2100950 (C)	NP	NP	SaPI2
vSaα	Type I, set6-15, lpl1-9; type II, set16-26, lpl10-14, type III, set1-5, lpl2,7,8,11,13	III, 465424-489723	I, 461919-491326	I, 436162-466813	II, 416307-452099	NP	NP
vSaβ	Type I, splA-F, lukDE, ear, epidermin gene cluster; type II, splA-F, lukDE, ear, seg, sen, sei, sem, seo, epidermin gene cluster	I, 1902466-1938731	II, 1932523-1961464	I, 1854608-1881615	I, 1890800-1922552	NP	NP
vSaγ	set, eta, _psm_β	1173206-1193358	1208629-1230219	1132235-1153775	1133469-1153549	NP	NP
vSeγ	_psm_β	734975-737635	840908-843572
SCC_mec_	mecA, ermA, bleO, addD	I, 68085-34173 (E)	II, 87085-34158 (E)	II, 87119-34153 (E)	IV, 58278-34150 (E)	II, 2536574-2584194 (F)	NP	Types I, II, III, IVa, IVb
SSC_pbp4_	pbp4, merT, merB, merA, tagF	NP	NP	NP	NP	NP	32031-99960 (G)
φSa1	72 ORFs	NP	917453-962005 (H)	NP	NP	NP	NP	φ11, φETA
φSa2	lukS-PV, lukF-PV	NP	NP	NP	1575042-1529123 (I)	NP	NP	φSLT
φSa3	sak, sea, sep, seg2, sek2	NP	2126304-2083238 (J)	2049591-2005321 (J)	2088820-2046205 (J)	NP	NP	φ42
φCOL	Integrated in geh	354674-398267 (K)	NP	NP	NP	NP	NP	φL-54
φSPβ	Integrated in yeeE, LPXTG surface protein gene	NP	NP	NP	NP	1567637-1694334 (L)	NP
Tn_5801_	tetM, two unknown ORFs	NP	I, 436002-461791 (M)	NP	II, 415784-417798 (M)	NP	NP
vSe1	Type I, cadCD; type II, one unknown ORF	NP	NP	NP	NP	I, 2250348-2267496 (N)	II, 2255537-2259992 (N)
vSe2	srtA, two LPXTG surface protein genes	NP	NP	NP	NP	NP	1558081-1519667 (O)

Seven pathogenicity genomic islands (νSa), in positions conserved across all sequenced genomes, have been identified in S. aureus (Table 1 and Fig. 1). These islands carry approximately one-half of the S. aureus toxins or virulence factors, and allelic variation of these genes, along with presence or absence of individual νSa, contributes to the pathogenic potential of this species (Table 2). For example, island νSa3 is unique to S. aureus MW2 and carries allelic forms of enterotoxin genes sel2 and sec4, which may contribute to its increased virulence. On the other hand, S. aureus MRSA252 (21) has a novel island, SaPI4, that contains homologs of pathogenicity proteins found in previously characterized νSa1 and νSa2 islands (Table 1) but does not carry known virulence genes. Our analysis identified a novel genomic island, νSaγ (νSeγ), that is found in all S. aureus and S. epidermidis genomes (Table 1 and Fig. 2). The S. epidermidis νSeγ allele contains genes for a cluster of four members of the phenol-soluble modulin (PSM) family, a potential virulence factor of S. epidermidis (38, 50). The S. aureus νSaγ allele contains a cluster of two PSM genes and a small secondary cluster of exotoxin genes similar to those in νSaα. Our analysis of S. epidermidis RP62a and ATCC 12228 also identified two integrated plasmids, νSe1 and νSe2 (Table 1 and Fig. 2; Supplemental Table 2), which contain prophage integrase genes in a structure similar to that for S. aureus genome islands. While neither νSe1 or νSe2 carries virulence factors found in S. aureus, the νSe1 island in RP62a contains genes for cadmium resistance and the νSe2 island in ATCC 12228 encodes a second strain-specific sortase (encoded by srtC) not found in other staphylococci and two strain-specific LPXTG cell surface attachment proteins with likely roles in adhesion to host tissue (Supplemental Table 2).

Circular representation of the sequenced S. aureus and S. epidermidis genomes. Each concentric circle represents genomic data for S. aureus (A) and S. epidermidis (B) and is numbered from the outermost circle to the innermost circle. The outermost circles indicate the genome coordinates in base pairs. The second and third circles represent the predicted S. aureus COL and S. epidermidis RP62a ORFs on the plus and minus strands, respectively, colored by role categories: salmon, amino acid biosynthesis; light blue, biosynthesis of cofactors and prosthetic groups and carriers; light green, cell envelope; red, cellular processes; brown, central intermediary metabolism; yellow, DNA metabolism; green, energy metabolism; purple, fatty acid and phospholipid metabolism; pink, protein fate and synthesis; orange, purines, pyrimidines, nucleosides, and nucleotides; blue, regulatory functions; grey, transcription; teal, transport and binding proteins; black, hypothetical and conserved hypothetical proteins. The fourth (strain COL), fifth (strain Mu50), sixth (strain N315), and seventh (strain MW2) circles in S. aureus and the fourth (strain RP62a) and fifth (strain ATCC 12228) circles in S. epidermidis indicate genome islands involved in virulence (red or yellow), regulatory loci (agr) (blue-green), prophage (black), SSC_mec_ (blue), SSC_pbp4_ (orange), integrated plasmids (pink), STAR elements (brown), transposable elements (dark green), and CRISPR regions (light blue). The 8th (strain Mu50), 9th (strain N315), and 10th (strain MW2) circles in S. aureus and the sixth (strain ATCC 12228) circle in S. epidermidis represent the number of SNPs per 5 kb compared to S. aureus strain COL and S. epidermidis strain RP62a. Gold ticks, 1 to 75 SNPs; red ticks, 76 to 200 SNPs; dark green ticks, 201 to 300 SNPs; blue ticks, more than 301 SNPs. Complete DNA sequence and annotation for S. aureus MW2, N315, and Mu50 and S. epidermidis ATCC 12228 were obtained from GenBank accession numbers BA000033, BA000018, BA000017, and AE015929, respectively.

Novel integrated plasmids and genome islands in S. aureus and S. epidermidis. Shown are the schematic diagrams of integrated plasmids of νSe1 in S. epidermidis RP62a (A), νSe2 in S. epidermidis ATCC 12228 (B), νSaγ in S. aureus COL (C), and νSeγ in S. epidermidis RP62a (D). ORFs are marked in the direction of transcription as arrows and are colored according to functional categories as indicated. Positions within the respective genomes are indicated as genome coordinates on the ends of each schematic. Putative functions of selected ORFs accompany the gene locus number. Putative functions of all ORFs and their locus numbers are presented in Supplementary Table 2.

TABLE 2.

