The first report of emerging mobilized colistin-resistance (mcr) genes and ERIC-PCR typing in Escherichia coli and Klebsiella pneumoniae clinical isolates in southwest Iran

Introduction

Gram-negative rods, especially Enterobacteriaceae, are considered as the main organisms that cause nosocomial infections.^1,2 E. coli and K. pneumoniae are among the main strains associated with such infections. E. coli causes some major infections, including bacteremia, neonatal meningitis, urinary tract infections, intra-abdominal infections, sepsis and gastroenteritis in developing countries. K. pneumoniae causes early acquired lobar pneumonia, wound, soft tissue and urinary tract infections.³

Recently, the emergence of multidrug-resistant (MDR) phenotype as the consequence of antibiotics overconsumption in the treatment of human and animal diseases caused a global challenge to health systems. MDR was defined as acquired nonsusceptibility to at least one agent in three or more antimicrobial categories.^4–6

Carbapenems are the choice drugs to treat such infections. However, with increasing global incidence of carbapenem resistance, colistin is now widely used as the last resort antibiotic for the treatment of carbapenem-resistant Enterobacteriaceae. Among carbapenem-resistant K. pneumoniae isolates, the rate of resistance to colistin has been reported to be 15% which is very worrying.⁷

Colistin uses lipopolysaccharide (LPS) as the bacterial target and disrupts the negative charge of the outer membrane.⁸ Several studies proposed two main mechanisms for colistin resistance including LPS modifications and LPS removal. Other possible mechanisms are overexpression of efflux pumps, overproduction of capsule polysaccharide and production of colistinase.⁸

Resistance to colistin in gram-negative bacteria is attributed to a reduction in binding affinity to the outer membrane of LPS. Two-component systems (TCSs) including PhoPQ and PmrAB, as the regulatory systems, reduce the negative charge of lipid A and subsequently reduce the binding affinity of colistin to LPS.^9,10

Recently, a plasmid transmitted resistance has been reported as the mobilized colistin resistance (mcr), and has been designated as mcr-1, which is the most prevalent mcr type.¹¹ The mcr-1 gene is a phosphatidylethanolamine transferase, which adds this residue to lipid A component of the LPS, leading to a lower binding affinity of colistin to its target.¹² The mcr-2 gene is a rare variant of mcr. The locations of mcr-1 and mcr-2 genes are on conjugative plasmid which may accelerate the transmission between strains and bacterial species, leading to breakdown of the therapeutic regimen.^13–16 The reports on the role of mcr-1 and mcr-2 genes indicated the increase in the resistance of E. coli and K. pneumoniae isolates.⁷

Several studies have reported the variants such as mcr-3, mcr-4 and mcr-5 in E. coli and Salmonella isolates.¹⁷ Furthermore, a new plasmid-mediated colistin-resistance gene called mcr-7.1 has been identified in K. pneumoniae in China.¹⁸

There are different methods to evaluate the clonality and genetic diversity of Enterobacteriaceae such as pulsed-field gel electrophoresis, multilocus sequence typing, multiple-locus variable-number tandem repeat analysis and enterobacterial repetitive intergenic consensus sequences PCR (ERIC-PCR).¹⁹

In this study, we investigate the prevalence of mcr-1 and mcr-2 genes in colistin-resistant E. coli and K. pneumoniae clinical isolates. Also, the clonality of colistin-resistant E. coli and K. pneumoniae strains was evaluated by ERIC-PCR-based dendrogram. ERIC-PCR is cost-effective, easy to perform and a rapid method, which can be applied for comparison of clusters generated.²⁰

Materials and methods

Bacterial isolates

A total of 470 isolates of E. coli and K. pneumoniae were detected among 712 Enterobacteriaceae isolates. These samples were isolated from different clinical specimens such as blood, urine, wound, etc. except stool. These isolates were collected from the laboratories of teaching hospitals collaborated with Ahvaz Jundishapur University of Medical Sciences (Ahvaz, Iran) from January to June 2017. This study was approved by the Institutional Review Board (IRB) and Ethics Committee of the Ahvaz Jundishapur University of Medical Sciences. The isolates were transferred to the Medical Microbiology Department and were subcultured on blood agar and eosin methylene blue agar (EMB Agar) media (Merck, Germany) for double checking and taking pure culture. After incubation of plates at 37°C for 18–24 hrs, the isolated colonies were identified by Gram staining and standard biochemical tests such as triple-sugar iron (TSI), Simmonʼs citrate, sulfide indole motility (SIM), urea, ornithine and lysine decarboxylase tests (Merck).²¹

Antimicrobial susceptibility testing (AST)

