Mupirocin-Resistant Staphylococcus aureus in Iran: A Biofilm Production and Genetic Characteristics

The spread of mupirocin-resistant Staphylococcus aureus strains in hospitals and communities is a universal challenge. Limited data is available on the genetic features of high-level mupirocin resistant- (HLMUPR-) S. aureus isolates in Tehran. In the present research, we investigated 48 high-level mupirocin resistance S. aureus by antimicrobial activity, virulence analysis, biofilm formation, multilocus sequence typing (MLST), and staphylocoagulase (SC) typing. All the HLMUPR strains were positive for mupA gene. The frequency of multidrug resistance was 97.9%. Twenty-one (43.8%) were toxinogenic with 14 producing pvl (29.2%), 5 tst (10.4%), and two eta (4.2%). Among the HLMUPR isolates, biofilm production was detected in 45 (89.6%) isolates with complete dominance clfB, clfA genes, and a noticeably high frequency fnbA (95.8%), followed by fnbB (93.8%), eno and icaD (each 83.3%), sdrC (81.3%), ebps (79.2%), icaA (75%), sdrD (66.7%), fib (60.4%), sdrE (50%), cna (41.7%), and bap (4.2%). Coagulase typing distinguished isolates into four genotypic patterns including III (50%), II (27.1%), and type IVa (22.9%). A total of three clonal complexes (CCs) and 4 sequence types (STs) including CC/ST22 as the most prevalent (52.1%), CC8/ST239 (20.8%), CC/ST8 (16.7%), and CC/ST5 (10.4%) were identified in current work. According to our analysis, nonbiofilm producer isolates belonged to CC8/ST239 (6.3%) and CC/ST8 (4.2%). Fusidic acid-resistant isolates belonged to CC/ST45 (n = 3) and CC8/ST239 (n = 1). Observations highlighted the circulation of the CC/ST22 HLMUPR S. aureus strains with strong biofilm-production ability in our hospitals, indicating the possibility of transmission of this type between community and hospital.


Introduction
As the most frequent pathogen, Staphylococcus aureus causes various diseases ranging from skin and soft tissue infections (SSTIs) to bacteremia, endocarditis, and osteomyelitis in hospitals and communities. Recent evidence has highlighted a dramatic increase in hospital-acquired S. aureus infections worldwide with significant morbidity and mortality [1][2][3]. The high prevalence of antibiotic resistance due to inappropriate antibiotic use and the remarkable ability of this bacterium to acquire resistance to multiple antimicrobials, particularly mupirocin-resistant S. aureus (mupR-S. aureus), complicated the treatment of related infections. Moreover, the scenario worsens with resistance to commonly used antibiotics, biofilm-forming ability, and various virulence factors for S. aureus [4][5][6].
Mupirocin is a topical antibiotic with excellent activity in treating SSTIs and eradicating MRSA carriage in nursing homes, healthcare workers, patients, and control of MRSA outbreaks. However, an increase in mupirocin resistance has been reported widely in many areas of the world following its extensive and widespread use [7][8][9][10]. Mupirocin, an analog of isoleucyl, interferes with protein synthesis by competitive inhibition of the bacterial isoleucyl-tRNA synthetase (IRS). Resistance to mupirocin has been described in two categories: (i) high level that occurs via plasmid-borne gene mupA and or the newly described mupB gene, coding for an alternate isoleucyl-tRNA synthetase (IleS), and (ii) low level, which results from point mutations in the chromosomal encoding isoleucyl-tRNA synthetase (ileS) gene. Data worldwide indicate that lack of proper and timely identification, distribution of molecular types, and managing of highlevel mupirocin resistance S. aureus infections (HLMUPR) could lead to decolonization failure, increase in carriage rate, and subsequently wide range of staphylococcal infections [10,11]. Diagnosis of HLMUPR in S. aureus clinical isolates is a great challenge due to the distribution of different types related to the type of infection, their antimicrobial profile, and biofilm formability [10]. On the other hand, timely diagnosis of HLMUPR in S. aureus clinical isolates is of great significance for formulating appropriate treatment plans, shortening hospitalization time, and reducing the disease burden [11]. Since extensive studies have not been performed in connection with HLMUPR S. aureus infection and the high importance of its timely diagnosis and treatment, therefore, it has become urgently needed to figure out the antibacterial activity, carriage of virulence determinants, biofilm-forming ability, and specific genes that may be involved in biofilm formation and molecular types in HLMUPR strains isolated from various clinical samples. In order to produce a more comprehensive description of the molecular epidemiology and phenotypic properties of HLMUPR S. aureus in Iran, we characterized 390 S. aureus isolates collected from several geographically dispersed Tehran hospitals over the past two years. The frequency of fourteen selected genes involved in biofilm production was assessed to identify genes involved in biofilm production. Coagulase (coa) typing and multilocus sequence typing (MLST) were applied to determine the molecular types of the HLMUPR strains.

