Plasmid-Encoded Transferable mecB-Mediated Methicillin Resistance in Staphylococcus aureus

During cefoxitin-based nasal screening, phenotypically categorized methicillin-resistant Staphylococcus aureus (MRSA) was isolated and tested negative for the presence of the mecA and mecC genes as well as for the SCCmec-orfX junction region. The isolate was found to carry a mecB gene previously described for Macrococcus caseolyticus but not for staphylococcal species. The gene is flanked by β-lactam regulatory genes similar to mecR, mecI, and blaZ and is part of an 84.6-kb multidrug-resistance plasmid that harbors genes encoding additional resistances to aminoglycosides (aacA-aphD, aphA, and aadK) as well as macrolides (ermB) and tetracyclines (tetS). This further plasmidborne β-lactam resistance mechanism harbors the putative risk of acceleration or reacceleration of MRSA spread, resulting in broad ineffectiveness of β-lactams as a main therapeutic application against staphylococcal infections.

During cefoxitin-based nasal screening, phenotypically categorized methicillin-resistant Staphylococcus aureus (MRSA) was isolated and tested negative for the presence of the mecA and mecC genes as well as for the SCCmec-orfX junction region. The isolate was found to carry a mecB gene previously described for Macrococcus caseolyticus but not for staphylococcal species. The gene is flanked by β-lactam regulatory genes similar to mecR, mecI, and blaZ and is part of an 84.6-kb multidrug-resistance plasmid that harbors genes encoding additional resistances to aminoglycosides (aacA-aphD, aphA, and aadK) as well as macrolides (ermB) and tetracyclines (tetS). This further plasmidborne β-lactam resistance mechanism harbors the putative risk of acceleration or reacceleration of MRSA spread, resulting in broad ineffectiveness of β-lactams as a main therapeutic application against staphylococcal infections.
S taphylococcal cassette chromosome mec (SCCmec)mediated β-lactam resistance resulting from production of an additional penicillin-binding protein (PBP) 2a drastically limits the treatment options in cases of hospital-and community-related infections by staphylococci, leading to increased illness, death, and socioeconomic costs (1,2). Besides methicillin-resistant coagulase-negative staphylococci, notorious for foreign body-associated infections, methicillin-resistant Staphylococcus aureus (MRSA) strains are a global public health priority, despite some countries in Europe reporting stabilizing or decreasing MRSA rates (3)(4)(5). Since the initial reports of MRSA in 1961, several epidemic waves have resulted in threats of healthcare-, community-, and livestock-associated MRSA (6)(7)(8)(9).
For staphylococci, 2 PBP 2a-encoding genes, mecA and mecC, including several allotypes, have been described as chromosomally located genetic bases for phenotypic methicillin resistance (10)(11)(12)(13)(14). In contrast, mecB, originally described as mecA m , was reported as part of a probable primordial form of a methicillin resistance gene complex often found in a transposon mec complex (Tn6045) in Macrococcus caseolyticus, a colonizer of animal skin (15,16). Just recently, a mecD gene, most closely related to mecB, has been detected in bovine and canine M. caseolyticus isolates (17).
The impact of plasmidborne resistance for staphylococci is abundantly demonstrated for β-lactamase-mediated penicillin resistance. Resistance rates are >60% in human S. aureus isolates from the general population and >90% from hospital-related cases, regardless of the clinical background (18,19). In contrast to frequent interstrain and interspecies transmission of resistance plasmids by conjugation or transduction, only a relatively low rate of spontaneous horizontal transfer of SCCmec elements is assumed, resulting in still-manageable and controllable MRSA rates if prevention measures are adequate (20)(21)(22)(23)(24). However, transferable methicillin resistance might bear the consequence of an almost complete loss of β-lactam drugs as the most efficient class of antibacterial drugs for treatment of staphylococcal infections. Here, we report both a plasmid-encoded, and thereby transferable, methicillin resistance encoded by mecB and the occurrence of this gene in an isolate of the genus Staphylococcus.

Genetic Analysis
We extracted genomic DNA from S. aureus isolate UKM4229 using the QIAamp DNA Mini Kit (QIAGEN, Hilden, Germany) according to the manufacturer's instructions. We isolated plasmid DNA with the PrepEase Mini-Spin Plasmid Kit (Affymetrix USB, Santa Clara, CA, USA) following the protocol standards. For both plasmid and genomic DNA, we applied lysostaphin (20 µg/mL) (Wakchemie, Steinbach, Germany) for bacterial cell lysis. We performed multilocus sequence typing and spa gene typing initially as described elsewhere (25,26) and confirmed our results later by analysis of whole-genome sequencing (WGS) and DNA microarray data (discussed later in this article). We analyzed DNA sequences using RidomStaph-Type and SeqSphere+ (Ridom GmbH, Münster, Germany). Applying DNA microarray analysis (IdentiBAC Microarray; Alere Technologies GmbH, Jena, Germany), we identified resistance and virulence determinants and checked genotyping results.

