Genetic Characterisation of Colistin Resistant Klebsiella pneumoniae Clinical Isolates From North India.

Background: Increasing use of colistin has led to the world-wide emergence of mobile colistin resistant gene (mcr). The present study aimed to identify and characterise mcr and other drug-resistant genes in colistin resistant Klebsiella pneumoniae clinical isolates.
Methods: Twenty-two colistin resistant K. pneumoniae were analysed for mcr and other drug-resistant genes, efflux pumps, and virulence genes, and for their biofilm forming ability. Pulsed-field gel electrophoresis (PFGE) and multi-locus sequence typing (MLST) were performed for all mcr-1 positive isolates. S1-PFGE and Southern hybridisation were performed for localisation of mcr-1 and bla NDM.
Results: Nineteen colistin resistant K. pneumoniae harboured mcr-1 and 3 had mgrB disruption. All isolates harboured bla OXA-48-type and ESBL genes; eight strains (five with mcr-1 and three with mgrB disruption) co-harboured bla NDM. Efflux pumps genes AcrAB and mdtK were detected in all 22 and tol-C in 21 isolates. Virulence-related genes entB and irp-1 were detected in all 22, mrkD in 20, and fimH-1 in 18 isolates; 11 isolates were strong biofilm producers. PFGE clustered mcr-1 positive isolates into eight groups based on ≥90% similarity; MLST revealed diverse sequence types, predominant being ST-15 (n = 4) and ST-16 (n = 4). Both mcr-1 and bla NDM were localised on plasmid and chromosome; mcr-1 was present on IncFII type and bla NDM on IncFIB and IncA/C type plasmids. Conclusions: Colistin resistance in K. pneumoniae was predominantly mediated by mcr-1. Co-existence of colistin, carbapenem, and other drug-resistant genes along with efflux pumps indicates towards enormous genomic plasticity in K. pneumoniae with ability to emerge as super-spreader of drug-resistance.

Introduction

Increasing prevalence of multi-drug resistant (MDR) Gram-negative bacteria (GNB) is a serious public health concern since they are susceptible only to few antibiotics (Laws et al., 2019). The World Health Organization (WHO) has listed carbapenem-resistant Klebsiella pneumoniae among the priority pathogen group as it poses great threat to human health (WHO, 2017). K. pneumoniae belongs to the Enterobacteriaceae family and is a common nosocomial pathogen responsible for significant morbidity and mortality. Virulence factors such as capsular polysaccharides, lipopolysaccharide (LPS), siderophores and adherence factors help K. pneumoniae to circumvent host immune response and increase its pathogenicity. Biofilm formation also plays a significant role in drug resistance and inflammation resulting in persistent infections (Navon-Venezia et al., 2017).

Colistin is the last resort drug of choice for treatment of lethal infections caused by carbapenem resistant GNB. Colistin is a cationic polypeptide antibiotic that binds to the negatively charged phosphate group of LPS of GNB, which results in disarrangement of cell membrane. Ultimately, there is a loss of cell membrane integrity resulting in increased permeability of the cell, leakage of cell contents, and finally cell lysis (Baron et al., 2016). The re-introduction of colistin in clinical practice has resulted in its increased reports of resistance in GNB. Resistance to colistin is either chromosomal or plasmid mediated. Mobile colistin resistant gene (mcr-1) located either on chromosome or on plasmid encodes phosphoethanolamine transferase. Since the first report of mcr-1 in late 2015, ten different mcr variants (mcr-1 to mcr-10) have been reported (Wang et al., 2020).

In this study, we investigated the presence of mcr in colistin resistant K. pneumoniae strains. Such strains were also examined for the presence of other drug-resistant genes and also for virulence and efflux pumps genes and for their ability to form biofilm. Analyses of clonal relatedness and strain typing were performed in mcr-1 positive isolates. Further, characterisation of plasmids harbouring both mcr-1 and bla NDM was also performed.

Materials and Methods

Bacterial Strains

The study was conducted at Sanjay Gandhi Postgraduate Institute of Medical Sciences (Lucknow, India), a 900 bed tertiary care referral hospital in North India. Twenty-two colistin resistant K. pneumoniae isolates recovered from various clinical samples like pus, blood, endotracheal aspirate, tissue, and sputum between October 2016 and March 2017 were included in the study. All the isolates were identified using biochemical tests and MALDI-TOF MS (BioMérieux, Marcy l’Étoile, France). Prior to testing, all the isolates were stored in brain heart infusion broth (Becton, Dickinson and Company, Sparks, USA) supplemented with 20% glycerol (Sigma-Aldrich, MO, USA) at −80°C.

Demographic and Clinical Data

Demographic and clinical data of patients were obtained from the hospital information system available in the hospital intranet.

