Molecular characteristics of community-acquired methicillin-resistant Staphylococcus aureus strains isolated from outpatients with skin and soft tissue infections in Wuhan, China

This study aims to investigate the antimicrobial susceptibility, molecular characteristics and virulence genes of community-acquired methicillin-resistant Staphylococcus aureus (CA-MRSA) isolates with skin and soft tissue infections (SSTIs). Outpatients with SSTIs visiting five medical and health institutions were enrolled from 2011 to 2013. Available S. aureus isolates were characterized by antimicrobial susceptibility testing, and detection of PVL genes. For CA-MRSA isolates, we performed typing of staphylococcal cassette chromosome mec (SCCmec), multi locus sequence typing (MLST) and carriage of 27 virulence genes. A total of 203 S. aureus strains were isolated from 1400 outpatients with SSTIs, and 21 (10.3%) were CA-MRSA isolates. The positive rate of PVL genes among S. aureus, CA-MRSA and methicillin-susceptible S. aureus (MSSA) isolates were 39.4%, 71.4% and 35.7%, respectively. CA-MRSA strains had greater sensitivity to non-β-lactam antimicrobial agents. All CA-MRSA isolates belonged to SCCmec IV and V, accounting for 47.6% and 52.4%, respectively. ST59 was the most common lineage accounting for 76.2%; ST59-SCCmec IVa-PVL-positive clone was found to be the predominant clone, accounting for 38.1%. All CA-MRSA isolates were found to be positive for one or more virulence genes, 28.6% of isolates carried PVL, seb, sek, seq, hla, hlb, hld and hlg-2. CA-MRSA infections were relatively uncommon in outpatients with SSTIs, but they carried many virulence genes, ST59-SCCmec IV a-PVL-positive clone was the predominant clone in Wuhan, China.

INTRODUCTION

Community-acquired methicillin-resistant Staphylococcus aureus (CA-MRSA) is an emerging threat worldwide, which is capable of causing severe, invasive infections, and most frequently skin and soft tissue infections (SSTIs), particularly in patients visiting hospitals (Fritz et al.2009; DeLeo et al.2010; Ray, Suaya and Baxter 2013). Over the past decade, there has been a dramatic increase in the number of patients with SSTIs in the outpatient setting. Between 1993 and 2005, the number of emergency department visits for SSTIs increased by more than 200% (Pallin et al.2008). Another study from the USA also found the number of S. aureus-SSTI incident (per 100 000 people) doubled from 57 in 2001 to 117 in 2009 (Suaya et al.2014). Increases in the number of SSTI incident appear to be driven by the dramatic increase in CA-MRSA skin infections (DeLeo et al.2010; Ray, Suaya and Baxter 2013).

It is crucial to have a better understanding of the epidemiology of CA-MRSA in order to establish public health interventions for MRSA control. The potential shift of important clones at the local and international levels is of great interest. Previous studies demonstrated that the most prevalent clones of CA-MRSA from SSTIs had unique geographic distributions. Clones with USA300 are mostly reported in the USA (Talan et al.2011), ST80 in Europe, ST59 in the Asia-Pacific region and ST30 worldwide, including in the USA, Europe, western Pacific area, Japan and Hong Kong(Ho et al.2008; Chuang and Huang 2013). In mainland China, data regarding CA-MRSA isolates from SSTIs is limited, with few reports from Wenzhou and Beijing (Wu et al.2010; Yao et al.2010; Zhao et al.2012; Yu et al.2015), most studies have shown that the currently widespread predominant clone in CA-MRSA isolates from SSTIs is the ST59-SCCmec IV. Some studies have shown that the ST239-MRSA-SCCmec III clone was also found to be the most prevalent clone among S. aureus isolates associated with SSTIs both in communities and hospitals. The most prevalent clone of CA-MRSA from SSTIs had unique geographic distributions, and the epidemic of CA-MRSA is evolving (DeLeo et al.2010). Whether this clone of CA-MRSA is still dominant or has been replaced in patients with SSTIs in our city needs to be determined. Furthermore, patients with CA-MRSA SSTIs have a high incidence of treatment failure and recurrence. Two recent studies involving urgent and primary care clinics found that 35% of patients with CA-MRSA SSTIs experienced treatment failure and 78% of patients reported a recurrent infection (Labreche et al.2013). Here, we performed a prospective observational study to investigate the epidemiology, antimicrobial susceptibility and molecular characteristics of CA-MRSA isolates with SSTIs in Wuhan, Hubei province, in central China. Moreover, as the virulence factors play an important role in the pathogenesis of CA-MRSA, we also investigated the prevalence of virulence factors. PVL is a major virulence factor in CA-MRSA, S. aureus SSTIs and severe deep-seated infections are frequently associated with strains harboring the PVL (Boyle-Vavra and Daum 2007). So we discussed the association of PVL gene expression with antimicrobial susceptibility, which may provide further insights related to clinical outcomes with these infections.

