Carriage and within-host diversity of mcr-1.1-harbouring Escherichia coli from pregnant mothers: inter- and intra-mother transmission dynamics of mcr-1.1

ABSTRACT Exchange of antimicrobial resistance genes via mobile genetic elements occur in the gut which can be transferred from mother to neonate during birth. This study is the first to analyse transmissible colistin resistance gene, mcr, in pregnant mothers and neonates. Samples were collected from pregnant mothers (rectal) and septicaemic neonates (rectal and blood) and analysed for the presence of mcr, its transmissibility, genome diversity, and exchange of mcr between isolates within an individual and across different individuals (not necessarily mother–baby pairs). mcr-1.1 was detected in rectal samples of pregnant mothers (n = 10, 0.9%), but not in neonates. All mcr-positive mothers gave birth to healthy neonates from whom rectal specimen were not collected. Hence, the transmission of mcr between these mother-neonate pairs could not be studied. mcr-1.1 was noted only in Escherichia coli (phylogroup A & B1), and carried few resistance and virulence genes. Isolates belonged to diverse sequence types (n = 11) with two novel STs (ST12452, ST12455). mcr-1.1 was borne on conjugative IncHI2 bracketed between ISApl1 on Tn6630, and the plasmids exhibited similarities in sequences across the study isolates. Phylogenetic comparison showed that study isolates were related to mcr-positive isolates of animal origin from Southeast Asian countries. Spread of mcr-1.1 within this study occurred either via similar mcr-positive clones or similar mcr-bearing plasmids in mothers. Though this study could not build evidence for mother–baby transmission but the presence of such genes in the maternal specimen may enhance the chances of transmission to neonates.


Introduction
The human intestine accommodates a complex dynamic microbial community including antimicrobial-resistant Gram-negative bacteria harbouring resistance genes such as bla TEM , bla CTX-M , qnrS, etc. in both healthy and sick populations [1,2].The intestine allows for the inter-and intra-species exchange of antimicrobial resistance genes (ARGs) primarily via different mobile genetic elements (MGEs) [3].MGEs (plasmids, transposons, etc.) aid exchange of genes.Exchange of genes in the gut can occur due to antibiotic exposure or even without its use [3,4].However, the use of antibiotics exerts selective pressure that facilitates the overgrowth of resistant bacteria.
Early in life, the newborn acquires microbes from the mother and the immediate environment [3,5,6].Antibiotic consumption during pregnancy increases the risk of selection of antibiotic-resistant bacteria in the maternal microbiota which in turn may be passed on to the newborn during birth [2].The acquisition of resistance genes from the mother or the hospital environment can have serious consequences for the newborn particularly those who are preterm or low birth weight [7].Translocation of the bacteria, which includes those that are resistant, from the neonatal gut via the mysentric lymph node to the bloodstream can lead to sepsis [8].
Members of the Enterobacterales family such as Escherichia coli, commensals of the human gut, can serve as a common reservoir of different ARGs [3,9].They have been found to exchange antibiotic-resistant plasmids and virulence genes among themselves [3].In comparison to the Bacteroides or Firmicutes, E. coli are fewer in gut but have the ability to acquire and exchange genes with other commensal and transient bacteria [10].Several resistance genes (bla NDM , bla CTX-M , bla CMY , qnrS, etc.) conferring resistance to different groups of antibiotics have been noted in E. coli [3,11].
In 2015, plasmid-mediated colistin resistance gene, mcr-1, emerged in China and E. coli served as the reservoir of this gene [12,13].Other bacteria such as Salmonella sp., Shigella sp., Klebsiella sp., Enterobacter cloacae, etc. were also reported to harbour mcr genes [14].mcr-1 has been found in multiple plasmid types, viz.IncI2, IncHI2, IncX4, IncP, and IncF, signifying the efficient spread of this gene among various organisms and countries [15].Apart from livestock, poultry, and aquaculture, the occurrence of mcr genes has been detected in humans [14] both in isolates causing infection and in colonizers, with a recent report of mcr-1 gene in the gut of healthy human [15,16].The presence of transmissible mcr gene in the healthy human gut, particularly in pregnant mother might expose neonates to bacteria harbouring the gene [3].This would result in the emergence of resistance in neonates even in the absence of colistin exposure, ultimately leading to treatment failure.
Reports of mcr and their transmission dynamics in human gut is limited, more so in pregnant mothers and neonatal population.With the increased use of colistin as a last line resort and incidences of sepsis due to resistant bacteria, an understanding of colistin-resistant bacterial carriage and transmission in the gut is essential.Herein, we studied (i) carriage of mcr in the gut of mothers and sick newborns (not necessarily mother-baby pairs), and among neonatal blood isolates, (ii) inter-and intra-patient transmission dynamics of mcr and comparison of mcrbearing organisms with susceptible isolates from the same sample, and (iii) phylogenetic relationships among the study isolates and with similar isolates across different Southeast Asian countries.