Antimicrobial Susceptibility Testing

Minimum inhibitory concentrations (MICs) were determined by broth microdilution method (BMD) in cation adjusted Mueller–Hinton broth following Clinical and Laboratory Standards Institute (CLSI) guidelines except colistin for which European Committee on Antimicrobial Susceptibility Testing breakpoints were followed (CLSI, 2017; EUCAST, 2017). Isolates were considered MDR if they were resistant to at least one antibiotic of three different classes among those tested (cephalosporins, carbapenems, fluoroquinolones, aminoglycosides, and polymyxins) according to Magiorakos et al. (2012).

DNA Isolation, Detection of Antibiotic Resistance, Efflux Pump and Virulence Genes

DNA was extracted from overnight grown culture using Wizard Genomic DNA Purification Kit (Promega, WI, USA). Genomic DNA quality was measured by NanoDrop ND-2000 spectrophotometer (Thermo Fisher Scientific, DE, USA). The integrity of genomic DNA was analysed by agarose gel electrophoresis. The extracted DNA was stored at −20°C.

The presence of mcr genes (mcr-1 to mcr-8) was analysed by conventional PCR, and the amplified products were confirmed by sequencing. The mcr positive isolates were also examined for the presence of carbapenemases (bla IMP, bla KPC, bla NDM, bla VIM, and bla OXA-48 type), extended spectrum β-lactamases (ESBLs; bla TEM, bla SHV, and bla CTX-M), 16S rRNA methyltransferases (armA and rmtA-F). List of primers is given in .

Chromosomal mutations were analysed in isolates negative for mcr. Conventional PCR was performed using specific primers ( ) to detect mutations in mgrB, phoP/phoQ, pmrA, and pmrB. The PCR products were purified, Sanger sequenced, and analysed to determine the mutations responsible for colistin resistance.

K. pneumoniae strains positive for mcr were also screened by conventional PCR for the presence of genes encoding for multidrug efflux pump systems like ArcAB, TolC, and MdtK, and virulence determinants such as regulator of mucoid phenotype (rmpA), type 1 and type 3 adhesins (fimH-1 and mrkD), iron siderophores (aerobactin synthase, lucC), bacteriocin biosynthesis [enterobactin (entB), and yersiniabactin (irP-1)], and serum resistance-associated outer membrane lipoprotein (traT).

Capsular Typing

Capsular typing based on wzi gene sequence was done as reported previously (Brisse et al., 2013). The PCR products were Sanger sequenced, and wzi alleles were identified, and corresponding capsular polysaccharide types (KL-types) were determined by comparing our wzi sequences with those available on the Klebsiella PasteurMLST sequence definition database (https://bigsdb.pasteur.fr/).

Biofilm Assay

Biofilm assay was performed by O’Toole and Kolter’s protocol with little modification (O’toole and Kolter, 1998). Briefly, 1 µl of overnight grown culture was inoculated into 100 µl of fresh tryptone soya broth (TSB) in 96 well sterile flat bottom polystyrene plates. After overnight incubation at 37°C, the cultures in wells were discarded. The wells were washed gently with water followed by air drying for 15 min. Biomass was stained with 125 µl of 0.1% (w/v) crystal violet for 20 min. Plates were rinsed off, air dried, and the dye bound to adherent biomass was eluted with 30% acetic acid. Absorbance was measured using automated microplate reader (MultiskanGO, Thermo Scientific, MA, USA) at 570 nm. Tests were performed in triplicate, and results were averaged. The results were interpreted according to Stepanovic et al. (2000). K. pneumoniae ATCC strain, ATCC 700603 was used as positive control whereas E. coli K-12 was used as negative control.

Clonal Diversity and Strain Typing

Clonal diversity among 19 mcr-1 positive K. pneumoniae isolates was examined by pulsed field gel electrophoresis (PFGE) according to previously reported protocol (Qin et al., 2014). Banding patterns were analysed using BioNumerics software version 7.6 (Applied-Maths, Sint-Martens-Latem, Belgium). Salmonella serotype Branderup strain (H9812) digested with XbaI (Promega, WI, USA) was used as reference strain.

Multi-locus sequence type (MLST) of 19 mcr-1 positive K. pneumoniae isolates was analysed as described previously (Diancourt et al., 2005). The seven housekeeping genes were amplified and sequenced. The sequence type (ST) was assigned by determining the allele number for each of the housekeeping genes using the database maintained by Pasteur Institute at http://bigsdb.pasteur.fr/klebsiella/klebsiella.html/.

Conjugation Experiment and Plasmid Replicon Typing

Horizontal gene transfer ability of bla NDM and mcr-1 was determined using liquid mating assay for five K. pneumoniae isolates that harboured both mcr-1 and bla NDM. E. coli J53 was used as recipient strain, and transconjugants selection was performed on MacConkey agar plates containing meropenem (2 µg/ml) or colistin (1.0 µg/ml) as applicable and sodium azide (100 µg/ml). Transconjugants were tested for mcr-1 or bla NDM by PCR and antimicrobial susceptibility. Plasmid DNA was isolated from transconjugants following Kado and Liu method (Kado and Liu, 1981). PCR-based replicon typing (PBRT) was done to determine the plasmid incompatibility group (Carattoli et al., 2005).

