Profiling of Antimicrobial Resistance Genes and Integron from Escherichia coli Isolates Using Whole Genome Sequencing

This study is designed to investigate Escherichia coli for the antibiotic resistance genes (ARGs) and integrons from healthy as well as diarrhoeic/diseased animals/birds’ faecal samples. A total of eight samples were selected for the study; from each animal, two samples were taken, one from healthy animals/birds and one from diarrhoeic/diseased animals/birds. Antibiotic sensitivity testing (AST) and whole genome sequencing (WGS) was performed for selected isolates. The E. coli isolates showed resistance to moxifloxacin, followed by erythromycin, ciprofloxacin, pefloxacin, tetracycline, levofloxacin, ampicillin, amoxicillin, and sulfadiazine (4/8, 50.00% each). The E. coli isolates were 100% sensitive to amikacin, followed by chloramphenicol, cefixime, cefoperazone, and cephalothin. A total of 47 ARGs from 12 different antibiotic classes were detected among the eight isolates by WGS. The different classes of antibiotics included aminoglycoside, sulphonamide, tetracycline, trimethoprim, quinolone, fosfomycin, phenicol, macrolide, colistin, fosmidomycin, and multidrug efflux. The class 1 integrons were detected in 6/8 (75.00%) isolates with 14 different gene cassettes.


Introduction
E. coli in both animals and humans can frequently cause gastro-intestinal infection [1]. Since it may cause serious infections in humans and animals and, on the other hand, makes up a sizeable portion of the autochthonous microbiota of the various hosts, E. coli holds a special place in the field of microbiology. The greatest concern is the possibility of virulent and resistant E. coli spreading between humans and animals via numerous routes. E. coli is a significant source of resistance genes that may be to blame for both human and veterinary medical treatment failures [2].
The evolution and selection of bacteria that are resistant to antibiotics is significantly influenced by the use of antibiotics in animals for both medicinal and non-therapeutic purposes. As a result of chromosomal changes or horizontal gene transfer (HGT) of ARGs between related or unrelated bacteria, antimicrobial resistance (AMR) is currently the most pressing issue in bacteria, especially in members of the Enterobacterales family. This allows pathogenic bacteria to resist antimicrobial treatments and persist in hostile environments such as the gastrointestinal tract [3]. In HGT, transformation, transduction, and conjugation are the way by which genetic exchange takes place between two organisms, and in the presence or absence of antibiotics, certain genomic processes occur [4]. AMR, particularly, multidrug resistance (MDR), has become increasingly widespread in clinical isolates, including E. coli isolates from animals [5]. Sadly, the situation with bacteria developing resistance is worsening every day, and we have actually reached the stage where we can say that antimicrobial resistance is a major issue on a global scale [6].
Mobile genetic elements (MGEs) play a major role in HGT between the organism, host, and environment [7]. The horizontal transfer of AMR genes is often facilitated by MGEs such as plasmids, transposons, integrons [8], and integrative and conjugative elements (ICE) [9]. Integrons are a platform for acquiring open reading frames (ORFs) and converting them into functional forms via site-specific recombination [10]; thus, integrons can capture genes' cassettes and express them as proteins ( Figure S1) [11]. The bacterial save our souls (SOS) and stringent responses activate the integron integrase, which is initiated in response to both DNA damage and nutrition deprivation [12]. By frequently integrating in plasmids or transposons, integrons can spread horizontally in bacterial populations ( Figure S2) [13]. Integrases are classified into many classes based on their aminoacidic sequence. Classes 1, 2, and 3 (intI1, intI2, intI3) were the first to be identified as being associated with MGEs, while class 4 (intI4) was associated with chromosomal integration [14].
As many antibiotic resistance genes (ARGs) are responsible for AMR in pathogens, high throughput and robust screening technology such as next generation sequencing (NGS) is required. The NGS technologies greatly benefit monitoring that rely on the characterization of genetic data [15]. These technologies make DNA and RNA sequencing significantly faster and less expensive than Sanger sequencing, which was previously employed.
The aim of this study is to determine the AMR genes present in the E. coli and to compare the numbers of genes present in the healthy and diarrhoeic/diseased animals/birds, because when an animal has diarrhoea, there is a presence of both pathogenic and nonpathogenic bacteria in the gut. Furthermore, integrons are also included in the study, because as a MGE, they play an important role in the transfer of genes from pathogenic bacteria to non-pathogenic bacteria.

