Molecular characteristic of mcr-1 producing Escherichia coli in a Chinese university hospital.

BACKGROUND: Colistin has been considered as a last-line treatment option in severe infections caused by multidrug-resistant (MDR) gram-negative pathogens. However, the emergence of the mobile colistin resistance gene (mcr-1) has challenged this viewpoint. The aim of this study is to explore the prevalence of mcr-1 in Escherichia coli (E. coli) in a Chinese teaching hospital, and investigate their molecular characteristics.
METHODS: A total of 700 E. coli isolates were used to screen mcr-1 by PCR and sequencing in a Chinese university hospital from August 2014 to August 2015. Susceptibility test of mcr-1-producing isolates was determined by Vitek -2 Compact system. 26 virulence factors (VFs), phylogenetic groups, Multi-locus sequence typing (MLST), and DNA Fingerprinting (ERIC-PCR) of strains were investigated by PCR.
RESULTS: Four (0.6%) mcr-1 producing E. coli isolates were found in this study. The results of antibiotic susceptibility test showed that all four isolates were resistant to colistin, ciprofloxacin, levofloxacin, cefazolin, and trimethoprim/sulfamethoxazole, and were susceptible to amikacin, ertapenem and imipenem. In addition, all 4 isolates exhibited high-level resistance to aztreonam, cefotaxime and gentamicin. The numbers of VFs contained in mcr-1 positive isolates were no more than 4 in our study. MLST result demonstrated that these isolates were assigned to two sequence types: ST156 and ST167. The result of phylogenetic analysis showed that four mcr-1-positive isolates belong to two phylogenetic groups: A and B1 group. ERIC-PCR showed that four mcr-1 positive strains were categorized into three different genotypes.
CONCLUSIONS: Our study demonstrated a low prevalence of mcr-1 in E. coli clinical isolates in a Chinese teaching hospital, and we have gained insights into the molecular characteristics of these mcr-1-positive strains. Increasing the surveillance of these infections, as well as taking effective infection control measures are urgently needed to take to control the transmission of mcr-1 gene.

Background

In recent years, colistin has been considered as an effective therapeutic option for the rapid increasing of multidrug-resistant (MDR) gram-negative pathogens [1, 2]. However, the prevalence of the mobile colistin resistance gene (mcr-1) in animals and human beings worldwide has challenged this viewpoint [3, 4]. Resistance to polymyxins is mainly caused by the modification to bacterial outer membrane, which was usually considered as chromosomally mediated resistance [5, 6].

Since it was initially found, plasmid-mediated mcr-1 has been detected widely [3, 7]. Nowadays, mcr-1-producing bacteria have been reported in many regions in China [4, 8]. Mcr-1 was firstly found in Escherichia coli (E. coli), and now it has been spreading to other Enterobacteriaceae [9]. Several reports showed that the mcr-1 gene could coexist with other resistance genes (such as CRE/ESBL) in E. coli and Klebsiella pneumoniae, which probably lead to the emergence pan-drug resistant and increase the difficulty of treatment [8, 10]. Therefore, the emergence and spread of mcr-1 gene among human beings should be given close attention. The aim of this study was to evaluate the prevalence of mcr-1 in E. coli clinical isolates in a Chinese teaching hospital, and to investigate the molecular characteristics of these strains.

Methods

Bacterial strains

A total of 700 E. coli clinical isolates were collected from the clinical laboratory of Fujian Medical University Union Hospital (Fuzhou, Fujian province, China) from August 2014 to August 2015. It is a 2200-bed tertiary care teaching hospital with approximately 95,000 hospital admissions per year, located in southeastern China. All isolates were identified by GNI card of the Vitek system (BioMèrieux, Missouri, France).

Antibiotic susceptibility testing

Antimicrobial susceptibility testing was performed using the AST-GN16 of Vitek-2 Compact system (Bio Mérieux, France). The antimicrobial agents tested included: tigecycline (glycycline); ertapenem and imipenem (carbapenems); cefazolin; cefoxitin; cefepime, and cefotaxime (cephalosporins); aztreonam (monobactam); amikacin and gentamicin (aminoglycosides); ciprofloxacin and levofloxacin (quinolone); piperacillin/tazobactam; trimethoprim/sulfamethoxazole. The results were interpreted by the Clinical and Laboratory Standards Institute (CLSI) [11]. The MIC of colistin was determined using agar dilution method, and the result was interpreted according to European Committee On Antimicrobial Susceptibility Testing (EUCAST) guidelines [12]. E. coli ATCC 25922 was used as a quality control.

