Discovery of mcr-1 Harboring Incl2 Plasmids from Clinical Isolates of Multiclonal E. Coli prevalent in Pakistan.

Background: Colistin is the last resort antibiotic against multiple drug-resistant (MDR) bacteria found in clinical infections; however, the emergence of plasmid-mediated mcr-1 gene annulled the ecacy of Colistin. This study was planned to determine the prevalence of mcr genes in clinical isolates collected in Pakistan. The molecular types and plasmids of mcr-1 bearing isolates were analysed. Methods: A total of 545 E. coli isolates collected from two major cities of Pakistan were screened for colistin-resistance and mcr genes from June 2018 to September 2019. All positive strains were subjected to antimicrobial susceptibility testing, ESBL, and MBL detection via DDST and CDT. ESBL genes detection, molecular typing, conjugation experiment, plasmid replicon typing, S1 PFGE, and southern hybridization were performed. Results: Four (0.73%) strains of mcr-1 positive isolates were susceptible to meropenem, fosfomycin, and chloramphenicol, including one that showed moderate level resistance to chloramphenicol and fosfomycin. All four strains were ESBL positive and harbored the blaCTX-M-15 gene, while three of the isolates also harbored the blaTEM-1 gene. Molecular typing revealed that four isolates belonged to diverse clonal types, mostly (75%) from avian pathogenic E. coli lineage. The mcr-1 gene was present on ~ 60 kb Incl2 plasmid, which was successfully transconjugated. Conclusion: The mcr-1 gene's detection in diverse clonal types from MDR clinical isolates enforces priority basis large scale surveillance studies followed by corrective actions to prevent the spread of the mcr-1 gene in clinical, poultry, and environmental setting. initiated with the screening of 545 E. coli isolates for colistin resistance. All isolates were collected from Microbiology laboratories of PIMS Islamabad (n= 260), HMC (n= 105), RMI (n=94), and KTH (n= 86) Peshawar - tertiary care hospitals in Pakistan from June 2018 to September 2019. The isolates obtained from human clinical samples included urine (n= 345), blood (n= 109), stool (n= 18), and pus (n= 73). The growth capacity of isolates was initially assessed on CLED and


Background
Antibiotic resistance is a serious concern worldwide. The situation is escalating due to the misuse and overuse of antibiotics.
Different antibiotic resistance mechanisms are developed by bacteria, among which the ESBL and MBL production are the most prominent. Most of the MDR bacteria show resistance to more than one drug [1]. Colistin, a cationic polypeptide antibiotic, was considered the last drug of choice against these superbugs [2]. Various chromosomal mediated colistin resistance such as the two-component system pmrAB, phoPQ, and regulator mgrB have to be known in Enterobacteriaceae that alter the lipopolysaccharide (LPS) of the bacterial cell wall, thus losing its a nity of attachment toward polymyxins [3]. However, it was not as high threatening due to their mode of vertical transmission [4]. In November 2015, Liu et al. for the rst time discovered a plasmid-mediated mobilized colistin resistance gene from E. coli and Klebsiella Pneumoniae. The gene product belongs to phosphoethanolamine (PEtN) transferases enzymes that add PEtN to the phosphate group of lipid A. This results in the loss of Colistin binding ability to LPS as lipid A more cationic [5]. Later, the colistin resistance mcr-1 gene was discovered in 47 countries [6]. It was initially discovered on Incl2 type of plasmid, but later on, it was found on other plasmid incompatibility types from different bacterial isolates collected from humans, animals, food, environment, insect, and water. Besides, other variants of mcr such as mcr-2, mcr-3, mcr-4, mcr-5, mcr-6, mcr-7, mcr-8, and mcr-9 have been reported from different regions of the world [7].
In Pakistan, mcr-1 harboring E. coli was detected in migratory birds and later on isolated from clinical samples and broiler in Faisalabad City of the Punjab Province of Pakistan [8][9][10][11][12]. In this study, we look for mcr genes (mcr-1 to mcr-5) in 545 E. coli isolates from human samples collected in two major cities Islamabad and Peshawar of Pakistan. We have determined the Antibiotic resistance pro le, ESBL genes, ST, plasmid Incompatibility type, and genetic context of mcr-1 harboring isolates.