S1-PFGE and Hybridisation

S1 PFGE and Southern hybridisation were performed for five strains that harboured both bla NDM and mcr-1. Bacterial DNA was prepared in agarose plugs, digested with S1 nuclease (Promega, WI, USA), and the linearised plasmid was then separated through PFGE. The gel was stained with ethidium bromide and transferred to nylon membrane (Hybond N, Amersham, UK) followed by hybridisation with digoxigenin labelled probes specific to mcr-1 or bla NDM. Probe labelling and signal detection were done by DIG DNA Labeling and Detection Kit (Roche Diagnostics, GmbH, Germany).

Results

Bacterial Isolates and Patient Details

Twenty-two colistin resistant K. pneumoniae isolates recovered from 22 (male 18) patients were analysed; 12 patients were from post-operative intensive care unit (ICU) and four from critical care medicine, three from nephrology, two from paediatric gastroenterology, and one from haematology wards. Most of the isolates were recovered from endotracheal aspirate (45.4%, 10/22), followed by blood (27.3%, 6/22) and sputum (9.1%, 2/22). All isolates except one were recovered after 48 h of admission. Among the 12 post-operative ICU patients, 66.7% (8/12) succumbed to their infection. Co-morbidities were present in 86.4% (19/22) of patients. Hypertension was present in 36.4% (8/22), followed by acute kidney injury (13.6%, 3/22), type-2 diabetes, and chronic liver disease in 9.1% (2/22) each. The clinical details of all patients are given in .

Table 1

Demographic and clinical features of patients infected with colistin resistant Klebsiella pneumoniae.

Isolate ID	Sex/Age	Specimen	Days from admission to isolation of CRkp	Ward	Diagnosis	Type of infection	Co-morbidity	Outcome
CRkp1	M/40	Intra-abdominal fluid	26	Critical Care Medicine (CCM)	Alcohol pancreatitis	Intra-abdominal	Alcoholic, smoker, recurrent pancreatitis	Recovered and discharged
CRkp2	M/10	Tissue	8	Pediatric short stay unit	Wilm’s tumour on chemotherapy	Gangrene in left leg	Nil	Left leg amputation, recovered and discharged
CRkp3	M/82	Endotracheal (ET) aspirate	31	Medical-ICU	Septic shock, LRTI	LRTI	Hypertension, CAD, AMI	Recovered and discharged
CRkp4	M/36	ET aspirate	9	Post-operative ICU	CLD with bilateral pneumonia and septic shock	LRTI	Nil	Death
CRkp5	M/61	ET aspirate	3	CCM	Acute febrile illness	LRTI	AKI, Acute liver failure	Death
CRkp6	M/72	Femoral catheter tip	18	Nephrology ward	Septic shock, renal failure	Infected catheter tip	Hypertension	Death
CRkp7	F/36	Sputum	5	Post-operative ICU	Acute severe pancreatitis	Pneumonia	T2DM	Recovered and discharged
CRkp8	F/41	ET aspirate	25	Nephrology Ward	CKD, LRTI, septic shock	LRTI	Hypertension, Anemia, CKD	Death
CRkp9	F/26	ET aspirate	28	Nephrology ICU	Post-partum AKI, MODS, septic shock	LRTI	AKI, Anaemia	Death
CRkp10	M/34	Purulent discharge from left calf	39	Haematology ward	ALL (B cell)	Soft tissue infection/abscess	Nil	Recovered and discharged
CRkp11	M/58	ET aspirate	30	Post-operative ICU	Gunshot injury (face), LRTI, pyogenic meningitis, septic shock	LRTI	Multiple myeloma	Death
CRkp12	F/53	ET aspirate	14	Post-operative ICU	MODS, septic shock	LRTI	RHD (MS), PAH, CVA	Death
CRkp13	M/2	Blood	23	Paediatric gastroenterology ward	Septic shock	CLABSI	Neonatal cholestasis, enterocholitis	Recovered and discharged
CRkp14	M/59	ET aspirate	3	Nephrology ward	Septic shock, LRTI	LRTI	Hypertension, T2DM, CKD	Death
CRkp15	M/39	Blood	18	Post-operative ICU	Severe acute pancreatitis, intra-abdominal sepsis, multi-organ failure	Blood stream infection	Hypertension, alcoholic	Death
CRkp16	M/30	Blood	10	Pulmonary medicine ICU	Septic shock	Blood stream infection	Alcoholic liver disease, disseminated TB	Recovered and discharged
CRkp17	M/36	Blood	14	Post-operative ICU	Pneumonitis, septic shock	Blood stream infection	CLD	Death
CRkp18	M/53	Blood	8	Post-operative ICU	Hepatic encephalopathy, MODS	Blood stream infection	CLD	Death
CRkp19	M/82	Sputum	35	Medical -ICU	Septic shock, LRTI	LRTI	Hypertension, CAD, AMI	Recovered and discharged
CRkp20	M/70	ET aspirate	1	CCM	Systemic hypertension, COPD with type 1 respiratory failure	Ventilator associated pneumonia	Systemic hypertension	Recovered and discharged
CRkp21	M/19	ET aspirate	3	CCM	Severe acute pancreatitis	Ventilator associated pneumonia	Non-oliguric AKI	Recovered and discharged
CRkp22	M/55	Blood	11	Post-operative ICU	Gastric Carcinoma	septic shock	Hypertension	Death