Study design, collection of rectal swabs, and blood cultures at clinical site
In this study, samples collected during a multi-centric study named "Burden of Antibiotic Resistance in Neonates from Developing Societies (BAR-NARDS)" [2,17] have been used.BARNARDS study involved pregnant mothers and sick newborns who were admitted in IPGMER and SSKM Hospital, Kolkata.Rectal samples from all pregnant mothers but only sick neonates (no healthy neonates) were collected, so the mother-baby pairs were only present for sick neonates.Herein, we focused on analyzing carriage of the transmissible colistin resistance gene, mcr, and its transmission dynamics within the samples collected during the BARNARDS study.
During the study period (July to November 2017), women in labour or immediately postpartum were recruited, and rectal swabs were collected following consent.Neonates (inborn and outborn) when suspected with sepsis were enrolled, rectal swabs were collected and cultured isolates from blood were sent to laboratory.As a part of the BARNARDS study, socio-demographic information of the mothers and neonates were collected [17].A schematic representation of the study design has been depicted in Figure 1.

Processing of rectal swabs and blood cultures, and detection of transmissible colistin resistance gene, mcr
Processing of rectal swabs and blood culture have been described in Figure 1 and in supplementary methods.Briefly, rectal swabs were plated on vancomycin (10 mg/L) (MP Biomedicals, California, USA) supplemented chrome agar (CA) (BD BBL, MD, USA) and incubated (37°C, 18-24 h).Cultures from the primary inoculum of each plate and cultured isolates from neonatal blood were screened for mcr genes by polymerase chain reaction (PCR) as described previously [18].Sample positive for mcr genes were further enriched in Enterobacteriaceae Enrichment (EE) Mossel broth (37°C, 18-24 h) (BD BBL) and plated on to CA supplemented with/without colistin sulphate (2 mg/L) (MP Biomedicals).
Colonies with different colours were picked from plates with/ without colistin, and again screened for mcr genes.Amplified PCR products were Sanger sequenced and stocked for further analysis [12].Any mcr-negative colony of similar colour as the mcr-positive ones were collected from CA plates (without colistin) to check clonality with mcr-positive colonies.

Identification and susceptibility of mcrpositive/mcr-negative isolates
Identification and susceptibility testing of all mcrpositive/mcr-negative isolates were done by Vitek2 compact system (BioMérieux, Marcy l'Etoile, France).For minimum inhibitory concentration (MIC) of colistin, broth microdilution (BMD) was carried out using colistin sulphate, following CLSI guidelines.Results were interpreted according to CLSI (2020) [19].Molecular typing of mcr-positive/mcr-negative isolates More than one colony of same colour with mcr gene was stocked from each sample and were subjected to repetitive extragenic palindromic elements-PCR (rep-PCR) [20].Additionally, all mcr-negative isolates (of same colour as the mcr-positive ones) were subjected to rep-PCR to check for similar clones.Clonality among mcr-positive isolates were determined by pulsed-field gel electrophoresis (PFGE) in a CHEF-DRIII apparatus (Bio-Rad Laboratories, Hercules, and CA) using XbaI macro digestion and visually interpreted as per Tenovar criteria [21].With FP Quest software v4.5 (Biorad Laboratories Inc, Hercules, California, USA), a dendrogram was prepared using Dice coefficient and UPGMA (unweighted pair group method using arithmetic averages).Tolerance and optimization were set at 1.5% and isolates with ≥90% similarity were considered identical.The phylogenetic classification of the isolates was performed via phylogroup multiplex PCR as described previously [22].