E. coli Isolates
The E. coli isolates used in the present study were isolated in the previous study at the Department of Microbiology, Veterinary College, Kamdhenu University, Anand. There was a total of eight groups, which included diarrhoeic/diseased samples and samples from healthy cattle, buffalo, dogs, and poultry. From each group, one representative sample was selected for the present study.

Revival of E. coli Isolates
A total of eight E. coli isolates from different animals and poultry were selected. The isolates were preserved in glycerol at −40 • C and inoculated on MacConkey agar and incubated for 24 h at 37 • C ( Figure S3). Isolated lactose-fermenting, pink-coloured colonies were inoculated on Eosin-Methylene Blue (EMB) agar for confirmation using greenish metallic sheen ( Figure S4) and incubated for 24 h at 37 • C. Further, isolated pure culture colonies were inoculated on Brain Heart Infusion (BHI) agar ( Figure S5) and stored for further study. The sample names are listed in Table 1.

Library Preparation
The quality of the extracted DNA was checked using a Qubit 3.0 Fluorometer (Thermo Fisher Scientific, Waltham, MA, USA). The library preparation was performed using a Nextera XT DNA Library Preparation kit. After library preparation, quantity and quality were checked using a Qubit 3.0 Fluorometer and Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA, USA) using high sensitivity DNA 1000 chip, respectively ( Figure S8). Sequencing was performed using Illumina's reagent kit v3 to produce a 2 × 300 bp pair-end on an Illumina MiSeq benchtop sequencer (Illumina, San Diego, CA, USA).

Whole Genome Data Assembly
The quality of raw data was checked using FastQC (version 0.11.9), while the lowquality sequence data were removed using PRINSEQ-lite (version 0.20.4), where the minimum limit for Q score was 20. Cutadapt (version 3.3) was used for the adaptor trimming. De novo assembly was performed on the CLC genomic workbench 22, where the minimum contig length was set to 1000. The quality of the contigs data was checked using QUEST (version 5.0.2).

Detection of AMR Genes
For the detection of AMR genes from the contigs data, NCBI's AMRFinderPlus [16] was used. AMRFinderPlus can detect acquired antibiotic resistance, stress response, virulence genes, and genetic mutations that are known to confer antibiotic resistance. AMRFinderPlus relies on the NCBI's curated Reference Gene Database and curated collection of Hidden Markov Models (HMMs). The Reference Gene Catalog, database version 2022-05-26.1, was used.

Detection of Integron
The Integron Finder (version 2.0.1) tool [17] was used to detect the integrons in the sequenced samples.

Quantification and Quality Checking of Prepared Library
The Qubit 3.0 fluorometer was used for the quantification of the prepared libraries, and the Bioanalyzer 2100 (Agilent Technologies, Santa Clara, CA, USA) with the Agilent DNA 1000 kit was used to check the library size distribution. The library concentration for the samples ranges from 10.5 ng/µL (CH13) to 15.1 ng/µL (DH9). The average library size for the samples ranges from 1500 bp (DH9) to 1719 bp (PH1) ( Table 2).

Whole Genome Sequencing Analysis
The data were generated using Illumina MiSeq, and the coverage for each sample ranged from 100× to 173× (Table 3). The WGS assembly was performed using the CLC genomic workbench 22. In the final assembly, the contig numbers ranged from 41 to 262, with a minimum length of 1000 bp.

Detection of Integrons Using Integron Finder Tools
Integrons play an important role in the transfer of ARGs between different bacteria. In the present study, the class 1 integron was detected in six out of eight isolates (Table 6). There was a total 14 gene cassettes which were related to the class 1 integron viz. dhfrA17, aadA5, ant(2 )-Ia, aac(3)-VI, aac(6)-Ib, aadA, aadA1, aadA2, emrE, catB, cmlA1, bla OXA-1 , qacL, and qacEdelta1. None of the isolates were found to be positive for the class 2 integron.

