OUP user menu

Capsular serotype of Staphylococcus aureus in the era of community-acquired MRSA

Deena E. Sutter, Amy M. Summers, Christine E. Keys, Kimberly L. Taylor, Carl E. Frasch, LoRanee E. Braun, Ali I. Fattom, Margaret C. Bash
DOI: http://dx.doi.org/10.1111/j.1574-695X.2011.00822.x 16-24 First published online: 1 October 2011


Capsular polysaccharide (CP) plays an important role in the pathogenicity and immunogenicity of Staphylococcus aureus, yet the common serotypes of S. aureus isolated from US pediatric patients have not been reported. We investigated capsular serotype as well as methicillin susceptibility, presence of Panton–Valentine leukocidin (PVL), and clonal relatedness of pediatric S. aureus isolates. Clinical isolates were tested for methicillin susceptibility, presence of mecA, lukS-PV and lukF-PV, cap5 and cap8 genes by PCR, and for capsular or surface polysaccharide expression (CP5, CP8, or 336 polysaccharide) by agglutination. Genetic relatedness was determined by pulsed-field gel electrophoresis. All S. aureus isolates encoded cap5 or cap8. Sixty-nine percent of 2004–2005 isolates were methicillin-susceptible (MSSA) and most expressed a detectable capsule. The majority of MRSA isolates (82%) were unencapsulated, exposing an expressed cell wall techoic acid antigen 336. Pulsed-field type USA300 were MRSA, PVL-positive, unencapsulated strains that were associated with deep skin infections and recurrent disease. Over half (58%) of all isolates from invasive pediatric dermatologic infections were USA300. All pediatric isolates contained either capsule type 5 or capsule type 8 genes, and roughly half of the S. aureus clinical disease isolates from our population were diverse MSSA-encapsulated strains. The majority of the remaining pediatric clinical disease isolates were unencapsulated serotype 336 strains of the PVL(+) USA300 community-associated-MRSA clone.

  • Staphylococcus aureus
  • MRSA
  • pediatric
  • community acquired
  • capsule


The epidemic of community-associated methicillin-resistant Staphylococcus aureus (CA-MRSA) has spread throughout the United States, affecting all patient populations, including healthy children (Dietrich et al., 2004; Purcell & Fergie, 2005; Kaplan, 2006). Multiple investigations have demonstrated the most prevalent CA-MRSA clone in the United States to be pulsed-field type (PFT) USA300 (King et al., 2006; Tenover et al., 2006), the majority of which encode the toxin Panton–Valentine leukocidin (PVL). Other virulence factors include surface-exposed molecules such as capsular polysaccharide (CP). Expression of CP has been shown to enhance the ability of the bacteria to evade opsonophagocytosis, whereas CP-specific antibodies mediate type-specific opsonophagocytosis and bacterial killing by polymorphonuclear cells (Cunnion et al., 2001). CPs have been shown to be protective antigens in animal models (Lee et al., 1997), leading to their investigation as potential vaccine targets.

Serotyping studies, predominantly in adults, have demonstrated that the majority of clinically important S. aureus isolates possess a type 5 or a type 8 polysaccharide capsule. Two recent studies using isolates from the United States and the Netherlands demonstrated that 92% and 82% of S. aureus isolates were of either type 5 or type 8, respectively (Roghmann et al., 2005; Melles et al., 2008). The remaining isolates were unencapsulated and reacted strongly with type 336 polysaccharide (336PS) antibodies. The 336PS antigen is a variant of the native cell wall teichoic acid of S. aureus, and is the prevalent cell-surface antigen noted in serotyping of isolates that do not produce a polysaccharide capsule. In a recent study, three isolates of the most prevalent CA-MRSA clone, USA300, were identified as unencapsulated; however, the prevalence of the unencapsulated phenotype among clinical CA-MRSA infections and the serotype of this clone have not been reported (Montgomery et al., 2008). In addition, recent epidemiological data on capsular and surface polysaccharide types of both MRSA and methicillin-susceptible S. aureus (MSSA) disease isolates from pediatric patients in the US have not been reported.

We previously reported a marked increase in the rates of both MRSA and MSSA infections in children in the National Capital Area in 2001–2004 (Braun et al., 2005). The current study was undertaken to examine the molecular epidemiology of S. aureus during the early phase of this epidemic and investigate all MRSA and MSSA strains isolated from pediatric patients at our institution over the following 1-year period. Our goal was to examine the relationships between capsule genotype and serotype, clonal type, presence or absence of the genes encoding PVL, and clinical disease presentation, to include colonization vs. infection. Additionally, the 2004–2005 MRSA isolates were compared with a historical collection of pediatric MRSA isolates from 2001 to 2004 to assess for trends in the epidemiology of MRSA in our institution over a 4-year time period.