AST was performed using disk diffusion method (Kirby–Bauer) on Mueller–Hinton agar (Merck) plates according to the Clinical and Laboratory Standards Institute (CLSI) Guidelines.²² The tested antibiotic disks were colistin (10 μg), ciprofloxacin (5 μg), tetracycline (30 μg), imipenem (10 μg), ceftazidime (30 μg), azithromycin (15 μg) and amikacin (30 μg) (MAST Co., UK). The phenotype of Enterobacteriaceae was defined as MDR according to the International Expert proposal for Interim Standards Guidelines.⁶

Minimum inhibitory concentration (MIC)

Colistin MICs were measured by E-test strips (Liofilchem, Italy) and interpreted based on European Committee on Antimicrobial Susceptibility Testing (EUCAST, Ver. 6, 2016) guidelines. After measuring the MIC with E-test method, the E. coli and K. pneumoniae isolates with a MIC higher than 2 μg/ml were considered as “resistant”.^23,24 E. coli ATCC 25,922 was used as a quality control for antimicrobial susceptibility testing.

DNA extraction and PCR amplification

DNA extraction was directly carried out by using the G-spin™ Total DNA Extraction Kit (iNtRON Biotechnology, South Korea) according to the manufacturer’s instructions. The concentrations of extracted DNA from E. coli and K. pneumoniae isolates were measured by using a nanodrop (Thermo Fisher Scientific, Waltham, MA, USA).

Detection of mcr-1 and mcr-2 genes by polymerase chain reaction (PCR)

DNA amplification was performed using a thermocycler (Eppendorf, Hamburg, Germany), and the set of primers as listed in Table 1.^25,26

Table 1 Sequences of primers used for detection of mcr-1 and mcr-2 genes and ERIC-PCR

The master mix was prepared in a final volume of 25 μL containing 10× PCR buffer, 50 mM MgCl₂, 10 mM dNTPs, 10 μM of each primer, 5 U/μL of Taq DNA polymerase and 5 μL of extracted DNA as template. The DNA amplification was performed based on the following program: initial denaturation at 94°C for 5 mins, 25 cycles of denaturation at 94°C for 1 min, annealing at 51°C for 30 s, extension at 72°C for 30 s and a final extension at 72°C for 5 mins. DNAs from E. coli KP81 and E. coli KP37 strains harboring mcr-1 and mcr-2, respectively, were used as the positive controls. Genomic DNA from colistin-susceptible E. coli ATCC 25,922 was used as the negative control. The PCR products were separated by electrophoresis on a 1.5% agarose gel containing 0.5 µg/ml ethidium bromide. The bands were visualized under UV light using a gel documentation system (Protein Simple, Santa Clara, CA, USA).²³ PCR products were sequenced, checked with BLAST and annotated to GenBank.

Genbank accession number

The partial nucleotide sequences of mcr-1 gene have been deposited in the GenBank database under accession number of MH627973 (mcr-1 from E. coli strain EC9).

DNA fingerprinting

ERIC-PCR was applied using the ERIC-1 and ERIC-2 primers for colistin-resistant E. coli and K. pneumoniae isolates (Table 1).²⁷ The ERIC-PCR conditions were regulated based on the published report by Smith et al²⁸. Analyses of the DNA fingerprints were performed by using GelCompar II; version 6.5 (Applied Maths, NV, Keistraat, Belgium). A cutoff value of 80% similarity was applied to define the clusters. The similarity between the profiles was evaluated with the band matching Dice coefficient, and dendrograms for each species were produced by the unweighted pair group method with arithmetic averages (UPGMA). Based on the UPGMA dendrograms, identical strains were defined as isolates with >97% similarity, closely related isolates with ≥95% similarity and isolates with <95% similarity as unrelated strains.

Statistical analysis

The results were analyzed by using SPSS version 22 software (IBM Armonk, North Castle, NY, USA). Statistical analyses were carried out by applying the Mann–Whitney, Chi-square and Kolmogorov–Smirnov tests, with a statistical significant P-value <0.05.

Results

In the present study, among 712 examined Enterobacteriaceae isolates, 470 isolates were identified as E. coli (n=351, 74.7%) and K. pneumoniae (n=119, 25.3%). The remaining 242 isolates belonged to other Enterobacteriaceae species which were not in the scope of this study. These 470 isolates were collected from different clinical specimens and were screened for antibiotic resistance and presence of mcr-1and mcr-2 genes. The distributions of E. coli and K. pneumoniae isolates in the specimens were as follows: urine, 87.4% (n=411); tracheal, 4.9% (n=23); blood, 3.4% (n=16); wound, 2.1% (n=10); discharge, 1.3% (n=6); and cord spinal fluid, 0.8% (n=4). It is important to mention that most isolates were collected from urine specimens.

