Prevalence of fosfomycin resistance and gene mutations in clinical isolates of methicillin-resistant Staphylococcus aureus

Background Fosfomycin exhibits excellent in vitro activity against multidrug-resistant pathogens, including methicillin-resistant Staphylococcus aureus (MRSA). Increasing fosfomycin resistance among clinical MRSA isolates was reported previously, but little is known about the relative abundance of Fosfomycin resistance genes in MRSA isolates circulating in Taiwan. Methods All MRSA isolates, collected in 2002 and 2012 by the Taiwan Surveillance of Antimicrobial Resistance (TSAR) program, were used in this study. Susceptibility to various antimicrobial agents, including fosfomycin, was determined by broth microdilution. Genetic determinants of fosfomycin resistance, including fosB carriage and murA, glpT and uhpT mutations, were investigated using PCR and sequencing of amplicons. Staphylococcal protein A (spa) typing was also performed to determine the genetic relatedness of MRSA isolates. Results A total of 969 MRSA strains, 495 in the year 2002 and 474 in the year 2012, were analyzed. The overall in vitro susceptibility was 8.2% to erythromycin, 18.0% to clindamycin, 29.0% to tetracycline, 44.6% to ciprofloxacin, 57.5% to trimethoprim/sulfamethoxazole, 86.9% to rifampicin, 92.9% to fosfomycin and 100% to linezolid and vancomycin. A significant increase in the fosfomycin resistance rate was observed from 3.4% in 2002 to 11.0% in 2012. Of 68 fosfomycin-resistant MRSA isolates, several genetic backgrounds probably contributing to fosfomycin resistance were identified. Twelve isolates harbored the fosB gene, and various mutations in murA, uhpT, and glpT genes were noted in 11, 59, and 66 isolates, respectively. The most prevalent gene mutations were found in the combination of uhpT and glpT genes (58 isolates). The vast majority of the fosfomycin-resistant MRSA isolates belonged to spa type t002. Conclusions An increased fosfomycin resistance rate of MRSA isolates was observed in our present study, mostly due to mutations in the glpT and uhpT genes. Clonal spread probably contributed to the increased fosfomycin resistance.

Among MDROs, MRSA is a major human pathogen which causes various dangerous infections, such as bacteremia, endocarditis, and abscess, in both community and hospital settings [10]. Fosfomycin alone or combined with other antimicrobial agents exhibited favorable in vitro activity against MRSA [6,8,11,12], and more than 70% clinical cure was observed with fosfomycin administration for the treatment of MRSA infection [13,14]. However, S. aureus with fosfomycin resistance developed and rose rapidly by 30-70% in China [15]. The mechanism of bacterial resistance to fosfomycin could involve either a chromosomeassociated defective transport system or plasmidmediated fosfomycin-inactivating enzymes. First, two key transporter systems, GlpT and UhpT, mediated the entry of fosfomycin into bacterial cells [16]. Once mutations in the chromosomal glpT and uhpT genes occurred, reduction in permeability with subsequent conferred MRSA resistance to fosfomycin was observed [16][17][18]. Second, the murA gene mutants exhibited lower affinity for fosfomycin [19], conferring various degrees of drug resistance. Moreover, a few fosfomycin-modifying enzymes, including FosA, FosB, FosC, and FosX, catalyzed the inactivation of fosfomycin [20][21][22], and only FosB is produced by Grampositive bacteria [23]. Most of the previous studies investigated the mechanism of fosfomycin resistance among Gram-negative bacteria, and only limited information about the resistance mechanism of Grampositive pathogens, particularly MRSA, is available. Hence, in the present study, we aimed to survey the prevalence of fosfomycin resistance and the associated uhpT, glpT, murA, and fosB genetics in clinical isolates of MRSA in Taiwan.