ALL, Acute lymphocytic leukemia; AKI, Acute Kidney Injury; AMI, Acute myocardial infarction; CKD, Chronic kidney disease; CLABSI, Central line associated blood stream infection; CLD, Chronic liver disease; CVA, Cerebrovascular accident; ICU, Intensive Care Unit; LRTI, Lower respiratory tract infection; MODS, Multi-organ dysfunction syndrome; PAH, Pulmonary arterial hypertension; RHD (MS), Rheumatic heart disease (mitral stenosis); T2DM, Type 2 diabetes mellitus.

The antimicrobial susceptibility profile showed that all the isolates were MDR as they were non-susceptible to at least one antibiotic from three or more antibiotics classes. All 22 isolates were resistant to carbapenems (imipenem and meropenem), 3rd generation cephalosporins (ceftazidime and ceftriaxone), monobactam (aztreonam), aminoglycoside (gentamicin), and fluoroquinolones (ciprofloxacin). The MIC values for colistin ranged from 8 to ≥512 mg/L. The antibiotic susceptibility results of 22 isolates are summarised in .

Table 2

Antimicrobial susceptibility profile and molecular characterisation of 22 colistin resistant Klebsiella pneumonia.

Isolate	MIC (mg/L)								Mechanism of colistin resistance	Resistance genes	Virulence genes	Genes coding for efflux pumps	Capsular type	Biofilm forming capacity
	IMI	MEM	CT	CAZ	CRO	AZT	GEN	CIP
CRkp1	8	16	≥512	≥512	≥512	≥512	≥512	512	mcr-1	blaOXA-48, blaCTX-M, blaTEM, blaSHV, blaVIM	mrkD, FimH-1, Ent B, Irp-1,	mdtK, tol-C, Acr-AB	KL155	Strongly adherent
CRkp2	8	8	8	≥512	≥512	≥512	≥512	512	mcr-1	blaOXA-48, blaCTX-M, blaTEM, blaSHV, blaVIM	mrkD, FimH-1, Ent B, Irp-1, traT	mdtK, Acr-AB	KL112	Weakly adherent
CRkp3	8	16	16	≥512	≥512	256	128	≥512	mcr-1	blaOXA-48, blaCTX-M, blaTEM, blaSHV, blaVIM	Ent B, Irp-1	mdtK, tol-C, Acr-AB	KL51	Moderately adherent
CRkp4	256	128	≥ 512	≥512	≥512	≥512	≥512	512	mcr-1	blaOXA-48, blaCTX-M, blaTEM, blaSHV, blaVIM	mrkD, FimH-1, Ent B, Irp-1	mdtK, tol-C, Acr-AB	KL10	Strongly adherent
CRkp5	16	32	8	≥512	≥512	128	512	128	mcr-1	blaOXA-48, blaCTX-M, blaTEM, blaSHV	mrkD, FimH-1, Ent B, Irp-1	mdtK, tol-C, Acr-AB	KL155	Moderately adherent
CRkp6	4	8	8	≥512	≥512	≥512	≥512	512	mcr-1	blaOXA-48, blaCTX-M, blaTEM, blaSHV, blaVIM	mrkD, FimH-1, Ent B, Irp-1, traT	mdtK, tol-C, Acr-AB	KL30	Strongly adherent
CRkp7	8	8	16	512	128	32	4	16	mcr-1	blaOXA-48, blaVIM	mrkD, FimH-1, Ent B, Irp-1	mdtK, tol-C, Acr-AB	KL2	Moderately adherent
CRkp8	32	16	256	≥512	≥512	≥512	≥512	512	mcr-1	blaOXA-48, blaCTX-M, blaSHV, blaVIM	mrkD, FimH-1, Ent B, Irp-1, traT	mdtK, tol-C, Acr-AB	KL15	Strongly adherent
CRkp9	16	32	32	256	≥512	≥512	128	128	mcr-1	blaOXA-48,blaSHV, blaVIM	mrkD, FimH-1, Ent B, Irp-1, traT	mdtK, tol-C, Acr-AB	KL30	Strongly adherent
CRkp10	32	32	8	≥512	256	128	256	256	mcr-1	blaOXA-48,blaCTX-M, blaSHV, blaVIM	mrkD, FimH-1, Ent B, Irp-1	mdtK, tol-C, Acr-AB	KL149	Moderately adherent