Methods
Sample collection, isolation, and species con rmation: The present study was initiated with the screening of 545 E. coli isolates for colistin resistance. All isolates were collected from Microbiology laboratories of PIMS Islamabad (n= 260), HMC (n= 105), RMI (n=94), and KTH (n= 86) Peshawar -tertiary care hospitals in Pakistan from June 2018 to September 2019. The isolates obtained from human clinical samples included urine (n= 345), blood (n= 109), stool (n= 18), and pus (n= 73). The growth capacity of isolates was initially assessed on CLED and MacConkey agar. A single colony from each Petri plate was stored in LB media for maintaining strains. Species con rmation of E. coli was performed with 16S rDNA PCR using speci c primer mentioned in Table S1. The PCR products were sequenced, and species con rmation was performed using EZ Biocloud online software (https://www.ezbiocloud.net/identify).
Phenotypic and molecular detection of colistin resistance: Phenotypic detection of colistin resistance was performed applying the broth microdilution method. The Minimum inhibitory concentration results were interpreted according to CLSI guidelines [13]. The genomic DNA from resistant isolates was extracted with the conventional boiling method [14]. The DNA samples were subjected to PCR using mcr genes-speci c primers mentioned in Table S1. The expected size amplicons were visualized on 1% Agarose gel, which was sequenced and analysed with BLAST (https://blast.ncbi.nlm.nih.gov/Blast.cgi) Antimicrobial susceptibility testing: The mcr-1 harboring isolates were subjected to antimicrobial susceptibility testing using the broth microdilution method. The antibiotics used were Ampicillin, Cefotaxime, Chloramphenicol, Cipro oxacin, Fosfomycin, Cefoxitin, Gentamycin, Aztreonam, Amikacin, Meropenem, and Tetracycline. ESBL and MBL activity of resistant isolates were also determined by the Double Desk Synergy Test (DDST) and combined desk test (CDT), respectively. The results were interpreted according to CLSI guidelines [13].

PCR detection of ESBL genes:
ESBL genes blaTEM, blaSHV, blaCTX-M, and blaOXA variants were ampli ed from the genomic DNA of colistin-resistant isolates with ESBL genes-speci c primers mentioned in Table S1. The expected sizes were visualized on 1% Agarose gel. For further con rmation, PCR products of ESBL genes were sequenced and blast via the NCBI blast tool (https://blast.ncbi.nlm.nih.gov/Blast.cgi).

Molecular typing:
The multilocus sequence typing of four mcr-1 harboring isolates was carried out. The sequence result of 7 alleles, adk, fumC, gyrB, icd, mdh, purA, and recA, were determined from the MLST database (http://enterobase.warwick.ac.uk/species/index/ecoli). To further determine the genetic link among mcr-1 harboring isolates, the Xba l PFGE was performed according to PulseNet PFGE protocol [15]. DNA ngerprint analysis and sketch drawing were performed by BioNumerics v.8.0 (Applied Maths, Sint-Martens-Latem, Belgium). Clusters were examined by the unweight pair-group method with arithmetic mean (UPGMA) analysis. The Dice similarity coe cient was calculated with a position tolerance of 1.5%, and a dendrogram was created based on the UPGMA.

