The prevalence of genotypes that determine resistance to macrolides, lincosamides, and streptogramins B compared with spiramycin susceptibility among erythromycin-resistant Staphylococcus epidermidis

Coagulase-negative staphylococci, particularly Staphylococcus epidermidis, can be regarded as potential reservoirs of resistance genes for pathogenic strains, e.g., Staphylococcus aureus. The aim of this study was to assess the prevalence of different resistance phenotypes to macrolide, lincosamide, and streptogramins B (MLSB) antibiotics among erythromycin-resistant S. epidermidis, together with the evaluation of genes promoting the following different types of MLSB resistance:ermA, ermB, ermC,msrA, mphC, and linA/A’. Susceptibility to spiramycin was also examined. Among 75 erythromycin-resistantS. epidermidis isolates, the most frequent phenotypes were macrolides and streptogramins B (MSB) and constitutive MLSB (cMLSB). Moreover, all strains with the cMLSB phenotype and the majority of inducible MLSB (iMLSB) isolates were resistant to spiramycin, whereas strains with the MSB phenotype were sensitive to this antibiotic. The D-shape zone of inhibition around the clindamycin disc near the spiramycin disc was found for some spiramycin-resistant strains with the iMLSB phenotype, suggesting an induction of resistance to clindamycin by this 16-membered macrolide. The most frequently isolated gene was ermC, irrespective of the MLSB resistance phenotype, whereas the most often noted gene combination wasermC, mphC, linA/A’. The results obtained showed that the genes responsible for different mechanisms of MLSB resistance in S. epidermidis generally coexist, often without the phenotypic expression of each of them.

Coagulase-negative staphylococci (CoNS), particularly Staphylococcus epidermidis, belong to the microbiota of human skin and the mucosal membrane of the upper respiratory tract, and they express low pathogenic potential as commensals in healthy people (Voung & Otto 2002, Otto 2009). However, they can be responsible for several serious infections in immunocompromised patients, particularly those associated with biomaterials (e.g., catheters, prosthetics etc.), leading to bacteraemia and sepsis (Ziebuhr et al. 2006, Caesy et al. 2007, Schoenfelder et al. 2010, Castro-Alarcón et al. 2011). On the other hand, as a natural part of the microflora, drug resistant strains may be selected during antibiotic therapy, which is a potential source of the resistance genes for pathogenic strains, e.g., Staphylococcus aureus (Reyes et al. 2007, Otto 2013, Vitali et al. 2014.
Resistance to macrolide, lincosamide, and streptogramins B (MLS B antibiotics) in staphylococci is associated with the following three mechanisms: (i) target modification, (ii) efflux pumps, and (iii) enzymatic modification of antibiotics. The first macrolide-resistant staphylococcal strains were identified in the 1950s (Roberts 2004). Currently, a large number of strains exhibit resistance to these antibiotics via different mechanisms. It is known that macrolide-resistant strains often exhibit co-resistance to other MLS B antibiotics. The most common mechanism is the modification of ribosomes as a result of methylation of adenine within 23S rRNA ribosomal subunits by a methylase encoded by the erm genes (predominantly ermC). Conformational changes in the ribosome result in the reduced binding of all MLS B antibiotics; these strains are resistant to all MLS B antibiotics (the combination of quinupristin/dalfopristin loses bactericidal activity as the result of the development of resistance to quinupristin). The phenotypic expression of MLS B resistance can be either inducible (iMLS B ) (generally induced by 14 and 15-membered macrolides) or constitutive (cMLS B ) (Weisblum 1995). The active efflux of antibiotics is mediated by msr genes (mainly msrA) and is responsible for resistance only to 14 and 15-membered macrolides and streptogramins B (MS B ) phenotype (Reynolds et al. 2003). The third mechanism of resistance is based on the production of antibioticinactivating enzymes (e. g., phosphorylase encoded by mph or lin, the gene responsible for inactivation of lincosamides) (Chesneau et al. 2007, Achard et al. 2008. The aim of this study was to assess the prevalence of different MLS B resistance phenotypes among S. epidermidis, together with the evaluation of genes responsible for target modification (ermA, ermB, ermC), antibiotic efflux (msrA) or antibiotic inactivation (mphC, linA/A'). The evaluation of susceptibility to the 16-membered macrolide spiramycin was also performed. This paper was developed using the equipment purchased within agreement POPW.01.03.00-06-010/09-00 Operational Program Development of Eastern Poland 2007-2013, Priority Axis I, Modern Economy, Operations 1.3. Innovations Promotion.

