Abstract
The aim of the present study was to determine the rate and mechanisms of resistance to macrolides, lincosamides, and streptogramin B (MLSB) antibiotics of Staphylococcus aureus collected in Central Greece. Of the 2,893 S. aureus collected during 2012–2017, 1,161 isolates (40.2%) exhibited resistance to at least one of the MLSB agents. The rate of erythromycin resistance was statistically significantly higher in methicillin-resistant S. aureus (MRSA) (58.6%) than in methicillin-sensitive S. aureus (MSSA) isolates (20.7%) (p = 0.002). Two hundred seventy-five representative MLSB-resistant S. aureus, including 81 MSSA and 194 MRSA isolates, were further studied. Thirty-eight MSSA isolates carried ermC, 26 MSSA were positive for ermA, whereas 17 isolates carried msrA gene. Among MRSA, the ermA gene was identified in the majority of the isolates (n = 153). Thirty-seven MRSA isolates carried ermC; three isolates carried msrA, whereas the remaining MRSA was positive for two genes (ermA and ermC). Phylogenetic analysis showed that ST225, which belongs to CC5, was the most prevalent, accounting for 137 MRSA isolates. Higher genetic diversity was found in the group of MSSA isolates, which comprised of 13 sequence types. Whole-genome sequencing data showed that all ermA-positive S. aureus, with the exception of one ST398 isolate, harbored the ermA-carrying Tn554 transposon integrated into their chromosomes. Furthermore, Illumina sequencing followed by polymerase chain reaction screening identified that ermC, which was identified in a polyclonal population of MSSA and MRSA isolates, was carried by small plasmids, like pNE131. These findings highlighted the important role of high-risk clones and of mobile elements carrying resistance genes in the successful dissemination of MLSB-resistant staphylococci.
Introduction
Staphylococcus aureus is a major bacterial pathogen causing a variety of infections ranging from skin and soft tissue infections to life-threatening conditions, such as endocarditis, pneumonia, and septicemia. 1 In 1961, S. aureus developed methicillin resistance due to the acquisition of the mecA gene. Methicillin-resistant S. aureus (MRSA) has dramatically increased in recent years, 2 and various MRSA clones have been associated with hospital- and community-associated infections. 3 The emergence of MRSA strains has underlined the need for alternative therapeutic options. Macrolides, lincosamides, and streptogramin B (MLSB) are among the most common antibiotics used for the treatment of staphylococcal infections in humans and animals.
MLSB antibiotics are chemically distinct, but they have similar effects on bacterial protein synthesis. However, the extensive use of MLSB antibiotics has been accompanied by an increased number of resistant strains. 4 The mechanisms of resistance to MLSB antibiotics in staphylococci are target-site modification, active drug efflux, and drug inactivation. 5 The most widespread mechanism is the methylation of ribosomes, which is the target of MLSB antibiotics. Methylation of ribosomes confers resistance to MLSB antibiotics 6 and is mediated by erm genes.
ABC-F proteins, encoded by msr, vga, lsa, and sal genes, mediate resistance by active efflux or by ribosomal protection.7,8 The msr genes encode for resistance to 14- and 15-membered macrolides and streptogramin B, whereas strains positive for the vga, lsa, or sal genes are resistant to pleuromutilins, lincosamides, and streptogramin A.7,9
On the other hand, enzymatic inactivation confers resistance to structurally related antibiotics only. Esterases and phosphotransferases encoded by ere and mphC genes, respectively, confer resistance to 14- and 15-membered macrolides.10,11 In addition, lnu genes encode nucleotidyl transferases inactivating lincosamides only,9,12 whereas vgb genes encode enzymes hydrolyzing streptogramin B.8,12
The rate of MLSB-resistant staphylococci varies in different geographic regions and species. A previous study from our group has shown that 11.7% of S. aureus isolates, collected in University Hospitals of Patras and Larissa (Southern and Central Greece) during 1999–2001, were resistant to erythromycin, whereas the ermC gene was the predominant among MRSA isolates. 13 However, in the last decade, an increase of erythromycin-resistant (Ery-R) S. aureus isolates was observed in our country. Thus, the aim of the present study was to determine the rate and mechanisms of resistance to MLSB antibiotics of S. aureus collected in Central Greece during 2012–2017. Also, in representative strains, the complete nucleotide sequences of the mobile genetic elements associated with the spread of erythromycin resistance determinants were characterized by whole-genome sequencing (WGS) to elucidate a possible intraspecies dissemination of resistance mechanisms.