Integrase Gene No. of Gene Cassettes Present in intI1 Name of Gene Cassettes Present in IntI1 IntI1
IntI2

Antimicrobial Susceptibility Testing Using Disc Diffusion Method
Antibiotics have the ability to inhibit the growth of bacteria or kill the bacteria. However, when these bacteria can survive in the presence of antibiotics, this is known as antibiotic resistance. Phenotypically, we can check the susceptibility of a bacteria to antibiotics. There are various methods for this, and one commonly used is the disc diffusion method, which was used in the present study.
Among the collected E. coli isolates, six isolates showed resistance to ciprofloxacin. Similar studies have shown resistance ranging from 60% to 92% [18][19][20][21]. In this study, 75% of the isolates showed resistance to tetracycline. One study reported 100% resistance to tetracycline in the E. coli isolated from broiler chicks and calves [21,22]. In the case of erythromycin, the observed resistance was 87.50%, and other studies have shown resistance ranging from 80% to 100% in the E. coli isolates from diseased poultry birds [18,23,24].
The present study is designed to measure the differences between the antibiotic resistance of the E. coli isolates from healthy and diarrhoeic/diseased animals/birds. The diarrhoeic isolates showed more resistance compared to the healthy isolates. The E. coli isolates from poultry showed a different scenario, where both (healthy and diseased) isolates showed no such difference in the antibiotic resistance. In the present study, the E. coli isolates that recovered from diarrhoeic cattle showed resistance to 19 antibiotics, which was higher than that of any isolate.

ARG Detection among Different E. coli Isolates
In the present study, a total of 47 AMR genes were identified, of which 34 genes were without chromosomal mutations and 13 were with chromosomal mutation, which makes 13 other genes (Table S3). All the isolates showed resistance to multiple classes of antibiotics, ranging from 3 to 10. Only two isolates had six AMR genes. Other isolates had AMR genes in the range from 12 to 21. Based on the findings of the present study, all the isolates were multidrug resistant.
From the aminoglycoside class, the strA and strB gene were detected in 37.50% of the isolates. Other similar studies have found strA and strB in E. coli isolates from different animals [7,25,26]. The aac(6 )-Ib-cr5 gene was found only in E. coli isolates from diarrhoeic dogs, which confers resistance to aminoglycoside as well as fluoroquinolone, and similar studies have also detected aac(6 )-Ib-cr5 in dog and calf samples [27,28].
Resistance to sulphonamides occurs principally through the acquisition of the alternative dihydropteroate synthase (DHPS) gene sul, the product of which has a low affinity for sulphonamides [36]. In the present study, three sul genes were detected. Similarly, sulphonamide resistance genes were detected in the E. coli in previous studies [28,35,37]. In the present study, trimethoprim resistance genes (dfrA) were found. Similarly, previous studies have also found trimethoprim resistance genes in E. coli isolated from cattle and poultry [30,31,33].
Two tetracycline resistance genes (tetA and tetB) were detected among eight E. coli isolates, wherein the E. coli isolated from diseased poultry harboured the tetB gene, and all other isolates harboured the tetA gene. Similarly, previous studies detected tetracycline resistance genes in E. coli isolated from cattle and poultry [28,37,38].

Correlation between Phenotypic Resistance to Antibiotics and Presence of ARGs
In this study, streptomycin resistance genes (strA and strB) were found in three isolates, all of which showed phenotypic resistance to streptomycin. Thus, we can say that the genes present in the E. coli isolate are expressing themselves. Similarly, four isolates harboured gentamicin resistance genes (ant(2 )-Ia, aac(3)-VIa, aac(3)-IIe, aac(3)-IId, and aac(3)-IVa); among them, three isolates showed phenotypic resistance to gentamicin, and one isolate did not show phenotypic resistance.
In the case of β-lactam antibiotics, the bla CTX-M-15 gene confers resistance to ampicillin, amoxicillin, ceftriaxone, and cefotaxime. Three isolates harboured the bla CTX-M-15 gene and showed phenotypic resistance to the above-mentioned antibiotics. One isolate harboured the bla OXA-1 gene and showed phenotypic resistance to ceftriaxone. For tetracycline, six isolates were positive for the tetA and tetB gene, and all six isolates showed phenotypic resistance to tetracycline. Four isolates that were phenotypically resistant to co-trimoxazole had sulphonamide and trimethoprim resistance genes. Two isolates had resistance genes against sulphonamide and trimethoprim but did not show phenotypic resistance. Similar to chloramphenicol, two isolates possessed resistance genes and showed phenotypic resistance to chloramphenicol.
The isolates that possessed the ARGs and showed resistance indicated that the gene was expressed and showed resistance to a particular antibiotic for which it confers resistance. The isolates that did not show resistance to antibiotics but had ARG might be due to a lack of gene expression.