Virulence factors and genomic islands in six sequenced staphylococcal genomes

Virulence factor	S. aureus	S. epidermidis
COL	Mu50	N315	MW2	RP62a	ATCC 12228
Locus and gene	Island	Locus and gene	Island	Locus and gene	Island	Locus and gene	Island	Locus and gene	Island	Locus and gene	Island
Enterotoxin	SA1657 sea	SAV2009 sec3	vSa4	SA1430 sea d	MW1889 sea	φSa3
SA0907 seb	vSa1	SAV1824 seg	vSaβ	SA1817 sec3	vSa4	MW0759 sec2	vSa3
SA0887 sei	vSal	SAV1828 sei	vSaβ	SA1642 seg	vSaβ	MW1937 seg	φSa3
SA0886 sek	vSal	SAV2008 sel	vSa4	SA1646 sei	vSaβ	MW0051 seh
SAV1829 sem	vSaβ	SA1816 sel	vSa4	MW0760 sel	vSa3
SAV1825 sen	vSaβ	SA1647 sem	vSaβ	MW1938 sek	φSa3
SAV1830 seo	vSaβ	SA1643 sen	vSaβ	MW0052 seo d
SAV1948 sep	φSa3	SA1648 seo	vSaβ
SAV1601 sep	SA1761 sep	φSa3
SAV1827 seu	vSaβ	SA1645 yent1	vSaβ
SA1644 yent2	vSaβ
Exotoxin	SA1178 set1	vSaγ	SAV0422 set6	vSaα	SA1009 set1	vSaγ	MW1047 set1	vSaγ
SA0469 set1	vSaα	SAV0423 set7	vSaα	SA0357 set2	MW1048 set4	vSaγ
SA0468 set3	vSaα	SAV0424 set8	vSaα	SA1011 set3	vSaγ	MW0382 set16	vSaα
SA0478 set3	vSaα	SAV0425 set10	vSaα	SA1010 set4	vSaγ	MW0383 set17	vSaα
SA1180 set3	vSaγ	SAV0426 set11	vSaα	SA0382 set6	vSaα	MW0384 set18	vSaα
SA0474 set4	vSaα	SAV0427 set12	vSaα	SA0383 set7	vSaα	MW0385 set19	vSaα
SA1179 set4	vSaγ	SAV1168 set12	vSaγ	SA0384 set8	vSaα	MW0386 set20	vSaα
SA0473 set5	vSaα	SAV0428 set13	vSaα	SA0385 set9	vSaα	MW0387 set21	vSaα
SA0470 set d	vSaα	SAV0429 set14	vSaα	SA0386 set10	vSaα	MW0388 set22	vSaα
SA0472 set d	vSaα	SAV0433 set15	vSaα	SA0387 set11	vSaα	MW0389 set23	vSaα
SA0388 set12	vSaα	MW0390 set24	vSaα
SA0389 set13	vSaα	MW0391 set25	vSaα
SA0390 set14	vSaα	MW0394 set26	vSaα
SA0393 set15	vSaα	MW0345 set d
Exfoliative toxin	SA1184 et d	vSaγ	SAV1173 eta	vSaγ	SA1016 eta	vSaγ	MW1054 eta	vSaγ
Toxic shock syn- drome toxin	SAV2011 tsst	vSa4	SA1819 tsst	vSa4
Esterase	SA2345	SAV2350_d_	SA2140_d_	MW2271_d_	SE1941	SE1929
SA2549_d_	SAV2535_d_	SA2323_d_	MW2456_d_	SE2109_d_	SE2095
Serine protease	SA1869 splA	vSaβ	SAV1813 splA	vSaβ	SA1631 splA	vSaβ	MW1755 splA	vSaβ
SA1868 splB	vSaβ	SAV1812 splB	vSaβ	SA1630 splB	vSaβ	MW1754 splB	vSaβ
SA1867 splC	vSaβ	SAV1811 splC	vSaβ	SA1629 splC	vSaβ	MW1753 splC	vSaβ
SA1866 splD	vSaβ	SAV1810 splD	vSaβ	SA1628 splD	vSaβ	MW1752 splF	vSaβ
SA1865 splE	vSaβ	SAV1809 splF	vSaβ	SA1627 splF	vSaβ	MW0903 htrA	vSaβ
SA1864 splF	vSaβ	SAV1023 htrA	SA0879 htrA	MW1670 htrA
SA1028 htrA	SE2390 htrA	SE0722 htrA
SA1777 htr d	SE2401	SE0723
Staphylokinase	SAV1944	φSa3	SA1758	φSa3	MW1885	φSa3
Serine V8 protease	SA1057 sspA	SAV1048 sspA	SA0901 sspA	MW0932 sspA	SE1397 sspA	SE1543 sspA
Cysteine protease	SA1056 sspB	SAV1047 sspB	SA0900 sspB	MW0931 sspB	SE2390 sspB	SE0184 sspB
SA1970 sspB	SAV1046 sspC	SA0899 sspC	MW0930 sspC	SE2391 sspC	SE0183 sspC
SA1055 sspC
Lipase	SA2694 lip	SAV2671 lip	SA2463 lip	MW2590 lip	SE2336 lip	SE0245 lip
SA0317 geh (amino)a	SAV0320 geh	SA0309 geh	MW0297 geh	SE0018 geh	SE2403 geh
SA0390 geh (carboxy)a	SE2297 geh1, gehC	SE0281 geh1, gehC
SE2388 geh2, gehD	SE0185 geh2, gehD
Lipase/esterase	SA0712 lipA d	SAV0655 lipA d	SA0610 lipA d	MW0617 lipA d	SE0309 lipA	SE0424 lipA
Extracellular elas- tasc precursor	SE2252 sepA	SE2219 sepA
Leukotoxin D	SA1880 lukD	vSaβ	SAV1819 lukD	vSaβ	SA1637 lukD	vSaβ	MW1767 lukD	vSaβ
Leukotoxin E	SA1881 lukE	vSaβ	SAV1820 lukE	vSaβ	SA1638 lukE	vSaβ	MW1768 lukE	vSaβ
Synergohymeno- tropic toxin	MW1379 lukSPV	φSa2
Leukocidin F	SA1812 lukF	MW1378 lukFPV	φSa2
Leukocidin M	SA2006 lukM	SA1813 lukM	MW1942 lukM
Alpha hemolysin	SA1173 hly	vSaγ	SAV1163 hly	vSaγ	SA1007 hly	vSaγ	MW0955 hly	vSaγ
Beta hemolysin	SA2003 hlb	SAV2003 hlb e	SA1752 hlb e	MW1940 hlb e	SE2544 hlb	SE0008 hlb
Delta hemolysin	SA2022 hld	SAV2035 hld	SAS065 hld	MW1959 hld	SE1489 hld	SE1634 hld
Gamma hemolysin, component A	SA2419 hlgA	SAV2419 hlgA	SA2207 hlgA	MW2342 hlgA
Gamma hemolysin, component C	SA2421 hlgC	SAV2420 hlgC	SA2208 hlgC	MW2343 hlgC
Gamma hemolysin, component B	SA2422 hlgB	SAV2421 hlgB	SA2209 hlgB	MW2344 hlgB
Hemolysin III	SA2160_d_	SAV2170	SA1973_d_	MW2096_d_	SE1769_d_	SE1760_d_
Hemolysin	SA0762_d_	SAV0919	SA0780_d_	MW0664_d_	SE2258_d_	SE2226_d_
Hyaluronate lyase	SA2194 hysA	SAV2202 hysA	SA2003 hysA	MW2129 hysA
Thermonuclease nuclease	SA1357 nuc	SAV1324 nuc	SA1160 nuc	MW1211 nuc	SE0891 nuc	SE1004 nuc
SA0860	SAV0815	SA0746	MW0769	SE1570	φSPβ
Cell wall hydrolase	SA1264 lytN	SAV1247 lytN	SA1090 lytN	MW1130 lytN
Zinc metalloprotense	SA1281	SAV1262	SA1105	MW1145	SE0829_d_	SE0938_d_
Clp protease, proco- lytic subunit	SA0833 clpP	SAV0768 clpP	SA0723 clpP	MW0730 clpP	SE0436 clpP	SE0551 clpP
Clp protease, ATP binding subunit	SA0979 clpB	SAV0975 clpB	SA0835 clpB	MW0857 clpB	SE0564 clpB	SE0674 clpB
Clp protease, ATP binding subunit	SA1721 clpX	SAV1674 clpX	SA1498 clpX	MW1618 clpX	SE1238 clpX	SE1349 clpX
Clp protease, ATP binding subunit	SA0570 clpC	SAV0523 clpC	SA2336 clpC	MW2469 clpC	SE0165 clpC	SE0287 clpC
Staphylococcal protein A Spa	SA0095 spa	SAV0111 spa	SA0107 spa	MW0084 spa
Phenol-soluble modulin	SA1186 beta	vSaγ	NT02SA1161 beta b	vSaγ	NT01SA1111 beta b	vSaγ	MW1056 beta	vSaγ	SE0736 beta1	vSeγ	SE0846 beta1	vSeγ
SA1187 beta	vSaγ	NT02SA1162 beta b	vSaγ	NT01SA1112 beta b	vSaγ	MW1057 beta	vSaγ	SE0737 beta1	vSeγ	SE0847 beta1	vSeγ
SA2022 hld c	SAV2035 hld c	SAS065 hld c	MW2121 hld c	SE0738 beta1	vSeγ	SE0848 beta1	vSeγ
SE0739 beta2	vSeγ	SE0849 beta2	vSeγ
SE2397 beta1	SE0177 beta
SE2400 beta1	SE0174 beta1
SE0083 alpha
SE1489 delta c	SE1634 delta c

