Phenotypic and Genotypic Characterization of Macrolide, Lincosamide and Streptogramin B Resistance among Clinical Methicillin-Resistant Staphylococcus aureus Isolates in Chile

Macrolides, lincosamides, and type B streptogramins (MLSB) are important therapeutic options to treat methicillin-resistant Staphylococcus aureus (MRSA) infections; however, resistance to these antibiotics has been emerging. In Chile, data on the MLSB resistance phenotypes are scarce in both community-(CA) and hospital-acquired (HA) MRSA isolates. Antimicrobial susceptibility to MLSB was determined for sixty-eight non-repetitive isolates of each HA-(32) and CA-MRSA (36). Detection of SCCmec elements, ermA, ermB, ermC, and msrA genes was performed by PCR. The predominant clones were SCCmec I-ST5 (HA-MRSA) and type IVc-ST8 (CA-MRSA). Most of the HA-MRSA isolates (97%) showed resistance to clindamycin, erythromycin, azithromycin, and clarithromycin. Among CA-MRSA isolates, 28% were resistant to erythromycin, azithromycin, and 25% to clarithromycin. All isolates were susceptible to linezolid, vancomycin, daptomycin and trimethoprim/sulfamethoxazole, and over 97% to rifampicin. The ermA gene was amplified in 88% of HA-MRSA and 17% of CA-MRSA isolates (p < 0.001). The ermC gene was detected in 6% of HA-SARM and none of CA-SARM isolates, whereas the msrA gene was only amplified in 22% of CA-MRSA (p < 0.005). Our results demonstrate the prevalence of the cMLSB resistance phenotype in all HA-MRSA isolates in Chile, with the ermA being the predominant gene identified among these isolates.


Introduction
Methicillin-resistant Staphylococcus aureus (MRSA) is an important pathogen involved in both human and animal infections [1,2]. Although MRSA was initially described as producing healthcare-associated infections (HA-MRSA), the appearance of communityassociated MRSA infections (CA-MRSA) has been documented since the 1990s [3]. MRSA has shown a remarkable ability to develop resistance to a myriad of antibiotics, as well as to different disinfectants and heavy metals [4]. Vancomycin (VAN), a member of the glycopeptides, has been used as an important option to treat MRSA infections [5]. However, the risk of dissemination of vancomycin-resistant or non-fully susceptible strains suggests that this antibiotic should be used sparingly [6]. For this reason, macrolides (erythromycin [ERY]), lincosamides (clindamycin [CLI]), and streptogramins B (MLS B ) have emerged as important therapeutic options to tackle CA-MRSA infections [7,8]. However, the increased use of these antimicrobials has favored the emergence of resistance to these drugs [9][10][11]. To date, there are three main MLS B resistance mechanisms described: i) changes in the ribosomal target site, which confers cross-resistance to the entire MLS B group [12]. This mechanism is conferred by ribosomal mutations or methylation of the 23S rRNA target site, which are mediated by the erm genes (mainly ermA, ermB, and ermC) [13,14]. Another mechanism corresponds to ii) an efflux-pump encoded by msrA, which can drive out 14and 15-membered macrolides and streptogramin B, producing the MS B phenotype [15]. Finally, another mechanism iii) relies on drug inactivation and it only confers resistance to lincosamides due to an enzyme encoded by the lnu gene [11].
Significantly, the MLS B phenotype can be either constitutive (cMLS B ) or inducible (iMLS B ) [9]. Specifically, CLI, which is the MLS B agent used for the treatment of S. aureus infections, is a weak MLS B -resistance inducer and may lead to treatment failure due to false susceptibility results displayed in in vitro antimicrobial susceptibility tests [16]. Therefore, it is necessary to perform the CLI susceptibility test in the presence of a strong inducer, such as ERY [12]. Another key point is that antibiotic resistance genes that mediate the MLS B -resistance phenotype are found in mobile-genetic elements (MGEs) and, in consequence, may be horizontally transferred to susceptible strains [17]. In Latin America, the resistance rates to MLS B antibiotics have been reported to be 74% and 81% to ERY and CLI, respectively, among HA-MRSA isolates [10].
In Chile, S. aureus is one of the main etiological agents in health care-associated infections (HAIs) [18]. Specifically, it is the main cause of surgical wound infections (27%), and the second cause of pneumonia associated to invasive mechanical ventilation (21%). Likewise, it is involved in bloodstream infections (18%) and infections of the central nervous system (18%) [18]. Despite these data, the MLS B -resistance phenotype among HA-and CA-MRSA is still unknown among Chilean isolates. Therefore, the aim of our study was to detect and characterize the MLS B -and MS B -resistance phenotypes among HA-MRSA and CA-MRSA isolates collected between 2007 and 2017 from the S. aureus surveillance program of the National Institute of Public Health of Chile (ISP).