MATERIALS AND METHODS

Strain collection

We conducted a prospective prevalence study involving outpatients with SSTIs who visited the surgical clinic and dermatological clinic in three general hospitals (Wuhan No. 1 Hospital, Wuhan Fifth Hospital and Huangpi People's Hospital, Wuhan) and two Community Health Service Centers (Community Health Service Center Of Wansong Street and Community Health Service Center Of Xima Street, Wuhan). The study was carried out from 2011 through 2013.

The SSTIs in this study included abscess, carbuncles, furuncles, folliculitis, mastitis, infected wound and impetigo that required incision and drainage. Patients with severe SSTIs requiring hospitalization were not included in the study. The samples were collected using a pus swab from the SSTI site of patients. Only one specimen per patient was included. The specimens were transported in sampling tube with ice and sent to Wuhan Centers for Disease Prevention and Control for further analysis within 24 h of collection. Specimens were processed and cultured, S. aureus was identified at the microbiology laboratory using standard techniques. Meanwhile, the following data was collected using a standardized questionnaire from each patient: sex, age, present diagnose, site of infection, comorbidities, antibiotical use and risk factors for MRSA infection including recent surgery, wounds, catheter, hemodialysis and hospitalization history over the past years. MRSA isolates were initially identified using cefoxitin disks (30 mg) (Tianhe Ltd., Hangzhou, China) and further confirmed for the presence of the mecA gene by polymerase chain reaction (PCR) as described previously (McClure et al.2006).

CA-MRSA infection was diagnosed in an outpatient, who had none of the following risk factors for health-care-associated MRSA: hemodialysis, surgery, residence in a long-term care facility or treatment in hospital during the previous year, presence of a permanent catheter or percutaneous device at the time of culture, or previous isolation of MRSA (Chuang and Huang 2013).

The study was approved by institutional review board of Wuhan Centers for Disease Prevention and Control. The patient's written consent was obtained from all patients at the time of enrollment.

Antimicrobial susceptibility testing

The antimicrobial susceptibility profiles of all S. aureus isolates were determined in accordance with Clinical and Laboratory Standards Institute (CLSI 2010) guidelines, which used the Kirby–Bauer disk diffusion test for penicillin, cefoxitin, ampicillin, trimethoprim-sulfamethoxazole (TMP-SMX), doxycycline, ciprofloxacin, erythromycin, clindamycin, gentamicin, tetracycline and rifampicin (Tianhe Ltd., Hangzhou, China). Staphylococcus aureus ATCC 29213 was used as the control.

DNA extraction

DNA was extracted using DNA extraction kit (TaKaRa, Dalian, China) and used as the template in all PCR described in the following section.

Molecular typing methods

All the CA-MRSA isolates were characterized by means of staphylococcal cassette chromosome mec (SCCmec) types, multi locus sequence typing (MLST) and accessory gene regulator (agr) typing. MLST were performed as described by Enright et al. (2000). The seven housekeeping genes (arcC, aroE, glpF, gmk, pta, tpi and yqiL) were compared with known alleles from the MLST website (http://saureus.mlst.net) and the STs were determined based on this database. The SCCmec types were performed under the conditions as described by Milheirico, Oliveira and de Lencastre (2007b). The SCCmec IV isolates were further determined by the methods described by Milheirico, Oliveira and de Lencastre (2007a). The agr type was performed as described by Sakoulas et al. (2002).