Materials and methods

Study populations and strains

All S. aureus isolates recovered by the Clinical Microbiology Laboratory at Walter Reed Army Medical Center (WRAMC) from patients aged 0–18 years between the dates August 1, 2004 and July 31, 2005, excluding those from neonates in the NICU, were included. Isolates were actively identified by study personnel throughout the study period. For patients with more than one culture, only the first culture isolate was included, except in those patients with both MRSA and MSSA, in which case both isolates were included.

The patient population included children admitted to the pediatric ward and ICU, as well as outpatients seen at WRAMC and several smaller satellite clinics in the Washington, DC area. WRAMC is a tertiary care military hospital located in Northwest DC, with a small (approximately 15–20-bed capacity) inpatient pediatric ward and PICU (typically four-bed capacity) and approximately 6500 outpatient general pediatric visits per year.

Historical isolates were those pediatric MRSA specimens obtained and archived between 2001 and 2004 as part of routine laboratory surveillance in which one culture per patient per 12-month period was banked for any MRSA identified by the laboratory. These isolates were collected before the initiation of this study and did not include MSSA. The historical isolates were analyzed separately and compared only with MRSA isolates from the study period.

Microbiological methods

Isolates were recovered and identified using routine microbiological methods. Antibiotic susceptibility testing was performed at the time of initial culture recovery and then isolates were stored at −20 °C in a tryptic soy broth. Staphylococcus aureus isolates were thawed within 3 months of collection and subcultured onto sheep blood agar for testing.

Capsular phenotypic testing

Phenotypic expression of S. aureus CP or surface polysaccharide 336 was determined using slide agglutination and confirmation of serotype by immunodiffusion with rabbit-derived polyclonal type 5, type 8, and 336 antibodies was performed as described previously (Karakawa et al., 1985; Roghmann et al., 2005; Melles et al., 2008). Briefly, for each isolate, a single colony was picked and grown to the stationary phase for 24 h on Columbia salt agar containing 2% MgCl2, the optimal medium for capsule production in S. aureus. Several colonies were resuspended in 0.9% saline and tested for agglutination; 10 µL of each antiserum and 10 µL of the bacterial suspension were spotted onto a glass slide and rocked gently to mix. Positive agglutination was noted if visible clumps were present within 5–10 s. For immunodiffusion, crude S. aureus lysates were prepared by resuspending one loopful with lysostaphin, RNase and DNase and tested as described previously (Melles et al., 2008). Strains that reacted strongly with type 5 or type 8 antibodies were defined as capsulated type 5 and type 8 strains, respectively. Strains that did not react with either type 5 or type 8 capsular antibodies and that reacted strongly with type 336 antibodies were defined as unencapsulated type 336 strains.


DNA from each isolate of S. aureus was extracted using the Wizard Genomic DNA purification kit (Promega, Madison, WI) as per the manufacturer's recommended instructions, with the modification that 0.225 µg of lysostaphin (Ambi, Lawrence, NY) was added at the cell lysis step. PCR amplifications of S. aureus gyrA (to confirm successful DNA extraction), mecA (to confirm MRSA or MSSA status), lukS-PV and lukF-PV (to evaluate for the presence of PVL genes), cap5 (to assess for genes encoding capsule type 5), and cap8 (to assess for genes encoding capsule type 8) were performed for all isolates using the following primers: gyrA1 5′-AATGAACAAGGTATGACACC-3′, gyrA2 5′-TACGCGCTTCAGTATAACGC-3′ (Schmitz et al., 1998); mecA P4 5′-TCCAGATTACAACTTCACCAGG-3′, mecA P7 5′-CCACTTCATATCTTGTAACG-3′ (Oliveira & de Lencastre, 2002); luk-PV-1 5′-ATCATTAGGTAAAATGTCTGGACATGATCCA-3′, luk-PV-2 5′-GCATCAASTGTATTGGATAGCAAAAGC-3′ (Lina et al., 1999); cap5-1 5′-GGTTTGCTGAAAAACCAGTC-3′, cap5-2 5′-CCTCATATGCTCCTACATTT-3′ (Sau et al., 1997); cap8-1 5′-GCGCTACAAACATTAAGCAT-3′, cap8-2 5′-TTCTTAGCCTGCTGGCATC-3′ (Sau et al., 1997). PCR conditions were as follows: 25 cycles of 20 s at 94 °C, 20 s at 50 °C, and 50 s at 72 °C for amplification of gyrA, mecA, cap5, and cap8; 30 cycles of 30 s at 94 °C, 30 s at 55 °C, and 1 min at 72 °C for lukS-PV and lukF-PV. Negative (water) and positive controls were used in each reaction: COL (NCTC 8325) for gyrA and mecA (Cohen & Sweeney, 1970); MW2 for lukS-PV and lukF-PV (Arbeit et al., 1984); Lowenstein (ATCC 49521) for cap5 (Cunnion et al., 2001); and Wright (ATCC 49525) for cap8 (Cunnion et al., 2001). PCR products were run on a 1% agarose gel and visualized with ethidium bromide staining.