Molecular Confirmation of Methicillin Resistance
Using PCR, we tested for the presence of methicillin resistance genes mecA and mecC (27,28) as well as mecB. DNA sequences of PCR oligonucleotides are given in Table 1.
Oligonucleotides for mecB were made on basis of the plasmid pMCCL2 of M. caseolyticus (GenBank accession no. NC_011996.1). We performed PCR reactions using the following protocol for mecA: 5 min at 95°C; 40 cycles of 0.5 min at 95°C, 0.5 min at 55.5°C, and 0.75 min at 72°C; and final elongation of 7 min at 72°C. The protocol for mecB was 5 min at 95°C; 35 cycles of 0.5 min at 95°C, 0.5 min at 57°C, and 2.5 min at 72°C; and final elongation of 7 min at 72°C. The protocol for mecC: 5 min at 95°C; 40 cycles of 0.5 min at 95°C, 0.5 min at 59.3°C, and 2 min at 72°C; and final elongation of 7 min at 72°C.

Antibiotic Drug Susceptibility Testing
We determined the MIC of cefoxitin for S. aureus isolate UKM4229 by the reference broth microdilution method according to the International Organization for Standardization (ISO) 20776-1 guideline (https://www.iso.org/ standard/41630.html), as required by the European Committee on Antimicrobial Susceptibility Testing (EUCAST) and the Clinical and Laboratory Standards Institute (CLSI). Cefoxitin (Sigma Aldrich, Taufkirchen, Germany) was tested in 2-fold concentrations (0.25-128 µg/mL). We subcultured the isolate and incubated it overnight before testing.
We investigated the susceptibility profile of UKM4229 by determining MICs of various β-lactam and non-βlactam antibiotic drugs (Table 2) using the gradient diffusion method (Etest; bioMérieux) according to the manufacturer's instructions. As recommended, the inoculated plates were incubated at 35°C for 18 ± 2 hours. In addition, we tested oxacillin using conditions for increased expression of methicillin resistance, as reported for mecA isolates (30): Mueller-Hinton agar supplemented with 2% saline, incubation at 30°C, and prolonged incubation up to 48 h. We investigated the applicability of a commercial automated susceptibility testing device to recognize methicillin resistance due to presence of mecB in S. aureus by using the VITEK 2 system. We evaluated the in vitro activity of the endolysin HY-133 against UKM4229 using the broth microdilution method in accordance with ISO 20776-1 guidance (https://www.iso.org/standard/41630.html), as described elsewhere (31,32). In brief, we tested 2-fold final concentrations of HY-133 ranging from 0.06 µg/mL to 8 µg/mL using 1-5 × 10 5 CFU/mL suspension of UKM4229 in cation-adjusted Mueller-Hinton broth. The MICs were read after incubation at 35°C for 18 ± 2 h.
We performed all experiments in triplicate on different days and calculated the median MIC values. We used S. aureus ATCC 29213 as a quality control strain on every testing day. For the antibiotic drugs we used, the MICs for the quality control strain were within acceptable limits throughout the testing.

Whole-Genome Sequencing
For the PacBio RS II platform (Pacific Biosciences, Menlo Park, CA, USA), we extracted staphylococcal DNA using the Genomic-tip 20/G Kit (QIAGEN) according to the manufacturer's instructions, except that we  from other mecB database entries, the highest nucleotide identity was shared with the sequence of mecD (68.7%), whereas the reported allotypes of mecC and mecA were more distantly related (online Technical Appendix Figure  1). WGS revealed that the UKM4229 genome consists of a 2,851,374-bp circular chromosome and 2 different plasmids, a 20,725-bp plasmid (pSAWWU4229_2) and an 84,599-bp plasmid (pSAWWU4229_1; Figure); the latter carried mecB (GenBank accession no. PRJEB19527  Figure 2).
Emerging Infectious Diseases • www.cdc.gov/eid • Vol. 24 DNA microarray analysis and WGS revealed the isolate possessed the leucotoxin genes lukF, lukS, lukD, lukE, lukX, and lukY. The isolate belonged to capsule type 8, and the biofilm-associated genes icaA, C, and D were detected. Furthermore, the hlb-converting bacteriophage of immuneevasion cluster type G comprising the enterotoxin encoding genes sep, sak, and scn was present in the genome. Additional information about the virulence profile of this isolate is given in online Technical Appendix Table 2.