Materials and Methods
2.1. Isolation S. aureus, Genomic DNA Extraction, and Identification of HLMUPR Isolates. In the present study, 390 S. aureus strains from clinical samples of patients referred to the hospitals affiliated to Shahid Beheshti University of Medical Sciences from January 2018 to December 2019 were screened for HLMUPR. Identification of S. aureus was performed by microbiological and biochemical tests and the presence of the nuc gene using polymerase chain reac-tion (PCR) assay [12]. The DNA extraction was performed by the phenol-chloroform extraction method previously described by Goudarzi et al. [12]. The purity of the extracted DNA was evaluated by a NanoDrop-2000 spectrophotometer (Thermo Fisher Scientific, Wilmington, DE, USA) at 260/280, respectively. An A260/A280 purity ratio of~1.8 was considered for the PCR assay.
The study was ethically approved by the Ethics Committee of the Shahid Beheshti University of Medical Sciences in Tehran, Iran (IR. SBMU. MSP.REC. 1398. 353), and consent was obtained from consent participants. HLMUPR strains were detected by the broth microdilution method (MIC value was ≥512 μg/ml) according to the clinical and the laboratory standards institute (CLSI) guideline (CLSI 2019) and PCR of mupA gene [4]. HLMUPR strains that were obtained on or after 96 hours of admission to a hospital were considered as hospital onset (HO) while a positive culture prior to four days of hospitalization along with one or more of the following criteria: (i) a history of hospitalization, surgery, dialysis, or residence in a long-term care facility in 12 months prior to culture date or (ii) the presence of a central vascular catheter (CVC) within two days before S. aureus culture was considered as community onset (CO).
The microdilution was performed to determine minimum inhibitory concentration (MIC) titer for antibiotics vancomycin (VAN), tigecycline (TIG), and fusidic acid (FUS) according to the procedure detailed in our previous report [4]. Results for fusidic acid and tigecycline were interpreted based on the European Committee for antimicrobial susceptibility testing (EUCAST) recommendations (http:// www.eucast.org). The inducible and constitutive macrolidelincosamide-streptogramin group B (iMLS B and cMLS B ) resistance phenotypes were identified by erythromycin and clindamycin disks by the CLSI guideline (CLSI 2019). S. aureus ATCC 25923, ATCC 43300, and ATCC 29213 strains were used for antibiotic susceptibility testing quality control.

