Klebsiella pneumoniae Complex Harboring mcr-1, mcr-7, and mcr-8 Isolates from Slaughtered Pigs in Thailand

Dissemination of the mobile colistin resistance gene mcr in Enterobacterales among humans, animals, and the environment is a public health issue. We characterized mcr genes in the Klebsiella pneumoniae complex (KpnC) isolated from slaughtered pigs in Thailand. The 280 KpnCs consisted of K. pneumoniae (85%), Klebsiella quasipneumoniae (8.21%), and Klebsiella variicola (6.79%). mcr genes were detected in 6.79% (19/280) of KpnC isolates, consisting of mcr-8 (n = 9; 3.21%), mcr-7 (n = 7; 2.50%), mcr-7 + mcr-8 (n = 2; 0.71%), and mcr-1 + mcr-7 (n = 1; 0.36%). K. pneumoniae predominantly carried the mcr-7 and mcr-8 genes, while K. variicola and K. quasipneumoniae harbored mcr-7 and mcr-8, respectively. Six of the nineteen mcr-harboring KpnC isolates exhibited colistin resistance, and five had mcr-1 or mcr-8 transferable to an Escherichia coli recipient. Antimicrobial susceptibility analysis revealed that all mcr-carrying KpnC isolates were susceptible to carbapenems, cefotaxime, cefepime, amoxicillin/clavulanic acid, piperacillin/tazobactam, amikacin, and fosfomycin, and had high resistance to azithromycin. Multilocus sequence analysis demonstrated that the mcr-harboring KpnC isolates were genetically diverse. A ‘One-Health’ approach is useful to combat antimicrobial-resistant bacteria through coordinating the human, animal, and environmental sectors. Hence, continuous monitoring and surveillance of mcr-carrying KpnCs throughout the pork supply chain is crucial for ensuring public health.

The mcr genes have been reported worldwide in Enterobacterales (including E. coli, Salmonella, and K. pneumoniae) from various sources, especially the environment and in animals [18]. E. coli is the most prevalent species among the mcr-harboring isolates reported so far, accounting for approximately 91% of the total mcr-carrying isolates, followed by Salmonella enterica (~7%) and K. pneumoniae (~2%) [19]. The mcr gene has been detected in 47 different countries across six continents, including developed and developing countries. It was identified from human sources in 44 countries; livestock in 21 countries; meat and food products in 13 countries; and from other sources, including pets, exotic or wild animals, and the environment, in 11 countries [19].
Livestock, especially pigs, are considered a reservoir of colistin-resistant organisms because colistin is approved in pig production in several countries with different purposes, including for the control of Enterobacterales infections therapeutically, prophylactically, and even for growth promotion [20]. Pigs maintain the microorganism on the farm, contaminating their environment and cross-contaminating carcasses during slaughter, which poses a potential risk to humans through co-circulation in human and swine populations, while transmission or spillover into humans can occur via tasting or eating undercooked products or preparing meals at home [21]. The prevalence of mcr-1-harboring bacteria has been found to range from 0.35 to 36.00% in pigs; however, some studies reported very high positivity rates in pigs ranging from 13.20 to 98.00% [22]. Recent studies in Thailand showed the presence of several mcr genes (mcr-1, mcr-3, mcr-6, mcr-7, mcr-8, and mcr-9, together with co-occurrences of mcr-1 + mcr-3, mcr-1 + mcr-9, and mcr-3 + mcr-6 + mcr-7) in E. coli isolated from slaughtered pigs, as well as mcr-1 or mcr-3 being detected in either E. coli or K. pneumoniae isolated from humans [23,24]. Two mechanisms of mcr-1 transmission have been recognized: horizontal gene transfer (HGT) and clonal transmission [24]. The transferability of the mcr-carrying plasmid from E. coli, K. pneumoniae, Klebsiella aerogenes, Salmonella enterica, Enterobacter cloacae, or Cronobacter sakazakii isolates of either animal or human origin to humans by HGT was demonstrated by in vitro conjugation or transformation experiments, showing the successful transfer of an mcr-1 plasmid from animal or human origin into common human pathogenic Enterobacteriaceae and Pseudomonas aeruginosa [11,19,25]. Thus, mcr-1 in uncommon Enterobacteriaceae strains can be transferred into common Enterobacteriaceae, being further disseminated and circulated among environmental or human Enterobacteriaceae species. Clonal transmission has indicated that mcr-1-harboring isolates from food animals could be transferred to humans mainly through the food chain or direct contact, showing that the same E. coli STs carrying mcr-1 were detected in food animals (on farms and in slaughterhouses), animal products (in markets and supermarkets), and humans (healthy populations and patients). Core-genome, singlenucleotide polymorphism (SNP)-based phylogenetic analysis, and XbaI pulsed-field gel electrophoresis (PFGE) analysis further supported the commonality of mcr-1-harboring isolates among disparate samples [22].
We characterized KpnC harboring mcr isolated from slaughtered pigs in Thailand by showing the predominant mcr-7 and mcr-8 in these isolates, antimicrobial susceptibility patterns, the genetic diversity of KpnC harboring mcr isolates, and the horizontal transferability of these mcr genes. The results provide evidence that slaughtered pigs are a reservoir of mcr-7 and mcr-8 for subsequent dissemination. This research will strengthen the evidence-based knowledge of mcr-harboring KpnC in slaughtered pigs and will contribute to strategic planning for the control of overuse or misuse of antimicrobial drugs on farms and the prevention of this pathogen contaminating farmed livestock, as well as assisting surveillance on this organism.