Materials and Methods
Bacterial isolates and susceptibility testing
A total of 2,893 consecutive S. aureus isolates recovered from clinical samples, including bloodstream infections, wounds, pneumonia, arthritis, soft tissue, and skin infections of individual patients (hospitalized and outpatients) admitted in the University Hospital of Larissa (UHL) between 2012 and 2017, were included in the study. UHL is a 800-bed tertiary care hospital in Central Greece, serving a population of 1,000,000 habitants. Duplicate isolates from the same patients, even if the site of infection was different, were excluded. Identification to species level and susceptibility testing against various antimicrobial agents were performed by the automated system Vitek®2 (BioMerieux, France). Minimum inhibitory concentrations (MICs) were interpreted according to European Committee on Antimicrobial Susceptibility Testing (EUCAST) criteria (www.eucast.org). To confirm the inducible MLSB (iMLSB) phenotype, double-disk diffusion test was performed in all isolates that expressed resistance to erythromycin (MICs >2 mg/L) and susceptibility to clindamycin (MICs >0.5 mg/L). 13
Two hundred seventy-five MLSB-resistant S. aureus were selected to be further studied. These isolates were selected as representatives of different MLSB phenotype (inducible or constitutive or others), MRSA or methicillin-sensitive S. aureus (MSSA) phenotypes, susceptibility pattern to various antimicrobial agents, site of infection, age of patient, time of isolation, and ward of admission.
Detection of resistance determinants
Genes encoding for resistance to MLSB were detected by polymerase chain reaction (PCR), using specific primers for ermA, ermB, ermC, ermF, ermY, ermT, msrA, msrB, ereA, ereB, lnuA, lnuB, and mphC. 14 Both strands of the PCR products were sequenced using an ABI 377 sequencer (Applied Biosystems, Foster City, CA).
Multilocus sequence typing of S. aureus isolates
The isolates were typed by Multilocus Sequence Typing (MLST). MLST was performed by sequencing of the seven gene loci, arcC, aroE, glpF, gmk, pta, tpi, and yqi. 15 The data for S. aureus alleles and sequence types (STs) were obtained through the MLST database at https://pubmlst.org/saureus. The eBurst v.3 was used to clarify the related STs into clonal complexes (CCs). 16
Illumina sequencing
Eight erm-positive staphylococcal isolates were selected for WGS. These isolates were selected as representative of different STs, isolation periods, and origin.
Genomic DNAs from the clinical isolates were extracted using the DNASorb-B Kit (Sacace Biotechnologies Srl, Como, Italy). Multiplexed DNA libraries were prepared, using the Nextera XT Library Preparation Kit, and 300-bp paired-end sequencing was performed on the Illumina MiSeq platform (Illumina, Inc., San Diego, CA) using the MiSeq v3 600-cycle Reagent Kit. Initial paired-end reads were quality trimmed, using the Trimmomatic tool v0.33, 17 with the sliding window size of 4 bp, required average base quality ≥17, and minimum read length of 48 bases. Then, reads were assembled by use of the de Bruijn graph-based de novo assembler SPAdes v3.9.1, 18 using k-mer sizes 21, 33, 55, 77, 99, and 127. For sequence analysis and annotation, the BLAST algorithm (www.ncbi.nlm.nih.gov/BLAST), the ISfinder database (www-is.biotoul.fr/), and the open reading frame finder tool (www.bioinformatics.org/sms) were utilized. Comparative genome alignments were performed using the Mauve v2.3.1 program. 19
Detection of erm-carrying mobile elements
Based on the results from Illumina sequencing (see erm-associated mobile elements subhead under the Results section), eight PCRs targeting characteristic regions of erm-carrying mobile elements sequenced during this study were designed. The selected regions included: (i) the ermA-carrying transposon Tn554, (ii) the ermA-carrying transposon Tn6133, (iii) and the ermC-carrying plasmid pSau-2716Lar. All erm-positive clinical strains were screened for the presence of the regions described above by the use of specific primers (Supplementary Table S1; Supplementary Data are available online at www.liebertpub.com/mdr). PCR products were sequenced as described above.
