Background

S. epidermidis are currently the most frequently isolated species among coagulase-negative staphylococci (CoNS) from clinical materials [1,2,3,4]. However, also other species of this group such as S. haemolyticus, S. hominis, S. warneri and S. simulans have recently been gaining significance [5,6,7,8,9]. They substantially contribute to developing vascular and foreign body-related infections by adhering to tissues and forming biofilm, thus impairing the efficiency of antibiotic treatment. These opportunistic pathogens are primarily isolated from immunocompromised patients as well as from patients undergoing invasive medical treatment. CoNS strains also constitute a therapeutic challenge due to their increasing frequency of the acquisition of the resistance to commonly used drugs [1, 3, 10,11,12,13,14,15]. S. haemolyticus strains have been described as having extensive antibiotic resistance [16]. However, some CoNS species such as S. simulans are rarely isolated and generally overlooked during regular diagnosis [7]. Other species like S. warneri appear unexpectedly as etiological agents of serious infections [8, 17, 18].

Macrolides and lincosamides occupy one of the leading positions among antibiotics used in therapy, especially in an outpatient treatment. Their role has substantially increased due to the emergence of methicillin-resistant staphylococci and the fact that they also provide an alternative for people allergic to β-lactam antibiotics [19, 20]. As a result, various types of bacteria have acquired resistance to macrolides and this phenomenon has become a global issue. Due to the fact that the binding site of streptogramins B and lincosamides overlaps that of macrolides, the resistance to macrolides often leads to cross-resistance between macrolides, lincosamides and streptogramins B (MLSB resistance phenotype). Many genes limiting the activity of these antibiotics have been described. They affect the expression of several mechanisms of resistance i.e. the methylase production which is responsible for the methylation of the adenine of the 23S rRNA ribosomal subunit as well as active efflux and the production of antibiotic-inactivating enzymes [21, 22]. The phenotypic expression of resistance genes can either be constitutive or inducible and as a result cause additional therapy limitations [21, 23].

As there is no universal mechanism or set of genes determining MLSB resistance, it is difficult to determine the extent to which horizontal gene transfer contributes to the prevalence of this form of resistance when compared with clonal spread. It is known that genes determining MLSB resistance are often incorporated in chromosomal cassettes and might be carried in plasmids what indicates that there is a possibility of their horizontal transfer [24,25,26].

The mechanism of MLSB resistance as well as the genes responsible for this phenomenon have been thoroughly studied in S. aureus and S. epidermidis. However, little is known about the prevalence and the genetic bases of MLSB resistance genes in other CoNS species. In order to address this knowledge gap, the aim of this study is therefore to clarify the MLSB resistance mechanisms in rarely isolated CoNS species such as S. haemolyticus, S. hominis, S. warneri and S. simulans.

Methods

Bacterial strains

A total of 97 clinical isolates were included in the study. All isolates tested in the study were obtained by a diagnostic microbiology laboratory (Synevo Sp. z o.o.) of Łódź area, Poland since 2014 to 2016. The strains were added to the collection of Pharmaceutical Microbiology and Microbiological Diagnostics Department, Medical University of Łódź and belonged to the following species: S. epidermidis, S. hominis, S. haemolyticus, S. warneri and S. simulans. However, while all S. epidermidis and S. hominis strains were obtained from blood samples, the other CoNS species derived from a range of clinical specimens including blood, ear swabs, conjunctival sacs, samples from the genitourinary system, abscesses or skin lesions. Seventeen strains were identified as S. epidermidis, 23 as S. haemolyticus, 19 as S. hominis, 20 as S. warneri and 18 as S. simulans. The isolates were identified both by means of the MALDI-TOFF technique and genetic methods [27, 28]. The control strains used in PCR identification assays were as follows: Staphylococcus epidermidis ATCC 12228, S. haemolyticus ATCC 29970, S. warneri ATCC 27836, S. simulans ATCC 27848 and S. hominis subsp. hominis ATCC 27844.

Detection of phenotypic resistance to methicillin and the MLSB resistance phenotype

Resistance to methicillin was determined by means of disc diffusion assays using discs with cefoxitin 30 μg (Oxoid, UK) on Mueller-Hinton agar (Emapol, Poland) according to the rules proposed by the European Centre for Disease Prevention and Control (EUCAST). The diameter of the zone of inhibition was evaluated after an 18-h incubation at 35 °C. Interpretation of the diameters of inhibition zones was as follows - for S. epidermidis: sensitive when ≥25 mm, resistant when < 25 mm; for S. aureus and coagulase-negative staphylococci other than S. epidermidis: sensitive when ≥22 mm, resistant when < 22 mm. S. aureus ATCC 29213 was used as negative control. Staphylococcus aureus ATCC 43300 was used as positive control.