Microtiter Plate (MtP) Assay to Determine Biofilm
Formability. The ability of bacterial biofilm formation was assessed using MtP assay as Yousefi et al. [9] described. All tests were run in quadruplicate and rerun three separate times to confirm reproducibility. Staphylococcus epidermidis ATCC 35984 strain was used as a positive control strain for biofilm formation. TSB supplemented with 1% glucose was used as a negative control. Considering an optical density cut-off (ODc) to be represented by average OD of negative control + 3 × standard deviation ðSDÞ of negative control, strains were classified into the following categories: without biofilm (OD ≤ 0:059), weak (ODc < OD ≤ 2 × ODc), moderate (2 × ODc < OD ≤ 4 × ODc), and strong (OD > 4 × ODc) [9].
2.4. Screening for Virulence-Related Genes. For determination of the prevalence of Panton-Valentine leukotoxin (pvl), toxic shock syndrome toxin (tsst-1), and exfoliative toxin (eta and etb) genes, PCR reaction was carried out with specific oligonucleotide primers [12]. For definitive confirmation of USA300 strains, the existence of the arginine catabolic mobile (ACME) elements in all the isolates was investigated employing PCR assay as previously described by Boswihi et al. [13].
2.6. Molecular Typing Methods 2.6.1. coa Typing. Multiplex PCR was carried out to determine coagulase (coa) types (I-X) based on the method provided by Hirose and coworkers. In this assay, four sets were applied, including (i) set A that identified coa types I, II, III, IVa, IVb, Va, and VI; (ii) set B that identified coa types VII, VIII, and X; (iii) set C that identified coa types IX and Vb; and (iv) set D that differentiated IVa and IVb [15].
2.6.2. MLST. We used the procedure described by the MLST database (https://pubmlst.org/) and the Enright et al. method to characterize HLMUPR isolates; PCR assay was carried out using recommended primers of housekeeping genes (pta, arcC, tpi, aroE, gmk, yqiL, and glp) on the website. Sequence types (STs) were determined by submitting the allelic profile to the online MLST database website.

Statistical
Analysis. SPSS Statistics version 22.0 for Windows was used for statistical analysis. In addition, data were analyzed using the chi-square or Fisher's exact tests. P values less than 0.05 were deemed statistically significant.
All HLMUPR isolates were resistant to methicillin (MRSA). The distribution of resistance among HO and CO HLMUPR strains to drugs is provided in Table 1 Thoroughly, all isolates except one isolate showed resistance to ≥3 classes of antibiotics and considered multidrug resistance (MDR) isolates. Of the total isolates, 20, 12, and 2 isolates showed cMLS B , iMLSB, and MS phenotypes accounting for 41.7%, 25%, and 4.2%, respectively. Inducible and constitutive clindamycin resistance rate among HO-HLMUPR strains (14.6% and 27.1%) was higher than CO HLMUPR strains (10.4% and 14.6%). All the HLMUPR strains with MS phenotype were isolated from HO cases ( Table 2).

Phenotypic Biofilm Formation.
Of the 48 strains of HLMUPR studied, 43 (89.6%) isolates were biofilm producers in varying degrees. Twenty-five (52.1%) isolates were identified as strong biofilm producers, 10 (20.8%) were moderate producers, and eight (16.7%) were found to be weak producers. A statistically significant relationship between resistance to mupirocin at a high level and strong biofilm production (P value 0.015) among HLMUPR MRSA isolates was seen.