Five types of integrated prophage were identified, with at least one phage in every genome except that of S. epidermidis ATCC 12228 (Table 1 and Fig. 1). In S. aureus COL, a L54-like phage (30), which we have named φCOL, was integrated near the 3′ end of the lipase gene (geh). The φSa3 phage, which is integrated into the beta-hemolysin gene (hlb) of S. aureus N315, Mu50, MW2, MRSA252, and MSSA476, was not found in S. aureus COL. A single Bacillus subtilis φSPβ-like phage (29) was identified in S. epidermidis RP62a (Table 1; Supplemental Table 3; see Fig. S1 in the supplemental material), where it is inserted in att sites within yeeE. Comparative genome hybridization of multiple S. epidermidis clinical isolates (S. Gill, unpublished data) shows that acquisition of the φSPβ-like phage is unique to RP62a and likely a recent event. The φSPβ-like phage is a mosaic structure carrying multiple staphylococcal IS elements and genes encoding a staphylococcal nuclease and an RP62a-specific LPXTG surface protein, indicating that multiple recombination events have likely occurred following entry of the phage into RP62a.

Three types of SCC_mec_ islands (types I, II, and IVa) (23, 24) were previously identified among the S. aureus COL, N315, Mu50, and MW2 genomes (Table 1; Supplemental Table 4; see Fig. S2 in the supplemental material). The SSC_mec_ islands are characterized by a set of site-specific recombinase genes (ccrA and ccrB) which promote site-specific integration into an att site within orfX and a mecA gene which encodes resistance to methicillin (1, 23). Our analysis of S. epidermidis RP62a identified a type II SSC_mec_ which is 98% identical at the nucleotide level and identical in gene organization (with the exception of the region from pUB110 flanked by IS_431mec)_ to that of the S. aureus type II SSC_mec_. Acquisition of additional transposon and IS elements, such as Tn_554_, by SSC_mec_ types II and III corresponds with the need for S. aureus to survive the increased use of antibiotics in clinical environments. Previous structural analysis of identified SSC_mec_ elements suggests that the ccrAB genes may form an independent mobile SSC element that mediates staphylococcal interspecies transfer of antimicrobial or virulence genes (25, 26). Our analysis of S. epidermidis ATCC 12228 has identified such an SSC element, named SSC_pbp4_ (also identified by Mongkolrattanothai et al. [34]), which lacks mec but which contains two pairs of ccrA1 and ccrB genes along with multiple IS elements, a restriction-modification system (hsdS and hsdM), and genes encoding penicillin binding protein 4 (pbp4) and resistance to mercury and cadmium (see Fig. S1 in the supplemental material). The presence of two ccrAB pairs and multiple putative att sites in addition to orfX suggests that SSC_pbp4_ is the result of two independent insertion events. The existence of SSC_pbp4_ in S. epidermidis and a novel SSC_mec_-like element (SSC_far)_ in the genome of MSSA476 (21) suggests that similar SSC elements capable of transferring virulence factors between S. aureus and S. epidermidis may already exist within these species.

The seven types of IS elements identified in S. aureus and S. epidermidis are randomly distributed throughout their genomes (Supplemental Table 1; Fig. 1). A new staphylococcal IS element, IS_Sep1_, was identified in both S. epidermidis genomes. Composite transposons Tn_554_ and Tn_4001_ were identified in S. aureus N315 and Mu50 and S. epidermidis RP62a, respectively, but not in S. epidermidis ATCC 12228 or S. aureus COL.

Multiple copies of the GC-rich STAR (S. aureus repeat element) signature sequence (11) were found dispersed throughout intergenic regions of the S. aureus and S. epidermidis genomes (Fig. 1). STAR elements are more abundant in S. aureus, but in neither species are they associated with regions of atypical genome composition or with predicted mobile genes. A single copy of the extragenic CRISPR (clustered regularly interspaced palindromic repeats) DNA repeat element (20, 21) was identified near the dnaA gene at the replication origin of the S. epidermidis RP62a genome (Supplemental Table 5; Fig. 1). CRISPRs were not identified in S. epidermidis ATCC 12228 or in the S. aureus genomes.

Comparative genomics and evolution of virulence.

A comparison of the six staphylococcal genomes against each other revealed (i) a total of 454 species-specific genes that are common to the S. aureus COL, N315, Mu50, and MW2 genomes but not found in S. epidermidis RP62a or ATCC 12228, (ii) a total of 286 species-specific genes that are common to the S. epidermidis RP62a and ATCC 12228 genomes but not found in the S. aureus genomes, (iii) 332 strain-specific genes that are found in S. epidermidis ATCC 12228 but not in S. epidermidis RP62a, and (iv) 346 strain-specific genes that are found in S. epidermidis RP62a but not in S. epidermidis ATCC 12228 (Supplemental Table 6). A core set of 1,681 genes common among all strains and both species was also identified (Supplemental Table 6). The majority of the unique genes can be accounted for by the presence or absence of prophage and genomic islands. For example, the 127-kb φSPβ-like prophage in S. epidermidis RP62a (Table 1; Fig. 1; see Fig. S1 in the supplemental material) represents approximately 5% of the RP62a genome. Similarly, νSa1 in S. aureus COL represents 0.5% of the genome and carries the staphylococcal enterotoxin B gene (seb), a major virulence factor.

Comparative analysis of the S. aureus isolates suggested variations in the evolutionary history of the pathogenicity islands, some of which appear to have been created as a result of integration and subsequent mobilization of resident prophage into other members of this species (31, 42). Movement of these islands, such as mobilization of νSa1 (SaP1) by phage 80α (31, 42), into multiple S. aureus isolates may enable them to evolve and grow through the acquisition of additional virulence genes. For example, of the seven identified pathogenicity islands (Table 1), vSa1 and vSa2 share conservation in gene order in strains COL and MW2, respectively. In S. aureus COL, however, SEB (seb) is encoded in the same position as the toxic shock syndrome toxin (encoded by tsst) in MW2, suggesting either gene displacement or independent acquisition events through phages. In νSa4, COL and MW2 share an att site into which four ORFs, including a phage integrase gene and remnants of phage have inserted. The same att site in N315 and Mu50 is occupied by a νSa4 which contains additional genes, including the enterotoxin K (sel), enterotoxin C (sec3), and tsst genes. The similarity of the phage integrases suggests that the same phage integrated in all strains but that the sek, entC, and tsst genes have been lost from COL and MW2.

