Comprehensive genomic analysis of avian Escherichia coli from Noakhali uncovers multidrug resistance, metal resistance, and zoonotic signatures Authors: Md. Habib Ullah Masum, Kazi Chamonara, Mohammad Sharif Uddin, Imam Hossain, Suhag Chandra Roy, Md. Imam Hossain, Md. Razib Hosen, Afifa Siddiqua, Abdullah Al Mukarrom Published: 2026 ABSTRACT Colibacillosis caused by avian pathogenic Escherichia coli (APEC) results in significant poultry losses and financial constraints globally, particularly in Bangladesh, emphasizing the urgent need for effective surveillance and control strategies. The present study employed whole-genome sequencing (WGS) of E. coli isolates from avian hosts and their surroundings to obtain genomic and evolutionary insights. Subsequently, the strains exhibited high genome completeness (> 99%) and coarse consistency scores (> 98) for genome assembly metrics. Further, genome annotation profiles showed a broad range of antimicrobial resistance (AMR) determinants, including resistance–nodulation–division (RND), major facilitator superfamily (MFS), and small multidrug resistance (SMR) multidrug efflux pumps. The coexistence of such AMR determinants within mobile genetic elements (MGEs) indicates a pattern of horizontal gene transfer and a possible dissemination pathway for the multidrug resistance phenotype. Several virulence-associated gene (VAG) clusters in the genomes suggest potential virulence profiles. A significant number of genes conferring heavy metal resistance and detoxification were identified in the genomes, including arsenic, copper, magnesium, tellurite, and zinc resistance, indicating extensive metal stress tolerance in the strains. Subsequent pangenome and phylogeny analyses uncovered significant similarities between strains derived from avian and human clinical isolates, suggesting a potential for zoonotic transmission. The findings highlight genetic association and potential public health implications of APEC and environmental E. coli (EEC) strains from poultry.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-026-54331-4. Introduction Avian pathogenic Escherichia coli (APEC) represents a major threat to poultry by causing extraintestinal infections that adversely impact both poultry health and production 1. Colibacillosis, characterized by localized and systemic infections, results from APEC infection in different avian species. This condition may result in considerable mortality among poultry, with death rates soaring to as high as 20% 1. Poultry affected by colibacillosis might experience a decline in egg production of up to 20% and reduced hatchability. Moreover, this condition considerably heightens the risk of carcass rejection during slaughter, with rates potentially soaring to 43% 1,2. Colibacillosis is among the most economically consequential conditions in the global poultry sector, including Bangladesh. The poultry sector in Bangladesh has seen substantial expansion in recent decades, making a considerable contribution to the nation’s gross domestic product (GDP) and rural employment 3. However, an increasing number of infectious diseases, among which colibacillosis emerges as one of the most significant challenges 4. In Bangladesh, epidemics of colibacillosis are prevalent, particularly during the summer and humid seasons, when environmental stress compromises the poultry’s immune systems, increasing their susceptibility to infections 5,6. The extensive use of antibiotics for therapy and growth promotion contributed to the development of antibiotic-resistant strains of APEC and environmental E. coli (EEC). This poses a significant public health concern because of the potential for zoonotic transmission 6,7. Implementing long-term systematic monitoring will yield stronger evidence on the persistence, dissemination, and transmission dynamics of these strains 8. Considering the detrimental effect of colibacillosis on poultry production and the increasing risk of antibiotic resistance, it is crucial to comprehend the epidemiology and resistance patterns of APEC isolates within the Bangladeshi poultry industry 9–11. Human extraintestinal pathogenic E. coli (ExPEC) strains may inherit virulence and antimicrobial resistance (AMR) genes from APEC. The presence of APEC genetic components in poultry raises the possibility of zoonotic transmission, in which drug-resistant or more virulent ExPEC strains might evolve in humans 1. The spread of this disease poses a serious threat to human health and may occur via direct contact with animals, exposure to animal