Discussion
The findings showed that 12.3% of S. aureus isolates were confirmed as HLMUPR and all carried mupA gene. Dadashi et al.'s systematic review and meta-analysis study reported an 8.1% prevalence of HLMUPR MRSA clinical isolates lower than our reported rate. They also reported the prevalence rate of HLMUPR MRSA isolates in Europe, Asia, and the USA, accounting for 8.0%, 12.1%, and 5.9%, respectively [11]. Moreover, Shittu et al. reported that the prevalence of HLMUPR MRSA isolates in Africa ranged between 0.5 and 38% [16]. However, much higher rates were also reported by studies conducted from India (26.1%) [5], the USA (19.3%) [17], and Korea (5.7%) [6]. Reasons for the high resistance rate in Asian countries and our research could be irrational use, various policies in the prescription of these antibiotics, easy access to antibiotics, low cost of the drugs, and spreading of dominant types in these areas. Recent available data has shown there is an increase in the preva-lence of mupA-positive MRSA isolates. In this connection, Monecke et al. noted a sharp increase in the prevalence of these isolates during the study period from 1.1% in 2000-2015 to 15.9% in 2016 and 17.6% in 2017 [18]. This high prevalence of mupA gene is probably because this gene, linked to mupirocin resistance at a high level, is contained within a plasmid whose acquisition was reasonably easy through horizontal gene transfer [10,11]. In the study, the prevalence of iMLS B was found to be 25% which was near to reported rate from Nepal (21%) [19], higher than those reported in Brazil (7.9%) [20], and lower than the rates reported from Jordan (76.7%) [21]. We observed that 41.7% of isolates had cMLS B phenotype. This was lower than the similar studies carried out in Iran (82.9%) [22] and higher than the rate stated in Nepal [19] (38%) and Greece     [23]. It seems that the acquisition of erythromycin resistance in HLMUPR MRSA maybe dynamic and associated with injudicious use of macrolides in patients infected with HLMUPR MRSA isolates [10,11].
According to the earlier published data, mupirocinresistant strains are more likely to be resistant to fusidic acid. In our survey, of the forty-eight strains, four isolates were fusidic acid-resistant (8.3%). These results differ from those found in Canada (7%), Ireland (19.9%), Greece (62.4%), and Australia (7%) [24,25], as well as a study carried out in Iran during four years on 726 tested S. aureus isolates, and a low prevalence of resistance to fusidic acid was reported (3%) [26].
In the current research, 29.2% of isolates carried pvl gene. Gambian-based research of the prevalence of PVL S. aureus in the 11 years reported a rate of 61.4% (180/293) of PVL among S. aureus isolates [27]. Different frequency of PVL-positive HLMUPR strains was reported in researches conducted from Iran (27.9%) [28], Ireland (17%) [25], and China (15.1%) [29]. Possible factors involved in the observed high prevalence of pvl gene include study population, design, and circulating clones in hospitals and communities.
We detected the tst-encoding gene in 10.4% of tested isolates. Nevertheless, Shahini Shams-Abadi et al. showed a relatively high prevalence of tst gene among S. aureus strains tested (21.3%) [30]. Differences are probably attributed to the origin of isolates, dissemination of specific clones, and sample type.
As mentioned earlier, three different coa types were detected in our study. Our data indicated low variability and heterogeneity of coa gene. Similarly, a previous study conducted by Omar et al. indicated three coa types and ten subtypes [31]. Based on published data by Afrough et al. in Iran [32], six coa types with the majority of the pattern I C 1 (21.7%) were detected among S. aureus isolates. Consistent with our finding, the results of coa typing of 100 S. aureus clinical isolates by Salehzadeh et al. noted two genotypes and three subtypes [33]. Contrary to Hirose et al.'s study [15], which displayed a high prevalence of coa type II (91.9%), VII (3.9%), and I (1.7%), in our study, coa type III was the most predominant coa type among S. aureus isolates (50%), followed by type II (27.1%) and type IVa (22.9%). Similar results in some studies indicate the predom-inance of coa type III in MRSA isolates [31,34]. The presence of coa type III in the present research and other recent studies suggests that this coa type is actively circulating in Tehran province's healthcare setting.
According to the evidence, there is high variability in biofilm-related genes among the MRSA strains and specific genotypes [35]. Earlier studies reported different prevalence rates of biofilm formation among S. aureus with data ranging from 43 to 88% [2]. In the present study, the prevalence of biofilm producers was high (89.6%) which was more significant reported in China (66%) [36] and Iran (62.9%) [9]. It is well established that biofilm formation depends on different factors, but it seems that resistance to mupirocin at a high level could affect the accessory gene regulator system involved in the expression of adhesion factors and biofilm induction.