Detection of virulence genes

All the S. aureus isolates were detected for lukS-PV–lukF-PV gene (which encode the PVL components S and F) with primers PVL-1 (5^′-ATCATTAGGTAAAATGTCTGGACATGATCCA-3^′) and NPVL-2 (5^′-GCATCAAGTGTATTGGATAGCAAAAGC-3^′) (Jarraud 2002). All the CA-MRSA isolates were screened by PCR for 27 specific virulence genes, as described in the following (Becker, Roth and Peters 1998; Monday and Bohach 1999; Jarraud 2002; Omoe et al.2002; Lovseth, Loncarevic and Berdal 2004; Sergeev et al.2004; Cremonesi et al.2005), comprising sequences specific for staphylococcal enterotoxins (sea, seb, sec, sed, see, seg, seh, sei, sej, sek, sel, sem, sen, seo, sep and seq), toxic shock syndrome toxin 1 (tst), exfoliative toxin genes (eta and etb), hemolysin genes (hla, hlb, hld, hlg and hlg-2), leukocidin genes (lukE-lukD and lukM) and epidermal cell differentiation inhibitor (edin). They were tested by six multiplex PCR. The multiplex PCR included multiplex I (sea, seb, seg and sen), II (sec, seh, sej, sek and seo), III (sed, see, sei, sel, sep and seq), IV (eta, etb, lukE-lukD and lukM), V (hla, hld, hlg-2 and edin) and VI (tst, hlb, hlg and sem).

Statistical analysis

Statistical analyses were performed using SPSS, version 16. We used the Student t-test for normally distributed variables and the χ² test for the homogeneity of proportions for categorical data. All hypotheses were two-tailed and were considered significant if P < 0.05.

RESULTS

Population diversity

The results of population diversity are listed in Table 1. A total of 1400 outpatients aged 1 to 91 (median, 48) years with SSTIs were recruited, about 56.4% of patients were males, 38.8% of patients had infected wounds and 35.3% had abscess. A total of 203 (14.5%) S. aureus were isolated, 21 (10.3%) were CA-MRSA and 182 (89.7%) were MSSA isolates. Among CA-MRSA patients, 61.9% were males, 47.6% had abscess. Among MSSA patients, 58.2% were males, 29.7% had abscess and 29.1% had infected wounds. The infection type among CA-MRSA group were different from MSSA group (χ²Ã‚ = 18.7, P < 0.01), no significance was found between two groups with age, sex and infection location.

The prevalence of PVL genes

Among the S. aureus isolates, 39.4% (80/203) carried PVL genes. Both MSSA and CA-MRSA isolates can harbor PVL. The positive rate of PVL genes among CA-MRSA isolates (71.4%, 15/21) was higher than among MSSA isolates (35.7%, 65/182), it was considered statistically significant (χ²Ã‚ = 10.1, P < 0.01). The mean age among PVL-positive patients was younger than PVL-negative group (36.5 ± 19.7 versus 43.3 ± 20.6, t = –2.4, P < 0.05). The infection type and location among PVL-positive patients were different from PVL-negative, such as abscess, furuncles, carbuncles, folliculitis and mastitis were more frequently harboring the PVL. No significance was found between two groups with sex (Table 1).

Antimicrobial susceptibility

The results of antimicrobial susceptibility are listed in Table 2. The CA-MRSA isolates exhibited high rates of resistance to β-lactam antibiotics, but had greater sensitivity to non-β-lactam antimicrobial agents. MSSA isolates were susceptible to most antimicrobial drugs. All the CA-MRSA isolates were resistant to penicillin and ampicillin, which was much higher than MSSA isolates (P < 0.01), whereas to TMP-SMX, doxycycline, gentamicin, tetracycline and rifampicin were less than 10.0%, there was no significant difference between them.

Among the S. aureus isolates, PVL-positive group was also more resistant to β-lactam antimicrobial drugs, erythromycin, ciprofloxacin and gentamicin than the PVL-negative group, this was all considered statistically significant (P < 0.01). There was no significant difference for TMP-SMX, clindamycin, tetracycline and rifampicin. Among the MSSA isolates, antimicrobial susceptibility with comparison between PVL-positive and PVL-negative was similar to comparison between the PVL-positive S. aureus isolates and PVL-negative group.

Molecular characteristics

The results of the multiple PCR analysis for all 21 CA-MRSA isolates are shown in Table 3. All 21 CA-MRSA isolates were characterized through SCCmec typing. Two types were found; 11 isolates were SCCmec V and 10 isolates were SCCmec IV, of which two SCCmec subtypes were found: SCCmec IVa and IVh. All isolates were classified as agr I and III. The agr I, accounting for 90.5% (19/21) was the most common group. According to MLST analysis, the CA-MRSA isolates were classified into five ST types, including ST59, ST630, ST2580, ST88 and ST398. ST59 was the most common lineage accounting for 76.2% (16/21).

