Molecular epidemiology and virulence factors of methicillin‐resistant Staphylococcus aureus isolated from patients with bacteremia

Abstract Background The various virulence factors of methicillin‐resistant Staphylococcus aureus bacteremia (MRSAB) are associated with a high mortality rate worldwide. Further studies are warranted to confirm the significant relationship between the strains and virulence genes. Here, we prospectively investigated the molecular characteristics underlying the genotypes and virulence factors of MRSA isolated from patients with bacteremia. Methods We collected 59 MRSA isolates from adult patients with bacteremia. Antimicrobial susceptibility results were obtained with the Vitek2 automated system. Genotypes were identified with multi‐locus sequence typing (MLST) and pulse‐field gel electrophoresis (PFGE), and 21 virulence genes were detected with polymerase chain reaction (PCR). Results The 59 MRSA isolates mainly comprised ST5 (n = 31, 52.5%) and ST72 (n = 22, 37.2%). Most ST5 isolates and all ST72 isolates were clustered into one and two PFGE groups, respectively. The mean number of virulence genes was higher in ST5 than in ST72. Sel was more frequently detected in ST5 than in ST72, whereas sec and sed were found only in ST5. ST5 had significantly higher resistance against many antibiotics than ST72. Conclusion Most MRSA isolates causing bacteremia were ST5 (CC5) and ST72 (CC8), and those belonging to the same STs were divided into only a few PFGE groups. ST5 was associated with higher antibiotic resistance and staphylococcal superantigen toxin genes, than ST72, which may be related to its higher virulence.

patients without an indwelling catheter or percutaneous device. 20 Healthcare-associated (HA) infections included those that did not meet these criteria. 21 This study was approved by the Institutional Ethics Review Board of Chungnam National University Hospital (IRB No. 2018-05-040).
No informed consent was acquired because all isolates were generated and analyzed as a part of microbiological diagnostics and therapeutic purpose.

| Multi-locus sequence typing (MLST)
Multi-locus sequence typing was conducted for all isolates as previously described. 22 The sequence types (STs) for each isolate were determined by comparing the sequence of each locus with the reference sequence in the S aureus MLST database (https ://pubml st.org).
Through eBURST, the isolates with similar STs that shared identical alleles at more than 6 of the 7 loci were grouped into a clonal complex (CC) and the evolutionary origin of strains was determined from the primary founder in each CC. Primary founder was assigned to the ST that had the largest number of single-locus variants (SLVs) in

| Pulse-field gel electrophoresis (PFGE)
Pulse-field gel electrophoresis was performed for the analysis of the genetic similarity between all MRSA isolates according to the guidelines of the Korea Centers for Disease Control and Prevention (KCDC). In brief, the chromosomal DNA of MRSA was prepared in agarose plugs and cleaved with 50 U SmaI enzyme.
The samples were subjected to electrophoresis on a 1% agarose gel in 0.5% Tris-Borate-EDTA buffer at 14°C using CHEF DR-III Cluster analyses were performed using BioNumerics 7.6 (Applied Math) with dice correlation for band matching at a 1.5% position tolerance and the unweighted pair group method with an arithmetic average (UPGMA) and similarity coefficient of 80%. 1

| Detection of virulence genes
We selected a list of virulence genes based on their prior association with MRSAB. To identify the presence of virulence factors, PCR was performed. The presence of superantigens was examined with multiplex PCR, as previously described. 24

| Statistical analysis
To compare the characteristics of ST5 and ST72, analyses were performed using the Fisher exact test. Odds ratios (ORs) and 95% confidence intervals (CIs) were calculated using the logistic regression model and a two-tailed P value < .05 was considered statistically significant. Analyses were performed using SPSS version 21.0 (SPSS Inc).

| RE SULTS
A total 59 MRSA isolates were collected. The mean (± standard de-   Figure 2).

| Phenotypic and molecular characteristics
We observed differences in the antibiotic susceptibility test results and virulence genes between genotypes (STs), especially ST5 and ST72, which were the two major clones isolated from patients with MRSAB. The results are shown in Table 2 rates against many antibiotics, including clindamycin, erythromycin, fusidic acid, gentamicin, rifampicin, and tetracycline, than ST72.
However, no significant difference was observed between the susceptibility of ST5 and ST72 to vancomycin.
We failed to observe any significant differences in the antibiotic susceptibility results of vancomycin MIC and the retained virulence genes between PFGE groups classified within the same STs.

| D ISCUSS I ON
In this prospective study, we identified ST5 and ST72 as the major strains of MRSA involved in causing bacteremia. Previous studies have reported various ST strains for each region. In North America, CA-MRSA, defined as USA300, was reported as ST8. 32 In Western Europe, PVL-positive strains, including ST80, were common. 33 In Japan, ST5/ST764 are known as major HA-MRSA. 17 ST5 and ST72 have been reported to be the major HA-and CA-MRSA in South Korea, and ST72 MRSA was widespread in community and hospital. [34][35][36] Our results confirmed these results. We found that 86.4% of ST72 were HA-MRSA, and the ratio of ST72 to entire isolates (37.3%) was higher than that reported in a previous study (22.4%). 15 These results indicate that ST72 has already emerged as a major strain in hospital environment.
Even with the high discriminatory power of PFGE, the isolates belonging to the same ST were divided into only a few PFGE groups.
We suggest that the bacteremia-causing ST5 and ST72 strains of MRSA may be endemic without any new influx.
We observed significant differences in the antibiotic resistance patterns and virulence genes harbored between STs, especially ST5 and ST72. ST5 had more virulence genes and higher resistance rates against antibiotics than ST72. The sel genes were more frequently detected in ST5 than in ST72, and sec and sed were found only in ST5. The genes sec and sel were reported to be associated with ST5 in a previous report. 15 These staphylococcal superantigen genes are known to play a critical role in the progression of S aureus infection. 37 Therefore, ST5 strains carrying more staphylococcal superantigens may be highly virulent. 15 We analyzed the mortality difference between ST5 and ST72 and failed to determine any statistical significance. However, the number of patients that died within 30 days was higher in ST5 group than in ST72 group. A previous study also reported lower mortality for ST72 than for ST5. 7,15 This study has some limitations. First, the exclusion of many patients may result in a bias analysis. Second, the number of isolates was insufficient to obtain statistical significance. Third, additional SCCmec typing needs to be carried out to identify whether ST5 and In conclusion, most MRSA isolates causing bacteremia were ST5 (CC5) and ST72 (CC8), and those belonging to the same STs were divided into only a few PFGE groups. The higher antibiotic resistance rate and staphylococcal superantigen toxin genes (sec, sed, and sel) in ST5 than in ST72 may be associated with its higher virulence capacity.