Statistical analysis
For analyzing the correlation between year when the samples were collected, and the percentage of MLSB-resistant S. aureus, we used the method Pearson's correlation. For analyzing the possible difference between the percentage of MLSB-resistant MSSA and MLSB-resistant MRSA, we used the Mann–Whitney U test. In both tests, values of p < 0.05 were considered as statistically significant. The computation was done by the program Review Manager (RevMan) Version 5.3. 20
Nucleotide sequence accession numbers
The nucleotide sequences of the erm-carrying fragments described in this study have been deposited in GenBank under accession numbers MH423311, MH423312, and MH454238-MH4543. Whole-genome assemblies of S. aureus isolates were deposited in NCBI under accession number PRJNA487708.
Results
MLSB-R staphylococci
Of the 2,893 S. aureus, 1,161 isolates (40.2%) exhibited resistance to at least one of the MLSB agents (Table 1). The majority of MLSB-resistant S. aureus (1,077 isolates) exhibited resistance to erythromycin and clindamycin, and expressed either the consecutive MLSB-phenotype (cMLSB) or the inducible MLSB-phenotype (iMLSB) (765 and 312 isolates, respectively). Among the remaining MLSB-resistant isolates, 84 were only resistant to erythromycin (MSB phenotype). No isolate with resistance to clindamycin (L phenotype) was detected. Table 1 describes the rate of MLSB-resistance and the distribution of phenotypes among MRSA and MSSA during the 6-year study period; the rate of MLSB-resistance was statistically significantly higher in MRSA than in MSSA (p = 0.002). Additionally, there was a statistically significant positive correlation between the isolation year and the percentage of MLSB-resistant S. aureus (Correlation Coefficient = 0.839, and p = 0.037).
Rate of Resistance and Distribution of Macrolides, Lincosamides, and Streptogramin B-Resistant Phenotypes Among Methicillin-Resistant Staphylococcus aureus and Methicillin-Sensitive Staphylococcus aureus During the Study Period
MRSA, methicillin-resistant S. aureus; MSSA, methicillin-sensitive S. aureus; MLS, macrolides, lincosamides, and streptogramin.
In addition, according to the susceptibility results, 53.8% of the MLSB-resistant S. aureus also exhibited resistance to levofloxacin, 52.0% to tobramycin, 32.4% to tetracycline, 28.7% to fusidic acid, 9.5% to gentamicin, 4.4% to trimethoprim–sulfamethoxazole, and 2.2% to rifampicin. None of the MLSB-resistant isolates expressed resistance to mupirocin.
To determine the resistance mechanisms, 275 MLSB-resistant S. aureus, including 81 MSSA and 194 MRSA isolates, were selected to be further studied. Among the MSSA isolates (n = 81), 22 expressed the cMLSB phenotype, 42 expressed the iMLSB phenotype, and 17 the MSB phenotype (Table 2). Most of MSSA isolates carried the ermC gene (n = 38), whereas 26 isolates were positive for the ermA. The 17 MSSA isolates expressing the MSB phenotype carried the msrA gene. These data are in agreement with the findings of previous studies showing that ermC gene was common among MSSA as well as coagulase-negative staphylococci.21,22
Characteristics of Macrolides, Lincosamides, and Streptogramin B-Resistant MSSA Isolates
CC, clonal complexes; ST, sequence types.