CRkp11	32	64	256	≥512	≥512	≥512	≥512	512	mcr-1	blaOXA-48, blaNDM, amp-c, blaCTX-M, rmtC	rmpA, Ent B, Irp-1, traT	mdtK, tol-C, Acr-AB	KL10	Strongly adherent
CRkp12	32	64	32	≥512	≥512	512	512	512	mcr-1	blaOXA-48, blaNDM, amp-c, blaCTX-M, blaSHV, rmtC	rmpA, mrkD, Ent B, FimH-1, Irp-1	mdtK, tol-C, Acr-AB	KL18	Strongly adherent
CRkp13	256	256	128	≥512	≥512	≥512	≥512	512	mcr-1	blaOXA-48, blaNDM, blaTEM, blaCTX-M, blaSHV, rmtB	mrkD, FimH-1, Ent B, Irp-1	mdtK, tol-C, Acr-AB	KL155	Strongly adherent
CRkp14	8	16	16	≥512	≥512	≥512	128	256	mcr-1	blaOXA-48, blaVIM, blaCTX-M, blaTEM, blaSHV	mrkD, FimH-1, Ent B, Irp-1	mdtK, tol-C, Acr-AB	KL2	Weakly adherent
CRkp15	32	64	16	≥512	≥512	≥512	≥512	512	mcr-1	blaOXA-48,blaVIM, blaCTX-M, blaTEM, blaSHV	mrkD, Ent B, Irp-1, traT	mdtK, tol-C, Acr-AB	KL149	Moderately adherent
CRkp16	32	16	64	≥512	≥512	≥512	≥512	512	mcr-1	blaOXA-48,blaCTX-M, blaTEM, blaSHV	mrkD, FimH-1, Ent B, Irp-1	mdtK, tol-C, Acr-AB	KL51	Weakly adherent
CRkp17	64	128	32	≥512	≥512	≥512	≥512	512	mcr-1	blaOXA-48, blaNDM, blaCTX-M, blaTEM, blaSHV, armA	mrkD, FimH-1, Ent B, Irp-1	mdtK, tol-C, Acr-AB	KL10	Strongly adherent
CRkp18	32	16	64	≥512	≥512	≥512	≥512	512	mcr-1	blaOXA-48, blaNDM, blaCTX-M, blaSHV, blaTEM	mrkD, FimH-1, Ent B, Irp-1	mdtK, tol-C, Acr-AB	KL30	Strongly adherent
CRkp19	4	8	64	≥512	≥512	≥512	≥512	512	mcr-1	blaOXA-48, blaCTX-M, blaTEM, blaSHV	Ent B, Irp-1	mdtK, tol-C, Acr-AB	KL2	Moderately adherent
CRkp20	≥512	8	16	64	64	64	128	32	mgrB	blaOXA-48, blaIMP, blaNDM, blaSHV, blaCTX-M, armA	mrkD, FimH-1, Ent B, Irp-1	mdtK, tol-C, Acr-AB	KL18	Moderately adherent
CRkp21	≥512	32	8	64	128	256	256	128	mgrB	blaOXA-48, blaVIM, blaNDM, blaSHV, blaTEM, blaCTX-M, armA	mrkD, Ent B, Irp-1	mdtK, tol-C, Acr-AB	KL30	Strongly adherent
CRkp22	512	128	4	64	128	128	4	8	mgrB	blaOXA-48, blaVIM, blaNDM, blaSHV, blaTEM, blaCTX-M, armA	mrkD, Ent B, Irp-1	mdtK, tol-C, Acr-AB	KL10	Moderately adherent

CT, Colistin; IMI, Imipenem; MEM, Meropenem; CAZ, Ceftazidime; CRO, Ceftriaxone; AZT, Aztreonam; GEN, Gentamicin; CIP, Ciprofloxacin.

PCR Based Detection of Resistant Genes

Nineteen (86.4%) of 22 colistin resistant isolates harboured mcr-1, and the remaining three (13.6%) had mgrB disruption. Eight (36.4%) strains harboured bla NDM; five and three of them were positive for mcr-1 and mgrB disruption respectively. All 22 isolates carried bla OXA-48-type gene; bla VIM was detected in 13 (59.0%) isolates and bla IMP in one (4.5%) isolate. Twenty (90.1%) isolates harboured both bla CTX-M and bla SHV, whereas bla TEM was detected in 15 (68.2%) isolates. 16S r-RNA methyl transferase was detected in seven (31.8%) isolates (armA in four, rmtB in one, and rmtC in two isolates). Distributions of resistance genes in different combinations are given in .