Genome islands as vectors of virulence determinants in the staphylococci contrasts with what is observed in the other low-GC-content gram-positive pathogens for which we have complete genomes available. For example, the Listeria monocytogenes genomes demonstrate a high degree of synteny, with variations in the genomes due to extensive SNPs (35). This suggests that the adaptation to infectious diseases in this species relies on small but specific genomic differences. In Bacillus anthracis, the majority of differences across strains are also in SNPs (but to a smaller degree than in Listeria's_)_ and in the presence or absence of the anthrax toxin genes carried on plasmid pX02. By comparison, S. aureus seems to demonstrate variable capabilities of virulence, depending on a combination of both genome islands in the form of phage and pathogenicity islands, as well as the presence of SNPs (see below). These differences may be the most significant factor contributing to the successive acquisition of resistance, as well as virulence factors.

Acquisition of virulence factors also appears to occur as a result of plasmid-mediated gene transfer between staphylococci and other low-GC-content gram-positive pathogens. For example, our analysis of the S. epidermidis RP62a and ATCC 12228 genomes revealed the presence of a cap operon (capABC) and gamma-glutamyl transpeptidase gene (Fig. 3) similar to that found on the B. anthracis pX02 plasmid, where it encodes the polyglutamate capsule, which is essential for B. anthracis virulence. Experimental verification of a functional polyglutamate capsule in S. epidermidis remains to be done, but polyglutamate may play a role in the formation of S. epidermidis biofilms. Phylogenetic analysis of cap genes in the operon (Fig. 3) indicates that the acquisition of this locus may have been the result of a plasmid-mediated transfer event from an ancestor of the bacilli to S. epidermidis. However, note that a number of species-specific metabolic functions, such as acetoin dehydrogenase and polyphosphate synthesis, that are encoded by complete operons in S. epidermidis could also be the result of gene loss by a common ancestor.

Phylogenetic analysis and organization of the Cap operon in S. epidermidis RP62a. Homologs of S. epidermidis RP62a Cap operon (E) were identified by BLASTP of the WU-BLAST formatted database of all complete bacterial genomes. Each gene in the Cap operon was aligned against respective homologs with ClustalW, and phylogenetic trees were generated with Belvu for gamma-glutamyltranspeptidase (A), CapA (B), CapC (C), and CapB (D). Organization of the S. epidermidis RP62a and B. anthracis Cap operons is shown in schematics E and F, respectively. Positions within the respective genomes are indicted as genome coordinates on the ends of each schematic. BLASTP e values determined from homolog search results are shown between each ORF in schematics E and F. GenBank accession numbers are in parentheses and accompany all matches on the phylogenetic tree.

A total of 22,888, 22,160, and 19,599 SNPs were found in the genomes of S. aureus Mu50, N315, and MW2, respectively, compared to that of strain COL. Of these SNPs, 6,447, 6,088, and 5,390 resulted in a nonsynonymous (NS) change in amino acid sequence in strains Mu50, N315 and MW2, respectively (Table 3 and Fig. 1). A total of 10,297 SNPs were found in the genome of S. epidermidis ATCC 12228 compared to that of strain RP62a. Of these SNPs, 2,579 resulted in an NS change in amino acid sequence. In S. aureus, SNPs are clustered in genome islands and, when these are grouped by function, it is found that there are a higher number of SNPs making up the cell envelope than performing other functions (∼20% of total SNPs for Mu50, N315, and MW2) (Fig. 1; see Fig. S3 in the supplemental material). In S. epidermidis, although the majority of SNPs are found in genes for hypothetical proteins, a significant number (12.5% of the total SNPs for ATCC 12228) are in genes encoding proteins with cell envelope functions (Fig. 1; see Fig. S4 in the supplemental material). Variations in cell envelope or surface proteins, such as the LPXTG/NPQTN proteins and fibronectin binding proteins (encoded by fnbA and -B) likely reflect their immunogenicity and the high level of protective antibodies against these proteins which are present in human sera (18). Changes in amino acid sequence within highly immunogenic domains of these proteins may enable the bacteria to evade attack by the immune system.

TABLE 3.

SNPs in S. aureus b and S. epidermidis c

Type of SNP	No.d in
S. aureus	S. epidermidis ATCC 12228
Mu50	N315	MW2
Total	22,888	22,160	19,599	10,297
In-gene	17,966	17,352	15,358	7,938
Intergenic	4,922	4,808	4,241	2,359
Synonymous	11,519	11,264	9,968	5,359
Codon position 1	531	518	490	239
Codon position 2	2_a_	2_a_	4_a_	1_a_
Codon position 3	10,986	10,744	9,474	5,119
Nonsynonymous	6,447	6,088	5,390	2,579
Codon position 1	2,850	2,710	2,436	1,281
Codon position 2	2,072	1,927	1,739	826
Codon position 3	1,525	1,451	1,215	472

Although much is known about staphylococcal virulence, very little is known about the metabolism of staphylococci. Previous studies on the metabolism and physiology of these organisms have been limited, but the complete genome sequence has allowed for an increased understanding of the basic biology of these species. In addition to the previously identified pathways for the synthesis of various amino acids, we have identified pathways (Fig. 4) for the synthesis of the amino acids leucine, valine, aspartate, isoleucine, glycine, and methionine. Pathways that would enable growth on a range of simple and complex sugars via the glycolytic pathway, the phosphate pathway, and the tricarboxylic acid cycle were also identified. The mevalonate pathway, required for the synthesis of isopentenyl-PP, essential for cell wall biosynthesis, as well as menaquinones and ubiquinones, needed for electron transport, were also identified.

Overview of metabolism and transport in S. aureus and S. epidermidis. Pathways for energy production, metabolism of organic compounds, and synthesis of carotenoids are shown. Orange text, processes unique to S. aureus; green text, processes unique to S. epidermidis. Transporters are grouped by substrate specificity as follows: inorganic cations (green); inorganic anions (pink); carbohydrates and carboxylates (yellow); amino acids, peptides, amines, and purines and pyrimidines (red); and drug efflux and other (black). Question marks indicate uncertainty about the substrate transported. Export or import of solutes is designated by the direction of the arrow through the transporter. The energy-coupling mechanisms of the transporters are also shown: double-headed arrow, solutes transported by channel proteins; two arrows, secondary transporters, indicating both the solute and the coupling ion; single arrow, transporters with an unknown energy coupling mechanism. ATP-driven transporters are indicated by the ATP hydrolysis reactions. Components of transporter systems that function as multisubunit complexes that were not identified are outlined with dotted lines. Where multiple homologous transporters with similar substrate predictions exist, the number of that type of transporter is indicated in parentheses.

S. aureus is primarily an inhabitant of mucous membranes, and S. epidermidis is primarily an inhabitant of the skin surface; in both environments the organisms are likely to encounter osmotic stress. With respect to transport, S. aureus and S. epidermidis possess seven and eight predicted sodium ion/proton exchangers, respectively. Both organisms are well adapted for osmotic stress, with six transport systems for proline, glycine betaine, or other probable osmoprotectants (Fig. 4). Other transporters related to osmoregulation include the MscL and MscS mechanosensitive ion channels and two Trk potassium ion channels.