Transmissibility of plasmid-mediated mcr
Transfer of mcr was carried out in Az r E. coli J53 (recipient) by solid-mating conjugation method.Transconjugants (TCs) were selected on Luria Bertani (LB) (BD BBL) agar plates supplemented with sodium azide (100 mg/L) (Sigma-Aldrich, St Louis, MO, USA) and colistin (2 mg/L).TCs retrieved were subjected to confirmation of mcr and other resistance genes by PCR, followed by assessment of colistin MIC by BMD.
Following this, plasmid typing was performed using PCR-based replicon typing (PBRT) [23] in both wild-type (WT) and TCs.
Whole genome sequencing (WGS) of mcrpositive/mcr-negative isolates Genomic DNA from mcr-positive/mcr-negative isolates was processed for paired-end WGS on an Illumina MiSeq.Selected isolates (based on genomic DNA quantity) were subject to additional long-read sequencing on a MinION (Oxford Nanopore Technology, UK) as described previously [24].Further analysis of resistance, virulence, plasmid types, and sequence types was carried out by different online pipelines (Supplementary methods).
Comparative phylogenetic analysis of mcrpositive/mcr-negative isolates A core genome alignment and maximum likelihood phylogenetic tree of isolates was performed using Roary (v3.12.0) and IQ-TREE (v2.0) with the GTR substitution model and 1000 bootstrap replicates.Phylogenetic trees were mid-rooted and annotated using iTOL(v5.7).Snippy (v4.6.0) was used to assess the number of single nucleotide polymorphisms (SNPs) between isolates of the same ST using an available hybrid genome as the local reference genome withmincov 10 applied and trimmed paired end fastq used as input (-R1 -R2).

Phylogenomic comparison of study isolates with mcr-1 from Southeast Asia
A literature search was conducted during July-August 2021, WGS data of mcr-bearing E. coli from Southeast Asia was retrieved and fastq were assembled into genomes as described [24].All genomes were screened for the presence of mcr-1.ClermonTyping was used to determine in silico phylogroups [26,27], Prokka (v1.14.5) was used to annotate the assemblies and Panaroo (v1.2.8; -clean-mode [moderate], -core parameters applied) was used to create the core genome alignment.IQ-TREE (v2.0) was used to generate the maximum likelihood phylogenetic tree with parameters as described previously [24].

Results
Identification of bacterial isolates with mcr mcr-1 was found in rectal samples of ten pregnant mothers (n = 10/1169), indicating a low prevalence (0.9%).These 10 mothers had healthy babies and hence the paired neonatal sample was not collected.None of the other sick neonates from whom rectal samples were collected harboured mcr-1.Sanger sequencing of mcr-1 revealed it to be mcr-1.1 and was found only in Escherichia coli.E. coli lacking mcr-1.1 from those samples that harboured mcr-positive isolates were also analysed.Fourteen mcr-1.1positiveE. coli and 25 mcr-negative E. coli were isolated from 10 mothers (Table 1).Few mothers had more than one isolate carrying mcr-1.1.Other colonies found in the sample were Klebsiella pneumoniae, Pseudomonas aeruginosa, and Acinetobacter baumannii, but none carried mcr-1.

