Emergence of Mobile Colistin Resistance (mcr-8) in a Highly Successful Klebsiella pneumoniae Sequence Type 15 Clone from Clinical Infections in Bangladesh.

There is a marked paucity in our understanding of the epidemiology of colistin-resistant bacterial pathogens in South Asia. A report by Davies and Walsh (Lancet Infect Dis 18:256–257, https://doi.org/10.1016/S1473-3099(18)30072-0, 2018) suggests the export of colistin from China to India, Vietnam, and South Korea in 2016 was approximately 1,000 tons and mainly used as a poultry feed additive. A few reports forecast that the prevalence of mcr in humans and livestock will increase in South Asia. Given the high prevalence of blaCTX-M-15 and blaNDM in India, Bangladesh, and Pakistan, colistin has become the invariable option for the management of serious infections, leading to the emergence of mcr-like mechanisms in South Asia. Systematic scrutiny of the prevalence and transmission of mcr variants in South Asia is vital to understanding the drivers of mcr genes and to initiate interventions to overcome colistin resistance.

A pilot antimicrobial resistance (AMR) survey was conducted from 21 October 2016 to 23 September 2017 at Dhaka Medical College Hospital (DMCH), which included 1,097 culture-positive clinical specimens. The project was approved by the Ethical Review Committee of DMCH (MEU-DMC/ECC/2017122). K. pneumoniae was recovered on chromogenic UTI containing vancomycin (10 mg/liter) and identification by matrix-assisted laser desorption ionization-time of flight mass spectrometry (MALDI-TOF MS; MALDI Biotyper; Bruker Daltonics, Inc., Billerica, MA, USA). MICs of relevant antimicrobials were determined by agar dilution and the MIC of colistin by broth microdilution. Susceptibility patterns of antimicrobials were interpreted according to EUCAST breakpoints. Sequencing was performed using Illumina MiSeq (Illumina Inc., San Diego, CA) and Nanopore (Oxford Nanopore Technologies, Oxford, UK) platforms. We adopted a hybrid strategy to assemble draft genomes using Unicycler (v0.4.0) (see Text S1 in the supplemental material). Pangenome analysis was performed using Roary (v3.12.0). A maximum likelihood phylogenetic tree was built using FastTree (v2.1.0) and visualized using Phandango and iTOL (v5.3). Intraclade single-nucleotide polymorphisms (SNPs) were identified using Snippy (v4.4.5). Plasmid size was confirmed by pulsed-field gel electrophoresis (PFGE) of S1 nuclease DNA digests and mcr-8.1 probing. Conjugation assays were performed using Escherichia coli J53 as the recipient (9). Serial passaging of MCR-positive K. pneumoniae (MCRPKP) was performed in a colistin-free medium up to 12 days, and genomic DNA (gDNA) was extracted on days 0, 3, 6, 9, and 12. Plasmid stability was FIG 1 Phylogenetic tree of K. pneumoniae ST15 identified in this study (n ϭ 29). Shown is a maximum likelihood (ML) phylogenetic tree constructed using a pangenome alignment. Strains were grouped together based on the similarity of genes and the presence of genes in the accessory genome using Roary (v3.12.0). Epidemiologically important resistance genes are indicated by orange cells, accessory genes by blue cells, and the absence of genes by white cells. NICU, neonatal intensive care unit; ICU, intensive care unit. (Continued on next page) Emergence of mcr-8.1 in Bangladesh assessed by relative abundance of mcr-8.1 compared to that of housekeeping genes (HKGs) using quantitative PCR (Bio-Rad, USA) (Text S2). The in vitro growth rate of E. coli J53 and transconjugants (TDM697b, TDM782, and TDM914b) was determined by optical density (OD) in 30-min intervals for 24 h using FLUOstar Omega (BMG Labtech Ltd., Aylesbury, UK). The growth rate of each transconjugant was compared to that of E. coli J53 by unpaired two-tailed t test using GraphPad Prism (v7.04) (Text S3).
Description of cases with infections by MCRPKP. In this study, 3 K. pneumoniae isolates (3/1,097, 0.3%) were phenotypically resistant to colistin. MCRPKP isolates were recovered from the urine of two patients admitted under urology and the blood of a third patient in the neonatal intensive care unit (NICU). Case 1 (BD_DM_697) was a 55-year-old male with benign enlargement of the prostate with diabetes mellitus and a history of catheterization for 13 days. Case 2 (BD_DM_782) was a 63-year-old male patient with a left renal tumor, a history of catheterization for 15 days, and hematuria. These patients were discharged on days 20 and 35 of hospitalization, respectively. Case 3 (BD_DM_914) was a 5-day preterm low-birth-weight neonate with late-onset neonatal sepsis who died within 18 days after hospital admission. We did not observe any overlapping of hospital stay among the MCRPKP cases. MCRPKP isolates were coresistant to amoxicillin-clavulanate, piperacillin-tazobactam, cephalosporins (ceftazidime and cefotaxime), ciprofloxacin, levofloxacin, gentamicin, trimethoprim-sulfamethoxazole, and colistin and susceptible to carbapenems, amikacin, fosfomycin, and tigecycline. Although MCRPKP cases initially were shown to be treated with inappropriate antimicrobials, we have no data on whether the antibiotic therapy was subsequently changed based on the sensitivity report from the local laboratory.