Probably the major difference between S. aureus and S. epidermidis in terms of transport is the absence of three PTS sugar transporters, for mannitol, sorbitol, and pentitols, and an ABC family maltose transporter from S. epidermidis. Both species have a variety of transporters for inorganic cations and anions (Fig. 4), but it appears that iron acquisition is a serious priority. Six complete or partial ferric iron ABC uptake systems, four additional orphan ferric iron binding proteins, and two ferrous iron FeoB uptake systems were identified in S. aureus. In comparison, there are three complete or partial ferric iron ABC uptake systems, two additional orphan ferric iron binding proteins, and two ferrous iron FeoB uptake systems in S. epidermidis. Both staphylococcal species include a large number of predicted drug efflux systems, including the previously described NorA fluoroquinolone transporter and a probable ortholog of the Lactococcus lactis LmrP multidrug efflux protein.

Pathogenicity and S. epidermidis as an emerging pathogen.

Genome-wide analysis of the six staphylococcal genomes revealed that approximately 11 and 7% of the total ORFs are predicted to encode cell surface proteins (Table 4) and secreted virulence factors (Table 2), respectively. Many surface proteins with essential roles in host colonization, biofilm formation, and evasion of host defense mechanisms have a common C-terminal LPXTG/NPQTN cell wall attachment motif and companion sortase processing enzymes (encoded by srtA and -B), which are conserved in all gram-positive bacterial pathogens (33, 44, 47). Our analysis of the LPXTG/NPQTN surface proteins revealed that only the accumulation-associated protein (encoded by aap) and most members of the Sdr gene family (sdrCDEFG) are functional homologs in both species. Those unique to each species likely reflect key differences in host tissue specificity and multifactorial adherence mechanisms used by S. aureus and S. epidermidis. S. epidermidis ATCC 12228 encodes a novel third sortase (encoded by srtC) not found in other staphylococci and most closely related to sortases of L. lactis and Streptococcus suis. Many of the secreted proteins (Table 2) have roles in multiple mechanisms for invasion of host tissue and evasion of host defense systems. The relative abundance of virulence factors in S. aureus compared to S. epidermidis reflects the propensity of S. aureus to cause fulminant and sometimes life-threatening infections, as opposed to the more subacute or chronic infections caused by S. epidermidis. For example, members of the enterotoxin and exotoxin (53) families (Tables 1 and 2) which function as superantigens and inducers of a proinflammatory cytokine response are unique to S. aureus and have not been identified in characterized isolates of S. epidermidis.

TABLE 4.

Surface proteins in S. aureus and S. epidermidis

Locus	Function_a_	Functional name	Gene(s)	SD_b_ present	LPXTG_c_	YSIRK_d_ present
S. aureus surface proteins_e_
SA0856	a	Clumping factor A	clfA	+	LPDTG	+
SA2652	a	Clumping factor B	clfB	+	LPETG	+
SA2511	a	Fibronectin binding protein A	fnbA	LPETG	+
SA2509	a	Fibronectin binding protein B	fnbB	LPETG	+
MW2612_g_	a	Collagen adhesion	cna	LPKTG
SA0608	a	SdrC	sdrC	+	LPETG	+
SA0609	a	SdrD	sdrD	+	LPETG	+
SA0610	a	SdrE	sdrE	+	LPETG	+
SA0095	a,b	Protein A	spa	LPETG	+
SA0050	e	Methicillin resistance surface protein	pls	+	LPDTG	+
SA2676	e	Cell wall surface anchor protein	sasA	LPDTG
SA2150	e	Phosphoglucomutase (FmtB/Mrp)	sasB	LPDTG	+
SA1806	e	Cell wall surface anchor protein	sasC	LPNTG	+
SA0119	e	Cell wall surface anchor protein	sasD	LPAAG
SA1140	d	Cell wall surface anchor protein	sasE (isdA)	LPKTG	+
SA2668	e	Cell wall surface anchor protein	sasF	LPKAG
SA2505	e	Cell wall surface anchor protein	sasG (aap)	LPKTG	+
SA0024	e	5′ nucleotidase family protein	sasH	LPKTG
SA1781	g	Cell wall surface anchor protein	sasI (harA)	LPKTG	+
SA1138	d	Cell wall surface anchor protein	sasJ (isdB)	LPKTG	+
SA2381_h_	d	Cell wall surface anchor protein	sasK	LPKTG
SA1141	e	Cell wall surface anchor protein (NPQTN)	isdC	NPQTN
SA2002	a	Extracellular adherence protein	eap, map
SA0858	a	Extracellular matrix and plasma binding protein	empbp
SA1472	a	Cell wall-associated fibronectin binding protein	ebh	+
SA1164	a	Fibrinogen binding-related protein	fib
SA1168	a	Fibrinogen binding protein	efb
SA1522	a	Elastin binding protein	ebp
SA1062	a	Biofunctional autolysin	atl
S. epidermidis surface proteins_f_
SE0026	a	SdrF (truncated) (not all strains have SdrF)	sdrF	+
SE0207	a	SdrG	sdrG	+	LPDTG
SE1487	a	SdrH	sdrH	+
SE1316	e	Cell wall surface anchor protein	sesA	LPLAG	+
SE2162	e	Cell wall surface anchor protein	sesB	LPNTG
SE2264	e	Cell wall surface anchor protein	sesC	LPATG
SE2392_i_	c	Cell wall surface anchor protein	sesD (bhp)	LPQTG	+
SE0719	e	Cell wall surface anchor protein	sesE	LPETG	+
SE2398	e	Cell wall surface anchor protein	sesF (aap)	LPDTG	+
SE1482_i_	e	Cell wall surface anchor protein	sesG	LPDTG	+
SE1483	e	Cell wall surface anchor protein	sesH	LPETG
SE1654_i_	e	Cell wall surface anchor protein	sesI	LPETG
SE1500_j_	e	Cell wall surface anchor protein	sesJ	LPKTG
SE1501_j_	e	Cell wall surface anchor protein	sesK	LPNTG
SE1011	a	Cell wall associated fibronectin binding protein	ebh	+
SE1048	a	Elastin binding protein	ebp
SE0636	a,f	Bifunctional autolysin	altE
SE0775	a	Fibrinogen binding protein	fbe

The most likely candidate for a bona fide virulence factor in S. epidermidis is the family of small cytokine-stimulating peptides (22 to 44 amino acids in length) previously identified as PSM (38) (Table 2). Members of the PSM family are present in other staphylococci, including S. aureus, but our analysis has revealed that they are more numerous in S. epidermidis, where they appear to have expanded as a result of gene duplication within the νSeγ genome island (Fig. 2).

Expression of staphylococcal virulence factors and cell surface adhesion proteins is regulated by two previously identified regulatory loci, the accessory gene regulator locus (agrABCD) (36) and the staphylococcal accessory regulator family (sarA, etc.) (8), which respond to environmental or host stimuli through a quorum-sensing mechanism to coordinate adherence, tissue breakdown, and further invasion. Our analysis has identified 15 additional two-component regulatory systems (Supplemental Table 7) that are similar to agr and conserved in both S. aureus and S. epidermidis, which is surprising considering the differences in adhesins and virulence factors expressed by these species. Identification of possible functional homologs of the agr locus in the genomes of Clostridium acetobutylicum, Enterococcus faecium, Enterococcus faecalis, Lactobacillus plantarum, and Listeria monocytogenes suggests a conservation of regulatory and quorum-sensing mechanisms among the low-GC-content gram-positive pathogens.