Bacterial isolates
All MRSA strains, collected in 2002 and 2012 through the TSAR program from different hospitals in Taiwan, were used in this study. The principles of isolate collection by the TSAR program had been described clearly in a previous study [8]. Duplicate isolates were excluded, and a total of 969 MRSA isolates, 495 collected in 2002 and 474 in 2012, were analyzed. All these strains were identified as S. aureus by performing Gram staining, a catalase-activity test, and a coagulase latex agglutination test (automated VITEK-2 system, Biomerieus, France). Methicillin resistance was ascertained using agar disk diffusion (Kirby-Bauer), according to the guidelines established by the Clinical and Laboratory Standards Institute (CLSI) [24]. The study was approved by the Ethical Committee of the National Taiwan University Hospital (NTUH-IRB No. 201504056RINB).

Antimicrobial susceptibility
The antimicrobial susceptibility to clindamycin, ciprofloxacin, erythromycin, linezolid, rifampicin, trimethoprim/sulfamethoxazole, tetracycline, fosfomycin, and vancomycin was determined using a broth microdilution method according to the CLSI recommendations [24], and the results were interpreted using the criteria for S. aureus provided by the CLSI [24]. Staphylococcus aureus ATCC 29213 was used as the internal control for each run of the susceptibility test.

Genetic analysis
DNA of 68 fosfomycin-resistant MRSA isolates was harvested using a DNA Extraction System kit (Viogene, New Taipei City, Taiwan) according to the manufacturer's instructions. The presence of fosB was detected by PCR using the previously described primers [25], and the full nucleotide sequence of three genes (murA, uhpT, and glpT) was determined by combing direct sequencing and primer walking with the individual PCR products. Primers, used in the present study, are shown in Table 1. The PCR and sequencing procedures were described in prior studies [15,25].

Molecular typing
Staphylococcal protein A (spa gene) typing was performed for 68 fosfomycin-resistant MRSA isolates. The highly variable X region in spa was amplified by PCR using the previously described primers [26]. The purified PCR products were sequenced, and the spa types were confirmed by analyzing the nucleotide sequence utilizing BioNumerics Version 6.5 (Applied Maths NV, Sint-Martens-Latem, Belgium) [27].

Statistical analysis
Categorical variables were expressed as percentages, and Fisher's exact test with two-sided comparison was utilized for the assessment of statistical significance.

Susceptibilities to tested antimicrobial agents
The overall susceptibilities to various antibiotics was demonstrated in Table 2. Susceptibility rates to clindamycin, erythromycin, tetracycline, and trimethoprim/sulfamethoxazole increased from 2002 to 2012; however, the susceptibility rates to rifampicin and fosfomycin between 2002 and 2012 decreased statistically significantly, shown in Table 2. All tested MRSA isolates remained 100% susceptible to linezolid and vancomycin. Among the fosfomycinresistant MRSA isolates, lower susceptibility rates to clindamycin, ciprofloxacin, erythromycin, and rifampicin were noted compared with those of the fosfomycin-susceptible group. In contrast, the susceptibility rates to trimethoprim/ sulfamethoxazole and tetracycline within the fosfomycinsusceptible group was lower ( Table 3).
Sixty-six of 68 fosfomycin-resistant MRSA strains contained one of the six different mutations (type I-VI glpT ) found in the glpT gene with the majority containing a type I glpT mutation (60 isolates). Each of these mutations caused amino acid substitutions within the GlpT protein. Furthermore, four different mutations (type I-IV uhpT ) were recognized in the uhpT gene of the 59   Table 4.