DNA extraction

Several colonies were suspended in 50 µl of sterile distilled water for preparing genomic DNA of the isolates, and then the bacterial suspension was heated at 100 °C for 10 min as described previously [13].

MCR-1 detection

mcr-1 gene was screened in E. coli clinical isolates by PCR using primers as previously described [4]. All of the PCR products were sequenced and then compared with known sequences listed in the GenBank (http://www.ncbi.nlm.nih.gov/blast/).

Detection of virulence factor genes

Twenty six virulence factors (VFs) genes associated with extraintestinal virulence [14, 15] were detected using a multiplex PCR method as previously described [15]. These genes were as follows: adhesions (papAH, papEF, papC, papG allele I, papG II/III, papG allele II, sfa/focDE, afa/draBC, fimH, gafD, sfaS, focG and nfaE), toxins (hlyA, cnf1 and cdtB), siderophores (fyuA and iutA), protections and invasions (kpsMTII, kpsMTIII, traT, cvaC, kpsMT and K1/K5), miscellaneous (rfc and PAI). The PCR products were sequenced and then compared with known sequences listed in the GenBank (http://www.ncbi.nlm.nih.gov/blast/).

Phylogenetic analysis

The phylogenetic groups (A, B1, B2, and D) of mcr-1 producing E. coli isolates were identified by a triplex PCR as previously described [16].

Multi-locus sequence typing (MLST)

Mcr-1 positive strains were analyzed by multilocus sequence typing (MLST), which was based on 7 standard housekeeping genes (adk, fumC, gyrB, icd, mdh, purA, recA) (http://mlst.ucc.ie/mlst/mlst/dbs/Ecoli) [17].

DNA fingerprinting

Enterobacterial Repetitive Intergenic Consensus Sequences PCR (ERIC-PCR) was applied to typing mcr-1 producing E. coli isolates with the primers ERIC-1 and ERIC-2 [18]. DNA fingerprints were compared by visual inspection, ERIC profiles were regarded as different if there were different bands on visual inspection [19].

Results and discussion

In this study, four isolates (0.6%) were confirmed to carry mcr-1 gene, which is lower than previous study [4]. The age of the patients ranged between 38 and 80 years. These mcr-1 producing strains were isolated from two different wards (Table 1). Two strains were isolated from the same patient. The clinical data of patients with mcr-1 positive E. coli infection were shown in Table 2.

Table 1

Main characteristics of the mcr-1 E. coli

Isolates	Data	Ward	Specimen	Phylogenetic groups	MLST	ERIC pattern	VFs	Antibiotic resistance
E321	2014.8	Colorectal surgery	Drainage-fluid	B1	ST156	1	traT, iutA	COL, CFZ, FOX, CIP, LVX, SXT, TGC
E684	2015.1	Colorectal surgery	Secretion	B1	ST156	2	fimH, traT, iutA	COL, CFZ, CTX, FEP, ATM, GEN, CIP, LVX, SXT
E921	2015.4	Hepatobiliary surgery	Secretion	A	ST167	3	fyuA, traT, iutA	COL, CFZ, CTX, ATM, GEN, CIP, LVX, SXT
E1005	2015.5	Hepatobiliary surgery	Drainage-fluid	A	ST167	3	fyuA, cvaC, traT, iutA	COL, CFZ, CTX, ATM, GEN, CIP, LVX, SXT

CFZ cefazolin, FOX cefoxitin, CTX cefotaxime, FEP cefepime, TZP piperacillin/tazobactam, ATM aztreonam, IPM imipenem, ETP ertapenem, AMK amikacin, GEN gentamicin, CIP ciprofloxacin, LVX levofloxacin, SXT trimethoprim/sulfamethoxazole, TIG tigecycline, COL colistin, MLST multi-locus sequence typing, ERIC enterobacterial repetitive intergenic consensus, VFs virulence factor genes

