IncX4 Plasmid Carrying the New mcr-1.9 Gene Variant in a CTX-M-8-Producing Escherichia coli Isolate Recovered From Swine

We studied a commensal colistin-resistant Escherichia coli isolated from a swine cecum sample collected at a slaughter, in Portugal. Antimicrobial susceptibility phenotype of E. coli LV23529 showed resistance to colistin at a minimum inhibitory concentration of 4 mg/L. Whole genome of E. coli LV23529 was sequenced using a MiSeq system and the assembled contigs were analyzed for the presence of antibiotic resistance and plasmid replicon types using bioinformatics tools. We report a novel mcr-1 gene variant (mcr-1.9), carried by an IncX4 plasmid, where one-point mutation at nucleotide T1238C leads to Val413Ala substitution. The mcr-1.9 genetic context was characterized by an IS26 element upstream of the mcr-pap2 element and by the absence of ISApl1. Bioinformatic analysis also revealed genes conferring resistance to β-lactams, sulphamethoxazole, trimethoprim, chloramphenicol and colistin, corresponding to the phenotype noticed. Moreover, we highlight the presence of mcr-1.9 plus blaCTX-M-8, a blaESBL gene rarely detected in Europe in isolates of animal origin; these two genes were located on different plasmids with 33,303 and 89,458 bp, respectively. MCR-1.9-harboring plasmid showed high identity to other X4-type mcr-1-harboring plasmids characterized worldwide, which strongly suggests that the presence of PMCR-encoding genes in food-producing animals, such as MCR-1.9, represent a potential threat to humans, as it is located in mobile genetic elements that have the potential to spread horizontally.


INTRODUCTION
Since the report of a plasmid-mediated colistin resistance (PMCR) mechanism, designated MCR-1, in Escherichia coli and Klebsiella pneumoniae isolated from animals, food and humans in China, further reports exposed the global dissemination of mcr-type gene in various bacterial species isolated from a wide range of different sources (Caniaux et al., 2017). In Portugal, PMCR has also been detected in a wide range of different sources and species, including humans, food-producing animals and meat, and in the environment (Campos et al., 2016;Figueiredo et al., 2016;Jones-Dias et al., 2016;Kieffer et al., 2017;Manageiro et al., 2017;Tacão et al., 2017;Mendes et al., 2018). Noteworthy, are the recent report of two cases presumably associated with the travel of patients from Portugal, one involving animals: a patient repatriated to France after hospitalization for 2 months in Portugal, in 2015 (Beyrouthy et al., 2017), and a New York state patient returning from Portugal in 2016 after staying on a farm with chickens and pigs (Gilrane et al., 2017).
More worrisome is the presence of mcr genes in Enterobacteriaceae carrying other resistance determinants namely, extended-spectrum β-lactamases (ESBL)-and/or carbapenemase-encoding genes. Since the first report of colocalization of mcr-1 and ESBL-in 2016 in bovines in France, an increase encoding genes in the proportion of mcr-1 genes among ESBL-producing E. coli in animals has been noticed, suggesting that the use of extended-spectrum cephalosporins may have simultaneously favored the spread of mcr-1 (Haenni et al., 2016). Here we describe the first detection of a novel mcr variant, hereafter-named mcr-1.9, identified in a commensal E. coli LV23529 isolated from a swine cecum sample collected at a slaughter, in Portugal.

Bacterial Isolate
Escherichia coli LV23529 was isolated in 2015 from a swine cecum sample collected at a Portuguese slaughter, during an evaluation study of commensal E. coli recovered from swine samples for antimicrobial susceptibility testing.

Antimicrobial Susceptibility Testing
Minimum inhibitory concentrations (MICs) were determined by microdilution method as previously described (Manageiro et al., 2017). In order to assess decreased susceptibility of the strain, interpretation of the results was done according to the epidemiological cut-off values recommended by the European Committee on Antimicrobial Susceptibility Testing (EUCAST 1 ).

Screening and Characterization of PMCR-and ESBL-Resistance Mechanisms
Molecular Detection of mcr-1 and bla ESBL -Encoding Genes Following phenotypic characteristics, PMCR-and ESBLresistance mechanisms were searched and identified by molecular methods, as previously described (Manageiro et al., 2017).

