Detection of the phenicol–oxazolidinone–tetracycline resistance gene poxtA in Enterococcus faecium and Enterococcus faecalis of food-producing animal origin in China

Sir, Oxazolidinones, including linezolid and tedizolid, are effective antimicrobial agents for treating infections caused by MDR Gram-positive bacteria, including VRE.1,2 Linezolid is the first commercially available oxazolidinone that can inhibit protein synthesis by binding to the peptidyltransferase centre of the bacterial 23S rRNA. After introduction of linezolid, the resistance mechanism that emerged rapidly was related to mutations in genes coding for the 23S rRNA. The transferable oxazolidinone resistance determinants, cfr and optrA, have been reported in enterococci in several regions worldwide.3–6 Very recently, another transferable oxazolidinone resistance gene, poxtA, was reported in an MRSA of clinical origin and a porcine Enterococcus faecium isolate in Italy.7,8,poxtA encodes a ribosomal protection protein of the ABC-F family showing 32% identity to OptrA, which can decrease susceptibility to phenicols (chloramphenicol and florfenicol), oxazolidinones and tetracyclines.7 Here we evaluate the prevalence of poxtA in E. faecium and Enterococcus faecalis of food-producing animal origin in China. A total of 208 E. faecium (122 from swine and 86 from chicken) and 137 E. faecalis (79 from swine and 58 from chicken) strains were isolated from 360 faecal swabs (swine, n = 210; chicken, n = 150), which were collected from 36 food-producing animal farms in 12 provinces of China in 2017 (Table S1, available as Supplementary data at JAC Online). All strains were screened for the presence of poxtA by PCR with primers poxtA-F (5′-GGTGGATTTACCGACACCGT-3′) and poxtA-R (5’- GACCAGTGGAAATGCCCGTA-3′). Positive PCR products (943 bp) were sequenced by Chengdu Sangon Biological Engineering Technology & Services Co., Ltd (Chengdu, China). The results showed that 18 (8.65%) E. faecium (13 from swine and 5 from chicken) and 1 (0.73%) E. faecalis of swine origin were poxtA positive (Table 1). Antimicrobial susceptibility testing for poxtA-positive strains was performed using the broth microdilution method according to the CLSI recommendations.9 All 19 poxtA-positive enterococci were found to have decreased susceptibility to linezolid (MICs 2–16 mg/L) and tedizolid (MICs 0.5–2 mg/L) (Table 1). All strains displayed MDR profiles and were resistant to chloramphenicol, florfenicol, erythromycin and doxycycline, but remained susceptible to ampicillin, fosfomycin and vancomycin. The 19 poxtA-positive enterococci strains were isolated from eight provinces, indicating that the poxtA gene was widespread in enterococci of food-producing animal origin in China. Very recently, Hao et al.10 reported that the poxtA gene was present in 57.9% (66/114) of the florfenicol-resistant enterococci isolated from two swine farms in Henan Province of China in 2018, which might result from the wide use of florfenicol and doxycycline in swine farms. poxtA-positive E. faecium and E. faecalis strains characterized in this study LZD, linezolid; TZD, tedizolid; CHL, chloramphenicol; FFC, florfenicol; DOX, doxycycline. poxtA-positive E. faecium and E. faecalis strains characterized in this study LZD, linezolid; TZD, tedizolid; CHL, chloramphenicol; FFC, florfenicol; DOX, doxycycline. Genomic DNAs of poxtA-positive strains were extracted by using a MiniBEST Bacteria Genomic DNA Extraction Kit (TaKaRa, Dalian, China). WGS was performed using the Illumina Hiseq platform (400 bp paired-end reads with 200-fold average coverage). The draft genomes were assembled using the software SPAdes_3.12.0. The acquired antimicrobial resistance genes were identified using ResFinder 3.1 (https://cge.cbs.dtu.dk/services/ResFinder/). Various resistance genes were identified in the 19 poxtA-positive enterococci strains (Table 1). All strains contained the tetracycline resistance genes tet(L) and tet(M). Eighteen strains harboured the phenicol resistance genes fexA and/or fexB. It is worth noting that five strains also had the concomitant presence of the transferable oxazolidinone resistance gene optrA. MLST STs were identified using MLST 2.0 (https://cge.cbs.dtu.dk/services/MLST/). Eighteen poxtA-positive E. faecium strains belonged to 14 different STs (Table 1), indicating the genetic diversity of the poxtA-containing E. faecium. poxtA-positive E. faecalis strain SC14-4 belonged to ST692. SNPs from genomes of the 18 poxtA-positive E. faecium strains were called and a phylogeny based on the concatenated alignment of the high-quality SNPs was inferred by using CSI Phylogeny 1.4 (https://cge.cbs.dtu.dk/services/CSIPhylogeny/) with parameters as defaults. Phylogenetic analysis showed that most strains were not clonally related (Figure S1). Nevertheless, the E. faecium strains SC3-1, SC3-2 and SC3-4 from the same chicken farm were clonally related and shared the same MLST profile, ST650 (Table 1 and Figure S1). A similar observation was made for the porcine strains HN7 from Henan and CQ9 from Chongqing that were clustered together by CSI Phylogeny and MLST (ST610) (Table 1 and Figure S1). An alignment of the deduced 19 PoxtA amino acid sequences revealed that 18 enterococci strains harboured WT PoxtA identical to that from Staphylococcus aureus AOUC-0915,7 whereas E. faecium HN11 harboured a PoxtA variant with a single substitution, R256H. The contigs that contained poxtA were analysed by using the BLAST programs (http://blast.ncbi.nlm.nih.gov/Blast.cgi). All strains harboured a conserved 2.6 kb contig that contained poxtA and partial IS1216E in the left and right ends. The conserved genetic context of poxtA flanked by IS1216E in diverse enterococci strains indicated that IS1216E was responsible for the horizontal transfer of poxtA.7 Conjugation or transformation experiments were performed using rifampicin-resistant E. faecalis JH2-2 as the recipient strain with selection on brain heart infusion agar plates containing 25 mg/L rifampicin and 8 mg/L florfenicol. The poxtA gene in 10 strains was successfully transferred into JH2-2, decreasing susceptibility to linezolid, tedizolid, chloramphenicol, florfenicol and doxycycline (Table S2). The poxtA-harbouring plasmids were sequenced using the Illumina Hiseq platform (200-fold average read depth) combined with the Nanopore PromethION platform (300-fold average read depth) (Biomarker, China). Eight different plasmids that varied from 30 to 70 kb were assembled using Canu v1.5, seven of which co-harboured poxtA, fexB, tet(L) and tet(M) (Table S2). The diversity of poxtA-harbouring plasmids might promote the dissemination of poxtA, as well as fexB, tet(L) and tet(M). The prevalence of the phenicol–oxazolidinone–tetracycline gene poxtA in diverse enterococci of food-producing animal origin is of concern for public health, given that poxtA-containing enterococci from food-producing animals might spread to humans through food consumption. The wide use of florfenicol and doxycycline in livestock could promote the spread of poxtA. The dissemination of poxtA in enterococci from humans and food-producing animals requires more attention. The poxtA gene sequences as well as those of eight poxtA-harbouring plasmids reported in this study have been deposited in GenBank and assigned accession numbers MK292893–MK292895, CP038162, CP038165, CP038167, CP038169, CP038171, CP038173, CP038175 and CP038176. This work was supported by the National Key Research and Development Program of China (grant no. 2018YFD0500300), the International Collaborative Program of Sichuan Province (grant no. 2018HH0027) and the Scientific Research Foundation of Sichuan University (grant no. 2017SCU12006). None to declare.