Discussion
Recent studies have shown that mobile SCCmec elements have been imported more frequently by different S. aureus clonal lineages than previously assessed (35). Nevertheless, in contrast to the huge diversity of non-MRSA S. aureus clonal lineages (36), comparatively few clonal lineages still dominate the global MRSA population (37). However, an increased transferability of methicillin resistance by a plasmid-encoded course of action would have the capacity to drastically change the MRSA epidemiology. In staphylococci and other members of the phylum Firmicutes, plasmids have contributed enormously to the emergence and spread of antimicrobial resistance, and plasmid-encoded penicillin resistance has reached or exceeded 80% of clinical staphylococcal isolates (38).
In M. caseolyticus, mecB genes have been found within the chromosome as part of an SCCmec element as well as on a plasmid (15,16,39). For S. aureus UKM4229, it was shown that the mecB carrying plasmid pSAWWU4229_1 was distantly related to a macrococcal plasmid (pMCLL2 of M. caseolyticus JCSC5402), substantiating a possible gene transfer between the two genera. Because macrococcal and staphylococcal species may share the same hosts, mammalian skin and food, an exchange of mobile genetic elements between members of both closely related genera is likely and transmission to mammal-adapted staphylococci is generally to be feared (3). Genotyping of S. aureus UKM4229 revealed spatype t091, which is relatively common, as 0.92% of the >370,000 submitted spa sequences assigned to ≈17,000 spa types (as of February 2017) of the RIDOM SpaServer database (http://spa.ridom.de/spatypes.shtml) belong to this spa type.
Routine phenotypic methods for susceptibility testing cannot distinguish between methicillin resistance determinants; thus, mecB-encoded methicillin resistance can remain undiscovered. Moreover, mecB detection is not part of molecular screening approaches. Certain clonal lineages of S. aureus, including MRSA, have emerged as zoonotic pathogens colonizing farm and wild animals (40). Tetracycline resistance frequently observed in staphylococci associated with husbandry is another indication for a possible livestock origin of the isolate (41). A putative livestock source of the mecBencoding plasmid underlines the importance of the One Health concept in combating the spread of antimicrobial drug resistance.
Although the mecB isolate has been tested susceptible toward several agents of non-β-lactam antibiotic drug classes, the generally increased risk, compared to that of a SCCmec transfer, should be taken into consideration in that a mecBencoding plasmid will be transmitted through horizontal gene transfer to other staphylococcal strains, even to already multidrug-resistant strains. In S. aureus, 2 major means of horizontal gene transfer for plasmids have been described: conjugation and bacteriophage transduction. Here, pSAW-WU4229_1 did not harbor the typical genes responsible for conjugation or mobilization, which is, however, a common lack in S. aureus, affecting ≈95% of plasmids (42). In contrast, for most staphylococcal plasmids, a transfer through bacteriophage generalized transduction has been suggested (43,44). Further studies are warranted to underpin this putative threat and to investigate how a plasmidborne methicillin resistance would affect the SCCmec-based methicillin resistance. For UKM4229, the WGS data revealed that the SCCmec chromosomal attachment site (attB) locus and the neighboring orfX (rlmH) gene were intact, and no integration of an SCCmec element was found.
The mecB isolate was tested to be susceptible to ceftobiprole and ceftaroline. Although cephalosporins with anti-MRSA activity are still active against the majority of MRSA isolates, nonsusceptibility has been already associated with certain MRSA lineages ranging between 3.9% and 33.5% of all MRSA isolates (45)(46)(47)(48).
The discovery of plasmid-encoded methicillin resistance in S. aureus of probably macrococcal origin in a healthcare setting reveals a novel level of risk of the transfer of broad β-lactam resistance in staphylococci. Further studies are needed to clarify the real prevalence of mecB-caused methicillin resistance among MRSA and methicillin-resistant coagulase-negative staphylococci in human and animal populations, whether mecA and mecC genes could be found integrated on plasmids, and how the answers to these questions may affect human and animal health.
supervising the routine diagnostics leading to the isolation and presumptive identification of the isolate, as well as Melanie Bach and Martina Schulte for excellent technical assistance.
This work was supported in part by the BMBF within the frameworks of the Infect Control 2020 consortium [03ZZ0802H to K.B. and G.P.], #1Health-PREVENT (01KI1727A to K.B. and A.M.) and the German Center for Infection Research TTU 08.807 (8037808809 to K.B. and G.P.); and by the European Regional Development Fund within the EurHealth-1Health project (EU/INTERREG VA-681377 to K.B.).

About the Author
Dr. Becker is a professor of medical microbiology at the Institute of Medical Microbiology, University Hospital Münster, Münster, Germany. His research is focused on the epidemiology, pathogenesis, diagnosis, prevention, and therapy of staphylococcal infections. In particular, he has done extensive research on the characterization of MRSA and the staphylococcal small colony-variant phenotype.