Occurrence and Characterization of mcr-1-Positive Escherichia coli Isolated From Food-Producing Animals in Poland, 2011-2016

Abstract

The emergence of plasmid-mediated colistin resistance (mcr genes) threatens the effectiveness of polymyxins, which are last-resort drugs to treat infections by multidrug- and carbapenem-resistant Gram-negative bacteria. Based on the occurrence of colistin resistance the aims of the study were to determine possible resistance mechanisms and then characterize the mcr-positive Escherichia coli. The research used material from the Polish national and EU harmonized antimicrobial resistance (AMR) monitoring programs. A total of 5,878 commensal E. coli from fecal samples of turkeys, chickens, pigs, and cattle collected in 2011-2016 were screened by minimum inhibitory concentration (MIC) determination for the presence of resistance to colistin (R) defined as R > 2 mg/L. Strains with MIC = 2 mg/L isolated in 2014-2016 were also included. A total of 128 isolates were obtained, and most (66.3%) had colistin MIC of 2 mg/L. PCR revealed mcr-1 in 80 (62.5%) isolates recovered from 61 turkeys, 11 broilers, 2 laying hens, 1 pig, and 1 bovine. No other mcr-type genes (including mcr-2 to -5) were detected. Whole-genome sequencing (WGS) of the mcr-1-positive isolates showed high diversity in the multi-locus sequence types (MLST) of E. coli, plasmid replicons, and AMR and virulence genes. Generally mcr-1.1 was detected on the same contig as the IncX4 (76.3%) and IncHI2 (6.3%) replicons. One isolate harbored mcr-1.1 on the chromosome. Various extended-spectrum beta-lactamase (bla _SHV-12, bla _CTX-M-1, bla _CTX-M-15, bla _TEM-30, bla _TEM-52, and bla _TEM-135) and quinolone resistance genes (qnrS1, qnrB19, and chromosomal gyrA, parC, and parE mutations) were present in the mcr-1.1-positive E. coli. A total of 49 sequence types (ST) were identified, ST354, ST359, ST48, and ST617 predominating. One isolate, identified as ST189, belonged to atypical enteropathogenic E. coli. Our findings show that mcr-1.1 has spread widely among production animals in Poland, particularly in turkeys and appears to be transferable mainly by IncX4 and IncHI2 plasmids spread across diverse E. coli lineages. Interestingly, most of these mcr-1-positive E. coli would remain undetected using phenotypic methods with the current epidemiological cut-off value (ECOFF). The appearance and spread of mcr-1 among various animals, but notably in turkeys, might be considered a food chain, and public health hazard.

Keywords: IncHI2; IncX4; WGS; aEPEC; colistin resistance; food animal; mcr-1.

FIGURE 1

Colistin MIC distribution and occurrence of the mcr-1 and mcr-2 gene among 128 Escherichia coli selected based on colistin MIC > 2 mg/L (2011–2013) and ≥2 mg/L (2014–2016). ^∗ only isolates with MIC_colistin ≥ 2 mg/L were tested for mcr-1 and mcr-2 genes. Different MIC values (>4 and 8, 16, and >16) are result of changed MIC panel in plates. EUMVS2 plate (2–4 mg/L) was used in 2011–2013, EUVSEC plate (1–16 mg/L) in 2014–2016.

FIGURE 2

Occurrence of isolates meeting the selection criteria (MIC > 2 mg/L for isolates in 2011–2013 and MIC ≥ 2 mg/L for isolates in 2014–2016) and mcr-1 positive commensal and ESBL/ampC producing E. coli from all tested sources (turkeys, chickens, pigs, and cattle), 2011–2016. The occurrence of colistin resistant and mcr-1 positive isolates when using exclusively EUCAST ECOFF was included (dashed line).

FIGURE 3

Occurrence of mcr-1-positive E. coli in Poland isolated from turkey and chicken fecal samples. ^∗ isolates collected in multiannual governmental programs (according to of the Council of Ministers Decisions), ^∗∗ isolates deriving from official monitoring according to Decision 2013/652/EC and thus not encompassing samples from turkeys and broilers in 2015; n1 and n2 indicate the total number of isolates tested for MIC determination from turkeys and chickens (both broiler and laying hens), respectively. The occurrence of mcr-1 positive strains when using exclusively EUCAST ECOFF during selection of isolates was also included (dashed line).

FIGURE 4

Geographical distribution (commune level) of turkey sampling. Single turkey caeca was collected at slaughter and farm of origin was retrieved for verification of randomization of samples used for isolation of indicator E. coli in 2014 and 2016.

FIGURE 5

Geographical distribution (commune level) of broiler sampling. Single broiler caeca was collected at slaughter and farm of origin was retrieved for verification of randomization of samples used for isolation of indicator E. coli in 2014 and 2016.

FIGURE 6

Geographical distribution (district level) of mcr-1 positive E. coli isolation sites.

FIGURE 7

Phylogeny of mcr-1-positive E. coli (sequence types, year and source of isolation, and map of phenotypic resistance and resistance genes). Full and empty square mean presence and absence of antimicrobial resistance (AMR) gene, respectively, whereas empty space means non-defined or not tested. (A) Resistance profiles (black): ampicillin, azithromycin, cefuroxime, ceftazidime, chloramphenicol, ciprofloxacin, gentamicin, colistin, nalidixic acid, meropenem, sulphonamides, tetracycline, tigecycline, and trimethoprim (Sensititre EUSVEC MIC panel). (B) Aminoglycoside resistance genes (green): aac(3)-IIa, aac(3)-IId, aadA1, aadA2, aadA5, aadA24, aadB, aph(3′)-Ia, aph(3′)-Ic, aph(3″)-Ib, and aph(6)-Id. (C) Beta-lactam resistance genes (blue): bla_TEM–1A, bla_TEM–1B, bla_TEM–1C, bla_TEM–30, bla_TEM–135, bla_TEM–52C, bla_CMY–2, bla_SHV–12, bla_OXA-1,bla_CARB-2, and bla_CTX–M–1. (D) Quinolone resistance genes: plasmid-mediated quinolone resistance (PMQR, brown): qnrS1, qnrB19; mutations in quinolone resistance determining regions (QRDR, violet): gyrA S83L, gyrA D87N, gyrA D87Y, parC S80I, parC S80R, parC S57T, parC E84G, parC E84K, parE L416F, and parE L460D. (E) Sulphonamide resistance genes (dark brown): sul1, sul2, and sul3. (F) Tetracycline resistance genes (light brown): tet(A), tet(B), and tet(M).

FIGURE 8

Phylogenetic relationship of contigs with mcr-1 and IncX4 plasmid replicon extracted from the genomes of 61 E. coli isolates by source and year of E. coli isolation. Complete sequence of IncX4 plasmid with accession number KU761327.1 from GenBank was used as the reference.