Socio-demographic details of mothers possessing mcr-1.1-positive E. coli
Mothers who harboured mcr-1.1 gave birth to term, singleton, healthy babies, with no birth complications.Since, the babies were healthy, rectal swabs were not collected from these neonates and hence, not assessed for the presence of mcr-1.1.No mother received antibiotic treatment, nor visited traditional, or private healthcare or were hospitalized or travelled outside the local province prior to enrolment.Most mothers (n = 7) were from urban areas while three mothers were from rural areas.

Antibiotic susceptibility of E. coli isolates collected from mcr-1.1-positive maternal sample
Both mcr-positive and mcr-negative isolates were susceptible to most of the antibiotics tested but were resistant to ampicillin and trimethoprim-sulfamethoxazole (Table 1).Few were additionally resistant to cefuroxime, ceftriaxone, cefoperazone/sulbactam, cefepime, nalidixic acid, and ciprofloxacin.However, mcr-positive isolates were also resistant to colistin (MIC: 8-16 mg/L), while mcr-negative isolates were susceptible (0.015-2 mg/L) (Table 1).
Molecular typing of E. coli from mcr-1.1positive maternal sample rep-PCR showed 11 distinct band patterns in 14 mcrpositive isolates, while band patterns were all diverse in case of 25 mcr-negative isolates.All the distinct isolates were proceeded with WGS.
Carriage of virulence genes within mcr-positive and mcr-negative isolates was nearly same (Table 1).gad (glutamate decarboxylase) was prevalent among both groups of isolates, followed by terC (tellurite resistance gene).Presence of different serum and complement resistance genes were noted in many mcr-positive and few mcr-negative isolates (Table 1).However, IN-MR569R1 and few mcr-negative didn't show the presence of any virulence genes at all.Most of the study isolates belonged to phylogroup A (mcr-positive = 8 and mcr-negative = 11) and B1 (mcr-positive = 4 and mcr-negative = 10) (Table 1).

Transmissibility and characterization of mcr-1.1 plasmid
Conjugal transfer of mcr-1.1 was successful with cotransfer of bla TEM, and qnrS in various combinations among transconjugants (TCs).Colistin MIC of TCs and wild-type isolates were found to be similar (Table 1).
mcr-1.1 was carried on IncHI2 plasmid (∼216-240 kb), except for one (IN-MR569R1) which harboured mcr-1.1 in IncX4 (∼33 kb) (Table 1).Plasmid types found in TCs as evaluated by PBRT corroborated with WGS data.All IncHI2 plasmids belonged to pST4 (Table 1).Out of 14 mcr-positive isolates, 13 complete plasmids including the IncX4 plasmid were assembled from long-read sequencing.Two groups of IncHI2 plasmids were noted, ∼216 kb (plasmid 1, P1) found in multiple isolates collected from three samples and the other ∼240 kb (plasmid 2, P2) found in six different samples (Table 1, Figure 3a,b).All P1s (n = 7) when aligned against pCFSAN061771 (Egypt, accession: CP042898.1),were found to be identical to each other (Figure 3a) with an average nucleotide identity (ANI) of 93.07% to pCFSAN061771 (455 single nucleotide variants, SNVs; 8 insertions, and 5 deletions).P1 from IN-MR750R1 had an ANI >99.88% to pCFSAN061771 and was slightly larger (∼217 kb) (Figure 3a), showing some variations.Similarly, all P2s (n = 5) when aligned against a plasmid of an E. coli from Saudi Arabia  3b).mcr-1.1 in P1 and P2 was bracketed between ISAplI upstream and downstream within a composite transposon, Tn6330, with variations in orientation and truncation of ISApl1 in few isolates (Figure 4).Genetic environment of mcr-1.1 of study isolates when compared with global mcr-1, revealed a variation in the arrangement of the genes, that is, ISApl1-mcr-1.1-ISApl1-pap2instead of ISApl1-mcr-1.1-pap2-ISApl1(Figure 4).mcr-1.1 when present in IncX4 plasmid (∼33 kb) did not have ISAplI or other IS elements in the vicinity, rather IS26 was noted in the same plasmid (Figure 4).IncX4 plasmid when aligned with other previously described plasmid sequences, produced multiple identical hits on PLSDB with mcr-1 isolates from China, Laos, and Vietnam.