Also, 55.95% (n=263) of isolates were from female, while 44.04% (n=207) of isolates were from male patients. The prevalence of E. coli and K. pneumoniae in different hospital wards is listed in Table 2. Outpatient clinic with 258 isolates (54.9%) and infectious ward with 2 isolates (0.4%) showed the highest and lowest rates, respectively.

Table 2 Prevalence of E. coli and K. pneumoniae in different hospital wards

On the basis of the results of AST, most isolates (81.3%) showed high resistance to ceftazidime. However, the most susceptibility among E. coli and K. pneumoniae isolates was attributed to colistin (81.5%). Antibiotics resistance patterns of E. coli and K. pneumoniae isolates which have been presented in Figure 1 show that 68.5% of the isolates are MDR phenotype.

Figure 1 Antimicrobial resistance pattern of E. coli and K. pneumoniae isolates.Abbreviations: CAZ, ceftazidime; T, tetracycline; AT, azithromycin; CIP, ciprofloxacin; IMI, imipenem; AK, amikacin; COL, colistin.

Among 470 isolates, 13.6% (n=64) were identified as colistin resistant by disk diffusion method. Among colistin-resistant isolates, 59.4% (n=38) were E. coli and 40.6% (n=26) were K. pneumoniae. These isolates were then examined by E-test to evaluate the MICs. By using the E-test, only 30 isolates were confirmed as colistin resistant. In this study, a significant difference was between MIC and disk diffusion methods (P=0.001).

Based on the PCR results, 1.7% (n=8) of E. coli and K. pneumoniae strains carried mcr-1 gene (Figure 2); however, none of the isolates harbored mcr-2 gene. All mcr-1 positive isolates were collected from urine specimens. There was not any statistically significant relationship between the presence of mcr genes with the aforementioned hospital wards (P=0.13) as well as with the gender of the patients (P=0.92).

Figure 2 Gel electrophoresis of mcr-1 gene (309 bp) encoding for colistin resistance. M: 100 bp DNA marker, line 1: positive control (E. coli KP81), line 2: negative control (E. coli 25,922), line 3: EC9, line 4: EC10, line 5: KP10, line 6: EC16, line 7: EC21, line 8: EC25, line 9: EC26, line 10: KP21, line 11: negative patient sample (E. coli isolates lack mcr-1 gene).

The majority of mcr-1-positive E. coli and K. pneumoniae isolates were resistant to colistin, with MIC values in the rates of >2 µg/ml. Two mcr-1-positive isolates were reported to be colistin susceptible by the E-test method but were resistant to colistin by the disk diffusion method.

The determination of genomic diversity of 38 strains of phenotypic colistin-resistance E. coli and 26 strains of K. pneumoniae was carried out using the ERIC-PCR fingerprinting method. By analyzing the ERIC-PCR profiles, the 38 E. coli isolates were categorized into 36 ERIC-types, including 14 main clusters and 34 singletons. These 36 distinct types were then called A01 to A36. Moreover, 26 K. pneumoniae isolates were categorized into 23 ERIC types (called B01 to B23), including 11 main clusters and 20 singletons (Figure 3). The results of ERIC-PCR showed that the colistin-resistant strains in this study were categorized into different genotypes. However, 2 ERIC-types named ERIC-type A05 and ERIC-type A23 had the same genetic background. Complex fingerprint patterns were obtained for all the strains. Generally, the electrophoretic analysis of the PCR reaction products has revealed that the number of bands in particular electrophoretic paths ranged from 3 to 15. The sizes of the PCR products ranged from 100 bp to about 1,500 bp.

Figure 3 ERIC-PCR-based dendrogram. The analysis of ERIC-PCR profiles was performed using the Dice coefficient with optimization set at 0.5% and tolerance at 0.5%. (A) ERIC-PCR profiles from 38 phenotypic colistin-resistance E. coli isolates (B) and 26 K. pneumoniae isolates.

Figure 3 (Continued).