Table 2

Clinical data of patients with mcr-1 positive E. coli infection

Isolates	Patients	Gender	Age (years)	Underlying diseases	Length of hospital stay (days)	Treatments used	Outcomes
E321	Patient 1	Female	80	Malignancy, hypertension, pulmonary tuberculosis	51	TZP	Survived
E684	Patient 2	Female	57	Perineal infection, hypertension	37	TZP	Survived
E921	Patient 3	Male	38	Hypertension, pancreatitis, diabetes	26	MEM	Survived
E1005	Patient 3	Male	38	Hypertension, pancreatitis, diabetes	26	MEM	Survived

TZP piperacillin/tazobactam, MEM meropenem

Mcr-1 was usually found to be co-localized with other resistance genes on plasmids, such as ESBL genes and carbapenemase genes [20], which might increase the emergence of pan-drug resistance. In our study, the results of antimicrobial susceptibility test showed a high drug resistance in the mcr-1-producing isolates. All of the mcr-1 positive isolates were resistance to at least 3 different kinds of antibiotics (Table 1).

All four mcr-1 positive strains detected in our study were resistant to colistin and the MICs ranged from 4 to 16 μg/ml. It will be worrisome once mcr-1 coexists with other resistant genes, especially carbapenemase genes because of limited therapeutic options [20]. Previous studies revealed that mcr-1 co-produced with carbapenem-resistant genes in E. coli [8, 21]. Fortunately, all of them were susceptible to carbapenems (IPM and ETP), which probably indicated that no carbapenem-resistant genes coexisted with mcr-1. Result of ERIC-PCR (Fig. 1) showed that four mcr-1 positive strains were categorized into three different genotypes, one of which contained 2 strains (from the same patient). These isolates which have different patterns suggest that they were non-clonal transmission. In a previous study, two mcr-1 positive E. coli isolates from a single fowl were belonging to phylogenetic B1 and D group [22]. The mcr-1-producing isolates in this study were belonged to phylogenetic groups A and B1, which were mainly distributed among human commensal E. coli isolates [23]. The mcr-1 producing isolates were assigned by MLST to two different sequence types: ST156 and ST167 (Table 1), which was similar to previous reports in other studies from China [8, 22]. E. coli ST156 has been found that it has connection with different ESBL genes [24, 25]. ST167 was belonged to ST10 complex and regarded as prevalent ST among ESBL-producing E. coli from human and animal sources [26]. In addition, E. coli ST167 was reported to be closely related to bla NDM, which needed closely concern of spreading [27]. The similar molecular characterizations illustrated that mcr-1 positive isolates detected from the same department in our study were clonally related.

Fig. 1

ERIC-PCR products from four mcr-1 positive isolates. M mark, lane 1 E321, lane 2 E684, lane 3 E921, lane 4 E1005

VFs in E. coli were associated with colonization, bacterial fitness and virulence [28]. VFs include five main groups: (1) adhesions; (2) toxins; (3) siderophores; (4) capsule production and (5) protections and invasions. Clinical E. coli strains often carry multiple VFs, and isolates belonging to groups A and B1 often have less VFs than those belonging to phylogroups B2 and D [28]. To the best of our knowledge, there is no study concerning about VFs in mcr-1 producing E. coli. In our study, mcr-1 producing isolates contained less than 4 different VFs (Table 1). Only five different kinds of VFs had been detected in our mcr-1 positive isolates, which included fimH, fyuA, traT, iutA and cvaC. fimH is one of the most commonly VFs present in E. coli, which encodes the adhesion subunit of type 1 fimbriae and related to colonization [15]. Lee et al. reported that fyuA, traT, and iutA were found to be independent predictors for pathogenicity. Meanwhile, traT and iutA were thought to be closely related to ESBL genes [29]. Pitout et al. found that cvaC was only present in non-CTX-M-producing isolates [30]. Previous reports suggested that antibiotic resistance has negative association with virulence factors [31], which could be interpreted by the loss of VFs associated with mutation to resistance [32].

It is noteworthy that two mcr-1 positive E. coli strains were isolated from the same patient but at different time (Table 1). Results of MLST and ERIC-PCR revealed that these isolates had identical genetic background. Result of antimicrobial susceptibility test showed that they had similar antibiograms. We speculate that the two isolates probably originated from a same source.

In conclusion, we have revealed a low prevalence of mcr-1 in E. coli clinical isolates in a Chinese teaching hospital, and presented detailed molecular characteristics of these isolates. The presence of mcr-1 in E. coli clinical isolates suggests that it will pose a threat to public healthcare. Effective infection control measures are urgently needed to take to control the transmission of mcr-1 gene.