Bacterial Strains and Identification
In total, 280 KpnC isolates were collected and isolated from the carcasses of slaughtered pigs across 10 provinces in Thailand during 2014 and 2015. Four slaughterhouses were randomly selected from each province, with 50 swab samples randomly collected from each slaughterhouse, resulting in a sample size of 2000 swab samples. One side of the carcass was swabbed for a total area of 400 cm 2 using a single swab. The swab samples were immediately stored on ice in zip-lock bags throughout transportation to the microbiological laboratory. Isolation and identification of K. pneumoniae were carried out using a 10-fold serial dilution of swab samples achieved using buffered peptone water (BPW). Each diluted BPW was spread onto MacConkey agar and incubated at 37 • C for 24 h. The presumptive Klebsiella species for up to five colonies were confirmed using conventional biochemical tests described elsewhere [26]. All Klebsiella isolates were stored at −80 • C in a laboratory freezer until used in this study.
All isolates from the freezer were cultured on MacConkey agar, and their DNA was prepared using ZymoBIOMICS™ DNA Miniprep kits (Zymo Research Corp., Irvine, CA, USA) following the manufacturer's instructions. Multiplex polymerase chain reaction (PCR) identification of species in KpnC (K. pneumoniae, K. variicola, and K. quasipneumoniae) was carried out as described previously [27], with certain modifications involving replacing the former PCR primers for K. pneumoniae with primers for the K. pneumoniae waaQ gene, as described previously [28]. The PCR program involved initial denaturation at 95 • C for 3 min, followed by 35 cycles of denaturation at 95 • C for 30 s, and annealing and extension at 65 • C for 1 min. PCR of the Kp50233 gene was used for confirmation of the K. pneumoniae-positive samples [29].

Antimicrobial Susceptibility Testing
All techniques were performed and interpreted according to the 2021 Clinical and Laboratory Standards Institute (CLSI) guidelines [33]. KpnC isolates positive for mcr were further investigated for antimicrobial susceptibility using the disk diffusion method to provide evidence-based guidance for further optimizing effective antimicrobial treatment options and surveillance for the emergence of antibiotic drug resistance. Antimicrobial disks used in the assay were loaded with ampicillin, gentamicin, amikacin, amoxicillin/clavulanic acid, piperacillin/tazobactam, cefepime, cefotaxime, ceftazidime, ertapenem, imipenem, meropenem, ciprofloxacin, levofloxacin, chloramphenicol, tetracycline, fosfomycin, nitrofurantoin, azithromycin, or trimethoprim. E. coli ATCC 25922 was used as a control. The minimal inhibitory concentration (MIC) of colistin was determined using the broth microdilution method.

Multilocus Sequence Typing (MLST)
To explore the genetic diversity of KpnC isolates from pigs in this study, MLST was performed according to the Pasteur scheme of Institut Pasteur (https://bigsdb.pasteur. fr/klebsiella/klebsiella.html accessed on 24 September 2021). The PCR products of the seven housekeeping genes were purified using an E-Z 96 Cycle Pure Kit (Omega, GA, USA) following the manufacturer's instructions. Sanger DNA sequencing of the purified PCR products was performed by Apical Scientific Sdn Bhd, Selangor, Malaysia. MLST alleles and the resulting sequence types (STs) were identified using the Institute Pasteur MLST database (https://bigsdb.pasteur.fr/klebsiella/klebsiella.html accessed on 24 September 2021). Novel alleles or allelic profiles were submitted to the curator of the database to assign an allele or ST number. The PHYLOViZ 2.0 software [34] was used to analyze STs and their clonal complexes (CCs) [35].