Resistance to macrolides, lincosamides and streptogramins B was determined by means of disc diffusion assays with erythromycin (15 mg) and clindamycin (2 mg) discs (Oxoid, UK) according to EUCAST. The MLSB resistance mechanisms were detected by D-test, by way of placing above mentioned discs 15–20 mm from each other. After the preparation of bacterial suspensions in line with the 0.5 McFarland standard, the bacteria were inoculated on the surface of Mueller-Hinton agar. Subsequently, the discs were added and the plates were incubated at 35 °C for 18 h. Interpretation of the diameters of inhibition zones was as follows - for erythromycin: sensitive when ≥21 mm, resistant when < 18 mm; for clindamycin: sensitive when ≥22 mm, resistant when < 19 mm.

Both resistance to erythromycin and clindamycin indicated the presence of a constitutive type of MLSB resistance (cMLSB). The appearance of a flattening zone around the clindamycin confirmed the presence of inducible resistance (iMLSB). Isolates that were resistant to erythromycin and sensitive to clindamycin exhibited an MS phenotype. S. aureus ATCC 29213 was used as a control strain.

Detection of the mecA gene and macrolide, lincosamide and streptogramin B resistance genes

Genomic DNA was isolated from the tested isolates within the study by means of Genomic Micro AX Staphylococcus Gravity (A&A Biotechnology, Poland) commercial set according to the manufacturer’s protocol. PCR was conducted with primers synthesized by the Genomed Sequencing Laboratory, Poland, specific for the mecA [29], erm(A), erm(B), erm(C), lnu(A), msr(A), msr(B) [23], mph(C) [30], ere(A) and ere(B) [31] genes. DNA amplification was performed in the thermal cycler (Biometra, Germany). The reaction products were detected by electrophoresis (70 V, 1.5 h) in 1% (w/v) agarose gels containing Midori Green DNA stain (NIPPON Genetics EUROPE, Germany). A CCD camera (Syngen, Poland) was used in order to obtain results. The sizes of the amplification products were determined by the usage of a commercial molecular size marker DraMix or DNA Marker 2 (A&A BIOTECHNOLOGY, Poland). The presence of tested of genes was determined on the basis of the sizes of the amplification products with specific primer. S. epidermidis ATCC51625 harboring the mecA gene and 3 S. epidermidis strains: S. epidermidis 3718INL harboring erm(B) and mph(C) genes, S. epidermidis 1486IG harboring erm(A), erm(C) and mph(C) genes as well as S. epidermidis 1923KIINL harboring the msr and lnu(A) genes were used as positive controls [32].

Statistical analysis

Statistical differences were conducted using chi-square (χ2) test. P value of < 0.05 was considered as significant.

Results

Phenotypic resistance to methicillin as well as the presence of the mecA gene in all isolates were determined. The resistance to methicillin were identified in 67% of the isolates. All S. epidermidis and S. hominis isolates were resistant to methicillin and harbored the mecA gene. Eighteen S. haemolyticus isolates were phenotypically resistant to methicillin and harbored the mecA gene whereas two strains had mecA gene without phenotypic resistance. The mecA gene was also determined in 6 S. warneri and 5 S. simulans isolates, however, phenotypic resistance was confirmed respectively only in four and one strains belonging to abovementioned species.

The MLSB resistance mechanism was found in isolates belonging to all species investigated in the study. However, the prevalence of this resistance was not evenly distributed (Table 1).

Table 1 The number of isolates with specific phenotypes of studied CoNS species
Full size table

Some isolates were sensitive to MLSB antibiotics: 2 S. epidermidis strains (both of them resistant to methicillin), 8 S. warneri isolates (three resistant to methicillin) and 10 S. simulans strains (one was resistant to methicillin). The MLSB resistance mechanism was common in S. hominis, S. haemolyticus and S. epidermidis isolates – respectively in 100, 78 and 59% of isolates. In most cases it was a constitutive type. The inducible type manifesting a D-shape on Mueller-Hinton agar was observed in several isolates of CoNS included in the study, however, in none of the S. haemolyticus. Only in S. warneri this type of resistance existed in a greater number of strains than the constitutive type. In some S. epidermidis, S. haemolyticus and S. warneri isolates the MSB phenotype occurred.