Discussion
A unique mechanism of action of fosfomycin made crossresistance to other antibiotic classes less common, which motivated physicians to reevaluate its ability to destroy drug-resistant pathogens, including MRSA [16]. In our study, elderly patients seemingly had the tendency of acquisition of fofsomycin-resistant MRSA infections, and the vast majority of those resistant strains were isolated from those hospitals located in central or southern Taiwan. The background mechanisms of this phenomenon need further investigation. Although the MRSA isolates exhibited high in vitro susceptibility to fosfomycin higher than 90%, a significant increase in fosfomycin resistance rate during past decades (from 3.4% in 2002 to 11.0% in 2012) was observed in Taiwan. Among the fosfomycin-resistant isolates, a higher resistance rate to clindamycin, ciprofloxacin, erythromycin, and rifampicin was noted; however, trimethoprim/sulfamethoxazole and tetracycline displayed more favorable susceptibility. Type I uhpT and type I glpT mutations predominantly caused fosfomycin resistance in our MRSA isolates, and the vast majority of isolates belonged to spa type t002.
Little was known about the MRSA fosfomycin resistance mechanism in epidemiological research, and very few literature reports were previously published [15,28,29]. Of those studies, a large-scale surveillance conducted by Fu et al. [15] in China demonstrated a 13.4% (9/67) fosB-positive rate with two-thirds (6/9) belonging to ST5. A research study by Etienne et al. [28] revealed that 18 of 39 (46.2%) S. aureus isolates, containing the fosB gene, were highly resistant to fosfomycin, but only one MRSA isolate had the fosB gene (Zhejiang, China) [29]. In our present study, approximately one-fifth of the MRSA isolates with fosfomycin resistance carried the fosB gene with the dominant t002 spa typing (11/12), which belonged to ST5. Our finding was similar to that described in Fu's report; it implicated the presence of clonal spread among the fosB-positive MRSA isolates, despite the previously reported triviality of fosB [30].
MurA, a target enzyme involved in the biosynthesis of bacterial peptidoglycan, could be inactivated by fosfomycin via its binding to the active site of the enzyme [16]. However, mutations of the murA gene resulted in amino acid substitutions, rendering susceptible clinical isolates resistant to fosfomycin [16]. Fu et al. [15] illustrated that a murA mutation played an unclear role in the fosfomycin resistance in their study, and a type II murA mutant was the most common among all murA mutations. Our results were different in this regard. The difference in the source of clinical specimens in these two studies might indicate that the mechanisms of fosfomycin resistance are different in Taiwan and mainland China.
The vast majority of the MRSA isolates in the present study possessed at least one of glpT and/or uhpT mutations, implicating that the genes, encoding transporter mutants, contributed to fosfomycin resistance substantially. This result contrasted to the findings of the preceding study [15]. The prevailing subtype of mutations in the glpT and uhpT genes was also different from that reported by Fu et al. [15]. Forty-eight fosfomycin-resistant MRSA isolates with dual resistance mechanisms (glpT and uhpT mutations) belonged to spa type t002, again implying clonal spread of fosfomycin-resistant MRSA.
The most prevalent and the second most common spa types were t002 and t037 in our study, respectively, revealing that fosfomycin resistance in MRSA isolates were correlated to some spa types. A similar correlation was noticed in other countries, including Sweden, Korea, China, Iran, Africa, Canada, and Brazil [31]. An international, or even intercontinental spread of specific fosfomycin-resistant MRSA clones may be occurring.
In the present study, the susceptibility of the MRSA isolates to various antibiotics was similar to that reported in the previous studies from Taiwan [8,27] but different from that in other countries [32][33][34]. Variation in drug susceptibility between geographic areas might be due to the presence of different prevalent MRSA clones and the difference in antibiotic selective pressure.
The major limitation of the present study is that it was conducted using the clinical MRSA isolates in Taiwan; thus, worldwide generalization of the results should be made carefully.

Conclusions
In conclusion, our study illustrated that the fosfomycin resistance rate of the MRSA isolates increased significantly in the past, and mutations in the glpT and/or uhpT genes were key for inducing fosfomycin resistance. These findings indicated to physicians that they should prescribe fosfomycin cautiously for treating MRSA infections empirically. Furthermore, t002 was the most frequently seen spa type among the fosfomycin-resistant MRSA isolates. This was comparable to that in other countries globally. Therefore, it is necessary to continuously monitor fosfomycin resistance and its mechanisms.