Molecular Characterization of MRSA Isolates
All HA-MRSA (32) and CA-MRSA (36) isolates were resistant to FOX and mecA positive. For HA-MRSA, the Staphylococcal Cassette Chromosome mec (SSCmec) analysis revealed the presence of the Type I and Type II elements in 27 (84.4%) and 5 (15.6%) isolates, respectively. In addition, in all isolates classified as HA-MRSA, the absence of the pvl gene was confirmed. On the other hand, in all CA-MRSA (36), the pvl gene and the type IV SSCmec cassette were detected. Of these, 24 (66.7%) harbored the cassette subtype SSCmec IVc, whereas 11 (30.5%), and 1 (2.8%) amplified for the subtypes IVa and IVb, respectively; therefore, they were confirmed as CA-MRSA.

Antimicrobial Susceptibility Testing
The antibiotic resistance profiles were determined for both HA-MRSA and CA-MRSA isolates ( Table 2). All isolates (32) of HA-MRSA were resistant to macrolides and to CLI. Moreover, 2 isolates (2/32) (6.3%) were also resistant to CHL and 1 isolate (1/32) (3.1%) to RIF. In the case of CA-MRSA, 9 isolates (9/36) (25%) were resistant to ERY, AZM and CLR, and one isolate was resistant to ERY and AZT (2.8%), but all were susceptible to CLI, CHL, and RIF ( Table 2). All HA-MRSA, and CA-MRSA isolates were susceptible to LZD, VAN, DAP, and SXT (Table 3). Furthermore, the iMLS mechanism was detected in none of the two groups of MRSA isolates.  The HA-MRSA group showed more extended resistance profiles than CA-MRSA. Among the HA-MRSA, the most prevalent resistance profile was CLI-ERY-AZM-CLR, with 90.6% of isolates. On the other hand, in the CA-MRSA group, the most prevalent antibiotic resistance profile was ERY, AZM, and CLR, with 25% of isolates.

Prevalence of msrA and erm Genes
The ermA gene was amplified in 28 (87.5%) HA-MRSA isolates compared with 6 (16.7%) in CA-MRSA (p < 0.001). Additionally, the ermC gene was found in 2 (6.3%) of HA-MRSA and in none of CA-MRSA isolates (p > 0.05), and the ermB gene was detected in none of the isolates. On the other hand, msrA was detected in 11 (30.6%) of the CA-MRSA isolates, but in none of the HA-MRSA (p < 0.005) ( Table 4).