Distribution of genes and their phenotypic manifestation varied between the analyzed species. The sets of genes encoding macrolides, lincosamides and streptogramins B resistance compared with phenotypic MLSB resistance in isolated tests were presented in Table 2.

Table 2 Phenotypic manifestation and genes encoding the MLSB resistance as well as the methicillin resistance in tested isolates belonging to the species: S. epidermidis, S. hominis, S. haemolyticus, S. warneri and S. simulans
Full size table

The erm(C) gene predominated in all isolates with the MLSB resistance mechanism and this was statistically significant in all species (p values were as follows: 0.00001 for S. epidermidis, 0.01 for S. hominis, 0.00001 for S. haemolyticus, 0.0001 for S. warneri and 0.0002 for S. simulans). The erm(C) gene was present solely or together with other resistance genes in 12 out of 15 resistant strains S. epidermidis. Similarly to S. epidermidis, MLSB resistant S. hominis and S. haemolyticus isolates also accumulated many genes. The erm(A) gene was present in S. epidermidis and S. hominis isolates, while lnu(A) only in S. hominis and S. haemolyticus. There were no such genes in other species. The MLSB resistance mechanism appeared also in 3 S. hominis isolates and 7 S. haemolyticus isolates that do not possess the erm gene. Among the examined S. simulans and S. warneri isolates the MLSB resistance mechanism was less frequent and associated with fewer number of genes. However, the erm(C) gene was also predominant. Additionally, there were detected: the msr gene family in isolates of S. simulans, while the msr and mph gene families in S. warneri. In the examined S. warneri isolates, the resistance to macrolides and streptogramins B (MSB) was frequently associated with the presence of the mph(C) and msr(B) genes. The existence of mph(C) and msr(B) genes in isolates with MSB resistance mechanism was statistically significant (p values were as follows: 0.006 for mph(C) and 0.0007 for msr(B)). The sole erm(C) gene was found in 24 out of 97 strains (25%). However, the most prevalent gene was msr(A) which was present in 37 of all isolates (38%). In S. warneri and S. haemolyticus it was occasionally identified as the sole gene. The msr(A) gene often coexisted with msr(B), mph(C) and erm(C). Such a set of genes was statistically significant for S. haemolyticus (p value = 0.0001). The prevalence of mph(C), was not evenly distributed among species. It was present in 20 S. haemolyticus isolates, 7 S. hominis, 4 S. warneri and 2 S. epidermidis whereas it was not detected in S. simulans species.

In most cases, MLSB resistance genes coexisted with the phenotypic methicillin resistance encoded by the mecA gene. However, no mecA was detected in 8 S. warneri, 4 S. haemolyticus and 2 S. simulans isolates harboring the macrolides, lincosamides and streptogramins B genes.

Discussion

In spite of different chemical structures, macrolides, lincosamides and streptogramins B exhibit a similar mechanism of action leading to cross-resistance (MLSB) between them [21,22,23]. The study examines molecular bases of phenotypic manifestation of resistance to macrolides and lincosamides among rarely isolated coagulase-negative staphylococci. This form of resistance is becoming increasingly prevalent and is regarded as being determined by various genes. It may result from the presence of the erm family genes encoding the methylase which is responsible for the methylation of adenine in the 23S rRNA ribosomal subunit. For instance, it is known that the erm(A) gene is prevalent in S. aureus, the erm(B) gene in β-hemolytic streptococci and erm(X) in corynebacteria and propionibacteria. Resistance to macrolides and streptogramins B (MSB phenotype) can also be determined by the presence of genes responsible for active efflux e.g. the msr(A)/msr(B) genes in Staphylococcus spp., the mef gene in Streptococcus spp. or Haemophilus influenzae. The third mechanism of resistance to antibiotics of the MLSB group is the production of enzymes inactivating a specific substance. These enzymes might be encoded by genes such as mph(C) and lnu(A) in Staphylococcus spp. or by ere(A) and ere(B) in Escherichia and other genera of Enterobacteriaceae [33,34,35,36,37,38,39,40].

Resistance to MLSB antibiotics in staphylococci has already been the subject of many studies. However, these studies were mostly limited to the most frequently isolated staphylococci i.e. S. aureus and S. epidermidis [13, 34, 35, 41,42,43]. Recently, also other staphylococcus species such as S. hominis, S. haemolyticus, S. warneri and S. simulans, have appeared as etiological factors of serious human infections. Due to their resistance to antibiotics and ability to form biofilms, these species have become an often underestimated clinical issue. However, the progress in diagnostic laboratory techniques has allowed to see CoNS as a set of specific species instead of a homogeneous group.