Transconjugation and PCR based replicon typing:
In order to detect the transferability of four colistin-resistant isolates, trans-conjugation experiments were performed. These isolates were selected as the donor, and E. coli EC 600 (Nal R , Rif R ) were taken as recipients. The experiment was performed, as described previously [16]. The transconjugants were analysed using antibiotic susceptibility testing and mcr-1 speci c PCR, as described earlier. The conjugation transfer rates were calculated by dividing the number of transconjugants by the number of donors.
To further investigate the plasmid incompatibility type responsible for mcr-1, the plasmid DNA of transconjugants was extracted by the alkaline lysis method [17]. The plasmid type was determined by using the PBRT 2.0 kit (MBK0078, Diatheva, Italy). The PCR products of eight multiplex PCR were visualized on 2% Agarose gel, and the results were analysed according to the manufactured instructions.
Page 4/13 S1 PFGE and southern blotting: S1 PFGE and southern hybridization of the successfully trans-conjugant E. coli EC600 were performed to determine the plasmid location. For S1 PFGE, the bacterial isolates were embedded into 1% low melting agarose plugs. Following cell lysis and washing, the plugs were pre-incubated in 200µl of 1x S1 buffer at 37° c for 30 minutes. S1 digestion was performed by ten units of S1 enzymes (Thermo scienti c) in 200 µl of 1x S1 buffer for 10 min at 37° c. The digestion was stopped by removing enzymes and adding 200 µl of 0.5M EDTA and left for 10 min at room temperature. Before running on the gel, the plug was incubated in 200 µl of TE buffer for 30 minutes. The products were electrophoresed on the CHEF mapper PFGE system (Bio-Rad USA) for 21 h at 6 V/cm with initial switch time 2.5 sec and nal switch time 60 sec. The plasmids from the gel were transferred onto the nylon membrane via the capillary transfer technique [18]. Southern hybridization of plasmid DNA with a digoxin-labelled mcr-1-speci c probe was performed according to kit instructions (Roche Diagnostics, Mannheim, Germany).
PCR mapping of mcr-1 genetic context: To explore the genetic context of mcr-1, whether pHNSHP45 [5] like key genetic component surrounds it, we designed seven pairs of primers targeting ParA, nikB, tnpA, mcr-1, hp, pilN, and vird4 genes. The plasmid DNA for PCR mapping was extracted by the alkaline lysis method and quanti ed by one drop [17]. The primers used for mapping are mentioned in Table S1. The PCR products were visualized on 1 % Agarose gel and subsequently sequenced and blasted via the NCBI blast tool.

Results
Bacterial isolation and colistin resistance screening: Among the 545 E. coli isolates, four isolates (PKE051, PKE141, PKE196, and PKE211) showed resistance to Colistin on the broth microdilution method. Upon molecular con rmation through PCR, all the four isolates harbored mcr-1 genes, while the other mcr genes were absent. Data about the isolates, location, patients, demography, samples, source, and colistin MIC values are presented in Table 1. Antibiotic susceptibility testing and resistant genes: All colistin-resistant isolates were MDR but were susceptible to meropenem and were MBL negative. Three isolates showed susceptibility to Chloramphenicol, Cipro oxacin, and Fosfomycin, while one among them was susceptible to Tetracycline. All of the isolates were ESBL positive via DDST. The blaCTXM-15 gene was found in all of the isolates, while three isolates had the blaTEM-1 gene. The complete antibiogram data are presented in Table 2.

Transconjugation and PBRT:
Assuming that the mcr-1 gene is present on the plasmid, a transconjugation experiment was performed. All of the strains were successfully conjugated with E. coli EC 600. The conjugation rates are presented in Table 3. Single colony from the selected plates were analysed for antimicrobial susceptibility testing and resistant genes. The transconjugants showed resistance to Colistin, and the presence of the mcr-1 gene was con rmed via PCR, while the ESBL genes were not detected in transconjugants.
To further determine the Incompatibility type of mcr-1 harboring plasmid, the PBRT was performed from the transconjugants. The results con rmed that all of the four isolates had Incl2 type of plasmid, as shown in gure 3. S1 PFGE and southern blotting: The S1 PFGE results of transconjugants indicate the transfer of a plasmid of size near to 60kb from donor strains. The presence of the mcr-1 gene on the plasmid was con rmed from Southern hybridization, as shown in gure 4.

PCR mapping of mcr-1 context:
In all colistin-resistant isolates, six genes in the nearby region of mcr-1 were ampli ed and sequenced. However, the tnpA gene, in terms of amplicon size, showed an unexpected result. The uncertainty was cleared using the other pair of primers for tnpA loci (absence of IsApl1). This time the expected bands were visualized, and the sequencing result showed missing IsApl1. The genetic contexts of four mcr-1 isolates in comparison with pHNSHP45 are shown in gure 5.