waste, consuming raw or undercooked meat, or touching surfaces that contain meat 12,13. Considering these varied transmission pathways, whole-genome analysis (WGS) of E. coli is crucial for comprehending its virulence characteristics, tracing infection origins, and discovering genetic determinants linked to AMR. Comprehensive genetic insights are essential for efficient monitoring, outbreak investigation, and the formulation of targeted measures to reduce the infection 14,15. This study aimed to annotate and comparatively analyze three Escherichia coli strains (APEC and EEC) isolated from poultry anatomical sites and their surrounding environments within the same poultry farm to characterize their antimicrobial resistance profiles and explore potential transmission of resistance genes across different farm-associated niches. The findings are expected to offer significant insights into the genomic characteristics of these strains, thereby enhancing the understanding of APEC epidemiology in Bangladesh and South Asia. Methods Ethical statement The Ethics Approval Committee of Noakhali Science and Technology University (NSTUEC) has reviewed and granted authorization for the study protocol (Reference number: NSTU/SCI/EC/2025/331). The sampling method was rigorously aligned with the ARRIVE 2.0 guidelines 16. The research was conducted in accordance with the guidelines set forth in the “Guide for the Care and Use of Laboratory Animals” issued by the National Institutes of Health (NIH). All methods adhered to the relevant regulations and legislation. The research was performed in a region devoid of endangered or protected species. Sample collection, processing and antibiotic resistance profiling A cross-sectional study was implemented in the Noakhali area of Bangladesh, using three E. coli isolates chosen from a previous study 17. Isolates were obtained from anatomical regions (B-23), faeces (F-3), and water (W-5). Cloacal swabs were inoculated onto MacConkey agar, whereas water and faecal samples were supplemented in peptone water prior to plating. Presumptive E. coli colonies, exhibiting dark pink coloration on MacConkey agar and metallic green on eosin methylene blue (EMB) agar, were purified and validated by triple sugar iron (TSI), motility indole urea (MIU), and citrate assays. Confirmed isolates were preserved at − 80 °C in Luria–Bertani (LB) medium supplemented with glycerol. Antimicrobial susceptibility (AST) was evaluated using the Kirby–Bauer method with seven antibiotics pertinent to poultry. Antibiotics often used in the poultry industry in Bangladesh were evaluated, including gentamicin, ampicillin, nitrofurantoin, levofloxacin, ciprofloxacin, trimethoprim-sulfamethoxazole, and tetracycline. E. coli ATCC 25922 was used as the quality control strain. The inhibition zone diameters (mm) were evaluated according to CLSI (2021) guidelines to categorize isolates as resistant, intermediate, or susceptible 18. Whole genome sequencing, assembly and strain characterization The total DNA from the isolates was extracted using the QIAamp® DNA Mini Kit (QIAGEN, Germany) and then quantified using a Qubit 4 fluorometer (Thermo Fisher Scientific, USA). Before DNA sequencing, a native barcoding kit (SQK.NBD.114.24) was used for library preparation, and sequencing was performed on the Oxford Nanopore MinION™ Mk1C platform with the FLO-MIN114 (R10.4.1) flow cell. Data was obtained using MINKNOW (1.11.5). The Guppy (v.6.3.2) in high accuracy mode was used to basecall MinION™ Mk1C sequencing data (FAST5 files) to generate pass reads (FASTQ format) with a mean quality score > 9. The adapter and barcode sequences were deleted using Porechop (v.0.2.4) 19, and Nanofilt (v.2.8.0) was used to remove low-quality bases (q < 10) 20. Porechop was applied with stringent parameters: adapter and end trimming thresholds of 100, removal of reads with internal adapters, and the requirement of two matching barcodes. Raw reads were processed with NanoFilt, with a minimum quality score of 12, and reads were retained between 1,300 and 1,700 bp. The first 20 bases of each read were removed to eliminate potential low-quality sequence at the start (Table S1). Flye (v.2.9.4) was used to assemble quality-trimmed reads into contigs via de novo assembly 21. The assembly was filtered with a minimum contig length of 1,000 bp and a minimum coverage of 10×, targeting 100% genome coverage. Subsequent genome polishing was performed with Racon (v.1.5.0) 22. The bacterial species were confirmed using the Type (Strain) Genome Server (TYGS) 