Moreover, 147 of the MRSA isolates expressed the cMLSB phenotype, whereas 44 MRSA expressed the iMLSB phenotype (Table 3). The three remaining MRSA expressed the MSB phenotype and carried the msrA gene. The ermA gene was identified in the majority of MRSA (n = 153), of which 133 isolates expressed the cMLSB phenotype. Thirty-seven MRSA isolates, expressing either the cMLSB (n = 13) or the iMLSB (n = 24) phenotypes carried the ermC gene. One MRSA isolate was positive for two genes, ermA and ermC, encoding erythromycin resistance. The later genotype has been rarely reported.13,21
Characteristics of Macrolides, Lincosamides, and Streptogramin B-Resistant Methicillin-Resistant Staphylococcus aureus Isolates
Population structure of erm-positive S. aureus isolates
The population structure of the 275 MLSB-resistant S. aureus isolates is shown in Tables 2 and 3. The majority of MLSB-resistant MRSA and MSSA belonged to the CC5. The group of 81 MSSA isolates was distributed to 13 STs, with STs 121 (n = 17), 145 (n = 10) and 282 (n = 10) accounting for 37 of the isolates. Most of ST282 isolates were associated with ermA, whereas most of the isolates belonging to ST121 carried the ermC gene. All ST145 MSSA isolates included the msrA gene. On the other hand, less genetic diversity was found in the group of MRSA isolates. Twelve different STs were identified among the group of the 194 MRSA isolates. ST225, which belongs to CC5, was the most prevalent, accounting for 137 MRSA isolates. The majority of ST225 MRSA isolates were associated with the presence of ermA. MRSA attributed to ST225 are predominant among isolates recovered from hospitals in European countries.23,24 Thirty-one out of 37 ermC-carrying MRSA isolates were distributed in STs 1 (n = 11), 80 (n = 10), 239 (n = 6), and 728 (n = 4).
erm-associated mobile elements
To define the regions flanking the erm resistance genes, the genome of staphylococcal isolate representatives of different erm genes, sequencing types (STs), and origin was completely sequenced.
De novo assembly and analysis of sequencing data showed that, in five out of six sequenced ermA-positive S. aureus isolates (ST5 S. aureus Sau-2933Lar, ST7 S. aureus Sau-3805Lar, ST30 S. aureus Sau-3298Lar, ST225 S. aureus Sau-3323Lar, and ST225 S. aureus Sau-3535Lar), ermA was part of the transposon Tn554, 25 integrated into their chromosomes. The insertion of Tn554 at different sites was observed. However, in the ST398 S. aureus isolate Sau-3221Lar, the ermA gene was carried by the transposon Tn6133. Tn6133 was composed of Tn554 with an integrated 4,787-bp DNA sequence encoding VgaE conferring streptogramin A, pleuromutilin, and lincosamide resistance. 26 Tn6133 has been previously detected in several ST398 MRSA isolates. 26 Screening by PCR and sequencing identified that all ermA-positive S. aureus, except isolate Sau-3221Lar, harbored the ermA-carrying Tn554 transposon.
Furthermore, WGS data showed that both ermC-positive S. aureus isolates (ST1 S. aureus Sau-2716Lar and ST121 S. aureus Sau-3893Lar), sequenced during this study, included small plasmids (pSau-2716Lar and pSau-3893Lar) of ∼3 kb (Fig. 1). Both small plasmids comprised the repL and ermC genes, and the insertion sequence IS257. Previous studies have reported the presence of small (2.3–2.5 kb) ermC-carrying plasmids, like pNE131 27 in staphylococcal isolates. 28 Screening by PCR and sequencing confirmed that all ermC-positive S. aureus harbored repL-associated plasmids, which carried the ermC gene.

Linear maps of mobile elements carrying erm
Further analysis of WGS data
Analysis of WGS data by ResFinder 3.0 tool (https://cge.cbs.dtu.dk/services/ResFinder/) 29 revealed that the majority of the sequenced isolates included additional genes for resistance to aminoglycosides, β-lactams, fluoroquinolones, fusidic acid, and tetracyclines (Table 4). The presence of mecA, blaZ, aph(3′)-III, aadD, aac(6′)-aph(2″), dfrG, tetK, tetM, fusB, and cat genes was confirmed by the antimicrobial resistant phenotypes of the isolates harboring those genes; only the presence of norA and fosD genes was not associated with the phenotypes. However, susceptibility to spectinomycin was not tested to clarify the expression of ant(9)-Ia gene in the respective isolates. S. aureus isolates that were assigned to ST225 harbored the highest number of resistance genes, explaining the MDR phenotypes of isolates belonging to this clone.