SUBJECTS, MATERIALS AND METHODS
Bacterial strains -A total of 197 strains of S. epidermidis were obtained from the mucosal membranes of the upper respiratory tracts of patients with nonsmall cell lung cancer who underwent hospitalisation. Nasal and pharyngeal swabs were obtained on the second day of the patients' stays at the hospital. Among the strains, resistance to erythromycin was detected in 75 isolates.
Isolation and identification -Isolation and identification of bacterial strains were performed using routine microbiological tests. The following tests were used in the identification of CoNS: the coagulase test tube using rabbit plasma (Biomed, Poland) and API Staph strips (bioMérieux, France).
Identification of resistance to MLS B antibiotics -Susceptibility to MLS B antibiotics, including the detection of resistance mechanisms, was based on the D-test according to European Centre for Disease Prevention and Control (EUCAST) recommendations. In addition, disks containing lincomycin (15 mg) were used to identify the L-phenotype. Moreover, for detection of the effects of spiramycin on clindamycin susceptibility, discs containing spiramycin (100 mg) were applied next to clindamycin (2 mg). (MICs) to spiramycin -Detection of MICs to spiramycin was based on EUCAST recommendations using the double broth dilution method. In the absence of breakpoints for spiramycin in EUCAST, only the MICs were evaluated without grouping the strains as susceptible or resistant.

Determination of minimal inhibitory concentrations
Isolation of bacterial DNA -The DNA Genomic Mini Kit (A&A Biotechnology, Poland) was used to iso-late S. epidermidis DNA according to the manufacturer's guidelines. Table I. For the PCR reactions, PCR REDTaq ® Ready Mix TM PCR Mix with MgCl 2 (Sigma-Aldrich, USA) was used. The final volume of each PCR reaction was 25 ml and contained 12.5 ml of REDTaq Ready Mix, 1 ml of each forward and reverse primer (concentration between 0.1-1.0 mM), 1 ml of DNA (50-200 ng), and 9 ml of water. The reactions were performed using a Whatman Biometra thermocycler, whereas the PCR products were subjected to agarose gel electrophoresis (2% agarose, 1xTRIS-acetate-EDTA, 120 mV, 40 min). The gels were stained with ethidium bromide and the PCR products were visualised using a Wilbert Lambert transilluminator and compared with molecular size markers [Gene Ruler TM 100 bp DNA Ladder (Fermentas, Thermo Scientific, USA)].

Identification of genes by polymerase chain reaction (PCR) -The sequences of the primers and the conditions of the PCR reactions are presented in
Ethics -The study design and protocols were approved by the Ethical Committee of the Medical University of Lublin (KE-0254/75/2011).

RESULTS
The 75 S. epidermidis isolates expressed resistance to erythromycin with the following mechanisms of resistance: 27 (36%) strains exhibited cMLS B resistance, 14 (18.7%) strains exhibited iMLS B resistance, and 34 (45.3%) strains exhibited MS B resistance (Figure). Twenty-five isolates exhibited L-phenotypes and were determined to be either resistant to only lincomycin (24 strains) or resistant to lincomycin and clindamycin (1 strain).
The MICs of spiramycin among erythromycin-resistant S. epidermidis were evaluated as follows: > 128 mg/L for all cMLS B strains, from 4-> 128 mg/L for iMLS B strains, and from 1-4 mg/L for strains exhibiting the MS B   (Table II). Moreover, for the 11 (78.6%) strains exhibiting iMLS B phenotypes, the noninhibition zone around the spiramycin disc was found together with a D-shaped zone around the clindamycin disk.