Our comparison of the S. epidermidis genomes revealed that a key difference between the biofilm-nonproducing ATCC 12228 type strain and the biofilm-producing RP62a is the presence of the intercellular adhesion locus (icaABCD) and the cell wall associated biofilm protein (Bap) or Bap homologous protein (Bhp). The ica locus, which encodes the polysaccharide intercellular adhesin protein with a key role in biofilm formation and bacterial accumulation on host surfaces, is present in biofilm-associated isolates, such as S. epidermidis RP62a, but is frequently absent in commensal isolates, such as ATCC 12228 (54). Bap was previously identified in bovine mastitis S. aureus isolates, where it has key roles in adherence to polystyrene surfaces, intercellular adhesion, and biofilm formation (12, 13, 49), but has not been found in human clinical S. aureus isolates. However, Bhp was identified in S. epidermidis RP62a, where it may have a function similar to that of the Bap homolog. Homologs of Bap and Bhp were identified in other bacteria, including the enterococcal surface protein or Esp in Enterococcus faecalis, where it plays a similar essential role in biofilm formation (46). The Esp, Bap, and Bhp surface proteins may play functional roles in biofilm formation among mixed populations of these bacteria, as documented by the recent transfer of the vancomycin conjugative transposon, Tn_1546_, from clinical isolates of Enterococcus faecalis to S. aureus (52).

Conclusion

The most significant observation from our study was evidence for gene transfer between the staphylococci and bacilli. The cap operon, encoding the polyglutamate capsule, a major virulence factor in B. anthracis, has integrated in the genomes of both S. epidermidis RP62a and ATCC 12228, likely as a result of plasmid-mediated gene transfer between the two genera. Evidence of active gene transfer and movement of mobile elements between the staphylococci and other low-GC-content gram-positive bacteria is suggestive not only of continued evolution of virulence and resistance in S. aureus but also the transition of S. epidermidis from a commensal pathogen to a more aggressive opportunistic pathogen through the acquisition of additional virulence factors. Evidence of S. epidermidis strains producing enterotoxin C (49) indicates that this gene movement has already occurred and leads us to propose the genome sequencing of additional S. epidermidis clinical isolates to examine the role of gene transfer in the evolution of staphylococcal virulence.

Supplementary Material

Acknowledgments

This work was supported in part by NIH grants U-01AI45667 and R01-AI43567 (S. R. Gill).

We thank Michael Heaney, Susan Lo, Michael Holmes, Vadim Sapiro, Robert Strausberg, Owen White, William C. Nelson, Jeremy D. Peterson, and Tanja Davidson at TIGR for support with various aspects of this project and also Brian Wilkinson (Illinois State University) for providing the S. aureus COL isolate used for sequencing.

REFERENCES

1. Archer, G. L., and D. M. Niemeyer. 1994. Origin and evolution of DNA associated with resistance to methicillin in staphylococci. Trends Microbiol. 2**:**343-347. [Google Scholar]

2. Arciola, C. R., L. Baldassarri, and L. Montanaro. 2001. Presence of icaA and icaD genes and slime production in a collection of staphylococcal strains from catheter-associated infections. J. Clin. Microbiol. 39**:**2151-2156. [PMC free article] [PubMed] [Google Scholar]

3. Baba, T., F. Takeuchi, M. Kuroda, H. Yuzawa, K. Aoki, A. Oguchi, Y. Nagai, N. Iwama, K. Asano, T. Naimi, H. Kuroda, L. Cui, K. Yamamoto, and K. Hiramatsu. 2002. Genome and virulence determinants of high virulence community-acquired MRSA. Lancet 359**:**1819-1827. [PubMed] [Google Scholar]

4. Beres, S. B., G. L. Sylva, K. D. Barbian, B. Lei, J. S. Hoff, N. D. Mammarella, M. Y. Liu, J. C. Smoot, S. F. Porcella, L. D. Parkins, D. S. Campbell, T. M. Smith, J. K. McCormick, D. Y. Leung, P. M. Schlievert, and J. M. Musser. 2002. Genome sequence of a serotype M3 strain of group A Streptococcus: phage-encoded toxins, the high-virulence phenotype, and clone emergence. Proc. Natl. Acad. Sci. USA 99**:**10078-10083. [PMC free article] [PubMed] [Google Scholar]

5. Centers for Disease Control and Prevention. 1999. Four pediatric deaths from community-acquired methicillin-resistant _Staphylococcus aureus_—Minnesota and North Dakota, 1997-1999. Morb. Mortal. Wkly. Rep. 48**:**707-710. [PubMed] [Google Scholar]

6. Centers for Disease Control and Prevention. 1997. Staphylococcus aureus with reduced susceptibility to vancomycin-United States, 1997. Morb. Mortal. Wkly. Rep. 46**:**765-766. [PubMed] [Google Scholar]

7. Centers for Disease Control and Prevention. 2002. Vancomycin-resistant _Staphylococcus aureus_—Pennsylvania, 2002. Morb. Mortal. Wkly. Rep. 51**:**902. [PubMed] [Google Scholar]

8. Cheung, A. L., S. J. Projan, and H. Gresham. 2002. The genomic aspect of virulence, sepsis, and resistance to killing mechanisms in Staphylococcus aureus. Curr. Infect. Dis. Rep. 4**:**400-410. [PubMed] [Google Scholar]

9. Christensen, G. D., W. A. Simpson, A. L. Bisno, and E. H. Beachey. 1982. Adherence of slime-producing strains of Staphylococcus epidermidis to smooth surfaces. Infect Immun. 37**:**318-326. [PMC free article] [PubMed] [Google Scholar]

10. Christensen, G. D., W. A. Simpson, J. J. Younger, L. M. Baddour, F. F. Barrett, D. M. Melton, and E. H. Beachey. 1985. Adherence of coagulase-negative staphylococci to plastic tissue culture plates: a quantitative model for the adherence of staphylococci to medical devices. J. Clin. Microbiol. 22**:**996-1006. [PMC free article] [PubMed] [Google Scholar]

11. Cramton, S. E., N. F. Schnell, F. Gotz, and R. Bruckner. 2000. Identification of a new repetitive element in Staphylococcus aureus. Infect Immun. 68**:**2344-2348. [PMC free article] [PubMed] [Google Scholar]

12. Cucarella, C., C. Solano, J. Valle, B. Amorena, I. Lasa, and J. R. Penades. 2001. Bap, a Staphylococcus aureus surface protein involved in biofilm formation. J. Bacteriol. 183**:**2888-2896. [PMC free article] [PubMed] [Google Scholar]

13. Cucarella, C., M. A. Tormo, E. Knecht, B. Amorena, I. Lasa, T. J. Foster, and J. R. Penades. 2002. Expression of the biofilm-associated protein interferes with host protein receptors of Staphylococcus aureus and alters the infective process. Infect. Immun. 70**:**3180-3186. [PMC free article] [PubMed] [Google Scholar]

14. Delcher, A. L., D. Harmon, S. Kasif, O. White, and S. L. Salzberg. 1999. Improved microbial gene identification with GLIMMER. Nucleic Acids Res. 27**:**4636-4641. [PMC free article] [PubMed] [Google Scholar]

15. Delcher, A. L., A. Phillippy, J. Carlton, and S. L. Salzberg. 2002. Fast algorithms for large-scale genome alignment and comparison. Nucleic Acids Res. 30**:**2478-2483. [PMC free article] [PubMed] [Google Scholar]

