Virulence Characteristic and MLST-agr Genetic Background of High-Level Mupirocin-Resistant, MRSA Isolates from Shanghai and Wenzhou, China

The emergence and prevalence of high-level mupirocin-resistant, methicillin-resistant Staphylococcus aureus (MuH MRSA) is challenging the eradication of MRSA nasal carriage and the treatment of skin and soft tissue infections. To understand the potentially pathogenetic capacity and the genetic basis of MuH MRSA, it is important to have a detailed knowledge of the molecular traits of this organism. Fifty three MuH MRSA isolates were gathered from Shanghai (28 isolates) and Wenzhou (25 isolates) in China. These isolates, consisting of 27 different PFGE-SCCmec-spa patterns, were examined by PCR for 35 virulence genes and further typed using agr (accessory gene regulator) typing and MLST (multilocus sequence typing). All 53 strains were positive for the genes hlg/hlg variant and icaD, and negative for seb, sed, see, seh, eta, etb, hld, cap-5, and ACME-arcA. Compared with Wenzhou isolates, Shanghai isolates were more likely to carry seg (P = 0.002) and several other genes which were not found in Wenzhou strains such as sec, sei, tst (P<0.001 each), and pvl (P = 0.012), and less likely to contain sea (P<0.001), cna (P = 0.031), and efb (P = 0.045). MLST and agr typing showed that ST239-agr1, ST5-agr1, and ST239-agr2 were the common lineages in MuH MRSA isolates from these two different regions. Our results indicated that MuH MRSA strains from two different geographic regions of China have differences in distribution of some virulence genes, while their major MLST-agr genetic backgrounds were accordant.


Introduction
Methicillin-resistant Staphylococcus aureus (MRSA) is still a leading pathogen of nosocomial infections in China. Because of its ability to produce virulence factors causing a variety of infections and its serious multidrug resistance, MRSA is one of the most frightening bacteria.
The pathogenesis of S. aureus infections is involved in the expression of a wide range of virulence factors associated with attachment, persistence, evading/destroying host defenses, tissue invasion/penetration and toxin-mediated disease [1]. However, a majority of the serious S. aureus infections are caused by the combined actions of several virulence factors. Otherwise, clinical isolates linked to S. aureus infection may be naturally deficient in a scale of putative pathogenic determinants. Therefore, strains causing S. aureus disease have variable combinations of virulence genes [2].
Mupirocin is a very effective topical antibiotic used to treat staphylococcal skin and soft tissue infections and eliminate MRSA from colonized nasal passages. However, the resistance to mupirocin has occurred, and its spreading is worrisome, with mupA-mediated high-level resistance [3]. The mupA gene is usually located on a large conjugative plasmid capable of mediating the co-transfer of other resistance genes. If high-level mupirocin resistance could not be controled, a highly effective means of decolonization of MRSA may be lost. Thus the high-level mupirocin resistance MRSA (MuH MRSA) may spread widely, and causes a range of infections. Therefore it is necessary to know the traits of virulence of clinical MuH MRSA strains more comprehensively.
The purpose of this study was to investigate the prevalence of putative virulence genes in a clinical population of MuH MRSA isolates with known types of PFGE-SCCmec-spa from two different geographic areas in China, and explore the possible difference in virulence determinants combination between the strains from the both origins. Finally we further determined the genetic characteristics of these isolates by agr (accessory gene regulator) typing and multilocus sequence typing (MLST).

MuH MRSA Isolates
Fifty three MuH MRSA isolates were selected from 5 university hospitals collection in Shanghai (n = 28, 4 hospitals) and Wenzhou (n = 25, 1 hospital), China. This collection comprised 803 MRSA, which were isolated from various clinical specimens of individual inpatients from August 2005 to May 2008. Most of the MuH isolates were gathered from respiratory samples (86.8%). Intensive care units, surgical wards, burn wards and neurology wards were the major hospital units affected by MuH MRSA. All isolates had been previously described by PFGE, SCCmec and spa typing and these results are shown in Table 1 [4]. Because of being focused on bacteria, this study was exempted from review by the Ethics Committee of Shanghai First People's Hospital.

DNA Sequencing
One randomly chosen amplicon for the each gene tested was sequenced on an ABI 3730 sequencer (Applied Biosystems) by Shanghai Invitrogen Biotech to confirm that primers amplified the expected genes.

Statistical Analyses
Pearson's chi-square test or Fisher's exact test if necessary was used to compare distribution of the virulence determinants in clinical MuH MRSA strains investigated (SPSS version11.5). All statistical tests were two tailed, with P,0.05 considered statistically significant.