DISCUSSION
CoNS are potential reservoirs of antibiotic resistance genes, which can be transferred to S. aureus not only in vitro but also in vivo (Reyes et al. 2007, Otto 2013. Erythromycin resistance among CoNS was previously reported to result from a methylase encoded by different erm family genes that can be horizontally transferred to recipient strains (Zmantar et al. 2011, Vitali et al. 2014. Hence, surveillance of erythromycin resistance and MLS B resistance in CoNS at phenotypic and genetic levels can provide important information regarding their current epidemiology. Among the S. epidermidis strains studied, the most frequently identified gene in strains exhibiting both cMLS B and iMLS B phenotypes was ermC, which is consistent with previous reports (Reyes et al. 2007, Gherardi et al. 2009, Coutinho et al. 2010, Bouchami et al. 2011, Brzychczy-Wloch et al. 2013, Heb & Gallert 2014. Only a few S. epidermidis exhibiting cMLS B phenotypes possessed ermA and/or ermB. Similar data have been previously reported (Bouchami et al. 2011, Teodoro et al. 2012, Szczuka et al. 2016). Moreover, the presence of other erm genes (e.g., ermF) has been rarely detected in Staphylococcus spp (Roberts 2004). Notably, the distribution of erm genes depends on the bacterial species. For example, ermA is more characteristic of S. aureus, whereas ermB is more characteristic of beta-haemolytic streptococci (Roberts 2004, Buter et al. 2010, Meehan et al. 2014, Vitali et al. 2014. Moreover, among CoNS, the type of erm gene also depends on the geographical region of their isolation. For example, ermC was previously detected in 50% of the strains exhibiting MLS B resistance in Great Britain, whereas it was detected 90% of those in Denmark (Lim et al. 2002, Gatermann et al. 2007, Cetin et al. 2010, Bouchami et al. 2011) and in Mexico, ermA was reported as predominant in S. epidermidis (Castro-Alarcón et al. 2011).
The MS B S. epidermidis isolates examined contained an msrA gene encoding an ATP-dependent efflux pump, which actively removes 14-,15-membered MS B . The MS B phenotype observed in msrA-negative S. epidermidis strains may be the result of the presence of mphC, which encodes for a macrolide-modifying enzyme (Gatermann et al. 2007), thereby resulting in a "false-positive" MS B phenotype.
All S. epidermidis isolates with L-phenotypes generally contained the linA/A' gene. Data from Novotna et al. (2005Novotna et al. ( , 2007 also indicated a connection between the presence of the linA/A' gene and resistance to only lincomycin among staphylococci. The S. epidermidis strains  studied exhibited resistance to lincomycin, but susceptibility to clindamycin as a result of increased enzyme affinity for lincomycin (Achard et al. 2005). Resistance both to lincomycin and clindamycin may be a consequence of the presence of other lin family genes or vga(A) LC , which encodes a "new" variant of the SgA protein that is responsible for cross-resistance to streptogramins A and all lincosamides (Novotna & Janata 2006). Among the iMLS B and cMLS B S. epidermidis strains, the erm genes do not exist separately, but in combination with others (predominantly with mphC). Notably, other erm genes (e.g., ermF), which are rarely detected in Staphylococcus spp, may encode both the inducible or constitutive MLS B phenotypes (Roberts 2004). In MS B -positive S. epidermidis strains, the msrA genes predominantly coexist with ermC, mphC, and linA/A', and the coexistence of msrA and ermC has also been previously reported (Roberts 2004, Novotna et al. 2007, Wang et al. 2008, Teodoro et al. 2012. Moreover, the presence of the linA/A' gene in msrA-positive strains results in resistance to lincomycin. The S. epidermidis strains exhibiting L-phenotypes correlated with the presence of the linA/A' gene in most of the strains that also contained the ermC and mphC genes, whereas those strains did not contain the msrA gene. Notably, the ermC genes were also detected in both of the MS B and L-phenotype S. epidermidis strains -but without its expression -suggesting a defect in ermC expression.
Previous studies have reported (Leclercq 2002, Coutinho et al. 2010) that 16-membered macrolides (e.g., spiramycin) are not inducers of MLS B resistance in staphylococci. According to our data, spiramycin is able to induce resistance to clindamycin among the iMLS B S. epidermidis isolates examined. Moreover, iMLS B S. epidermidis strains, which contain ermC, exhibited resistance to spiramycin in vitro. These observations contradict previous reports that 16-membered macrolides remain active against staphylococci that exhibit iMLS B phenotypes (Leclercq 2002, Szczuka et al. 2016. Notably, resistance to spiramycin appears to be characteristic of iMLS B streptococci containing ermB (Leclercq 2002, Acikgoz et al. 2003.
The diversity of genes involved in different mechanisms that are responsible for the resistance of S. epidermidis to MLS B antibiotics suggests that the insensitivity of CoNS strains to these antibacterial drugs is not necessarily a unidirectional process and that the coexistence of various genes may influence the nature of their resistance.