Pulsed-field gel electrophoresis (PFGE)

PFGE was performed for all MRSA strains, all PVL-positive strains, and a random subset (approximately half) of the MSSA PVL-negative isolates using the smaI restriction enzyme and the BioRad CHEF Mapper XA System (Bio-Rad, Hercules, CA) using the digestion run and switch times of McDougal and colleagues (McDougal et al., 2003; Ross et al., 2005). Gels were stained with ethidium bromide (1 µg mL−1), with images obtained on a Gel Doc 2000 system (Bio-Rad). Analysis was performed using bionumerics fingerprinting software (Applied Maths, St-Martens-Latem, Belgium). The relatedness between S. aureus clones was established using the criteria of McDougal (2003) and confirmed by CDC criteria for PFTs USA100-1200.

Clinical data

Clinical data were obtained by reviewing inpatient and outpatient electronic medical records. Paper medical records were reviewed when available. For outpatients without available medical records, ICD-9 and CPT codes were used to determine diagnosis and any requirement for surgical drainage. Data recorded included age, gender, site of infection, requirement for hospitalization, whether incision and drainage was performed, and type of antibiotic therapy the patient received, and whether the patient had a history of recurrent S. aureus disease. Infections were considered community-associated unless they developed 48 h or more after admission, or if patients were hospitalized in the prior year, had indwelling catheters or other devices, or had any history of previous healthcare-associated MRSA infection.

Clinical diagnoses were stratified as follows: (1) superficial dermatologic skin and soft tissue infection (SSTI) (impetigo, folliculitis, or superinfection of atopic dermatitis without cellulitis); (2) primary invasive dermatologic SSTI (cellulitis, furunculosis, or subcutaneous abscesses) (DiNubile & Lipsky, 2004); (3) postoperative wound infections; (4) bacteremia; (5) eye, ear nose and throat (EENT) infection (suppurative otitis, mastoiditis, parotitis, or conjunctivitis); (6) cystic fibrosis sputum culture; and (7) colonization (strains obtained from skin, nares, tracheal aspirates, or urine in patients without clinical evidence of disease). Postoperative SSTIs were excluded from analyses of primary dermatologic infections, as these infections occurred secondary to breach of the skin via a surgical incision.

Statistical methods

Statistical analysis was performed using spss statistical software 16.0 (SPSS Inc., Chicago, IL). Categorical data were analyzed for associations between microbiological or molecular features of the isolates and clinical characteristics using χ2 (Fisher's exact test with low numbers), and logistic regression analysis (backward stepwise Wald) was performed to determine any independent risk factors. A P-value <0.05 was considered significant.


Ninety-one unique isolates were obtained from 89 patients during the 12-month study period (2004–2005). Two patients each had a second isolate included; these were children with inpatient MRSA infections who later had nasal swabs to assess carrier status and were colonized with MSSA. Patient age ranged from 2 weeks to 18 years, with a mean age of 7.4 years. Fifty-two isolates (57%) were from male patients. Thirty-nine isolates (44%) were obtained from inpatients. Thirty isolates (33%) were from patients who required admission specifically because of the S. aureus infection.

The clinical features and methicillin resistance of the 91 isolates are shown in Fig. 1. Twenty-one (75%) of methicillin-resistant isolates were community-associated.