Our findings indicate that the MLSB cross-resistance mechanism was present in 61% of all isolates whereas the MSB phenotype was observed in a further 19% of isolates. Li et al. (2015) reported that resistance to erythromycin was more frequent in CoNS isolates than in S. aureus and concerned more than half of the strains isolated from mastitis cases in cattle [44]. Similar results were obtained in studies on strains derived from humans. Goudarzi et al. (2016) noted that in isolates obtained from the nasal vestibule, the strains belonging to the CoNS group proved to be more resistant to erythromycin than S. aureus [41]. The phenotypic expression of MLSB resistance might be inducible and manifest in clinical resistance to lincosamides and streptogramin B induced by 14- and 15- membered macrolides, or constitutive which determines resistance to all MLSB group antibiotics [21, 23]. Analyzed in our study staphylococcal species form two groups. The first one contains S. epidermidis, S. hominis and S. haemolyticus. They were isolated from blood and relatively common in clinical laboratory. The greatest number of the isolates of these species exhibited constitutive type of the MLSB resistance. Similar conclusions were drawn by other researchers [35, 45,46,47]. However, some studies such as these conducted by Dizbay et al. (2008), Teodoro et al. (2012), Szczuka et al. (2016) and Li et al. (2015), indicated that the inducible MLSB resistance might also be commonly present in S. haemolyticus and S. hominis species [44, 48,49,50]. The iMLSB mechanism has been frequently reported to occur mostly in S. aureus [35, 50, 51]. The second group of species investigated in our study consists of S. warneri and S. simulans which are not often isolated from blood and their pathogenecity potential is weaker. Nonetheless, some of these isolates were also resistant and among S. warneri strains the the iMLSB resistance mechanisms were more frequent. The isolates included in our study were tested for the presence of genes responsible for encoding macrolides, lincosamides and streptogramins B resistance in staphylococci: erm(A), erm(B), erm(C), lnu(A), msr(A), msr(B) and mph(C). Additionally, the examined isolates were tested for the presence of ere(A) and ere(B) genes that are described as typical for the Enterobacteriaceae family but might also appear in staphylococci [52]. The presence of these genes in staphylococci might indicate transfer of genes between various species and genera of bacteria. The studies regarding CoNS emphasize that the MLSB resistance results mostly from methylase activity and is conditioned by the erm(C) gene [35, 45, 53,54,55]. This gene was frequently detected in the present study, particularly in S. epidermidis, S. hominis and S. simulans isolates. This gene also dominated in the examined by El-Mahdy et.al. (2010) S. epidermidis and S. simulans strains [56]. Gatermann et.al. (2007) found erm(C) as prevalent in S. haemolyticus [45]. In our findings erm(C) was detected as a sole gene in part of S. hominis isolates but in other isolates it coexisted with a range of other genes i.e. lnu(A), msr(A), msr(B) and mph(C). El-Mahdy et.al. (2010) found the msr(A) gene in S. hominis strains as dominant whereas the result obtained in the study conducted by Szczuka et.al. (2016) proved to be similar to ours - the erm(C) gene predominated [50, 56].