Clonal spread of mcr-8.1. In silico genome-wide analysis of MCRPKP detected a 1,698-bp open reading frame (ORF), encoding a phosphoethanolamine transferase, showing 100% nucleotide identity to mcr-8.1. The prevalence of K. pneumoniae among all clinical isolates from this study was 21% (228/1097), of which 13% (29/228) belonged to ST15. K. pneumoniae ST15 harboring mcr-8.1 was clustered in one clade (Fig. 1), suggesting the clonal spread of MCRPKP. SNP mapping found 110 and 107 SNPs in BD_DM_782 and BD_DM_914, respectively, compared to BD_DM_697, and 23 SNPs in BD_DM_782 compared to BD_DM_914. MCR-8 was described previously in K. pneumoniae ST1, of human origin, and K. pneumoniae ST42, of animal origin (14). K. pneumoniae ST15 has been regarded as a successful clone in disseminating bla CTX-M-15 globally (15). The draft genome sequences and S1 PFGE indicate that mcr-8.1 elements in K. pneumoniae were located on identical IncFIB(pQil) plasmids of ϳ113 kb (GenBank accession no. CP046384, CP046952, and CP046942) ( Fig. 2 and Fig. S1). The gene mcr-8.1 was stable after serial passaging without any antibiotic challenge. Compared to that at day 0, the abundance of mcr-8.1 versus HKG was static up to day 12 (Fig. S2). Yang et al. (16) reported that colistin susceptibility could be attenuated after serial passaging of mcr-1-positive strains in antibiotic-free medium. Our findings demonstrate that the IncFIB(pQil) plasmid harboring mcr-8.1 was remarkably stable, suggesting adaptive plasmid-host evolution (17). K. pneumoniae ST15 can be a vector capable of spreading mcr-mediated colistin resistance, particularly in a setting with suboptimal infection control practices (18).
Genetic context and dynamics of plasmids harboring mcr-8.1. Genome-wide analyses demonstrated that the plasmids recovered from the MCRPKP were almost identical to each other (Fig. S3). Complete plasmid sequences were determined for pKP782 (accession no. CP046384) and pKP914 (accession no. CP046952) by hybrid assembly, while pKP697 (accession no. CP046942) was not successfully closed. One copy of the IncFIB(pQil) plasmid with an identical resistance profile was recovered from each MCRPKP isolate and shared 99.72% nucleotide identity at 70% coverage with previously described plasmids (accession no. CP023922.1 and CP024040.1). However, those plasmids (CP023922.1 and CP024040.1) were absent from mcr-like genes (Fig. 2B). The genetic environment around mcr-8.1 in IncFIB(pQil)-MCR-8.1 (pKP697, pKP782, and pKP914) shares identity with a previously described mcr-8.1-containing plasmid isolated from pigs in China (accession no. MG736312.1), although the plasmids harboring mcr-8.1 in this study are truncated at the 5= end and IS903B at the 3= end was replaced by ISKpN14 (14) (Fig. 2C). It is possible that mcr-8.1 originally was transposed to the IncFIB(pQil) plasmid by an IS903B composite transposon ( Fig. 2B and C). An array of AMR genes (bla TEM-1b and bla CTX-M-15 ) was in a composite transposon, flanked by insertion sequences (Fig. 2B). Incidentally, all resistance components in IncFIB(pQil) plasmids in this study were bracketed by IS903B from nucleotide position 20590 to 64656, demonstrating the potential for the transposition of the entire intervening DNA segment. The conjugation assay confirmed the transferability of the plasmid containing mcr-8.1 to E. coli J53 with a frequency range of 3.1 ϫ 10 Ϫ2 to 8 ϫ 10 Ϫ2 . Phenotypically, the transconjugants were resistant to ampicillin, amoxicillin-clavulanate, 3rd-generation cephalosporins, trimethoprim-sulfamethoxazole, and colistin (Table S1).
The acquisition of a resistance plasmid may impose a fitness cost, depending on the host and plasmid backbones (19,20). We found a significantly lower growth rate over time in TDM697b and TDM914b relative to that of E. coli J53 (P Ͻ 0.0001) (Fig. S4), implying a significant fitness cost owing to the acquisition of a plasmid harboring mcr-8.1. Compared to that of E. coli J53, a lower growth rate was also observed in TDM782b; however, the fitness cost was not statistically significant (Fig. S4). This is the first report of transferable colistin resistance associated with human infections from Bangladesh. Given the acquisition of mcr-8.1 on a conjugative plasmid, with good stability in ST15, a successful high-risk clone of K. pneumoniae, there is a serious risk of dissemination of mcr-8.1 in South Asia.
Accession number(s). The nucleotide sequences of MCRPKP isolates are available under NCBI accession no. CP046939 to CP046947, CP046381 to CP046385, and CP046939 to CP046947.

SUPPLEMENTAL MATERIAL
Supplemental material is available online only. TEXT S1, DOCX file, 0.02 MB.