PCR amplification of mgrB in three K. pneumoniae isolates (CRkp20, CRkp21, and CRkp22) revealed a larger (~1000 bp) amplicon. Sequencing of the amplicons showed IS elements mediated mgrB disruption. The IS elements involved in mgrB disruption belonged to IS1-like (777 bp) in CRkp20 and IS5-like families, 1,066 bp and 1,196 bp in CRkp21 and CRkp22 respectively. None of the isolates had mutation in phoP/phoQ, pmrA, and pmrBgene.

Detection of Efflux Pump and Virulence Genes

All 22 colistin resistant strains harboured AcrAB, mdtK, and tol-C efflux pumps except one isolate that lacked tol-C ( ). Mucoid phenotype regulator, rmpA, was identified in two isolates. The siderophore associated genes, entB and irp-1, were present in all the isolates. Other virulence genes fimH-1, mrkD, and traT were detected in 16 (72.7%), 19 (86.4%), and six (27.3%) isolates respectively. The distribution of virulence genes is shown in .

Biofilm Forming Capacity

In-vitro biofilm forming ability assay indicated that all 22 isolates were biofilm producers; 11 (50%) were strong, eight (36.4%) were moderate, and three (13.6%) were weak biofilm producers ( ).

Wzi based capsular typing of colistin resistant K. pneumoniae indicated a high diversity as it predicted 10 different capsular polysaccharide serotypes (KL155 (n = 3), KL112 (n = 1), KL51 (n = 2), KL10 (n = 4), KL30 (n = 2), KL2 (n = 3), KL15 (n = 1), KL30 (n = 2), KL149 (n = 2), KL18 (n = 2).

Clonal Diversity and Molecular Typing

All 19 mcr-1 positive K. pneumoniae isolates were typeable by PFGE. The maximum and minimum genetic similarity observed between the isolates was 99 and 86.5% respectively ( ). Based on ≥90% similarity they were clustered into eight groups.

Figure 1

A dendrogram of the pulsed-field gel electrophoresis (PFGE) fingerprinting results and sequence types of 19 mcr-1 positive Klebsiella pneumoniae.

MLST analysis of 19 mcr-1 positive K. pneumoniae revealed eight different STs and their distributions were as follows: ST-15 (n = 4), ST-16 (n = 4), ST-231 (n = 3), and ST-147 (n = 3), ST-43 (n = 2) and one isolate each for ST-14, ST-11, and ST-23. The source of strains and their STs are shown in .

Conjugation experiments were performed for five mcr-1 positive K. pneumoniae, which also co-harboured bla NDM. PCR assay showed that mcr-1 was successfully transferred from four isolates (CRkp11, CRkp12, CRkp13, and CRkp17) by conjugation and failed to transfer in one isolate (CRkp18). All the transconjugants were phenotypically resistant to colistin but sensitive to imipenem and meropenem. PBRT showed transconjugants of CRkp11, CRkp12, and CRkp13 carried IncFII type plasmid, whilst transconjugants of CRkp17 carried an untypeable plasmid. Similarly, PCR assay showed that bla NDM was transferred successfully from all five isolates by conjugation, and PBRT results showed that transconjugants of CRkp11 and CRkp12 carried IncA/C type plasmid whilst transconjugants of CRkp13, CRkp17 and CRkp18 carried IncFIB type plasmid. Phenotypically, all bla NDM transconjugants were resistant to imipenem and meropenem but susceptible to colistin.

S1 PFGE and Southern Hybridisation

S1 PFGE followed by Southern hybridisation showed that mcr-1 was present both on plasmid and chromosome in three isolates (CRkp11, CRkp12, andCRkp13), whilst one each only on plasmid (CRkp17) and chromosome (CRkp18). The plasmid size in CRkp11, CRkp12, and CRkp17 was between ~138 and ~210 kb, whilst in CRkp13, mcr-1 was present on a small plasmid between ~33 and ~78 kb ( ). The S1 nuclease digested genomic DNA from five K. pneumoniae was also probed with digoxigenin labelled bla NDM, and the results showed that bla NDM gene was present both on plasmid (between ~45 and ~400 kb) and chromosome in all five isolates ( ).

GenBank Accession Numbers

The GenBank accession numbers assigned to nucleotide sequences of mcr-1 were MN652072-MN652090 and for nucleotide sequence of mgrB were MW389562–MW389564.