Discussion
In recent years, we have observed an increased resistance to antibiotics, especially in those used for the treatment of serious infections associated with health care. MLS B group are antibiotics commonly used to treat skin and soft tissue infections caused by CA-MRSA [11]. The present study reports percentages of resistance to antibiotics in the MLS B group ≥90% in HA-MRSA. This finding agrees with the results of previous studies carried out with strains collected in Chile [10,19]. Besides, 20% of strains of CA-MRSA were resistant to MLS B group. These results show lower rates of resistance to these antibiotics in comparison to the official reports of the National Institute of Public Health of Chile (20% v/s 29%, respectively). On the other hand, our results showed higher values than previous reports that included strains isolated in Latin America, among both HA-MRSA (81% for ERY and 74% for CLI) and CA-MRSA [9,10,[20][21][22][23][24].
Among the isolates included in this work, the predominant phenotype was the cMLS B phenotype. Molecular characterization of 68 MLSB-resistant MRSA revealed that among HA-MRSA, 87.5% were positive for ermA. However, in the CA-MRSA strains, 16.7% were positive for ermA, 6.3% for ermC, and 30.6% for msrA. The main mechanism of resistance to macrolides in CA-MRSA is mediated by the presence of the msrA gene, which results agree with previously published data [25].
Our results are in agreement with previous reports about the predominance of SCCmec type I-ST5 in HA-MRSA in Chile with classic resistance profiles of the Chilean/Cordobes clone that has a marked presence in hospitals of our country [10,26], and isolates of type IV-ST8 in CA-MRSA in Latin America, related to the USA-300 clone [10,19]. On the other hand, the dichotomy regarding the presence of MLS B or MS B resistance among HA-MRSA isolates highlights compared with CA-MRSA (97% vs approximately 25%, reaching statistical significance, p < 0.005). However, it is important to emphasize that these findings, which are consistent with the classic concept that hospital isolates of MRSA are multi-resistant and the community-based multi-susceptible and only resistant to β-lactams, should be monitored, since 20% of the isolates of CA-MRSA were resistant to antibiotics in this group, that is, 1 over 5 isolates were not widely susceptible. Accordingly, it is important to perform the proper laboratory detection of these phenotypes to analyze these isolates, since if the criterion of resistance to methicillin and broad susceptibility is the method of choice, other families, including those of the MLS B group, could obtain biased results.
All the strains analyzed are susceptible to VAN, LZD, DAP, and SXT, keeping these antibiotics as an alternative treatment within the therapeutic arsenal available in Chile, which is consistent with previous reports [10,18].
In summary, despite the higher frequency of the cMLS B phenotype than iMLS B in this study, we recommend performing the D test to identify clindamycin-induced resistance and guide therapeutic procedures in both HA-MRSA and CA-MRSA. Likewise, it is not recommended ruling out the submission of suspected CA-MRSA strains in surveillance programs based exclusively on the criterion of resistance only to β-lactams.

MRSA Isolates
Thirty-two non-repetitive HA-MRSA isolates recovered from eight Chilean cities between 2007 and 2017 (Table 5), and thirty-six CA-MRSA isolates collected in ten Chilean cities between 2012 and 2017 (Table 6) were included in this study. All isolates were selected from the biorepository maintained by the National Institute of Public Health of Chile (ISP), Santiago, Chile. All isolates were cryo-preserved at −80 • C in glycerol (50% v/v) and trypticase broth (2:1). The ISP criteria were used to define HA-MRSA and CA-MRSA [20].

Characterization of MRSA Isolates
The presence of mecA, pvl in MRSA isolates, and the detection and characterization of the SCCmec element were performed by PCR-based protocols, as previously described [30][31][32]. Sequence types (ST) were obtained according to Opazo-Capurro et al. (2019), using the Pasteur's scheme STs employing the bioinformatic tools available at the Center for Genomic Epidemiology (CGE) server (http://www.genomicepidemiology.org/, accessed on 13 March 2022) [33].

Molecular Detection of Antibiotic Resistance Genes
The detection of genes involved in the MLS B (ermA, ermB and ermC) and MS B (msrA) phenotypes were screened by conventional PCR according to protocols and primers previously described [34] (Supplementary Materials, Table S1).

Statistical Analyses
Pearson's chi-squared test was used to determine associations between antibiotic resistance profiles, MLS B resistance genes, and MRSA types (CA or HA-MRSA). This was achieved utilizing the IBM SPSS Statistics version 23.0 software (SPSS Inc, Chicago, IL, USA), establishing statistical significance at p < 0.05 [35].

Conclusions
In Chile, in isolates of HA-MRSA, there is an evident predominance of ST5-SCCmec I, a Chilean/Cordobes clone, characteristically multiresistant, which includes resistance to antibiotics from the MLS B group; and susceptible to SXT and RIF. On the other hand, at the community level (CA-MRSA), there is an emergency of ST8-SCCmec IV, related to clone USA 300. Thus, microbiological surveillance of these isolates at the nosocomial level is required to verify whether the Chilean/Cordobes clone will be replaced by this community clone in Chile, and to monitor whether the latter will continue to increase its resistance to non-beta-lactam antibiotics, such as those of the MLS B group.