16. De Lencastre, H., S. W. Wu, M. G. Pinho, A. M. Ludovice, S. Filipe, S. Gardete, R. Sobral, S. Gill, M. Chung, and A. Tomasz. 1999. Antibiotic resistance as a stress response: complete sequencing of a large number of chromosomal loci in Staphylococcus aureus strain COL that impact on the expression of resistance to methicillin. Microb. Drug Resist. 5**:**163-175. [PubMed] [Google Scholar]

17. Dyke, K. G., M. P. Jevons, and M. T. Parker. 1966. Penicillinase production and intrinsic resistance to penicillins in Staphylococcus aureus. Lancet i**:**835-838. [PubMed] [Google Scholar]

18. Etz, H., D. B. Minh, T. Henics, A. Dryla, B. Winkler, C. Triska, A. P. Boyd, J. Sollner, W. Schmidt, U. von Ahsen, M. Buschle, S. R. Gill, J. Kolonay, H. Khalak, C. M. Fraser, A. von Gabain, E. Nagy, and A. Meinke. 2002. Identification of in vivo expressed vaccine candidate antigens from Staphylococcus aureus. Proc. Natl. Acad. Sci. USA 99**:**6573-6578. [PMC free article] [PubMed] [Google Scholar]

19. Hanaki, H., and K. Hiramatsu. 1997. Evaluation of reduced vancomycin susceptibility of MRSA strain Mu50 with various conditions of antibiotic susceptibility tests. Jpn. J. Antibiot. 50**:**794-798. (In Japanese.) [PubMed] [Google Scholar]

20. Hiramatsu, K. 1998. The emergence of Staphylococcus aureus with reduced susceptibility to vancomycin in Japan. Am. J. Med. 104**:**7S-10S. [PubMed] [Google Scholar]

21. Holden, M. T., E. J. Feil, J. A. Lindsay, S. J. Peacock, N. P. Day, M. C. Enright, T. J. Foster, C. E. Moore, L. Hurst, R. Atkin, A. Barron, N. Bason, S. D. Bentley, C. Chillingworth, T. Chillingworth, C. Churcher, L. Clark, C. Corton, A. Cronin, J. Doggett, L. Dowd, T. Feltwell, Z. Hance, B. Harris, H. Hauser, S. Holroyd, K. Jagels, K. D. James, N. Lennard, A. Line, R. Mayes, S. Moule, K. Mungall, D. Ormond, M. A. Quail, E. Rabbinowitsch, K. Rutherford, M. Sanders, S. Sharp, M. Simmonds, K. Stevens, S. Whitehead, B. G. Barrell, B. G. Spratt, and J. Parkhill. 2004. Complete genomes of two clinical Staphylococcus aureus strains: evidence for the rapid evolution of virulence and drug resistance. Proc. Natl. Acad. Sci. USA 101**:**9786-9791. [PMC free article] [PubMed] [Google Scholar]

22. Hussain, M., M. Herrmann, C. von Eiff, F. Perdreau-Remington, and G. Peters. 1997. A 140-kilodalton extracellular protein is essential for the accumulation of Staphylococcus epidermidis strains on surfaces. Infect. Immun. 65**:**519-524. [PMC free article] [PubMed] [Google Scholar]

23. Ito, T., Y. Katayama, K. Asada, N. Mori, K. Tsutsumimoto, C. Tiensasitorn, and K. Hiramatsu. 2001. Structural comparison of three types of staphylococcal cassette chromosome mec integrated in the chromosome in methicillin-resistant Staphylococcus aureus. Antimicrob. Agents Chemother. 45**:**1323-1336. [PMC free article] [PubMed] [Google Scholar]

24. Ito, T., K. Okuma, X. X. Ma, H. Yuzawa, and K. Hiramatsu. 2003. Insights on antibiotic resistance of Staphylococcus aureus from its whole genome: genomic island SCC. Drug Resist. Updates 6**:**41-52. [PubMed] [Google Scholar]

25. Katayama, Y., T. Ito, and K. Hiramatsu. 2000. A new class of genetic element, staphylococcus cassette chromosome mec, encodes methicillin resistance in Staphylococcus aureus. Antimicrob. Agents Chemother. 44**:**1549-1555. [PMC free article] [PubMed] [Google Scholar]

26. Katayama, Y., F. Takeuchi, T. Ito, X. X. Ma, Y. Ui-Mizutani, I. Kobayashi, and K. Hiramatsu. 2003. Identification in methicillin-susceptible Staphylococcus hominis of an active primordial mobile genetic element for the staphylococcal cassette chromosome mec of methicillin-resistant Staphylococcus aureus. J. Bacteriol. 185**:**2711-2722. [PMC free article] [PubMed] [Google Scholar]

27. Krogh, A., B. Larsson, G. von Heijne, and E. L. Sonnhammer. 2001. Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes. J. Mol. Biol. 305**:**567-580. [PubMed] [Google Scholar]

28. Kuroda, M., T. Ohta, I. Uchiyama, T. Baba, H. Yuzawa, I. Kobayashi, L. Cui, A. Oguchi, K. Aoki, Y. Nagai, J. Lian, T. Ito, M. Kanamori, H. Matsumaru, A. Maruyama, H. Murakami, A. Hosoyama, Y. Mizutani-Ui, N. K. Takahashi, T. Sawano, R. Inoue, C. Kaito, K. Sekimizu, H. Hirakawa, S. Kuhara, S. Goto, J. Yabuzaki, M. Kanehisa, A. Yamashita, K. Oshima, K. Furuya, C. Yoshino, T. Shiba, M. Hattori, N. Ogasawara, H. Hayashi, and K. Hiramatsu. 2001. Whole genome sequencing of methicillin-resistant Staphylococcus aureus. Lancet 357**:**1225-1240. [PubMed] [Google Scholar]

29. Lazarevic, V., B. Soldo, A. Dusterhoft, H. Hilbert, C. Mauel, and D. Karamata. 1998. Introns and intein coding sequence in the ribonucleotide reductase genes of Bacillus subtilis temperate bacteriophage SPβ. Proc. Natl. Acad. Sci. USA 95**:**1692-1697. [PMC free article] [PubMed] [Google Scholar]

30. Lee, C. Y., and J. J. Iandolo. 1986. Lysogenic conversion of staphylococcal lipase is caused by insertion of the bacteriophage L54a genome into the lipase structural gene. J. Bacteriol. 166**:**385-391. [PMC free article] [PubMed] [Google Scholar]

31. Lindsay, J. A., A. Ruzin, H. F. Ross, N. Kurepina, and R. P. Novick. 1998. The gene for toxic shock toxin is carried by a family of mobile pathogenicity islands in Staphylococcus aureus. Mol. Microbiol. 29**:**527-543. [PubMed] [Google Scholar]

32. Lyczak, J. B., C. L. Cannon, and G. B. Pier. 2002. Lung infections associated with cystic fibrosis. Clin. Microbiol. Rev. 15**:**194-222. [PMC free article] [PubMed] [Google Scholar]

33. Mazmanian, S. K., H. Ton-That, and O. Schneewind. 2001. Sortase-catalysed anchoring of surface proteins to the cell wall of Staphylococcus aureus. Mol. Microbiol. 40**:**1049-1057. [PubMed] [Google Scholar]

34. Mongkolrattanothai, K., S. Boyle, T. V. Murphy, and R. S. Daum. 2004. Novel non-_mecA_-containing staphylococcal chromosomal cassette composite island containing pbp4 and tagF genes in a commensal staphylococcal species: a possible reservoir for antibiotic resistance islands in Staphylococcus aureus. Antimicrob. Agents Chemother. 48**:**1823-1836. [PMC free article] [PubMed] [Google Scholar]