Discussion

The extensive use of antibiotics for treating infectious diseases has led to the emergence of bacterial antimicrobial resistance. The microbes have benefitted enormously from overuse of antibiotics in clinical practice, also in agricultural and livestock. Emergence and dissemination of transmissible colistin resistance have severely compromised the use of colistin for treatment of infections caused by carbapenem resistant Enterobacteriaceae. In studies reported across the world, mcr-1 has been predominantly reported in E. coli whereas K. pneumoniae accounts for less than 5% of the total mcr positive isolates to date (Sun et al., 2018; Nang et al., 2019). In contrast to global data, studies from India indicate that colistin resistance is more common in K. pneumoniae than in any other bacterial species (Pragasam et al., 2016; Singh et al., 2018; Sodhi et al., 2020). Also, very few studies are available on genomic characterisation of colistin resistant isolates. Hence, we investigated the mechanism of colistin resistance in clinical K. pneumoniae isolates and performed genetic characterisation of these isolates to expand our knowledge on colistin resistant K. pneumoniae.

mcr mediated colistin resistance has been reported across the world, but only few such reports are available from India (Singh et al., 2018; Gogry et al., 2019). We found mcr-1 mediated colistin resistance in 19 K. pneumoniae isolates, whilst insertional inactivation of mgrB gene by IS elements in three isolates. Insertional inactivation of mgrB activates the PhoP/Q two component signalling system that upregulates the arnBCADTEF operon which adds 4-amino-4-deoxy-L-arabinose to lipid A resulting in colistin resistance (Cannatelli et al., 2014). Insertional sequences of IS1and IS5 family are most common IS elements responsible for inactivation of mgrB gene (Azam et al., 2021). It is noteworthy to mention that we found coexistence of mcr-1 and bla NDM in five K. pneumoniae isolates; however studies suggest they are more commonly found in E. coli as compared to K. pneumoniae (Delgado-Blas et al., 2016; Zheng et al., 2017). Among carbapenemases, bla OXA-48 was found to be present in all the isolates. In recent years, bla OXA-48 has increasingly been reported from India; a multi-centric study from India reported the presence of bla OXA-48 in 80% of the carbapenem resistant isolates (Shankar et al., 2019a). We also observed 39% of our carbapenem resistant isolates were bla OXA-48 producers (unpublished data). Among MBLs, bla VIM was present in 59.1% (13/22) isolates. The unusually high prevalence of bla VIM (50% of bla NDM positive isolates) was also reported previously from our centre (Rahman et al., 2018). Another study from North India reported bla VIM in 18.4% (52/282) of carbapenem resistant isolates (Garg et al., 2019). Aminoglycosides in combination with other antibiotics such as tigecycline are often used for treating infections caused by carbapenem and colistin resistant K. pneumoniae (Petrosillo et al., 2019). In the current study, 16S RNA methyltransferase genes were found to be present in seven isolates that also harboured mcr-1 and bla NDM, which indicates towards a grim situation. Among twenty-two patients, twelve (54.5%) succumbed to their disease. We found that the patient death as outcome was attributed to lower respiratory tract infection, blood stream infections, and septic shock caused by MDR K. pneumoniae.

The resistance nodulation division acrAB-tolC efflux pumps are reported in diverse members of the Enterobacteriaceae family including K. pneumoniae. In K. pneumoniae acrAB-tolC efflux pumps have been associated with resistance to β-lactams, fluoroquinolones, and tetracycline (Li et al., 2015). Similarly, Multi-Antimicrobial Extrusion mdtK efflux pumps have been also been reported in K. pneumonia (Li et al., 2015). In the present study, we detected acrAB, tolc, and mdtK in MDR K. pneumoniae. Our results are in concordance with previous studies where authors had shown the presence of drug-resistant genes and efflux pumps in MDR K. pneumoniae (Maurya et al., 2019; Ni et al., 2020).

The role of virulence factors in colonisation, invasion, and pathogenicity of K. pneumoniae is well known (Paczosa and Mecsas, 2016). Mucoid regulator gene, rmpA, is involved in capsule biosynthesis and often associated with hypervirulence was detected in two K. pneumoniae isolates (Cheng et al., 2010). The other important virulence factors are siderophores; they are low molecular weight iron scavenging molecules secreted by many GNB that affect the iron homeostasis in host (Page, 2019). In this study, all the colistin resistant K. pneumoniae harboured ent B and irp-1 siderophores, which are also known to contribute towards inflammation and bacterial spread during infection (Holden et al., 2016). Adhesin associated genes fimH, a type 1 fimbria adhesive subunit and mrkD, a type 3 adhesive subunit have been detected in 72.7 and 86.4 isolates respectively. mrkD is known to facilitate binding to extracellular matrix which is responsible for bacterial adherence to tissue and indwelling devices such as endotracheal tubes (Paczosa and Mecsas, 2016). Serum resistant outer membrane lipoprotein (traT) was detected in 27.3% isolates and reported to play a crucial role in bacterial pathogenesis by blocking the action of membrane attack complex (Miajlovic and Smith, 2014). K. pneumoniae is known to produce biofilm which provides a layer of protection by preventing antibiotic penetration and reducing their efficacy. In our study all the isolates were biofilm producer with 50% of them producing strong biofilm, which suggests that MDR K. pneumoniae strains are associated with biofilm production.