Transfer Experiments
Conjugation experiments were performed using sodium azideresistant E. coli J53 as a recipient strain. Transconjugants 1 http://mic.eucast.org/Eucast2/ were selected on McConkey agar supplemented with sodium azide (150 mg/L) and either cefotaxime (2 mg/L) or colistin (2 mg/L). Plasmid DNA was extracted from E. coli LV23529 using a NucleoBond Xtra Plus kit (Macherey-Nagel), and transformed into E. coli TOP10 OneShot chemically competent cells (Invitrogen), accordingly to manufacture's protocol. E. coli transformants were selected on MacConkey agar supplemented with 2 mg/L of colistin. PCR for bla CTX-M-8 or mcr-1-type and MICs of recipients and transformants were determined as mentioned above.
Genetic Context of mcr-1.9 Gene Colistin-resistant E. coli LV23529 was genotypically characterized by whole-genome sequencing (WGS), as previously described (Manageiro et al., 2017). Sequence reads were trimmed and filtered according to quality criteria, and de novo assembled into contigs by means of CLC Genomics Workbench 10.0 (Qiagen). The assembled contigs were analyzed and studied for the presence of antibiotic resistance, virulence genes and plasmid replicon types, serotype, multi-locus sequence type (ST) and fim-type, using bioinformatics tools 2 . The NCBI prokaryotic genome automatic annotation pipeline (PGAAP) was used for annotation.
Plasmid sequencing was also performed on a MiSeq Illumina platform using 150 bp paired-end reads, after plasmid DNA extraction from TLV23529 (mcr-1.9) using a NucleoBond Xtra Plus kit (Macherey-Nagel), and quantification using Qubit 1.0 Fluorometer (Invitrogen), as previously described (Manageiro et al., 2017). Sequence reads were trimmed and filtered according to quality criteria, and mapped against E. coli ATCC 25922 genome (NZ_CP009073). Unmapped reads (80.2%/total reads) were then used for plasmids structure construction by mapping assembly based on the genetic organization of the closest plasmid sequences obtained by BLASTn; this was followed by contig neighbor's prediction from assembly information using CLC Genomics Workbench 10.0 (Qiagen). NCBI Microbial genomes BLAST analysis tool 3 was used to search for plasmid sequences. Plasmid alignments and ORF representations were also done using EasyFig v. 2.2.3 (Sullivan et al., 2011).

Genomic Epidemiological Analysis
BacWGSTdb database was used for genotyping and source tracking bacterial pathogen (Ruan and Feng, 2016).