On the basis of MLST and SCCmec typing, six types were received, ST59-SCCmec IVa clone (42.9%) was found to be the most prevalent clone among CA-MRSA isolates, followed by ST59-SCCmec V (33.3%), ST630-SCCmec V (9.5%), ST2580-SCCmec IVh (4.8%), ST88-SCCmec V (4.8%) and ST398-SCCmec V (4.8%).

Virulence genes

All the CA-MRSA isolates were found to be positive for one or more toxin genes (median, 7; range, 1 to 11). A total of 10 toxin genes were not found, including sed, see, seh, sej, sen, seo, eta, etb, lukM and edin. Combinations of SE genes were highly conserved in CA-MRSA strains. A total of 16 (76.2%) isolates were found to be positive for more than one SE gene. The SE genotypes were classified into 10 genotypes. A total of 10 (47.6%) isolates harbored seb, sek and seq, all belonging to ST59.

All the CA-MRSA strains carried at least one hemolysin gene, the positive rates of hld, hla, hlg-2 and hlb were 90.5%, 81.0%, 76.2% and 71.4%, respectively. hla-hlb-hld-hlg-2 was the most common gene profile which was harbored by 14 strains (66.7%). A total of 28.6% of isolates carried PVL, seb, sek, seq, hla, hlb, hld and hlg-2, all belonging to ST59.

A total of 90.0% of SCCmec IV and 54.6% of SCCmec V isolates were PVL positive. A total of 86.7% of ST59 isolates carried PVL genes. ST398 was the lineage that carried only one toxin gene that was hld. ST2580 isolates carried 11 kinds of toxin genes, which included sec, sel, seg, sei, sem, seq, hla, hld, hlg and tst, in addition to PVL. ST630 isolates had hla, hlb, hld and hlg-2 gene, in addition to PVL.

In conclusion, ST59-SCCmec IVa-PVL-positive clone was found to be the predominant clone, accounting for 38.1%, followed by ST59-SCCmec V-PVL-positive clone (23.8%).

DISCUSSION

This prospective study of CA-MRSA strains in outpatients with SSTIs in Wuhan provided several findings. First, the prevalence of CA-MRSA was low among outpatients with SSTIs. Second, CA-MRSA was broadly resistant to almost all β-lactam antibiotics, whereas they had greater sensitivity to non-β-lactam antimicrobial agents. Third, ST59-SCCmec IVa-PVL-positive clone was the predominant clone among CA-MRSA isolates.

Over the past decade, CA-MRSA has become a notable public health problem, which has become a major cause of SSTIs (Edelsberg et al.2009). In a study of Atlanta outpatients (n = 389), presenting with SSTIs, among all S. aureus isolates, 63% were CA-MRSA (King et al.2006). In a study performed in Argentina, a total of 311 patients enrolled with SSTIs, and 70% had CA-MRSA (Lopez Furst et al.2013). Our study revealed that among 1400 outpatients with SSTIs, 14.5% were S. aureus isolates, of which 10.3% were CA-MRSA. These results are consistent with previous studies performed in China (Zhao et al.2012). Until recently, CA-MRSA infections were relatively uncommon in outpatients with SSTIs in Wuhan, China.

This strongly suggests that TMP-SMX and clindamycin are appropriate first-line therapeutic options for the treatment of SSTIs due to CA-MRSA in Europe (De Angelis et al.2011). The Infectious Diseases Society of America recommends that clindamycin (A-II), TMP-SMX(A-II) and doxycycline(A-II) are appropriate for empirical coverage of CA-MRSA in outpatients with SSTIs (Liu et al.2011).

In our study, CA-MRSA was broadly resistant to almost all β-lactam antibiotics, whereas they had greater sensitivity to non-β-lactam antimicrobial agents, especially TMP-SMX and doxycycline, whether the isolates harbored PVL, the rate of susceptibility to clindamycin was low. These findings may be explained by the high rate of antibiotic use in the Chinese community (Aires de Sousa et al.2003) and they are indicative of the presence of strong selective pressure from antimicrobial use in the community.

