Characterization of Community Acquired Staphylococcus aureus Associated with Skin and Soft Tissue Infection in Beijing: High Prevalence of PVL+ ST398

Adult community-associated methicillin-resistant Staphylococcus aureus (CA-MRSA) and methicillin-susceptible S aureus (CA-MSSA) skin and soft tissue infection (SSTI) in China is not well described. A prospective cohort of adults with SSTI was established between January 2009 and August 2010 at 4 hospitals in Beijing. Susceptibility testing and molecular typing, including multilocus sequence typing, spa, agr typing, and toxin detection were assessed for all S. aureus isolates. Overall, 501 SSTI patients were enrolled. Cutaneous abscess (40.7%) was the most common infection, followed by impetigo (6.8%) and cellulitis (4.8%). S. aureus accounted for 32.7% (164/501) of SSTIs. Five isolates (5/164, 3.0%) were CA-MRSA. The most dominant ST in CA-MSSA was ST398 (17.6%). The prevalence of Panton-Valentine Leukocidin (pvl) gene was 41.5% (66/159) in MSSA. Female, younger patients and infections requiring incision or drainage were more commonly associated with pvl-positive S. aureus (P<0.03); sec gene was more often identified in CC5 (P<0.03); seh gene was more prevalent in CC1 (P = 0.001). Importantly, ST59 isolates showed more resistance to erythromycin, clindamycin and tetracycline, and needed more surgical intervention. In conclusion, CA-MRSA infections were rare among adult SSTI patients in Beijing. Six major MSSA clones were identified and associated with unique antimicrobial susceptibility, toxin profiles, and agr types. A high prevalence of livestock ST398 clone (17.1% of all S. aureus infections) was found with no apparent association to animal contact.


Introduction
Methicillin resistant Staphylococcus aureus (MRSA) infection represents a significant cause of morbidity and mortality in both hospital and community settings. Community-associated MRSA (CA-MRSA) has become increasingly important as a cause of skin and soft tissue infections (SSTIs), particularly in patients presenting to emergency departments [1][2][3]. In China, HA-MRSA has been extensively studied during the past years with ST239-SCCmec III as the predominant clone [4,5]. However, data regarding CA-MRSA was limited, with mainly reporting from children. According to these studies, the prevalence of CA-MRSA from SSTI among children was 4%, and ST59-MRSA-IV-t437 was the most common clone among CA-MRSA isolates [6]. To our knowledge, no prospective clinical studies of SSTIs from adults caused by CA-MRSA have been reported in China.
The origin of the staphylococcal cassette chromosome mec (SCCmec) in CA-MRSA is still unknown. It is hypothesized that MRSA probably originated through the transfer of SCCmec into extant MSSA lineages with a genetic background common to MRSA clones. Therefore, it is very important to investigate the clonal structure of MSSA in each country or region, and to compare with MRSA. Moreover, as the virulence factors play an important role in the pathogenesis of community-acquired S. aureus infections, we also investigated the prevalence of virulence factors, including Panton-Valentine Leukocidin (pvl), staphylococcal enterotoxins C (sec) and staphylococcal enterotoxins H (seh) in these S. aureus isolates and their relationship with the genetic background.
In this study, we prospectively enrolled 501 SSTI patients from 2 teaching hospitals and 2 community hospitals in Beijing in order to investigate the prevalence, clinical and molecular characteristics of SSTIs in adults caused by S. aureus, with emphasis on the prevalence of CA-MRSA in adult SSTIs in China. To our surprise, we found that the livestock-associated clone ST398 in MSSA was the most prevalent with unique antimicrobial susceptibility, toxin profiles, and agr types.