Discussion
S. aureus produces numerous extracellular proteins which involve in the ability of this organism causing disease in the mammalian host. In this study, we detected six groups of pathogenic genes for 53 clinical MuH MRSA isolates from Shanghai and Wenzhou regions, China ( Table 2). The first of these were the genes involved in toxin mediated disease. Previous study showed toxins encoded by sea and sec tend to generate higher immune responses resulting in host tissue damage than do other enterotoxins [16]. However, we did not find the existence of sec in Wenzhou isolates. It is generally believed that the existence of the enterotoxin gene cluster (egc, containing seg and sei) is not related with severe infections, but probably contributes to the colonization potential of an S. aureus strain [16,17,18]. Because the toxins transcripted by the egc element appear to be generated in lower amounts compared to the other well-studied enterotoxins which may permit strains carrying the egc determinants to live together with healthy hosts [17]. However, Morgan's report showed the products of egc may play a part in some cases, especially in immuno-compromised patients [19]. Due to the genes seg and sei being located on egc element [20], the combined occurrence of the toxin gene pair can generally be observed. Notably, 25.0% (5/20) seg-positive isolates were not confirmed with the fixed seg-sei combination in our study. The sed and sej genes are encoded by a plasmid pIB485 [21]. However, the coexistence of the two determinants cannot also be certified with 2 sej-positive isolates ( Table 4). The possible explanation for these opposite results is the existence of still-unknown variants of sei and sed.
The second group consisted of the determinants involved in attachment (fnbA, fnbB, clfA, clfB, cna, bbp and ebpS). The expression products of these genes were termed MSCRAMMs (microbial surface components recognizing adhesive matrix molecules). A great majority of our strains harbored the genes encoding FnbA, ClfA, ClfB, Cna and EbpS (Table 3), which may make them possess the ability to bind to fibronectin, fibrinogen and fibrin [22], collagen substrates and collagenous tissues [23], and soluble tropoelastin [24]. There is evidence that Bbp is a key factor in bone and joint infections produced by S. aureus [25,26]. However, the positive rate of bbp implied most MuH MRSA strains of our collection might not have the ability to cause those diseases.
The third group included the genes pvl, hlg/hlgv, cap-5, cap-8, map and sbi. The toxins encoded by pvl and hlg are leukotoxic for neutrophils and macrophages [27]. Capsular polysaccharide (Cap) can protect the bacterium from phagocytic uptake and increases microbial virulence. Map may potentiate S. aureus survival by affecting protective cellular immunity [28,29]. Sbi has the ability to hinder phagocytosis and is implicated in blood coagulation [30]. In this study, we did not see any significant difference in the prevalence of this group of genes except pvl between the Shanghai and Wenzhou strains (Table 3).
Ssp can degrade host cell receptors and/or bacterial adhesins, and promote the spread and transmission of infection [31]. Sak may mediate bacterial invasion into the host tissues and enhance bacterial resistance to phagocytosis [32]. According to Table 3, most of our isolates may have the functions mentioned above.
The genes icaA and icaD belong to ica operon (icaADB and C), which is revealed to induce polysaccharide intercellular adhesin (PIA, associated with biofilm formation) synthesis in staphylococcus. However, expression of icaA alone leads only to low  production of PIA. It was demonstrated that coexpression of icaA with icaD will promote the biosynthesis of capsular polysaccharide [33]. In this study, a small part of our strains were negative for icaA (Table 3). This phenomenon may be the deficiency of icaA gene or the existence of point mutations in the primer binding sites causing a negative PCR reaction. It has been hypothesized that Efb might benefit the bacterium by interacting with fibrinogen and preventing the clotting process, thereby delaying the healing process [34]. sar (containing 3 transcripts designated sarA, sarB and sarC) is a global regulatory locus, and controls the production of many virulence factors in S. aureus. Among which the sarA encodes the major effector molecule [35]. ACME can encode an arginine deiminase pathway and an oligopeptide permease system that could enhance the ability of S. aureus, especially for USA300 clone, to grow and survive within the host [36]. Table 3 showed the genes efb and sarA were prevalent in most our strains; however, no isolate studied possessed the founction of ACME gene.
agr is another important global regulatory locus controlling the production of most staphylococcal exoproteins. In S. aureus, four different agr alleles have been described and agr is regarded as a slowly evolving genetic marker to investigate hospital-acquired MRSA [37]. A report by Van Leeuwen et al. [38] suggested that the agr-1 is the most prevalent agr group in MRSA isolates. We found similar in our isolates (71.7%, 38/53) ( Table 4) and by Liu et al. [39] (96.4%, 134/139) in Beijing MRSA strains. Besides agr-1, 28.3% (15/53) of our strains belonged to agr-2 or agr-3. Previous investigation has demonstrated the association between agr-3 and TSST-1 [40], whereas the findings of this study revealed that only 6 of 14 tst-positive strains were agr-3, and the remainder belonged to agr-1 (7 isolates) and agr-2 (1 isolates) (Table 4). Likewise, the data reported by Ben Nejma et al. [41] could also not reveal the relationship between agr-3 and this toxin.
For the determinants involved in toxin mediated disease, all the isolates from the two regions contained at least one gene of this group. However, there were 6 Shanghai strains and 4 Wenzhou isolates that did not carry any staphylococcal enterotoxin gene (Table 4). Table 4 shows there were 3 Shanghai isolates lacking genes involved in tissue invasion/penetration and the 'others' group genes (sarA and efb). In addition, 1 Shanghai isolate and 1 Wenzhou isolate were only absent from the genes divided into the 'others' group. In Shanghai and Wenzhou strains hosting #2 genes involved in toxin mediated disease, most of them also carried fewer other virulence genes (Table 4). In addition, Some MuH isolates yielded different virulence genes patterns even though they were of the same PFGE-SCCmec-spa-agr-ST type, and the same virulence determinants combination can belong to different genotype patterns (Table 4). In respect to these results, the possible explanation may be that some virulence determinants are located on mobile genetic elements and can be horizontally transmitted among bacteria. In summary, our study showed there were some differences in virulence profiles between MuH MRSA isolates from Shanghai and Wenzhou, and the differences mainly existed in the genes sea, sec, seg, sei, tst, cna and pvl. The results of MLST and agr typing displayed that the two regions' isolates were genetically less diverse.