Detection in Italy of a porcine Enterococcus faecium isolate carrying the novel phenicol-oxazolidinone-tetracycline resistance gene poxtA

Sir, Acquired resistance to linezolid in staphylococci and enterococci emerged shortly after its introduction into clinical practice. Initially, resistance mechanisms mediated by mutations affecting the ribosomal target were reported.1 Subsequently, transferable linezolid resistance determinants have emerged, including the cfr gene, encoding a 23S rRNA methylase,2 and the optrA gene, encoding a ribosomal protection protein of the ABC-F family.3 Very recently we described poxtA, a novel resistance gene detected in a highly linezolid-resistant MRSA of clinical origin, which also carried the cfr gene and a mutation in the rplC L3 ribosomal protein gene.4,5,poxtA encodes a protein that is 32% identical to OptrA and belongs to the same family of ABC-F resistance determinants. Its expression in Staphylococcus aureus, Enterococcus faecalis and Escherichia coli was shown to cause a decrease in susceptibility to phenicols, oxazolidinones and tetracyclines. The poxtA gene is associated with a putative mobile element containing two IS1216-like elements, which could contribute to its dissemination. A BLAST search in GenBank using poxtA as a query revealed the presence of identical genes in a few Gram-positive strains (of the species E. faecalis, Enterococcus faecium and Pediococcus acidilactici), mostly of animal origin, suggesting that selection of poxtA could have occurred in the animal setting owing to extensive use of phenicols in veterinary medicine.5 We decided to carry out a search for poxtA in enterococci of animal origin exhibiting resistance to florfenicol. To this end, a pig farm where florfenicol was used previously was investigated. Three pooled faecal samples were collected from piglets at weaning as described previously.6 The samples were inoculated in buffered peptone water supplemented with florfenicol (10 mg/L) and incubated at 44°C for 48 h. Then, they were inoculated in Slanetz and Bartley agar plates supplemented with florfenicol (10 mg/L) and the enterococci grown on this medium were identified by API rapid ID32 Strep (bioMérieux) and tested by PCR for poxtA and six other florfenicol resistance genes [fexA, fexB, cfr, cfr(B), cfr(C) and optrA] (for primers see Table S1, available as Supplementary data at JAC Online). A poxtA-positive E. faecium isolate (called S156), reidentified by MALDI-TOF (Vitek-MS, bioMérieux), was found. It was also positive for fexA and fexB but did not carry the cfr or optrA genes or show mutations involving the 23S rRNA or ribosomal proteins. It was also positive for two tetracycline resistance genes, tet(M) and tet(L) (see Table S1 for primers). MIC values and resistance genes of E. faecium S156, the two poxtA-carrying transconjugants and the recipient strain FFC, florfenicol; LZD, linezolid; TZD, tedizolid; TET, tetracycline. MIC values and resistance genes of E. faecium S156, the two poxtA-carrying transconjugants and the recipient strain FFC, florfenicol; LZD, linezolid; TZD, tedizolid; TET, tetracycline. Susceptibility tests, performed using Etest strips (Liofilchem, Roseto degli Abruzzi, Italy) for tedizolid and by standard broth microdilution for the other antibiotics, were interpreted according to clinical breakpoints (EUCAST version 8.1, 2018; CLSI M100-S27, 2017). They revealed resistance to florfenicol (MIC >128 mg/L), linezolid (MIC 8 mg/L), tedizolid (MIC 4 mg/L) and tetracycline (MIC 128 mg/L) and susceptibility to erythromycin (MIC 2 mg/L), penicillin (MIC 2 mg/L) and vancomycin (MIC 1 mg/L). Sequencing experiments, after amplification of the gene with external primers (Table S1), demonstrated 100% DNA identity between this enterococcal poxtA and the reference sequence of the gene from S. aureus AOUC-0915 (GenBank accession number MF095097).5 The transferability of poxtA from E. faecium S156 was investigated by filter-mating experiments using florfenicol (10 mg/L) for selection.7 The gene was successfully transferred to the E. faecium 64/3 recipient8 (transfer frequency 3.5 × 10−4 per recipient). Two transconjugants were randomly selected. Florfenicol and linezolid MICs for the two transconjugants were identical to those for the donor. Conversely, a tetracycline MIC of 64 mg/L was recorded for one of the two transconjugants, which was found to have received no tet resistance genes, whereas a tetracycline MIC of 128 mg/L was recorded for the other, which had received both tet(M) and tet(L). Both transconjugants were also positive for the fexB gene. The above results are summarized in Table 1. In S1-PFGE9 and hybridization assays, a poxtA-specific probe hybridized with two plasmids [one of ∼97 kb and one tentatively estimated as around 30 kb (the smallest size marker being 48.8 kb)], both detected in the donor as well as in the tet(M)- and tet(L)-positive transconjugant; in contrast, the other transconjugant displayed a single hybridization signal corresponding to the plasmid of around 30 kb (Figure S1). At the time of discovery of poxtA in a human MRSA isolate, the gene was also identified in Gram-positive cocci of animal origin by a database search.5 The present finding confirms that it can be harboured by enterococci of animal origin. Moreover, poxtA was apparently able to confer linezolid resistance by itself, also in the absence of other known oxazolidinone resistance mechanisms. Florfenicol, extensively used in livestock to prevent or cure bacterial infections, could promote the spread of poxtA-mediated linezolid resistance; this spread is especially likely in view of the gene’s location on conjugative plasmids, reported here. A wider screening for the presence of poxtA in florfenicol-resistant animal isolates would be helpful in clarifying the actual prevalence of this gene and the relevant hazard to public health. We are grateful to Pietro E. Varaldo for discussion and critical reading of the manuscript. This study was supported by internal funding. None to declare.