Nucleotide Sequence Accession Number
The pLV23529-MCR-1.9 and pLV23529-CTX-M-8 nucleotide sequences from this study were submitted to the NCBI GenBank Database with accession numbers KY964067 and KY964068, respectively. The new mcr-1.9 nucleotide sequence was submitted with accession number KY780959.
This Whole Genome Shotgun project has been deposited at DDBJ/ENA/GenBank under the accession SBIH00000000. The version described in this paper is version SBIH01000000.
Molecular characterization of the E. coli LV23529 isolate allowed the detection of bla CTX-M-8 and mcr-1-type genes.
The WGS assembly of E. coli LV23529 yielded 193 contigs (average 143.7-fold coverage), which together comprised 5,122,415bp, showing a GC content of 50.7%. The largest contig was 320,931 bp long; the N50 statistic, which stands for the minimum contig length of at least 50% of the contigs, was 113,197 bp. The average length of the obtained contigs was 26,541 bp. Overall, the genome sequence comprised 5,124 putative genes, among which 5,037 consisted of protein encoding sequences.
The mcr-1.9 genetic context was characterized by an IS26 element upstream of the mcr-pap2 element and by the absence of ISApl1 (Figure 1), which is in accordance with other studies about mcr-1 gene (Veldman et al., 2016;Sun et al., 2017). The mcr-1.9 gene can be mobilized within an ISApl1-flanked composite transposon (Tn6330), although many sequences have been identified without ISApl1 or with just a single copy (Snesrud et al., 2018). Indeed, it has been described that initially ISApl1 was presumably involved in the transposition of the mcr-1 cassette and then was lost, contributing for the stability of mcr gene on IncX4 plasmids (Sun et al., 2017;Snesrud et al., 2018).
The PMCR-encoding gene was found in an IncX4 plasmid (pLV23529-MCR-1.9), showing highest identities (>99.9%) with six IncX4-type mcr-1-harboring plasmids identified worldwide, in unrelated E. coli isolates, mainly collected from human patients (Figure 2 and Table 2). Indeed, all belonged to different MLST, which might suggest a resistance plasmid dissemination across strains (plasmid outbreak) rather than clonal transmission of MCR-1-type-producing strains. Furthermore, no E. coli LV23529 closely related isolates were detected among those currently deposited in the public database BacWGSTdb (Ruan and Feng, 2016), which reinforce the importance of the horizontal gene transfer in this study.
Like pLV23529-MCR-1.9, the six plasmids ( Table 2) doesn't have the ISApl1 element. Hence, similarities may suggest that the one-point mutation (T1238C) in mcr-1.9 occurred on the X4 plasmid, since mobilization of mcr-1 occurs as part of a composite transposon (Tn6330) and that structures lacking the downstream ISApl1 are not capable of mobilization (Snesrud et al., 2018). The IS26 upstream of the mcr-pap2 element is flanked by an 8bp direct repeat (Figure 3A), indicating that its insertion wouldn't seems to be related to the mcr-1.9 context, justifying the differences found with other IncX4 mcr-1-harboring plasmids. IncX4 plasmid has been widely implicated in the spread of MCR-1 gene in Europe (Caniaux et al., 2017). In Portugal, this plasmid type is circulating among diverse hosts (humans, pigs, poultry), being responsible for hospital-based outbreak caused by MCR-1 plus KPC-3-producing K. pneumoniae (Mendes et al., 2018), as well as for the diffusion of this PMCR at the farm level (Kieffer et al., 2017). Indeed, IncX4 plasmids seem to be efficiently transferred at different temperatures and different lackof-fitness burdens among bacterial hosts, which may facilitate the transfer of mcr-type among Enterobacteriaceae (Lo et al., 2014;Wu R. et al., 2018). The pLV23529-MCR-1.9 plasmid backbone contains all the core genes common to IncX plasmids involved in segregation, stability, replication, and conjugative transfer of the plasmid (Figure 3A), namely the IncX-type pilus synthesis operon (pilX1-pilX11). However, pLV23529-MCR-1.9 was mobilizable, but not self-transmissible. Of note, we found a one-point mutation (G64T), leading to Asp22Tyr substitution, in the PilX1, a peptidoglycan hydrolase involved in T-DNA plasmid transfer. This mutation might explain why the attempts to conjugate mcr-1.9 from E. coli LV23529 were unsuccessful (Chen et al., 2009).
Further plasmid analysis revealed the presence of two other plasmids: IncF [F2:A-:B-], IncR and the colicinogenic IncI1-ST113-carrying the bla CTX-M-8 (pLV23529-CTX-M-8, Figure 3B). Of note, the mcr-1.9-positive isolate, co-harboring bla CTX-M-8 and bla TEM-1 genes, is here reported for the first time in an E. coli isolate of animal origin. In fact, bla CTX-M-8 gene is rarely detected in Europe in isolates of animal origin (Börjesson et al., 2016), but in humans seems to be emerging (Eller et al., 2014). Indeed, a recent phylogenetic study suggested an increasing trend of co-existence and transmission of bla CTX-M and mcr-1 in both clinical medicine and veterinary medicine (Wu C. et al., 2018).
In conclusion, the presence of PMCR-encoding genes, such as MCR-1.9, in food-producing animals represents a potential threat to humans, as it is located in mobile genetic elements that have the potential to spread horizontally. As mentioned, in Portugal, PMCR is an emerging problem and its international spread is a worrying reality (Beyrouthy et al., 2017;Gilrane et al., 2017).

AUTHOR CONTRIBUTIONS
VM designed the study, performed the molecular experiments and bioinformatics analysis, interpreted the data, and wrote the manuscript. LC, RR, and EF performed the microbiological and molecular experiments. CS and LV performed the Illumina genome sequencing experiments. MC designed the study, wrote, reviewed and edited the manuscript. All authors read and approved the final manuscript.