PFGE is considered gold standard for molecular epidemiology of bacterial strains. PFGE data indicated that the clonal spread of K. pneumoniae was not responsible for colistin resistance. The isolates having more than 90% similarity most often were of same ST except in few cases where isolates of same STs clustered separately. Further, the MLST data showed that ST-15 and ST-16 were the most dominant clones followed by ST-231 and ST-147 amongst the mcr-1 positive K. pneumoniae. We found that ST15 K. pneumoniae isolate was associated with the presence of rmpA gene. Out of four ST-15 isolates, three harboured bla NDM and 16S rRNA methyltransferase [rmtB (n = 1) and rmtC (n = 2)]. All ST15 K. pneumoniae were associated with strong biofilm production, whilst the other dominant clone ST16 K. pneumoniae was moderate biofilm producers. Previous study from India also supports our data where authors had detected ST-231, ST-14, ST-147, ST-15, ST-16, ST-11, ST-23, and ST-43 in colistin resistant K. pneumoniae (Shankar et al., 2019b), whereas global data suggests the presence of heterogeneous STs in mcr-1 producing K. pneumoniae. The diversity in PFGE and ST was also supported by capsular serotyping which predicted eight serotypes based on wzi allele sequence. KL10 was the most common capsular serotype detected; in mcr-1 producing K. pneumoniae, KL10 capsulate serotype was associated with entB and irp-1 siderophores along with strong biofilm forming ability.

Conjugation experiments revealed that in four out of five K. pneumoniae isolates, mcr-1 was present on conjugative plasmid. Conjugative plasmids are self-transmissible and are often responsible for rapid spread of resistant traits. Three of the four mcr-1 transconjugants had IncFII type plasmid, which are conjugative plasmid with low copy number and size ranging between 45 and 200 kb (Rozwandowicz et al., 2018). The role of IncFII type plasmid in dissemination of mcr-1 is well known (Xavier et al., 2016; Wang et al., 2018). The mcr-1 harbouring IncFII plasmids were associated with ST15 K. pneumoniae. Conjugation experiments in the above five K. pneumoniae showed successful transfer of bla NDM to recipient E. coli J53 that suggests their location on conjugative plasmid. In two transconjugants bla NDM was present in IncA/C type plasmid whereas in three transconjugants bla NDM was present in IncFIB type plasmid. IncA/C type plasmids are broad host range, low copy number, and frequently found to be responsible for dissemination of bla NDM. Similarly, previous studies had shown that dissemination of bla NDM was linked to transferable IncA/C and IncFIB plasmids (Khan et al., 2017; Sugawara et al., 2019). IncF are considered as epidemic plasmids and linked with the global spread of K. pneumoniae ST258 (Rozwandowicz et al., 2018). The presence of multiple plasmids in MDR strains imparts fitness cost; however, it provides bacteria specific traits which help them to survive in stress conditions. S1-PFGE showed that majority of K. pneumoniae isolates harboured multiple plasmids. mcr-1 was present on plasmid of different sizes in these isolates. In three isolates, mcr-1 was present both on plasmid and chromosome. The chromosomal integration stabilises mcr-1 and enables it to be vertically transferred without the risk of plasmid loss. Co-existence of transferable bla NDM along with mcr-1 is a major threat to human health by compromising the available treatment options. Previous studies from USA, China, and Vietnam also reported the coexistence of mcr-1 and bla NDM in various members of Enterobacteriaceae and their potential to spread as extensively drug-resistant strains (Mediavilla et al., 2016; Feng et al., 2018; Jin et al., 2018).

In conclusion, K. pneumoniae has emerged as the most notorious pathogen among the members of Enterobacteriaceae. They are the reservoirs of diverse resistant traits and virulence genes. Moreover, their biofilm forming ability provides them survival and colonisation advantages. Co-existence of mcr-1 and bla NDM on the transmissible plasmids is a matter of concern as such plasmids possess significant risk of inter- and intra-species dissemination in the environmental and livestock pathogens. Therefore, strict epidemiological surveillance, infection control measures, and antibiotic stewardship are required to curb this menace of colistin resistance from dissemination.

Data Availability Statement

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found below: https://www.ncbi.nlm.nih.gov/genbank/, MN652072-MN652090 https://www.ncbi.nlm.nih.gov/genbank/, MW389562-MW389564.

Ethics Statement

The studies involving human participants were reviewed and approved by the Institutional ethics committee of Sanjay Gandhi Postgraduate Institute of Medical Sciences, Lucknow, India [2017-191-PhD-99(B)]. The patients/participants provided their written informed consent to participate in this study.

Author Contributions

KP conceptualized and supervised the study. SS collected the sample, performed experiments, and drafted the manuscript. AP, MR, and AS performed the experiments and edited the manuscript. SN and CS collected the patient information and provided the demographic data. All authors contributed to the article and approved the submitted version.

Funding

This study was supported by Science and Engineering Research Board (SERB) (EMR/2015/001804), Government of India.

Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.