First Detection of the mcr-1 Colistin Resistance Gene in Escherichia coli in Italy

LETTER

Polymyxins are old antibiotics that have recently regained popularity for treatment of severe infections caused by extensively drug-resistant (XDR) Gram-negative bacterial strains (1). As a likely consequence, emergence of polymyxin resistance is being increasingly reported in the clinical setting, especially among carbapenem-resistant Klebsiella pneumoniae isolates (2, 3). Acquired resistance to polymyxins is generally associated with chromosomal mutations (4, 5), but a new plasmid-mediated transferable resistance determinant, the mcr-1 gene, encoding a phosphoethanolamine transferase, has been described recently (6). The mcr-1 gene was originally detected in Enterobacteriaceae (mostly Escherichia coli) of animal and human origin in China (6) and subsequently also elsewhere (7–15), suggesting a broader distribution. In this communication, we report on the first detection of mcr-1 in colistin-resistant (COL-R) E. coli isolates from Italy.

A retrospective analysis of the laboratory databases from the clinical microbiology laboratories of two Italian hospitals (Florence, central Italy; Lecco, northern Italy) revealed overall stable and low rates of colistin resistance among E. coli clinical isolates in the period 2012 to 2015 (Table 1). In both laboratories, routine susceptibility testing had been performed with the Vitek 2 system (bioMérieux, Marcy l'Etoile, France), and interpretation had been performed according to the EUCAST breakpoints (www.eucast.org).

TABLE 1.

Rates of colistin resistance among E. coli clinical isolates from two different Italian settings

Year	Florence		Lecco
	Total no. of tested isolates	No. (%) of COL-R isolates	Total no. of tested isolates	No. (%) of COL-R isolates
2012	6,241	50 (0.8)	1,995	17 (0.9)
2013	8,927	62 (0.7)	2,456	19 (0.8)
2014	10,636	64 (0.6)	2,534	9 (0.4)
2015	9,373	66 (0.7)	2,368	24 (1)

Eleven isolates that had been reported as colistin resistant (COL-R) were available for investigation, and nine of them were confirmed to be COL-R by reference broth microdilution (16). The COL MIC was 8 μg/ml in all cases.

The presence of mcr-1 was screened for by PCR and sequencing as previously described (6), using also an additional pair of primers (CLR5-F1, 5′-ATGATGCAGCATACTTCTGTGTGG; CLR5-R1, 5′-TCAGCGGATGAATGCGGTGC) targeting the extremities of the mcr-1 gene. Multilocus sequence typing (MLST) was carried out as described previously (17).

Eight of the nine COL-R E. coli isolates were positive for the mcr-1 gene. In the sequenced region (positions 25 to 1576), with reference to the mcr-1 coding sequence (accession no. KP347127), the nucleotide sequences from all isolates were identical to that previously reported (6). Isolates positive for mcr-1 were detected from both centers, from inpatients in different wards, and also from two outpatients. Two mcr-1-positive isolates (LC-902/14 and LC-279/13) exhibited a multidrug-resistant phenotype, including expanded-spectrum cephalosporins, fluoroquinolones, and trimethoprim-sulfamethoxazole, and produced extended-spectrum-β-lactamase activity (Table 2).

TABLE 2.

Features of the mcr-1-positive E. coli clinical isolates^a

Isolate	Isolation date	Source	Ward	ST	MIC mg/liter (S/I/R)b
					AMC	CTX	CAZ	FEP	PIP/TAZ	ERT	IMP	MEM	GEN	CIP	SXT	COL
LC-279/13c	June 2013	Urine	Neurology	354	8 (S)	8 (R)	2 (I)	2 (I)	≤4 (S)	≤0.5 (S)	≤0.25 (S)	≤0.25 (S)	≤1 (S)	>4 (R)	>320 (R)	8 (R)
LC-705/14	August 2014	Urine	Medicine	131	8 (S)	≤1 (S)	≤1 (S)	≤1 (S)	≤4 (S)	≤0.5 (S)	≤0.25 (S)	≤0.25 (S)	≤1 (S)	≤0.25 (S)	>320 (R)	8 (R)
LC-902/14c	October 2014	Urine	Outpatient	602	4 (S)	>64 (R)	≤1 (S)	2 (I)	≤4 (S)	≤0.5 (S)	≤0.25 (S)	≤0.25 (S)	≤1 (S)	>4 (R)	>320 (R)	8 (R)
LC-968/14	November 2014	Surgical wound	Intensive care unit	10	16 (R)	≤1 (S)	≤1 (S)	≤1 (S)	8 (S)	≤0.5 (S)	≤0.25 (S)	≤0.25 (S)	≤1 (S)	>4 (R)	>320 (R)	8 (R)
LC-17/15	January 2015	Urine	Neurosurgery	95	4 (S)	≤1 (S)	≤1 (S)	≤1 (S)	≤4 (S)	≤0.5 (S)	≤0.25 (S)	≤0.25 (S)	≤1 (S)	≤0.25 (S)	>320 (R)	8 (R)
FI-4531	November 2015	Urine	Outpatient	648	8 (S)	≤1 (S)	≤1 (S)	≤1 (S)	≤4 (S)	≤0.5 (S)	≤0.25 (S)	≤0.25 (S)	≤1 (S)	≤0.25 (S)	>320 (R)	8 (R)
FI-4592	November 2015	Urine	Orthopedics	804	4 (S)	≤1 (S)	≤1 (S)	≤1 (S)	≤4 (S)	≤0.5 (S)	≤0.25 (S)	≤0.25 (S)	>16 (R)	≤0.25 (S)	>320 (R)	8 (R)
FI-4451	November 2015	Urine	Medicine	117	≤2 (S)	≤1 (S)	≤1 (S)	≤1 (S)	≤4 (S)	≤0.5 (S)	≤0.25 (S)	≤0.25 (S)	≤1 (S)	≤0.25 (S)	≤20 (S)	8 (R)

^aColistin MICs were determined by reference broth microdilution; susceptibility to other agents was determined with the Vitek 2 system. Isolates indicated as LC were from Lecco, and those indicated as FI were from Florence.

^bAbbreviations for antimicrobial agents are as follows: AMC, amoxicillin-clavulanic acid; CTX, cefotaxime; CAZ, ceftazidime; FEP, cefepime; PIP/TAZ, piperacillin-tazobactam; ERT, ertapenem; IMP, imipenem; MEM, meropenem; GEN, gentamicin; CIP, ciprofloxacin; SXT, trimethoprim-sulfamethoxazole; COL, colistin. S, susceptible; I, intermediate; R, resistant.

^cExtended-spectrum-β-lactamase (ESBL)-producing isolate.

The mcr-1-positive isolates belonged to several different sequence types (STs) (Table 2), some of which have not been previously associated with mcr-1. No clear epidemiological relationships could be traced among the mcr-1-positive isolates detected in each setting.

Taken together, the present findings revealed that E. coli strains carrying the mcr-1 gene are circulating in Italy, in the clinical setting, with a multifocal distribution and have done so at least since 2013. The fact that mcr-1 was detected in the majority of the COL-R E. coli isolates available for investigation suggests that it plays an important role as a polymyxin resistance determinant in this species. Emergence of mcr-1 in clinical isolates of E. coli is alarming, and a broader surveillance of this resistance determinant would be advisable, although colistin resistance in E. coli from clinical specimens remains very uncommon and no increasing trends were recently observed in our settings.