35. Nelson, K. E., D. E. Fouts, E. F. Mongodin, J. Ravel, R. T. DeBoy, J. F. Kolonay, D. A. Rasko, S. V. Angiuoli, S. R. Gill, I. T. Paulsen, J. Peterson, O. White, W. C. Nelson, W. Nierman, M. J. Beanan, L. M. Brinkac, S. C. Daugherty, R. J. Dodson, A. S. Durkin, R. Madupu, D. H. Haft, J. Selengut, S. Van Aken, H. Khouri, N. Fedorova, H. Forberger, B. Tran, S. Kathariou, L. D. Wonderling, G. A. Uhlich, D. O. Bayles, J. B. Luchansky, and C. M. Fraser. 2004. Whole genome comparisons of serotype 4b and 1/2a strains of the food-borne pathogen Listeria monocytogenes reveal new insights into the core genome components of this species. Nucleic Acids Res. 32**:**2386-2395. [PMC free article] [PubMed] [Google Scholar]

36. Novick, R. P. 2003. Autoinduction and signal transduction in the regulation of staphylococcal virulence. Mol. Microbiol. 48**:**1429-1449. [PubMed] [Google Scholar]

37. Okonogi, K., Y. Noji, M. Kondo, A. Imada, and T. Yokota. 1989. Emergence of methicillin-resistant clones from cephamycin-resistant Staphylococcus aureus. J. Antimicrob. Chemother. 24**:**637-645. [PubMed] [Google Scholar]

38. Otto, M. 2004. Virulence factors of the coagulase-negative staphylococci. Front. Biosci. 9**:**841-863. [PubMed] [Google Scholar]

39. Paulsen, I. T., L. Banerjei, G. S. Myers, K. E. Nelson, R. Seshadri, T. D. Read, D. E. Fouts, J. A. Eisen, S. R. Gill, J. F. Heidelberg, H. Tettelin, R. J. Dodson, L. Umayam, L. Brinkac, M. Beanan, S. Daugherty, R. T. DeBoy, S. Durkin, J. Kolonay, R. Madupu, W. Nelson, J. Vamathevan, B. Tran, J. Upton, T. Hansen, J. Shetty, H. Khouri, T. Utterback, D. Radune, K. A. Ketchum, B. A. Dougherty, and C. M. Fraser. 2003. Role of mobile DNA in the evolution of vancomycin-resistant Enterococcus faecalis. Science 299**:**2071-2074. [PubMed] [Google Scholar]

40. Paulsen, I. T., M. Firth, and R. A. Skurray. 1997. Resistance to antimicrobial agents other than β-lactams, p. 175-212. In K. B. a. A. Crossley, G.L. (ed.), The staphylococci in human disease. Churchill Livingstone, New York, N.Y.

41. Projan, S. J., and R. P. Novick. 1997. The molecular basis of pathogenicity, p. 55-82. In K. B. a. A. Crossley, G.L. (ed.), The staphylococci in human disease. Churchill Livingstone, New York, N.Y.

42. Ruzin, A., J. Lindsay, and R. P. Novick. 2001. Molecular genetics of SaPI1—a mobile pathogenicity island in Staphylococcus aureus. Mol. Microbiol. 41**:**365-377. [PubMed] [Google Scholar]

43. Sabath, L. D., S. J. Wallace, and D. A. Gerstein. 1972. Suppression of intrinsic resistance to methicillin and other penicillins in Staphylococcus aureus. Antimicrob. Agents Chemother. 2**:**350-355. [PMC free article] [PubMed] [Google Scholar]

44. Skaar, E. P., and O. Schneewind. 2004. Iron-regulated surface determinants (Isd) of Staphylococcus aureus: stealing iron from heme. Microbes Infect. 6**:**390-397. [PubMed] [Google Scholar]

45. Stevens, D. L. 2003. Community-acquired Staphylococcus aureus infections: increasing virulence and emerging methicillin resistance in the new millennium. Curr. Opin. Infect. Dis. 16**:**189-191. [PubMed] [Google Scholar]

46. Toledo-Arana, A., J. Valle, C. Solano, M. J. Arrizubieta, C. Cucarella, M. Lamata, B. Amorena, J. Leiva, J. R. Penades, and I. Lasa. 2001. The enterococcal surface protein, Esp, is involved in Enterococcus faecalis biofilm formation. Appl. Environ. Microbiol. 67**:**4538-4545. [PMC free article] [PubMed] [Google Scholar]

47. Ton-That, H., S. K. Mazmanian, K. F. Faull, and O. Schneewind. 2000. Anchoring of surface proteins to the cell wall of Staphylococcus aureus. Sortase catalyzed in vitro transpeptidation reaction using LPXTG peptide and NH2-Gly3 substrates. J. Biol. Chem. 275**:**9876-9881. [PubMed] [Google Scholar]

48. Vandenesch, F., T. Naimi, M. C. Enright, G. Lina, G. R. Nimmo, H. Heffernan, N. Liassine, M. Bes, T. Greenland, M. E. Reverdy, and J. Etienne. 2003. Community-acquired methicillin-resistant Staphylococcus aureus carrying Panton-Valentine leukocidin genes: worldwide emergence. Emerg. Infect. Dis. 9**:**978-984. [PMC free article] [PubMed] [Google Scholar]

49. von Eiff, C., G. Peters, and C. Heilmann. 2002. Pathogenesis of infections due to coagulase-negative staphylococci. Lancet Infect. Dis. 2**:**677-685. [PubMed] [Google Scholar]

50. Vuong, C., M. Durr, A. B. Carmody, A. Peschel, S. J. Klebanoff, and M. Otto. 2004. Regulated expression of pathogen-associated molecular pattern molecules in Staphylococcus epidermidis: quorum-sensing determines pro-inflammatory capacity and production of phenol-soluble modulins. Cell. Microbiol. 6**:**753-759. [PubMed] [Google Scholar]

51. Walsh, C. 1999. Deconstructing vancomycin. Science 284**:**442-443. [PubMed] [Google Scholar]

52. Weigel, L. M., D. B. Clewell, S. R. Gill, N. C. Clark, L. K. McDougal, S. E. Flannagan, J. F. Kolonay, J. Shetty, G. E. Killgore, and F. C. Tenover. 2003. Genetic analysis of a high-level vancomycin-resistant isolate of Staphylococcus aureus. Science 302**:**1569-1571. [PubMed] [Google Scholar]

53. Williams, R. J., J. M. Ward, B. Henderson, S. Poole, B. P. O'Hara, M. Wilson, and S. P. Nair. 2000. Identification of a novel gene cluster encoding staphylococcal exotoxin-like proteins: characterization of the prototypic gene and its protein product, SET1. Infect Immun. 68**:**4407-4415. [PMC free article] [PubMed] [Google Scholar]

54. Zhang, Y. Q., S. X. Ren, H. L. Li, Y. X. Wang, G. Fu, J. Yang, Z. Q. Qin, Y. G. Miao, W. Y. Wang, R. S. Chen, Y. Shen, Z. Chen, Z. H. Yuan, G. P. Zhao, D. Qu, A. Danchin, and Y. M. Wen. 2003. Genome-based analysis of virulence genes in a non-biofilm-forming Staphylococcus epidermidis strain (ATCC 12228). Mol. Microbiol. 49**:**1577-1593. [PubMed] [Google Scholar]

Articles from Journal of Bacteriology are provided here courtesy of American Society for Microbiology (ASM)