Many studies have noted the critical role of adhesionrelated genes in adherence to the host components and producing biofilm. The clfA, clfB, fnbA, and fnbB genes are essential for microorganism adhesion and might be related to biofilm formability. In the present study, the complete dominance was observed for clfA and clfB genes followed by fnbA (95.8%) and fnbB (93.8%). Similarly, the published results by Yousefi and coworkers in Iran revealed a 69.2% prevalence of biofilm production and the presence of fnbA, clfA, and icaA genes in all examined strains [9]. These rates were comparable to those in other researches [36][37][38][39][40].
Some published data documented that the ability of strains of S. aureus to produce biofilm has related to the synthesis of polysaccharide intercellular adhesion (PIA), which its production is under the control of the icaADBC operon. In this study, a high prevalence of icaA (75%) and icaD (83.3%) was found, which is in line with the findings of Azmi and colleagues [35] and support those of Sharma and coworkers [39] who reported that icaA/D found abundantly among biofilm-producing isolates. Contrarily, Azmi et al. [35] indicated that icaA and icaD genes were present in 16.5% and 83.5% of strains. These findings suggest that the ica-dependent system may not be an absolute mechanism in biofilm production, and these strains perhaps use other systems to produce biofilm.
According to our findings, CC/ST22 was a predominate genotype (52.1%). Forty percent of CC22 isolates were PVLpositive. This finding supports previous results reported by Shore et al. [25], Goudarzi et al. [12], and Monecke et al. [3]. Although biofilm formability and antibiotic resistance profiles in CC/ST22 isolates were found to be varied, resistance to mupirocin and biofilm formability in these strains have been described by several investigators [4,41].
CC8/ST239 was the second prevailing lineage detected accounting for 20.8% of tested isolates. In line with the present study, ST239 was confirmed as the most predominant genotype in China from 2005 to 2013 in a study conducted by Li et al. [42]. One of the four fusidic acid-resistant MRSA strains belonged to CC8/ST239; this agreed with the findings of Shore et al. from Ireland [25]. Concordant with our study, in a 13year study, Boswihi et al. in Kuwait displayed fusidic acidresistant isolates belonged to ST239-MRSA-III that none of them carried fusB or fusC genes [13]. Inducible clindamycin resistance was noted in 40% of the ST239 isolates. In support of this, Abimanyu and coworkers recently reported the emergence of ST239 MRSA strains with both mupirocin and inducible clindamycin resistance in India [43].
Our analysis confirmed CC/ST8 (ACME+ and PVL+) isolates that resembled the USA300. Earlier researches have noted the resistance to mupirocin and trimethoprimsulfamethoxazole among USA300 isolates [3,4,44,45], in conformity of results obtained from McDougal et al. in the USA, which identified USA300 clone MRSA isolates harboring mupA [44]. We found mupA gene in 12.3% of examined isolates. Contrary to a prior study from Iran, which described ST8 S. aureus isolates with resistance to vancomy-cin [45], none of CC/ST8 isolates described in this study was vancomycin-resistant. The emergence of the USA300 in the present study reflects its growing importance as an epidemic clone, which may be due to an import of this clone from Asian or European countries.
In this study, we noted a low prevalence of ST5 (10.4%) with 100% biofilm formability. González-Domínguez and colleagues analyzed 147 MRSA strains isolated from different clinical samples and found that HLMUPR-MRSA isolates belonged to ST125 (97.5%) and ST5 (2.5%) [46]. Present work showed that half of the isolates had a high resistance rate to trimethoprim and fusidic acid. These findings confirmed a study conducted by Shore et al. which indicated resistances to trimethoprim and tetracycline among CC5 isolates at a high level [25].

Conclusion
Altogether, our study provides evidence for the occurrence of four different lineages of HLMUPR-MRSA in Iran. The high prevalence of CC/ST22 with strong biofilm formation in our hospitals highlights special attention for implementing efficient control protocols and stricter precautions to stop the dissemination of these isolates in both communities to hospitals. There is a need for continuous monitoring of the genotypes of HLMUPR-MRSA isolates to prevent nosocomial outbreaks of these isolates.

Data Availability
All data generated or analyzed during this study are included in this published article.

Disclosure
The funding agency has no role in the design of the project, work execution, analyses, interpretation of the data, and manuscript writing and submission.