Discussion

The spread of antibiotics resistance to a wide variety of antibiotics such as beta-lactams, aminoglycosides and carbapenems is a global challenge to the health systems. Using colistin is regarded as the last resort for treating infections caused by MDR-gram negative rods, especially Enterobacteriaceae.^23,29 However, its nephrotoxicity and neurotoxicity impacts have reduced its application as a routine prescribed drug.³⁰

In this study, ERIC-PCR typing method showed 36 patterns for 38 clinical E. coli isolates. 89.4% (n=34) of these isolates displayed a separate ERIC profile. The uniqueness of ERIC patterns suggests that they had a nonclonal transmission. However, 10.5% (n=4) of the isolates had identical patterns, indicating the similar origin of dissemination. In addition, ERIC-PCR typing method showed 23 patterns for 26 clinical K. pneumoniae isolates. Which 76.9% (n=20) of these isolates displayed a single ERIC profile, whereas, 23% (n=6) showed shared patterns. The isolates of the ERIC types, namely A05 and B08, had similar antibiotic resistance patterns. The ERIC type A23 dissemination with similar mcr-1 gene could be causes anxiety in society. Genetic analysis using ERIC-PCR showed that majority of colistin-resistant isolates were clonally unrelated, suggesting that dissemination of these isolates was not due to a clonal outbreak.³¹

The results of this study showed a low rate of resistance against colistin for E. coli and K. pneumoniae (10.8% and 21.8%, respectively). The isolates were generally susceptible to amikacin; however, most of them were resistant to ceftazidime and tetracycline. In addition to colistin, most of the mcr-1-positive isolates were resistant to ceftazidime (87.5%) and were variably resistant to tetracycline (62.5%), azithromycin (62.5%), imipenem (50%) and ciprofloxacin (25%). Consistently, high resistance to ceftazidime and tetracycline was reported in previous studies.³² The susceptibility rates of these isolates to amikacin were 100%.³³

In our study, 53.1% (n=34) of the colistin-resistant isolates tested by disk diffusion were susceptible when examined by E-test. This observation suggested that the specificity and sensitivity of the E-test were higher than the disk diffusion method. In the E-test method, 10% of the K. pneumoniae strains were resistant to colistin, which was consistent with the study conducted by Singh et al (10.5%).³⁴

The recent reports have shown plasmid harboring novel colistin-resistance genes, mcr-1 and mcr-2, in some isolates such as E. coli and K. pneumoniae.^25,26 An E. coli isolate containing mcr-1 gene which was sensitive to colistin has been before reported by Liassine et al.³⁵ Before Liassine et al, Pham et al reported the first mcr-1-positive but colistin-susceptible isolate, a Shigella sonnei.³⁶

Our PCR results showed that among 470 E. coli and K. pneumoniae isolates, 6 E. coli isolates (1.2%) and 2 K. pneumoniae isolates (0.4%) had mcr-1 gene. Consistently, Quan et al²⁹. reported that among 1,494 E. coli and 571 K. pneumoniae isolates, only 20 (1%) isolates and 1 (<1%) isolate were mcr-1 positive, respectively. The mcr-2 gene was not detected in our study, consistent with other studies.^37–39

The results of previous studies have shown the presence of mcr-1 gene in 7 countries in Southeast Asia (China, Thailand, Laos, Japan, Vietnam, Cambodia and Malaysia).⁴⁰ In the studies conducted in Asia, the isolation rate of mcr-1 gene is about 1%, which is consistent with our results (1.7%) (Table 3).^25,29,41

Table 3 Comparison of the previous studies with the current study

In this study, among 8 isolates of positive mcr-1 gene, 2 isolates were colistin sensitive (25%), which was in agreement with the Newton-Foot et al study (26%).²³ In both the studies, PCR was used as the molecular method which does not examine the gene expression. The identification of a colistin-susceptible/mcr-1-positive isolate in this study indicates that the silent spread of this gene might happen. The gene expression as well as the expression of the gene by real-time PCR, in samples with higher MIC, could demonstrate the inactivation of mcr-1 gene in some strains.²³ These studies show the importance of approving phenotypic methods by molecular methods.

Among 30 colistin-resistant isolates tested by E-test, only 26.6% (n=8) were harboring the plasmid-mediated colistin-resistance mcr-1 gene. Therefore, other possible resistance mechanisms should be considered. These mechanisms are as follows: 1) inactivation of genes encoding proteins involved in the LPS biosynthesis (lpxA, lpxC, lpxD), 2) LPS modification by the regulatory role of the PhoPQ and PmrAB TCSs and 3) mgrB gene alterations.⁴²

Continuous monitoring is very important for determining the exact frequency of mcr-1 gene among gram-negative bacteria in both human and veterinary medicine. Reevaluation of polymyxins application in animals and implementation of large regular screening of animal isolates for mcr genes would be an important step in preventing the spread of these genes to the human isolates.²⁵

Conclusion

Spreading of Enterobacteriaceae strains harboring mcr containing plasmids could fail the colistin-included therapy regimen which is used as a last line of treatment against MDR gram-negative bacterial infections. Increased surveillance of colistin-resistance mechanisms for monitoring their acquisition and spread is vital. Persistent efforts to ensure the judicious use of colistin (and indeed all antibiotics) both in agriculture and in health-care systems are welcome.