23. The genome contamination and completeness were evaluated using several tools, including ConEst16S (https://www.ezbiocloud.net/tools/contest16s)24, CheckM (v1.2.4)25, QUAST (v.5.2.0)26, and BUSCO (v.6.0.0) 27. The average nucleotide identity (ANI) was assessed using the OrthoANI tool available on the Ezbiocloud server (https://www.ezbiocloud.net/tools/orthoaniu) 28. Phylogroup analysis of the strains was conducted using the EzClermont tool (https://pmc.ncbi.nlm.nih.gov/articles/PMC7656184/) 29. The structural and functional annotation of the assembled genomes was performed by RAST ( v.1.3) 30, and the Prokaryotic Genome Annotation Pipeline (PGAP) of NCBI (v.6.10)31. These analyses were conducted using the default parameters recommended by the respective servers. The Proksee server was used to demonstrate the labelled genome map 32. Antimicrobial resistance gene, mobile genetic elements, and virulence genes analysis The AMR genes were identified by comparing them with the Comprehensive Antibiotic Resistance Database (CARD, v.6.03). Only sequences with at least 80% amino acid identity and 80% query coverage were included in downstream analyses 33. Mobile genetic elements (MGEs) were identified by comparison with the mobileOG-db (v.1.1.3) database, requiring at least 80% amino acid identity and 80% query coverage. Only the highest-confidence matches were used for downstream analysis 34. Virulence-associated (VAGs), and metal resistance genes were assessed by Virulence Factor Database (VFDB, v.6.0) 35, and BacMet (v.2.0) 36 tools, respectively. For VAGs identification using the VFDB database, a 70% sequence identity threshold and an 80% query coverage cutoff were applied. In contrast, detection of metal resistance genes with the BacMet database utilized 80% sequence identity and 80% query coverage thresholds (Table S1). Pangenome analysis and pathogenicity profiling A rigorous pangenome evaluation and comparative genomic analysis were conducted to clarify the inherent diversity of the target strain. The aim was to identify distinct presence-absence variations (PAVs) in certain genes or gene families across strains from diverse geographical locations and origins. 37 avian and 21 human WGS of E. coli were obtained from the Bacterial and Viral Bioinformatics Resource Center (BV-BRC) (Data S1 and S2). Analyses included a total of 61 E. coli genomes using Roary (v.3.13.0) 37, which efficiently constructs extensive pangenomes and differentiates between core and accessory genes. This program was executed with BLASTP-based clustering at a 95% (-i) protein sequence identity threshold, and core genes were defined as those present in at least 99% (-cd) of the analyzed genomes (Table S1). Results Multidrug resistance profile The presumptive E. coli isolates showed a pronounced multidrug-resistance (MDR) profile. Specifically, isolates B-23 and F-3 were resistant to six of the seven antibiotics tested. In particular, isolate W-5 exhibited resistance to all antibiotics evaluated (Table S2). Whole genome sequence analysis and functional annotation Genomic analysis of the isolates B-23, F-3, and W-5 confirmed that these isolates belong to E. coli. Accordingly, the B-23, F-3, and W-5 were designated as “E. coli strain APEC_AN-25”, “E. coli strain EEC_F3”, and “E. coli strain EEC_W5”, respectively. Validation using the ANI score further confirmed that the isolates belong to E. coli, with ANI values > 98% when compared to reference genomes (E. coli str. K-12 substr. MG1655). Genome sequence analysis of the three E. coli strains showed high assembly quality, supported by coarse consistency metrics. Coarse consistency values for the APEC_AN-25, EEC_F3, and EEC_W5 were 98.98%, 98.09%, and 99.49%, respectively. Genome completeness scores were also high, with the APEC_AN-25 at 99.8%, EEC_F3 at 99.6%, and EEC_W5 at 99.7%. The genomic features of three E. coli strains—APEC_AN-25, EEC_F3, and EEC_W5, were compared to evaluate assembly quality and coding sequence composition. The number of contigs varied among the strains: the APEC_AN-25 contained 7 contigs, the EEC_F3 exhibited 2 contigs, and the EEC_W5 had 6 contigs. Genomic content analysis revealed that the predicted coding sequences (CDS) showed slight differences across strains. Following de novo genome assembly and annotation, the CDS ratios of E. coli strain APEC_AN-25, EEC_F3, and EEC_W5 were predicted to be 0.99, 0.98 and 0.99, respectively (Table 1). Phylogroup analysis of the isolates revealed that APEC_AN-25, EEC_F3, and EEC_W5 belonged to E. coli phylogroups