Transmission of mcr-1.1 within gut microbiome
Within the study population, intra-and inter-gut transmission of mcr-1.1 was studied.We compared isolates in terms of their typing patterns (PFGE), STs, SNPs, and mcr-1.1-bearingplasmids in individual or different maternal samples.We analysed transmission of mcr-1.1 from two aspects: (i) clonal spread, where mcr-positive isolates of same STs were isolated from different mothers, and (ii) plasmid-mediated spread, where mcr-carrying similar plasmids belonging to different STs were isolated from individual or different mothers (Figure 5).When assessing the clonal spread, mcr-negative isolates belonging to similar STs of mcr-positive isolates were also compared.
Isolates IN-MR674R1, IN-MR680R1, IN-MR683R1 collected from three mothers admitted within a time frame of 0-1 days (Figure 5) belonged to ST2705, phylogroup A, and shared >90% similarity (PFGE) (Table 1, Figure 2).These isolates clustered together in core genome phylogenetic tree and IN-MR674R1, IN-MR680R1 were 22 SNPs distant from each other (Figure 6).These results indicated a possible inter-mother transfer of clonal isolates with mcr-1.1.
To understand the transmission of mcr-1.1 through plasmids, we compared mcr-1.isolated from IN-MR727 also showed similarity with plasmids of E. coli from IN-MR725 and IN-MR750 (Table 1, Figures 3a, 4, and 5).These mothers were distantly related to each other as per core genome SNP phylogeny (Figure 6) but harboured a similar plasmid (Figure 5).Both horizontal transmission of mcr-1 through other E. coli/ bacteria among different mothers or independent acquisition of similar plasmids at different points of time beyond the hospital environment are possible.However, this cannot be definitively assessed by the findings of this study.In some mothers, more than one diverse E. coli isolate belonging to different STs were found -IN-MR725 (ST101, ST2178, ST9421) and IN-MR750 (ST394, ST515), exhibiting similar mcr-1.1-bearingIncHI2 plasmids (∼216 kb) with same genetic environment (Table 1, Figures 3a, 4, 5).This indicates a possible transmission of plasmid within an individual host (intra-mother transmission).Three other independent acquisitions of mcr-1.1bearingplasmid/ bacteria were noted, one with mcr-1.1 in IncX4 plasmid, and the other two with mcr-1.1 in IncHI2 having a genetic environment distinct from others that are discussed above (Figure 5).Few mcr-negative isolates identical to mcr-positive isolates in terms of STs and core genome were also isolated but they lacked IncHI2 and IncX4 (Table 1, Figures 5 and 6).
Southeast Asian mcr-1 collection harboured bla Study strains showed closeness with strains from various countries of Southeast Asia viz.Bangladesh, China, Thailand, Singapore, Vietnam, and Laos (Figure 7).Most of the study isolates showed resemblances with E. coli isolated from chicken (n = 5) or human carriage (n = 4) and few from pig and blow flies.This shows that mcr-positive isolates circulate among different origins (mainly food animals), emphasizing its presence within the food chain.