Figure 1

The distribution of clinical features and methicillin resistance for 91 pediatric Staphylococcus aureus isolates from 2004 to 2005. Clinical data (age, gender, site of infection, hospitalization, need for incision and drainage, antibiotic therapy, and history of recurrent S. aureus disease) were obtained for each pediatric S. aureus isolate by reviewing inpatient and outpatient electronic medical records and paper medical records when available. ICD-9 and CPT codes were used for diagnosis and requirement for surgical drainage for outpatients without available medical records. aColonization: nine nares swabs, five respiratory cultures, four urine cultures. The nares specimens were collected to assess for possible MRSA nasal carriage in children with recurrent MRSA disease or among their siblings, or in patients with suspected respiratory or cutaneous S. aureus disease that was later determined to be due to another pathogen. Respiratory specimens were endotracheal cultures from patients later confirmed to have a diagnosis other than S. aureus pneumonia. Four patients had S. aureus isolated from urine cultures; these were small quantities of mixed growth from children without pyuria. bEye, ear, nose, and throat: suppurative otitis, mastoiditis, parotitis, or conjunctivitis.


Thirty (33%) of all 91 isolates and 29/73 (40%) disease-associated (noncolonizing) isolates contained genes lukS-PV and lukF-PV (PVL positive). PVL was much more common among MRSA than MSSA isolates (25/28 or 89% vs. 5/63 or 7.9%, P<0.0001).

CP or surface polysaccharide type

Fifty-one isolates (56%) contained the genes for CP type 5 (cap5), and the remaining 40 (44%) encoded CP type 8 (cap8). Capsule expression was detected in 57 isolates (62%), all of which corresponded with the capsule genotype (i.e. no isolates encoding cap5 reacted with CP8 antibodies and vice versa). The remaining 34 isolates (38%) did not react with capsular antisera and reacted strongly with the 336 antibodies (serotype 336). Isolates with cap8 were more likely than those with cap5 to express capsule phenotypically (35/40 or 88% vs. 22/51 or 43%, P<0.0005). Among MRSA isolates, 25 (89%) encoded cap5 and three (11%) encoded cap8. The majority of MSSA isolates (52/63 or 83%) expressed capsule. However, in contrast, only 18% of all MRSA isolates expressed CP type 5 or 8; the remainder (82%) were type 336. Figure 2 demonstrates the characteristics of capsule genotype and phenotype among MRSA and MSSA isolates.

Figure 2

Distribution of capsule genotype and phenotype of methicillin-susceptible and methicillin-resistant isolates by clinical diagnosis (n=91). Vertical bars indicate the number of isolates that expressed CP (▪, expressed) or that reacted with 336PS antisera (Embedded Image, type 336) by clinical category.

PFGE clonal types

Fifty-eight isolates were evaluated by PFGE typing [all 28 MRSA, all five PVL(+) MSSA and 25 PVL(−) MSSA isolates randomly selected from the 58 PVL(−) MSSA isolates (43%)]. Twenty-four of 28 (86%) MRSA isolates were PFT USA300. A total of four USA300 subtypes were represented. Of the non-USA300 MRSA, one was USA100, one was USA1000 and the other two were unique PFTs (not meeting the criteria for relatedness to one of the main USA PFTs). Among MSSA isolates, few isolates were genetically related to the predominant United States MRSA clones (one isolate was USA300, one was USA500, two were USA700 and three were USA400). The remaining 23 MSSA isolates were genetically disparate, with 14 additional PFTs identified.

Clinical associations

Microbiological characteristics including methicillin susceptibility, presence of PVL, capsular genotype and phenotype were analyzed for associations with the following clinical characteristics: infection (vs. colonization), dermatologic infections, invasive SSTIs, and recurrent disease (Table 1). Other characteristics analyzed included requirement for hospitalization, requirement for surgical drainage for invasive SSTIs, and requirement of intravenous antibiotics; however, none of these characteristics showed statistically significant associations.

View this table:
Table 1

Association of methicillin resistance, PVL status, and presence or absence of capsular genes and expression with Staphylococcus aureus strains from pediatric patients over a 12-month period

StrainsMethicillin susceptibilityPVLCapsular genotypeCapsular expression
Clinical featureNMRSAMSSAP-valuePVL positivePVL negativeP-valuecap5cap8P-valueCP5 or CP8336PSP-value
All (%)9128 (30)63 (69)30 (33)61 (67)51 (56)40 (44)57 (63)34 (37)
Colonizing strains (%)181 (6)17 (94)0.011 (6)17 (9)0.0058 (44)10 (56)0.2715 (83)3 (17)0.06
Disease associated (%)7327 (37)46 (63)0.0129 (40)44 (60)0.00543 (59)30 (41)0.2742 (58)31 (42)0.06
    Recurrent disease (%)2111 (52)10 (48)0.0814 (67)7 (33)0.003 15 (71)6 (29)0.178 (38)13 (62)0.03
    Dermatologic infection (%)6427 (42)37 (58)0.0229 (45)35 (55)0.009 39 (61)25 (39)0.4735 (55)29 (45)0.29
        Deep dermatologic (%)3319 (58)14 (42)0.00124 (73)9 (27)<0.00124 (73)9 (27)0.01815 (45)18 (55)0.07
  • Independently associated on logistic regression analysis: disease associated P=0.022, 95% CI=1.4–88; recurrent disease P<0.001, 95% CI=2.3–19.6; dermatologic infection P=0.04, 95% CI=1.1–75; and deep dermatologic infection P=0.017, 95% CI=4.9–85.

