Transmissible colistin resistance encoded by mcr-1 detected in clinical Enterobacteriaceae isolates in Singapore

Abstract

Emerging Microbes & Infections (2016) 5, e87; doi:10.1038/emi.2016.85; published online 17 August 2016

Dear Editor,

Following the recent description of transmissible plasmid-mediated colistin resistance encoded by mcr-1 in clinical and veterinary Escherichia coli and Klebsiella pneumoniae isolates in China,¹ several groups have reported mcr-1 in colistin-resistant isolates from humans and food animals across Asia^{2, 3, 4} to Europe.^{5, 6, 7, 8} It is evident that mcr-1 has disseminated globally.

We performed a prospective study in January 2016 using 306 consecutive clinical Enterobacteriaceae isolates collected from blood, urine and miscellaneous samples (swabs and pus). The species investigated were E. coli (n=166), K. pneumoniae (n=87), Klebsiella oxytoca (n=4), Enterobacter spp. (n=22), Proteus spp. (n=10), Citrobacter spp. (n=9), Morganella morganii (n=5), Providencia rettgeri (n=1), Salmonella enteritidis (n=1) and Serratia marcescens (n=1). Isolates were PCR-screened for mcr-1¹ without prior knowledge of their antibiograms or colistin-resistance status. Three of these isolates (two E. coli and one K. pneumoniae) were mcr-1 positive, with their full-length mcr-1 gene matching the nucleotide identity of the first-described isolate exactly.¹ Multilocus sequence typing (MLST) was performed (http://bigsdb.web.pasteur.fr/index.html). There was no evidence of nosocomial transmission, as the two E. coli isolates were of different sequence types (STs; Table 1).

Table 1 Characteristics of mcr-1-positive clinical Enterobacteriaceae isolates in Singapore

Isolate	Specimen	Species	Date of isolation	MLST	β-lactamases	ISApl1 associated	mcr-1 transmissible via conjugation^a	MIC (mg/L)^b
								PB	COL	IMP	MEM	ETP	FOX	CAZ	AMP	PTZ	CIP	LEV	GEM	AMK
NM12	Urine	E. coli	11/01/2016	ST460^C	TEM-1	Yes	Yes, ≈10^−6	4	4	≤0.25	≤0.25	≤0.5	≤4	≤1	≥32	64	2	4	<1	<2
NM4	Urine	E. coli	14/01/2016	ST156	TEM-1	No	Yes, ≈10^−7	4	4	≤0.25	≤0.25	≤0.5	≥64	16	≥32	≤4	1	2	>16	<2
NM2	Urine	K. pneumoniae	13/01/2016	ST1535	TEM-1, CTX-M-15	No	Yes, ≈10^−3	24	8	≤0.25	≤0.25	≤0.5	≤4	16	≥32	≤4	1	2	<1	<2
Transconjugant NM12	—	E. coli	—	—	TEM-1	—	—	2	2	≤0.25	≤0.25	≤0.5	≤4	≤1	≥32	128	2	4	≤1	≤2
Transconjugant NM4	—	E. coli	—	—	TEM-1	—	—	4	6	≤0.25	≤0.25	≤0.5	≤64	16	≥32	≤4	1	2	≥16	≤2
Transconjugant NM2	—	E. coli	—	—	Not detected	—	—	8	6	≤0.25	≤0.25	≤0.5	≤4	≤1	8	≤4	≤0.25	≤0.12	≤1	≤2
J53	—	E. coli	—	—	—	—	—	0.5	0.25	≤0.25	≤0.25	≤0.5	≤4	≤1	8	≤4	≤0.25	≤0.12	≤1	≤2

Abbreviations: amikacin, AMK; ampicillin, AMP; ceftazidime, CAZ; ciprofloxacin, CIP; colistin, COL; cefotaximase, CTX; ertapenem, ETP; cefoxitin, FOX; gentamicin, GEM; imipenem, IMP; levofloxacin, LEV; minimum inhibitory concentration, MIC; meropenem, MEM; polymyxin B, PB; piperacillin-tazobactam, PTZ.

a The conjugation efficiency is calculated as the number of transconjugants per donor cell.

b The interpretation of results of susceptibility testing were based on European Committee on Antimicrobial Susceptibility Testing (EUCAST) guidelines. For colistin, a susceptible breakpoint of ≤2 mg/L and a resistant breakpoint of>2 mg/L was applied. Because no interpretative criteria are available for polymyxin B, colistin breakpoints were applied.

c The Institut Pasteur MLST scheme was utilized. The following new allelic profile was obtained: dinB 22; pabB 21; putB 23; trpB 127; icdA 61; polB 109; trpA 41; uidA 164, and was assigned as ST460.

The isolates were urinary specimens (Table 1). Our detection rate was estimated to be 0.98% (95% confidence interval of 0.3%–2.8%, Wilson score interval). This was very close to 1% mcr-1 prevalence in China.¹ Sensitivity testing was performed via E-test for polymyxin B and colistin and with the Vitek2 GNR257 card for all other antimicrobials. Phenotypically, the isolates were resistant to polymyxin B (minimum inhibitory concentration (MIC) 4 mg/L–24 mg/L) and colistin (MIC 4 mg/L–8 mg/L) but were sensitive to carbapenems. There was a variable sensitivity to third-generation cephalosporins and piperacillin-tazobactam (Table 1). PCR screening for carbapenemases (K. pneumoniae carbapenemase (KPC), metallo-β-lactamases (New Delhi Metallo-β-lactamase (NDM), verona integron-encoded metallo-β-lactamase, imipenemase), class D carbapenemases (oxacillinase (OXA)-23, OXA-48-like)) and broad- and extended-spectrum β-lactamases (BSBL and ESBLs) was performed.⁹ Only narrow-spectrum β-lactamase (TEM)-1 BSBL was detected in the E. coli isolates (Table 1). This was noteworthy because mcr-1-positive isolates have been found to be associated with cefotaximase (CTX)-M-like ESBLs.^{4, 5, 6} Furthermore, in our isolates, carbapenemase genes were not carried with mcr-1, which contrasts growing reports in which mcr-1 has been found together with bla_KPC-2 and bla_NDM carbapenemase genes^{7, 8} and results in colistin-resistant isolates that are also carbapenem resistant.^{7, 8} Using liquid-mating assays, mcr-1 was transferable to an E. coli recipient, J53, in all the clinical donor isolates. Transconjugants were selected on Luria Bertani agar containing 50 mg/L of sodium azide and 0.5 mg/L of colistin. All of the transconjugants were phenotypically resistant to colistin and polymyxin B. TEM-1 also transferred in two transconjugants (Table 1). PCR replicon typing¹⁰ indicated that the transconjugants of NM2 and NM12 could not be typed, while transconjugant NM4 carried IncF and IncI1. The genetic environment of mcr-1 is variable and is not always associated with an upstream ISApl1.⁷ Here PCR mapping and sequencing based on the initial genetic environment¹ showed that one isolate did not have a flanking ISApI1 (Table 1). This suggests that the dissemination of mcr-1 is likely facilitated by a diverse range of mobile genetic elements. We plan to commence full-genome sequencing in the near future to characterize the plasmids and mobile elements in detail. Because Singapore has limited farming and agricultural activity, mcr-1 is less likely to be acquired through contact with local livestock; although not yet conclusively proven, it appears that imported meat products and vegetables are more likely sources of mcr-1.^{11, 12, 13}

We thank the team of curators of the Pasteur Institute MLST and whole genome MLST databases for curating the data and making it publicly available at http://bigsdb.web.pasteur.fr/.