Enrollment of Patients and Isolation of Bacterial Isolates
From January 2009 to August 2010, consecutive outpatients with SSTIs were prospectively enrolled at the surgical clinic and dermatological clinics in two teaching hospitals (Peking Union Medical College Hospital and Beijing Chao-Yang Hospital) and two community hospitals (Beijing Pinggu Hospital and The First Hospital Affiliated to the People's Liberation Army General Hospital). The annual visits of surgical clinic and dermatological clinics of these four hospitals ranged from 4000 to 6,000. Community acquisition of S. aureus was defined as a positive culture from outpatients, or inpatients within 48 hours admission if they had no risk factors for healthcare-associated acquisition such as recent hospitalization, surgery, hemodialysis, the presence of any permanent in-dwelling catheter or percutaneous medical device, or residence in a long-term care facility. A Case Report Form (CRF) completed for each patient that included demographic information, clinical symptoms, laboratory findings, the type of infection diagnosed, all antibiotic use, and clinical outcome. Specimens were collected from infection sites of every patient enrolled and cultured on blood agar and Eosin methylene blue agar. According to the colony morphology and Gram stain, rapid methods were used for S. aureus identification [7]. MRSA isolates were initially identified using cefoxitin and oxacillin disks (30 mg, Oxoid, Cambridge) and confirmed for the presence of the mecA gene by polymerase chain reaction (PCR) as described previously [8].
Institutional Review Board approval was obtained before this study was begun. Written informed consent was obtained from all patients at the time of enrollment.

Antimicrobial Susceptibility Testing
Antimicrobial susceptibility profiles of S. aureus isolates were determined by the agar dilution method on Mueller-Hinton agar, according to the recommendations and definitions from Clinical and Laboratory Standards Institute (CLSI) [9]. Antimicrobial

Molecular Typing Methods
All of the S. aureus isolates were investigated by multilocus sequence typing (MLST), spa typing, and accessory gene regulator (agr) typing. Suspensions of overnight S. aureus cultures on blood agar were lysed by lysostaphin for phenol-chloroform extraction of the genomic DNA, which was reconstituted in 1 ml Tris-EDTA buffer for PCR reactions. MLST was carried out as described previously [10] comparing sequences of the PCR products to the MLST website (http://saureus.mlst.net) with clustering of related STs into clonal complexes (CCs) utilizing the online eBURST program [11]. Purified PCR products of spa were sequenced and repeat patterns and spa types were assigned from the spa database [12]. SCCmec typing was performed on the CA-MRSA isolates using the multiplex PCR protocol of Oliveira et al [13]. Nontypeable (NT) types were those differing from the standard types. International clones of SCCmec types I to V were used as quality controls. agr typing was performed as previously described method [14].

Statistical Analysis
The x 2 test with Yastes's correction or Fisher's exact test using SPSS, version 13.0 (SPSS, Chicago, IL, USA) were used for analyzing the quantitative variables. A P value of #0.05 was considered statistically significant. All susceptibility data and molecular test results were analyzed using WHONET software, version 5.6.

Clinical and Microbiological Characteristics of SSTIs
From January 2009 to August 2010, a total of 501 SSTI cases were enrolled in this study. Demographic details, clinical and laboratory features, and antibiotic usage are shown in Table 1. Of all 501 cases enrolled, S. aureus was the most frequent organism, accounting for 32.7% (164 isolates) of all SSTIs.
Overall, agr type I was identified as the predominant type, accounting for 64.6% (106/164) of all S.aureus isolates, followed by type III (14.0%), type II (13.4%), and type IV (7.9%). No significant difference on the prevalence of pvl was found among four agr types. Interestingly, none of the isolates with agr IV carried sec gene, while 59.1%-60.9% of the isolates with agr type II and III produced SEC (P,0.001). agr III had higher prevalence of seh gene than others (39.1% vs. 2.8%-7.7%, P,0.001). In addition, the isolates with agr III and IV carried more tsst-1 genes (15.4%-17.4%) than those with agr I (1.9%, P = 0.005).
The 5 patients with CA-MRSA infections occurred during work or daily activities without any identifiable risk factors. Two of them were farmers, who developed infection after trauma during work. The third was a female worker who had mastitis, the fourth was a retired woman who developed spontaneous abscesses on her check, and the last one was a 25-year old exchange student from Turkey with a spontaneous abscess on the abdomen. All of the five patients were treated with antibiotics and surgical drainage. Notably, although the antibiotics were later shown to be inactive against MRSA, all 5 patients cured 5 to 7 days after surgical intervention. The clinical and molecular characteristics of 5 CA-MRSA isolates are shown in Table 3.