  • Proportion of methicillin-resistant vs. proportion of methicillin-susceptible isolates associated with clinical feature.

  • χ2.

  • Proportion of PVL-positive vs. proportion of PVL-negative isolates associated with clinical features.

  • Limited to primary dermatologic infections (postoperative infections excluded).

Methicillin resistance and PVL were highly associated with infection (as opposed to colonization), SSTIs (particularly invasive SSTIs), and recurrent infections. Capsular genotype cap5 was more commonly encoded by isolates with those clinical presentations than cap8, but statistical significance was reached only for invasive SSTIs. Isolates that were encapsulated (CP5 or CP8) were less likely to be associated with recurrent infection than unencapsulated isolates.

Our ability to assess for independent associations between PFT and clinical presentations was limited by the high proportion of MRSA that were PFT USA300; 24 of 28 MRSA strains were indistinguishable, all being PFT USA300, methicillin-resistant, PVL(+), cap5(+), unencapsulated and CP 336. MSSA were substantially more diverse (18 PFT among 30 MSSA strains vs. five PFT among 28 MRSA).

Only a small number of nondermatologic invasive infections occurred in this study (N=5); thus, a statistical analysis of the associations between isolate and clinical characteristics was not meaningful; however, these isolates were notably all PVL(−) MSSA with varying capsular genotype (two cap5 and three cap8) and all five expressed a polysaccharide capsule.

Comparison of study year with historical pediatric MRSA

MRSA isolates from the study period were compared with historical controls (n=24) from the WRAMC clinical microbiology laboratory. Four isolates were available from 2001 to 2002, seven isolates from 2002 to 2003, and 13 from 2003 to 2004. Deep SSTIs, PVL(+) isolates, and PFT USA300 were all more common in the study period, 2004–2005, than in the previous 3 years (Fig. 3).

Figure 3

Comparison of the characteristics of MRSA isolates from 2004 to 2005 with historical isolates from 2001 to 2004. MRSA isolates from 2004 to 2005(n=91) are shown in black. Historical MRSA isolates from 2001 to 2004 (n=24) are shown in gray. The proportion of MRSA isolates that were PVL(+), USA300, and lacked capsule expression increased significantly between 2001–2004 and 2004–2005. *P<0.002; **P=0.0005.


Our study describes the epidemiology of all S. aureus isolates cultured from pediatric military dependents in the greater Washington, DC area during a 12-month period. This study is unique among previously published pediatric S. aureus epidemiology studies as our study design included all isolates collected over the study period including both MSSA and MRSA, whether colonizing or pathogenic, and we have included an analysis of capsule genotype and phenotype. We observed a high prevalence of CA-MRSA in our population, especially in deep complicated dermatologic infections. Our study also emphasizes the continued importance of methicillin-susceptible isolates in community-acquired S. aureus infections; our rates of MSSA exceeded the rates reported recently in some US locales (Moran et al., 2006).

We investigated the prevalence of PVL among our isolates and found PVL to be highly associated with invasive SSTIs. PVL has been linked epidemiologically to invasive dermatologic infections as well as necrotizing pneumonia and other life-threatening invasive infections, and the presence of PVL genes has been reported in several genotypically distinct epidemic clones (Boyle-Vavra & Daum, 2007). However, the connection between PVL and pathogenicity is debated as some widespread CA-MRSA clones are noted to be PVL(−) (Zhang et al., 2008). While PVL has been shown to have dermonecrotic and leukocytolytic activity, isogenic PVL(+) and PVL(−) strains failed to demonstrate a difference in pathogenicity in a mouse bacteremia and skin abscess model (Prevost et al., 1995; Voyich et al., 2006). In contrast, PVL(+) strains caused severe necrotizing pneumonia in a murine pneumonia model while isogenic PVL(−) strains did not (Labandeira-Rey et al., 2007). In our study, similar to other recent epidemiologic studies of MRSA, the strong association between PVL and deep dermatologic infections is confounded by the high association with epidemic clone USA300. However, we also identified a small number of non-USA300 strains in our study that were PVL(+), and all PVL(+) MSSA isolates (five strains from four different genotypic backgrounds) were from patients with furunculosis and abscesses consistent with the proposed association between PVL and invasive SSTI.