Discussion
The passage of antibiotic-resistant organisms from mother to the neonate is of concern.The pristine and immature gut of the neonate, particularly for those who are premature and low-birth weight, can allow the translocation of resistant organisms to the bloodstream leading to sepsis [28].mcr-1.1 was isolated from maternal rectal samples only, with very low prevalence (0.9%).No mcr was noted in gut samples or in cultured blood from neonates suspected with sepsis.Mothers with mcr-1.1 had healthy babies and since samples were not collected from healthy neonates, the presence of mcr in the healthy neonatal gut could not be assessed.Hence, the possibility of mother-to-baby transmission of mcr-harbouring E. coli could not be studied.
Carriage of mcr in this study is much lower than that reported from China (2.08-6.2%),Switzerland, and France, but higher than Netherlands (0.35%) [29].Colistin resistance is low in India and a study on neonatal blood isolates over a period of 12-years from the same unit also exhibited low prevalence of colistin resistance (2.8%) with no mcr genes [30].Such low resistance may be attributed to the limited use of colistin in clinical settings.Absence of selection pressure (colistin) probably restricted the spread of mcr-1 in clinical settings as studies from India documented fewer reports of mcr-1 from human clinical isolates compared to food items such as meat, poultry, and environmental isolates [31,32].Since 2019, restrictions on colistin usage in India within animal industries have attributed to low carriage of mcr-1 [32].
In India, mcr-1 was reported first in Klebsiella pneumoniae [33] and later in E. coli, Aeromonas sp., A. baumannii, etc. [34][35][36].Presence of different bacteria such as K. pneumoniae, P. aeruginosa, A. baumannii along with E. coli was found in the maternal samples.However, mcr-1.1 was found only in E. coli, implicating its restricted spread.
In this study, mcr-positive isolates belonged to phylogroup A and B1, were highly susceptible and carried few virulence genes.mcr-1-E.coli from Southeast Asian countries, also exhibited the prevalence of similar phylogroups.Though primarily restricted to the gut, A and B1 isolates do cause sepsis in debilitated or immunocompromised patients by translocating from gut to blood through the immature/compromised gut barrier [37].A and B1 E. coli being commensals, possess virulence genes necessary for colonization [27], and harbour different ARGs [38] which is also noted among the global mcr-1 isolates but study isolates exhibited low carriage of ARGs.
mcr-positive isolates were diverse and belonged to various STs as also observed in previous studies [39,40].Occurrence of mcr-1.1 among novel STs as found in this study, exhibited the emergence of new colistin-resistant clones.This highlighted the fact that the spread of mcr-1.1 is not through any particular clone or lineage.Southeast Asian mcr-1 isolates were also diverse irrespective of source of origin.Study isolates exhibited resemblances with isolates from Southeast Asian countries (Bangladesh, Vietnam, Singapore, Laos, etc.).mcr-1.1 was detected in IncHI2 and IncX4 replicons, though several other replicons were present in the mcr-positive study isolates.Plasmids of this study showed similarities with plasmids from Egypt and Saudi Arabia (IncHI2), and China, Laos, and Vietnam (IncX4), highlighting proficiency of these plasmids for the spread of mcr gene.Analysis of Southeast Asian mcr-1 isolates also exhibited the prevalence of these two plasmids harbouring mcr-1 gene.IncI2, IncHI2, IncP, and IncX4 are the predominant carriers of mcr-1 worldwide [27,39,40].The study isolates shared plasmid backbone similar to replicon type IncHI2.Some differences were noted due to the inversion/deletion of certain sequences.Association of IncHI2-pST4 with mcr-1.1 in the study isolates, corroborated with others [41,42], implying IncHI2-pST4 to be a dominant plasmid lineage contributing to the horizontal transfer of mcr-1.1.
IncHI2 and IncX4 plasmids were conjugative and associated with various MGEs such as ISApl1, Tn6330 (IncHI2), and IS26 (IncX4).IS26 being a hotspot for plasmid fusion has intensified the spread of mcr-1-harbouring IncX4 plasmids in the absence of ISApl1 or any transposon [43].ISApl1-mcr-1.1-pap2-ISApl1 is the known genetic background of the gene reported so far [44] but a different genetic environment: ISApl1-mcr-1.1-ISApl1-pap2has been detected within the study isolates which has not been reported previously.
Clonal spread of mcr-1.1-E.coli (ST2705) was observed in three mothers who were admitted in hospital during the same time highlighting the possibility of an inter-mother transmission of mcr-1.1-harbouringE. coli.Further, similar mcr-1.1-plasmidswere isolated from five maternal samples in diverse E. coli.Each isolate belonged to different STs, but their plasmid sequences revealed significant similarities.Prevalence of different clones with identical plasmids in several mothers indicate that the spread of mcr-1.1 is plasmid-mediated.However, this study could not definitively underline the fact whether the presence of similar plasmid within different mothers was due to horizontal transmission through bacteria in the hospital environment or independent acquisition of similar plasmids from other sources beyond the hospital environment.In contrast, the transmission of the plasmid in the gut is more evident in cases where similar mcr-1.1 plasmid has been isolated from different STs in individual mother.Other studies documented the spread of mcr-1 via plasmids occurring between diarrhoeal patients and from animals to human [39,40,45].We hypothesize that when more than one mcr-positive E. coli of distinct STs with similar plasmids were isolated from an individual mother (intramother transmission), transfer of mcr-1.1 via plasmids was more probable than an independent acquisition of two separate distinct mcr-possessing isolates.
Presence of mcr-1.1 in healthy pregnant mothers with no exposure to antibiotics might be due to the presence and persistence of mcr-1 gene in the food chain.In a recent study, mcr-1 has been found to enhance the commensal lifestyle of E. coli, which led to the maintenance of mcr-1-positive E. coli within the gut even in the absence of antibiotic pressure [46].Similarities of mcr-1-harbouring maternal study isolates with isolates from chicken (food animal) indicated the presence and circulation of mcr-positive isolates within the food chain.
In conclusion, studies highlighting the prevalence of mcr-1.1 in pregnant mothers are rare with no studies explaining within-host diversity of mcr-1.1harbouringE. coli.Presence of mcr-positive isolates with highly similar plasmids in the gut of healthy mothers (individual mother or among different mothers) indicated the involvement of plasmids.The only limitation of this study lies in the fact that transmission of mcr-1.1 from mothers to their respective neonate could not be studied as their babies were healthy and hence not included in the study.The presence of mcr-1.1 in susceptible E. coli of healthy individuals is worrisome since they remain undetected and may serve as a focal point for the spread of colistin resistance in community and in newborns from their colonized mothers.

