Sir, In a previous study, we identified a methicillin-resistant Staphylococcus epidermidis (MRSE) clone being present in nine different samples obtained from two cats hospitalized in a small animal clinic, an employee and the clinic environment. Members of this MRSE clone exhibited SCCmec type IV, dru type dt10a, multilocus sequence type ST5 and very similar or indistinguishable SmaI macrorestriction patterns. These MRSE isolates were resistant to aminoglycosides/ aminocyclitols, b-lactams, fluoroquinolones, lincosamides, macrolides, chloramphenicol/florfenicol, pleuromutilins, sulphonamides, tetracyclines and trimethoprim, and shared the same plasmid profile including one plasmid of 28 kb in addition to several smaller plasmids. The aim of this study was to investigate this 28 kbplasmidfor itsstructureandthepresenceofresistancegenes. Plasmids from all nine isolates were transferred into Staphylococcus aureus RN4220 by protoplast transformation with selection on regeneration plates supplemented with tiamulin (10 mg/L), trimethoprim (30 mg/L), tetracycline (20 mg/L), erythromycin (10 mg/L) or florfenicol (20 mg/L). The resistance phenotype of the transformants was determined by broth microand macrodilution according to documents VET01-A4 and VET01-S2 (formerly M31-A3) of the CLSI. The antimicrobial agents tested are listed in Table S1 (available as Supplementary data at JAC Online). Plasmid DNA from one representative transformant, namely that from the feline MRSE isolate 47, was cut separately with HindIII and BglII. The resulting fragments were cloned into pBluescript II SK+ and sequenced starting with the M13 universal and reverse primers. Sequence determination was completed by primer walking on both strands. Sequence analysis was performed with the basic local alignment tools BLASTN and BLASTP (http://www.ncbi. nlm.nih.gov/BLAST), the ORF Finder program (http://www.ncbi. nlm.nih.gov/gorf/gorf.html) and the IS finder program (https:// www-is.biotoul.fr//) (all URLs last accessed in July 2013). From each of the nine isolates, the 28 kb plasmid was successfully transferred into S. aureus RN4220 after selection on plates supplemented with tetracycline, tiamulin or trimethoprim. All the transformed plasmids showed indistinguishable restriction patterns with BglII, EcoRI and HindIII. All nine transformants exhibited increased MICs compared with the empty recipient S. aureus RN4220 (Table S1) for clindamycin, tobramycin, neomycin, minocycline, tetracycline, tiamulin, trimethoprim/sulfamethoxazole and trimethoprim. Sequence analysis of one representative plasmid, designated pSWS47, showed that it had a total size of 28743 bp and revealed 27 open reading frames, 9 of which were truncated (Figure 1). The mosaic structure of plasmid pSWS47 strongly suggested that this plasmid has been developed from parts of plasmids and transposons previously described in staphylococci, but also other Gram-positive bacteria, including pUR4128 from a methicillinresistant S. aureus (MRSA), pSERP from human MRSE, pERGB from human MRSA, pPECL-3 and pPECL-4 from Pediococcus claussenii, pMD5057 from Lactobacillus plantarum and pSAP016A from S. epidermidis (Figure 1). Similarity to segments of the aforementioned plasmids was often terminated by insertion sequences, such as IS431, IS257 or IS1310. This might point towards the ability of insertion sequences to mediate the integration of small plasmids into larger plasmids. During such interplasmid recombination processes, insertion sequences may also be truncated, as seen in the present case. Moreover, closer inspection of the 8 bp sequences usually found immediately upand downstream of complete IS431 and IS257 revealed that none of these 8 bp sequences represented direct repeats (Figure 1). This observation suggested that multiple recombination processes, in which these insertion sequences had been involved, might have taken place. This scenario may also be supported by the detection of two segments of the same repU gene disrupted by IS257 and another three intact plasmid replication genes, a plasmid recombination gene, a relaxase gene and two plasmid mobilization genes, originating at least in part from different ancestral plasmids (Figure 1). Plasmid pSWS47 carried five functionally active resistance genes, accounting for the resistance phenotype observed: vga(A) (clindamycin and tiamulin resistance), dfrK (trimethoprim resistance), aadD (neomycin and tobramycin resistance), tet(L) and tet(M) (tetracycline and tetracycline/minocycline resistance) (Figure 1). When comparing plasmid pSWS47 with other genetic elements carrying resistance genes, three interesting observations were made. First, plasmid pSWS47 also harboured an inactive blaZ b-lactamase gene. Sequence analysis revealed an additional adenine residue at position 21227, which resulted in a frameshift mutationintheblaZgene.Second,thestartcodonforthereadingframeof the 20 amino acid peptide in the tet(L) translational attenuator was changed from ATG to AGG in pSWS47. However, MIC testing (32 mg/ L) of the4079 bpHindIIIclonethatcarried the tet(L)geneand its regulatory region confirmed that the tet(L) gene is functionally active despite this mutation. This observation is in agreement with thedetection of functionally active tet(L) genes in Pasteurella multocida and Mannheimia spp. that lacked the entire translational attenuator. Third, a segment that corresponded to the plasmid pMD5057 from
Read full abstract