PVL might be the common factor responsible for the observed increased virulence, among the S. aureus and MSSA isolates. When the isolates harbored PVL, the resistance of antimicrobial agents increased, especially β-lactam antimicrobial drugs, erythromycin, ciprofloxacin and gentamicin, but the rate of susceptibility to TMP-SMX and doxycycline remained high. So for the outpatients with SSTIs in China, we recommend TMP-SMX and doxycycline for empirical coverage, but not clindamycin.

PVL is widely associated with the presence of SCCmec IV and sporadically with SCCmec V_T or V but not SCCmec types I, II or III (Boyle-Vavra and Daum 2007). In our study, all CA-MRSA isolates belong to SCCmec IV and V, however, 90.0% of SCCmec IV and 54.6% of SCCmec V isolates were PVL-positive. This finding suggests that PVL and SCCmec IV may confer a selective advantage for CA-MRSA pathogens (Vandenesch et al.2003).

As we have already mentioned, ST59 is the most prevalent clone in the Asia-Pacific region, and ST30 is more prevalent worldwide, including USA, Europe, western Pacific area, Japan and Hong Kong. Our data revealed that the ST59 clone was still the most prevalent clone among CA-MRSA isolates, and we did not find any ST30 isolates in Wuhan, China.

According to the study by Chambers (2005), PVL is the principal virulence factor responsible for the epidemic spread of many CA-MRSA strains. However, more than 80% CA-MRSA ST59 clones in Taiwan are SCCmec V_T -PVL-positive and SCCmec IV-PVL-negative (Huang and Chen 2011), most ST59 clones in Hong Kong belong to SCCmec IV-PVL-positive (Ho et al.2008). In our study, the majority of these clones are SCCmec IVa-PVL-positive, and the others are SCCmec V-PVL-positive. Thus, the CA-MRSA ST59 clones identified in mainland China might differ from those identified in Taiwan, and are probably similar to those in Hong Kong.

MRSA clonal type ST398 is usually associated with livestock contact, especially pigs (Pan et al.2009; Becker et al.2015). However, in our study, the patient with ST398 was a 25-year-old man admitted to a surgical department because of a carbuncle in his left leg. He lived and worked in urban regions, and as such, had no chance of coming into livestock contact. This finding is contradictory with Becker and Pan's observation, but is consistent with a similar study previously conducted in China (Zhao et al.2012).

All the CA-MRSA isolates were found to be positive for one or more virulence genes, with a median of seven virulence genes, most commonly PVL, the enterotoxins(seb, sek and seq) and hemolysin genes (hla, hlb, hld and hlg-2).

The positive rate of PVL genes in CA-MRSA infection was high (71.4%), which was also higher than the previous study in China (Yao et al.2010; Zhao et al.2012), likely because of its acquisition of the PVL genes, a virulence determinant with well-established linkage to SSTIs (Diep et al.2006). The distribution of some virulence genes, especially enterotoxin genes, was correlated with different MRSA lineages (Diep et al.2006; Loughman et al.2009). The molecular event that is likely responsible for the evolution of the epidemic phenotype among early MSSA strains and early MRSA strains is the acquisition of the staphylococcal pathogenicity island (SaPI) 3, which contains the genes for three potent enterotoxins, seb, sek, and seq (Yarwood et al.2002). In this study, we found that SaPI3 were most frequently detected in CA-MRSA isolates.

In summary, this study provides the baseline information of the epidemiology, antimicrobial susceptibility and molecular characteristics of CA-MRSA in outpatients with SSTIs in Wuhan. CA-MRSA infections were relatively uncommon in outpatients with SSTIs, but they carried many virulence genes. ST59-SCCmec IVa-PVL-positive clone was the predominant clone, when the S. aureus harbored mec or PVL genes, the resistance of some antimicrobial agents increased, we will pay attention to antimicrobial use for empirical coverage in the community.

We wish to thank the following institutions that participated in this study as major collaborators: Wuhan Centers for Disease Prevention and Control (Yan Xiong, Zhi Chen), Community Health Service Center of Wansong Street (Zhibing Yang) and Community Health Service Center of Xima Street (Zhizhong Xia), Jianghan district Centers for Disease Prevention and Control (Jianlong Liu), Huangpi district Centers for Disease Prevention and Control (Mo Han), Jiang'an district Centers for Disease Prevention and Control (Xiaohong Ding).

FUNDING

This work was supported by Public Health Project of Wuhan Health Bureau (No.WG12D03).

Conflict of interest. None declared.

REFERENCES