Resistance to macrolides and streptogramins B (MSB) caused by the msr(A)/msr(B) genes is responsible for active efflux of antibiotics out of bacterial cells. The mrs(A) gene was indicated as more prevalent in CoNS than in S. aureus [35, 41, 57]. In the presented study, msr(A) proved to be the most frequently detected gene. It was present mostly in groups with msr(B) and mph(C) which condition the synthesis of phosphotransferase. Similarly to the result obtained by Gaterman et al. (2007), the mph(C) gene was harbored almost exclusively together with other genes [45]. In our study it was determined as a sole gene in two isolates (S. haemolyticus and S. warneri) resistant to macrolides. Matsuoka et al. (2003) reported that in S. aureus the mph(C) gene occurs only if the msr(A) gene is also present. The sole presence of the mph(C) gene results in a low macrolide resistance [58]. The presence of the lnu gene was detected in S. haemolyticus and S. hominis isolates but not in the S. epidermidis, S. warneri or S. simulans isolates. This gene was frequently identified in S. haemolyticus strains by Novotna et al. (2005) and in S. hominis isolates by Szczuka et al. (2016) [50, 59]. The lnu gene is considered to be responsible for lincomycin resistance and might be present in strains susceptible to clindamycin [59, 60]. The presence of the ere genes facilitates the resistance to high concentrations of erythromycin by hydrolysis of the lactone rings [31]. The ere(A) and ere(B) genes were not detected in any of the isolates tested in the study. However, after PCR with primers applied for ere(A) we detected the presence of the product of size corresponding to sought gene in seven isolates of S. epidermidis. The result seemed to be reliable as in the studies there was not detected any of unspecific products in sensitive stains. What is more, the primers designed by Sutcliffe et al. (1996) have been already applied to detect this gene in staphylococci [31]. In a study made by Schmitz et al. (2000) that concerned 851 S. aureus isolates, ere(A) was not identified whereas ere(B) was present in less than 1% of all isolates [61]. The same author in other article described the presence of the ere(A) gene in one strain of S. aureus [62]. The lack of the ere genes in S. aureus was described by Otsuka et al. (2007) [63]. On the other hand, Li et al. detected the presence of the examined gene in six strains of CoNS and 21 strains of S. aureus of bovine milk origin [44]. To our knowledge, it would be the first time when the presence of the ere(A) gene in CoNS isolated from humans has been reported. Because of this fact, we have sequenced received PCR products in order to check the homology with ere(A) gene that was hitherto mainly detected in bacteria from Enterobacteriaceae family. Unfortunately, after the comparison of the sequences in the BLAST program available at http://www.ncbi.nlm.nih.gov/BLAST/ it has turned out that in all seven strains of S. epidemidis the obtained sequences are identical but with no homology with the sequence of the ere(A) gene. Moreover, the received sequences turned out to be homologous in 98% with the genome fragment of Staphylococcus capitis CR01. Our studies revealed that primers designed by Sutcliffe et al. should not be used for staphylococci [31].

Our findings show a correlation between the presence of specific genes or sets of genes and the phenotypic type of the MLSB resistance. The predominant type of the MLSB resistance in S. epidermidis, S. hominis and S. simulans was constitutive. This mechanism was encoded by the erm(C) gene as a sole gene or as a part of a group of the genes and it was statistically proved. However, erm(C) was also detected in isolates with both inducible MLSB and MSB resistance. Gatermann et al. (2007) reported that the presence of the erm(C) gene was associated with the inducible MLSB mechanism in S. hominis strains whereas mostly with a constitutive type in other species [45]. Similar results were obtained in the study by Szczuka et al. (2016), in which more than half of the 55 S. hominis isolates exhibited the iMLSB mechanism which was associated with the presence of the erm(C) gene [50]. In our study, the inducible MLSB mechanism was less frequent than constitutive but both were associated with the presence of the erm(C) gene.

Methicillin resistance was observed in all S. epidermidis and S. hominis, the majority of S. haemolyticus isolates and also in minority of S. warneri and S. simulans. Therefore, it is difficult to evaluate the scope of correlation between the resistance to β-lactam antibiotics and macrolides and lincosamides. However, the in-depth analysis of resistant strains indicate that genes encoding resistance to these antibiotics might be transferred in a common chromosomal cassette [64]. In a study of MRSA strains, Teodoro et al. (2012) reported that the erm(A) gene was most commonly present in TN554 as a part of SCCmec II or III cassettes whereas in methicillin resistant S. epidermidis, erm(C) was located mainly on SCCmec IV or nontypeable cassettes [49]. A phenomenon of the common occurrence of the MLSB and methicillin resistance was also analyzed by Castro-Alarco’n (2011) [65]. Bouchami et.al. (2011) indicated the frequent occurrence of the mec(A), erm(C) and msr(A) genes in S. epidermidis, S. haemolyticis and S. hominis isolates derived from blood [66]. Towards increasing resistance to methacillin in staphylococci, including CoNS, the range of antibiotics that are available for treatment of infections caused by these bacteria is getting more limited. The alternative can be macrolides, lincosamydes and streptogramins. Additionally, erythromycin and clindamycin that are recommended for the treatment of the infections of people with ß-lactam allergy [20]. Spreading of the resistance also to such antibiotics may considerably complicate the treatment in the future. Moreover, CoNS might become dangerous genetic reservoir of resistance genes for S. aureus.

Conclusion

Macrolide resistance is rapidly increasing. Therefore, it is reasonable to assume that genes encoding this type of resistance can be transferred between species and genera. However, due to the diversity of the genes involved it is too early to link them with specific mechanisms. Progress in medicine and the increasing contribution of opportunistic pathogens to infections in immunocompromised patients has shifted the profile of the bacterial species responsible for such infections. Apart from S. epidermidis, other species such as these described in our study have become more and more frequently isolated. As a result of antibiotic selection such strains which are simultaneously natural human skin microbiota might become dangerous genetic reservoir of resistance genes.