Figure 1 .
Figure 1.Schematic flow diagram of the study plan, summarizing sampling/laboratory/sequencing workflows.

Figure 3 .
Figure 3. Alignment of mcr-1.1-bearingIncHI2 plasmid sequences of the study isolates.(a) ∼216 kb: IN-MR750R1 and IN-MR750R4 were identical, hence in this figure only IN-MR750R1 has been included.(b) ∼240 kb: Due to poor assembly issue, IN-MR364R1 was excluded from this analysis.Yellow ochre-coloured arrows: different genes, green arrow: mcr-1.1.Shaded regions: percentage similarities.

Figure 4 .
Figure 4. Genetic environment of mcr-1.1 found among study isolates.Genes and their corresponding transcription orientations are indicated by horizontal arrows.Grey shaded region: homology, light yellow shaded region: inversion.

Figure 6 .
Figure 6.A core genome phylogenetic tree summarizing the isolates with short-read WGS available from this study.Due to contamination and poor assembly issue, IN-MR364R1 and IN-MR683R1 were excluded from this analysis.

Figure 7 .
Figure 7. Core genome phylogenetic tree of mcr-1-E.coli collected from published studies in Southeast Asia including those from the present study.Sequence type (ST), green shading: study isolates.IN-MR364R1 and IN-MR683R1 were excluded due to contamination and poor assembly.

Table 1 .
Colistin susceptibility and genotypic characterization of mcr-1.1-positiveEscherichia coli, their transconjugants (TCs), and other susceptible Escherichia coli from the same rectal specimen.