Intraspecies Discussion in Relation to Antimicrobial Susceptibility Testing
In the present study, the healthy buffalo and diarrhoeic buffalo samples show a high difference in the antibiotic resistance pattern. The healthy buffalo sample was found to be resistant to only 2 antibiotics, whereas the diarrhoeic buffalo samples were found to be resistant to 15 antibiotics.
The healthy cattle and diarrhoeic cattle samples also show a high difference in the antibiotic resistance pattern. The healthy cattle sample was resistant to 3 antibiotics, and the diarrhoeic cattle sample was found to be resistant to 19 antibiotics.
The healthy dog sample was found to be resistant to 5 antibiotics, whereas the diarrhoeic dog sample was found to be resistant to 15 antibiotics. The healthy poultry and the diseased poultry samples do not show much difference in the resistance against antibiotics. The healthy poultry samples show resistance to 9 antibiotics, and the diseased poultry samples show resistance to 11 antibiotics.
In the present study, the diarrhoeic/diseased samples were found to be more resistant to antibiotics when compared to the healthy samples, but in the poultry, there is not much difference shown between the healthy and diseased samples. Among all the diarrhoeic/diseased samples, the diarrhoeic cattle sample shows the highest resistance to a number of antibiotics (19 antibiotics). The healthy poultry samples show resistance against nine antibiotics, which is higher when compared to other healthy samples.

Intraspecies Discussion in Relation to ARGs Detected by AMRFinderPlus
In the present study, 6 ARGs were detected in the healthy buffalo isolate (BH8), whereas the diarrhoeic buffalo isolate (BD11) was found to be positive for 15 different ARGs. The healthy cattle (CH13) and diarrhoeic cattle (CD13) showed a difference in the detected ARGs. A total of 6 ARGs were found in the CH13 isolate, and a total of 18 ARGs were found in the CD13 isolate.
In the dog isolates, 12 ARGs were detected in the healthy dog (DH9) isolate, whereas the diarrhoeic dog isolate (DD9) was found to be positive for 21 ARGs. The healthy poultry (PH1) and diseased poultry (PD14) samples do not show much difference in the number of ARGs detected. In the healthy poultry isolate, 19 ARGs were detected, and in the diseased poultry isolate, 16 ARGs were detected.
Among the healthy isolates, the healthy poultry isolates carry the highest number of ARGs (19). The diarrhoeic dog sample carries the highest number of ARGs (21) among all the diarrhoeic/diseased samples. The diarrhoeic/diseased samples carried more ARGs when compared to the healthy samples, except the poultry samples, in which there is not much difference between the ARGs detected in the samples.
In the present study, among the isolates obtained from diarrhoeic/diseased animals or poultry, all the isolates were found to be positive for the class 1 integron (intI1), whereas among the isolates from the healthy animals/poultry, two isolates were positive for the class 1 integron (intI1). However, none of the isolates were positive for the class 2 integron (intI2).

Conclusions
This study was performed to determine the AMR genes in the E. coli isolates from healthy as well as diarrhoeic/diseased animals/birds. Overall, the results clearly show a difference in the number of AMR genes between the E. coli isolated from healthy and diarrhoeic/diseased animals/birds. The E. coli isolates from the healthy poultry harboured a greater number of AMR genes compared to the E. coli isolate of the diseased birds. In the present study, integrons are also detected, which plays an important role in the transfer of AMR genes between related or unrelated bacteria. Class 1 integrons were detected from all the isolates. In conclusion, whenever there is a diarrhoeic condition in the animal, at that time, normal flora comes into contact with more pathogenic bacteria that harbour many AMR genes, which can get transferred from the pathogenic bacteria to the normal flora of the animal with the help of the integron.
Supplementary Materials: The following supporting information can be downloaded at https:// www.mdpi.com/article/10.3390/genes14061212/s1, Figure S1: Structure of class 1 integron; Figure S2: Environmental resistome and the structure of class 1 integron; Figure S3: Lactose fermenting pink coloured colonies of E. coli on MacConkey agar; Figure S4: Greenish metallic sheen producing E. coli on Eosin-Methylene Blue agar; Figure S5: Isolated pure cultured colonies of E. coli on BHI agar; Figure S6: Petridishes showing antibiotic susceptibility pattern of E. coli by disc diffusion method (phenotypic method); Figure S7: Percentage of E. coli isolates showing sensitive, intermediate and resistance to drug; Figure S8: Representative Electropherogram of DNA library on bioanalyzer; Table S1: Result of Anti-microbial Susceptibility Test;   Data Availability Statement: The E. coli genome data can be found in the NCBI database with the accession number PRJNA891125 (Escherichia coli (ID 891125))-BioProject-NCBI (nih.gov) (accessed on 22 October 2022)).