Conjugation Assays
Conjugation assays were carried out using all isolates of mcr-harboring KpnC strains (donor) with streptomycin-resistant E. coli UB1637 (recipient), as described elsewhere [36]. All E. coli transconjugants were then selected on the basis of growth on MacConkey agar containing 1 µg/mL of colistin and 3200 µg/mL of streptomycin. Transconjugants were confirmed as E. coli using PCR [37] and for the presence of antimicrobial-resistance genes consisting of mobile colistin resistance (mcr-1 to mcr-9) genes, β-lactamase genes (bla TEM , bla SHV , and bla CTX-M ), and plasmid-mediated quinolone resistance (PMQR) genes using PCR, as described above. Colistin MIC values were determined as described above. Details concerning the donor strains used in conjugation assays are shown in Table 1.

Transferability of mcr Genes
To prove the ability of KpnC to transfer mcr-harboring plasmid to the different bacterial species, the 19 mcr-carrying KpnC isolates were subjected to conjugation assays using E. coli UB1637 (streptomycin-resistant) as the recipient. Of these, only five KpnC donors exhibited transferability. Among these transconjugants, mcr-8 (n = 4) from K. pneumoniae and mcr-1 (n = 1) from K. quasipneumoniae were successfully transferred, whereas mcr-7 was not detected in any of the transconjugants in this study (Table 3). All transconjugants had MIC values for colistin in the range 4-8 µg/mL (Table 3). In addition, donors of K. pneumoniae harboring oqxAB (n = 2) or bla SHV (n = 2) were co-transferred into recipient E. coli cells (Table 3).

Discussion
Colistin has been extensively used in pig production for the prevention and treatment of diarrhea (especially during the post-weaning period), and as a countermeasure to endemic diseases [38]. Currently, in Thailand, colistin application is prohibited for disease prevention in livestock [39]; however, the dissemination has been reported of a mobile colistin resistance gene such as mcr contributing to cross-contamination among carcasses during slaughter and post-slaughter processing [11,21,40]. The current study identified the mobile colistin genes mcr-8, mcr-7, mcr-7 + mcr-8, and mcr-1 + mcr-7 in KpnCs, listed in order of prevalence (most to least). Interestingly, this is the first report to demonstrate that K. variicola and K. quasipneumoniae from slaughtered pigs harbored mcr-7 or mcr-8 genes.
We have shown that certain KpnC isolates were able to transfer mcr-1 or mcr-8 to an E. coli UB1637 recipient, resulting in higher MIC values for colistin than those of the donors. In addition, the mcr-1 gene was transferred from K. quasipneumoniae to E. coli recipients with a higher colistin MIC value than for the donor strain. Increased resistance to colistin was observed in K. pneumoniae of swine origin and the recipient E. coli DH5α in terms of the mcr-1 gene [36]. These findings suggest that E. coli UB1637 is a host strain that is likely to support enhanced mcr expression compared to KpnC. Co-transfer of mcr-8 + bla SHV + oqxAB was found in two isolates. This suggested that, in this study, the spread of mcr or β-lactamase or PMQR genes could be transmitted to other pathogens and could spread to other hosts.
KpnCs (especially involving K. pneumoniae) from slaughtered pigs that harbored mcr genes were highly resistant to azithromycin. Macrolides are important antimicrobials for the treatment of infections in cattle and pigs and are commonly used in pig farms worldwide [45]. Resistance to macrolide of K. pneumoniae from pigs has also been reported in Thailand, with the most resistant gene being mdf(A) [44]. However, all our isolates were susceptible to carbapenem, amoxicillin/clavulanic acid piperacillin/tazobactam, and third-generation cephalosporins, which was consistent with other studies [46,47].
The current study showed that mcr isolates exhibited high diversity, as revealed by MLST analysis, and the STs of 26.3% (5/19) KpnC isolates were novel. The goeBURST analysis showed that ST999 had a single-locus variant of ST35 [30]. This ST has been identified in OXA-48-producing K. pneumoniae isolated from humans [48], whereas our isolate carried mcr-8. K. pneumoniae ST35, a global multidrug-resistant clone that has been isolated in many countries [49][50][51]. In Thailand, ST35 harboring mcr-3 have been found in humans [44]. The current study detected ST35 in slaughtered pigs, emphasizing that pigs and human K. pneumoniae isolates might be from the same source, and pig-borne transmission plays a crucial role in the transmission of mcr-carrying K. pneumoniae.

Conclusions
The study revealed colistin non-susceptible KpnC harboring mcr-1, mcr-7, and mcr-8 genes in slaughtered pigs in Thailand. Therefore, improvements are strongly recommended in food hygiene standards and biosecurity measures on farms and in slaughtering procedures. This evidence-based knowledge of mcr-harboring KpnC in slaughtered pigs can be applied in livestock policy planning, monitoring changes in population dynamics, and the development of surveillance strategies for prevention and control programs.