Comparison of Clinical Characteristics between pvlpositive and pvl-negative S. aureus SSTIs
Overall, 66 of 159 MSSA (41.5%) and 2 of 5 CA-MRSA (40.0%) isolates harbored the pvl gene. Female patients were found more commonly in pvl-positive group than in pvl-negative group (47.1% vs. 28.7%, P = 0.017), and the mean age among pvlpositive patients was younger than pvl-negative group (37.8616.4 vs. 45.5621.0, P = 0.014). In addition, the percentage of patients who needed surgical intervention was significantly higher in pvlproducing S. aureus SSTI than that in pvl-negative group (60.7% vs. 42.0%, p = 0.032). However, no significance was found between two groups with underlying diseases, types of infections, severity of illness and clinical outcomes, such as symptoms alleviation time and length of stay in hospital,

Comparison of Clinical and Molecular Features among the Major MLST Clones
Comparison of clinical features, toxins and agr types was performed among the six major clones (Table 4). More patients in ST59 group needed incision or drainage (p,0.001). Moreover, significant difference was found on antimicrobial susceptibility to erythromycin, clindamycin, tetracycline and gentamicin among the 6 CC clones (P values, 0.006-0.039, Table 4). ST59 was more resistant to erythromycin, clindamycin and tetracycline, while CC5 was more resistant to gentamicin.

Discussion
In this study, we have three novel findings: 1) the prevalence of CA-MRSA was low among adults with SSTIs; 2) six major MSSA clones (ST398, ST7, ST1, ST59, and ST5) were responsible for SSTIs in adults in Beijing, China; 3) ST398 was the most prevalent clone among MSSA isolates.
Our previous study showed that the major HA-MSSAs belonged to the same clones (ST398, ST7, ST1, ST59, and ST5) suggesting a similar epidemiology for both hospital acquired infections and community acquired infections [4]. We confirmed the presence of ST398 clone in the community [6,16]. Similar findings came from European researchers who found the percentage of ST398 in MRSA increased from 0 in 2002 to 30% in 2007 in Netherlands and from 13% in 2005 to 22.4% in 2008 in Germany [17,18], suggesting both regional variation as well as evolution in S. aureus ST398 over time.
ST398 is usually associated with livestock infection in pigs and people exposed to animals [19], however, we did not find any association between livestock contact and ST398 infection.Whole genome sequencing of European ST398 has revealed that it lacked virulence factors such as enterotoxins and phage-encoded toxins [20]. Our findings are novel because we find a great number of strains (64.3%) harbored pvl gene.  S. aureus strains with pvl are associated with abscess formation and tissue necrosis. Our study indicated that the prevalence of pvl in CA-MSSA infection was high (41.5%), which was different from the previous study on children where only 4.2% of strains carried pvl [6]. The difference in the distribution of major clones between adults and children may explain this difference in pvl prevalence between two studies. Similarly, past investigation has found the presence of pvl to be associated with a enhanced inflammatory response and localized infections [21], which is in concurrent with our findings that more pvl + patients needed incision or drainage. Another study from New York also found the presence of PVL as a significant predictor for incision and drainage for MSSA infection, and they found patients infected with pvl + S. aureus were significantly younger than those infected with pvl 2 S. aureus [22], similar to our study.
The agr locus belongs to the core variable genome and is thus linked to CCs. Our molecular testing is in agreement with the findings of other studies. For example, we observed that agr-I was the most common type and linked to ST22, ST45, ST7, ST188, and ST59; agr-II was present in CC5 and CC15; agr-III was associated with ST1 and CC30; and agr-4 was detected in CC121 [23,24].
In summary, this study provides the baseline information of the epidemiology and molecular characteristics of MSSA and CA-MRSA in adults with SSTIs in Beijing -currently most disease is caused by MSSA. We identified 6 major MSSA clones that were responsible for SSTIs in the community with ST398 the most prevalent clone. Though the prevalence of CA-MRSA in Beijing is low at present, the possibility that CA-MRSA could be imported from other countries, or the potential that MSSA may acquire the SCCmec elements makes it important for continuous surveillance of S. aureus infections.