In this study, we investigated the polysaccharide capsule of S. aureus, which may play a role in host invasion due to its antiphagocytic properties, and cell wall teichoic acids, which have been shown to act as adhesions to host cells (Karakawa et al., 1988; Weidenmaier et al., 2004, 2005). While capsular epidemiology has been described for adult or international pediatric populations (Arbeit et al., 1984; Sompolinsky et al., 1985; Na'was et al., 1998; Paul-Satyaseela et al., 2004; Melles et al., 2008; Lattar et al., 2009), our study comprises the only recent report of capsular and surface polysaccharide types of pediatric S. aureus isolates in a region of the United States.

Type 5 and type 8 capsule expression was only detected in 58% of the strains from disease isolates in our study; the remaining 42% of disease isolates were either cap5 or cap8 genetically, but did not express capsule in vitro and reacted strongly with type 336PS antisera. The unusually low frequency of encapsulated strains was related to the fact that PFT USA300 strains from invasive dermatologic infections were primarily type 336 strains that did not express a capsule phenotypically in vitro despite encoding the type 5 capsule gene. The lack of capsule production by USA300 was in contrast to that observed among other MRSA and MSSA PFTs that were cultured, stored, and tested in an identical fashion. Importantly, this is consistent with the observations of Montgomery (2008), who reported a lack of capsule production by three USA300 isolates studied in an in vivo rat model of pneumonia. Our study suggests that the lack of capsule expression is common among clinical isolates of the most prevalent CA-MRSA clone, USA300, and among our strains, this clone universally reacted with 336PS antisera. For most isolates unrelated to USA300, the expression of capsule genes cap5 or cap8 was identified, and few MSSA isolates were noncapsulated.

Our study had several limitations that should be mentioned. Although isolates were collected from the military population living dispersed throughout the greater Washington, DC area, the size of our study was limited by the number of isolates that were recovered from pediatric patients during the study period. In spite of this, our data provide a unique look at S. aureus isolates from children in a large metropolitan area without previously reported epidemiology, as well as unique information about the capsular expression of USA300 isolates. A second limitation is that of the many potential virulence factors of S. aureus, only capsule and PVL were studied in all isolates. It is difficult to assess the independent contribution of these virulence factors without controlling for other toxins and determinants of pathogenicity, especially in the setting of a highly clonal population of isolates. A third important limitation is inherent in the comparison of study isolates with historical MRSA isolates; although the historical isolates were collected under a laboratory protocol, it is possible that differences in the collection methods may have introduced an unexpected bias in the types and severity of MRSA infections from which isolates were obtained.

In conclusion, PVL-containing USA300 S. aureus was the predominant MRSA clone among pediatric patients in the national capital area, and the introduction of this epidemic clone coincided with a marked increase in CA-MRSA disease in our patient population. All pediatric isolates contained either capsule type 5 or capsule type 8 genes, and over half of the culture-positive S. aureus clinical disease in our population was caused by diverse MSSA strains, the majority of which expressed either type 5 or type 8 CP. The remaining pediatric clinical disease isolates were predominantly unencapsulated serotype 336 strains of the PVL(+) USA300 CA-MRSA clone.


Work performed at: Center for Biologics Evaluation and Review, Bethesda, MD, USA and Walter Reed Army Medical Center, Washington, DC, USA.

The opinions or assertions contained herein are the private views of the authors and are not to be construed as official or as reflecting the views of the Department of Defense or the US Food and Drug Administration. Some of the authors are military service members. This work was prepared as part of their official duties. Title 17 USC §105 provides that ‘Copyright protection under this title is not available for any work of the United States Government.’ Title 17 USC §101 defines a US Government work as a work prepared by a military service member or employee of the US Government as part of that person's official duties.


We thank Fred Tenover, Linda McDougal, and Jean Patel (Centers for Disease Control) for technical assistance and control isolates, George McCracken (University of Texas Southwestern Medical Center) for providing control isolates, Jean Lee for capsule primer sequences, Amruta Kale for providing laboratory assistance, and Steve Rosenthal, Linda Lewis, and Rebecca Prevot for critical manuscript review.


  • Editor: Jacques Schrenzel


View Abstract