Discussion
This study identi ed, for the rst time, mcr-1 harboring E. coli from Khyber Pakhtunkhwa province and the capital city of Islamabad, Pakistan. The prevalence of mcr-1 was 0.73%, while three out of four isolates were found in urine samples. To date, only a single case of mcr-1 harboring E. coli out of 29 ESBL positive clinical isolates from the Punjab province of Pakistan has been reported [11]. In another study, the mcr-1 gene in E. coli was reported from migratory birds and broiler in the Faisalabad region [9 12]. The mcr-1 gene has also been found in Klebsiella pneumoniae, Acinetobacter baumannii, and Pseudomonas aeruginosa isolated from clinical samples [19 20].
The four isolates we report here were MDR and ESBL positive and have the blaCTXM-15 gene, while three of them also harbor the blaTEM-1 gene. The co-existence of mcr-1 with ESBL has been frequently reported [21][22][23]. A recent study suggests that since 1980, there is an evolutionary link between ESBL and mcr-1; however, this statement needs further con rmation by tracing the mcr-1 gene in archived ESBL isolates, which might provide clues regarding the kinetics over time between ESBL and mcr-1 [24].
One study suggests that the co-occurrence of mcr-1 and blaCTXM-15 is perhaps due to intricate genetic actions taken under antibiotic pressure [25].
In the present study, the clonal assortment of E. coli is diverse, that is, two of the isolates belong to ST  belong to extraintestinal pathogenic E. coil, and avian pathogenic E. coli lineage, respectively. They are mainly involved in the global spread of blaCTXM-15 type extended-spectrum β lactamases [28 30]. A recent study from Pakistan reported mcr-1 harboring ST117 in the broiler [32]. ST117 belongs to avian pathogenic E. coli lineage and may form a reservoir for human extraintestinal pathogenic E coli and antimicrobial resistance [33]. The detection of mcr-1 in avian pathogenic and extraintestinal pathogenic E. coli in human clinical isolates is threatening because it supports the statement of mcr-1 transmission via the food chain [34].
All the four isolates have been successfully trans-conjugated to E. coli EC600. The S1 PFGE, southern blot, and PBRT results revealed that the mcr-1 gene was present on 60 Kb Incl2 plasmid. The Antibiotic susceptibility pro les of transconjugants showed that only the mcr-1 antibiotic-resistant gene was present on Incl2 plasmid. The similar incl2 plasmid harboring only mcr-1 antibiotic resistance gene had been reported in ST 155 isolated from a healthy broiler in Pakistan. The insertion sequence IsApl1 of mcr-1 in ST155 incl2 was missing, while the remaining genetic context was similar to pHNSHP45 [9]. The PCR based mapping of mcr-1 genetic context in our isolates revealed that almost all ampli ed genes were similar to pHNSHP45, except the truncated tnpA gene, missing IsApl1. Globally, plasmid analysis of various Incompatibility types has shown that IsApl1 is usually absent in IncX4 type plasmid while in Incl2 type plasmid, it is either absent or present [6]. The missing of IsApl1 might be due to the coevolution via the acquisition of IsApl1 for speedy relocation into other plasmids [35]. Alternatively, it might reinforce the constancy of mcr-1 on the plasmid [36] The missing of IsApl1 strengthens our statement of having the same plasmid in ST155 (detected earlier from healthy broiler) and in our isolates [9]. This nding of mcr-1 on ST117 and similar plasmid pro les with a previously detected plasmid from a healthy broiler suggests that both horizontal transfer and transfer via food chain might be responsible for disseminating colistin resistance mcr-1 gene in Pakistan.

Conclusion
We reported the mcr-1 gene in human clinical isolates belonging to avian pathogenic E. coli lineage. The plasmid analysis and sequence typing indicate mcr-1 spread via the food chain and horizontal transfer. It might have severe consequences in the community. Our data stresses the need for mcr-1 prevention in humans, food chains, veterinary, and the environment by involving health care professionals, researchers, and government authorities to take concrete steps to prevent colistin resistance among the population.     S1 PFGE and Southern blot, PKE051, PKE14, PKE196, and PKE211 are the isolates; M is the H9812 molecular size marker (1135 to 20.5kb) digested with Xbal enzymes. All strains show to 60kb plasmid on S1 PFGE gel and southern blot.