Resistance Genes and Virulence Factors of Eight Macrolides, Lincosamides, and Streptogramin B-Resistant Staphylococcus aureus Isolates Sequenced by Illumina Platform
Additionally, analysis of WGS data by VirulenceFinder 1.5 tool 30 showed that all sequenced were positive for the presence of several virulence genes (Table 4), which could be implicated in colonization and later infection of the host. 31 Diverse profiles of virulence factors were observed in isolates belonging to different clones. ST225 S. aureus isolates that were the most prevalent in our setting carried genes encoding Gamma-hemolysin (HlgA, HlgB, HlgC), Hlb hemolysin, which permeabilize cell membranes and lyse macrophages and lymphocytes and alter platelet morphology 31 and leukocidins (LukD and LukE) forming pores in leukocyte membranes that results in cell death. 32 Also, ST225 isolates encoded the staphylokinase, enterotoxins (SED, SER, SEG, SEJ, SEI, SEM, SEN, SEO, SEU), exhibiting superantigen activity that deregulates the immune system of the host, the staphylococcal inhibitor of complement, and metalloprotease aureolysin (Aur). Previous studies have shown that different combinations of virulence factors are carried by different phages. 33
Discussion
The present study reported the increasing rate of MLSB-resistant S. aureus in a tertiary care Greek hospital (UHL) during 2012–2017. During the period 1999–2001, 11.7% of S. aureus isolates were resistant to erythromycin, 34 whereas the rate of MLSB-resistant S. aureus increased from 35.6% to 45.0% during the period 2012–2017. This finding is in agreement with the data from a previous study reporting that 44.1% of S. aureus strains collected in a tertiary Athenian hospital (Greece) during 2003–2008, were resistant to MLSB antibiotics. 35 However, international studies focused on the prevalence of Ery-R staphylococci are limited. Otsuka et al. reported that 97% of MRSA and 34.6% of MSSA were resistant to one or more MLSB agents in a study conducted between 2001 and 2006. 36 Cetin et al. in a large collection of staphylococci in a Turkish hospital have found that 38.5% were resistant to MLSB antibiotics, whereas Uzun et al. reported that during 2011–2012, 79% of staphylococcal isolates were found to be Ery-R in a tertiary hospital in Ismir, Turkey.37,38
In our setting, the predominant phenotype was the cMLSB phenotype followed by the iMLSB and MSB phenotypes. In addition, molecular characterization of 275 representative MLSB-resistant S. aureus revealed that 65% were positive for the presence of ermA, 27.2% were positive for ermC, 7.2% were positive for msrA, whereas one strain carried both ermA and ermC. These results, which are in discordance with previous data from our group demonstrating the predominance of ermC among S. aureus, including MRSA, 34 might suggest a change in the epidemiology of MLSB-resistant S. aureus. Clonal relationships studied by MLST revealed that this increase was mainly due to the clonal spread of ST225 MRSA that belongs to the CC5, carrying ermA. Further analysis showed that ST225 strains included additional resistance genes, which can be implicated in the development of extensively MDR bacteria limiting therapeutic options, and several virulence factors (Table 4) that have been previously associated with increased pathogenicity and mortality of the bacterium.32,33 In addition, the ermA gene was also found in MSSA clones, which belonged to the CCs 30 and 5 (Tables 2 and 3). Almost all ermA-positive isolates, except the ST398 MRSA isolate Sau-3221Lar, harbored the transposon Tn554. On the other hand, ermC was identified in a polyclonal population of MSSA and MRSA, and was carried by small plasmids, such as pNE131. 27
In conclusion, over the past decade, an excessive and inappropriate use of antibiotics for human and animal treatment, as well as animal feed supplements for growth promotion has led to an increase in staphylococci acquiring crossresistance to MLSB antibiotics. Thus, characterization of MLSB-resistant S. aureus from Central Greece highlighted the important role of high-risk clones such as ST225, as well as intraspecies dissemination of mobile elements carrying resistance genes, in the successful spread of clinically important infectious agents, challenging infection control.
Footnotes
Acknowledgments
The authors are very grateful to I. Spiliopoulou for helpful suggestions. This work was supported by funding from the Research Committee of the University of Thessaly. It was also financed partially by the National Sustainability Program I (NPU I; grant no: LO1503) provided by the Ministry of Education Youth and Sports of the Czech Republic.
Disclosure Statement
No competing financial interests exist.
References
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