B1, E1, and A, respectively. The genome analysis showed significant parallels in the distribution of protein-encoding genes among the three E. coli strains—APEC_AN-25, EEC_F3, and EEC_W5. The highest count of protein-encoding genes with functional assignments was observed in EEC_F3, featuring 3,780 annotated genes (Table 1; Fig. 1b). This was followed by the APEC_AN-25, with 3,699 genes (Table 1; Fig. 1a), and the EEC_W5, with 3,664 genes (Table 1; Fig. 1c). Table 1The genomic features of the E. coli strain APEC_AN-25, EEC_F3 and EEC_W5.FeaturesStrainE. coli strain APEC_AN-25E. coli strain EEC_F3E. coli strain EEC_W5ANI score (%)98.9398.0299.40Coarse consistency (%)99.9898.0999.49Completeness (%)99.899.699.7Genome length (bp)5,141,1825,400,2665,085,109GC content (%)50.5750.5450.69Contig count726Contig N50 (bp)4,687,8255,257,3004,699,303Contig L50111CDS513766375067CDS ratio0.990.980.99PhylogroupB1E1AProtein-encoding genes with functional assignment369937803664Protein-encoding genes without functional assignment143815181403 Table 1The genomic features of the E. coli strain APEC_AN-25, EEC_F3 and EEC_W5.FeaturesStrainE. coli strain APEC_AN-25E. coli strain EEC_F3E. coli strain EEC_W5ANI score (%)98.9398.0299.40Coarse consistency (%)99.9898.0999.49Completeness (%)99.899.699.7Genome length (bp)5,141,1825,400,2665,085,109GC content (%)50.5750.5450.69Contig count726Contig N50 (bp)4,687,8255,257,3004,699,303Contig L50111CDS513766375067CDS ratio0.990.980.99PhylogroupB1E1AProtein-encoding genes with functional assignment369937803664Protein-encoding genes without functional assignment143815181403 The genomic features of the E. coli strain APEC_AN-25, EEC_F3 and EEC_W5. Fig. 1Genome features of the E. coli strain (a) APEC_AN-25, (b) EEC_F3, and (c) EEC_W5. The genome maps featuring CDS, rRNA, tRNA, tmRNA, and GC content. Genome features of the E. coli strain (a) APEC_AN-25, (b) EEC_F3, and (c) EEC_W5. The genome maps featuring CDS, rRNA, tRNA, tmRNA, and GC content. Antimicrobial resistance genes and mobile genetic elements analysis Analysis of key AMR and AMR-regulatory genes in the E. coli strains APEC_AN-25, EEC_F3, and EEC_W5 revealed a range of resistance factors across different antibiotic categories. A significant number of the identified genes encode multidrug efflux pumps belonging to the resistance-nodulation-cell division (RND), major facilitator superfamily (MFS), and small multidrug resistance (SMR) families. Significant RND efflux genes, including cpxA, gadX, gadW, mdtE, CRP, acrE, acrS, rsmA, acrD, baeS, mdtB, mdtA, acrA, acrB, and marA, were identified in the strains’ genomes. Similarly, the genes involved in MFS transporters such as emrB, emrA, emrR, mdtH, mdtG, mdtM, mdtN, mdtO, mdtP, floR, evgA, and hns (H-NS) were predominantly reported in all three strains (Fig. 2a-c, Data S3). Additionally, the existence of kpnE, kpnF, and qacL genes in the genomes suggests the acquisition of SMR-type efflux systems. In addition to efflux pump families, several direct AMR genes were identified. Specifically, pmrF, ugd, and eptA were detected in the analyzed genomes, each encoding enzymes involved in the polymyxin resistance (pmr)-associated lipid A modification pathway. Regulatory genes with mutational variants associated with antibiotic resistance were also observed. These included alterations in gyrA, which are commonly linked to fluoroquinolone resistance, as well as mutations affecting the AcrAB–TolC efflux system and the soxS regulatory pathway, both of which can enhance multidrug resistance phenotypes. Additionally, gene responsible for encoding antibiotic-modifying enzymes such as ant(3’’)-IIa (ANT(3’’)-IIa), which confers resistance to aminoglycosides, were identified. Structural modification-associated genes, including vanG, were also detected, indicating the presence of resistance mechanisms beyond efflux-mediated pathways (Fig. 2a–c, Data S3). Fig. 2The antimicrobial resistance (AMR) genes in the genome E. coli strains. Multiple AMR genes exist within the genomes of (a) APEC_AN-25, (b) EEC_F3, and (c) EEC_W5, represented by red arrows. The antimicrobial resistance (AMR) genes in the genome E. coli strains. Multiple AMR genes exist within the genomes of (a) APEC_AN-25, (b) EEC_F3, and (c) EEC_W5, represented by red arrows. Further analysis showed that the strains contained various MGEs involved in replication, recombination, repair, phage integration, excision, horizontal gene transfer, and genomic stability and defense. The genomic context of E. coli strains (APEC_AN-25, EEC_F3, and EEC_W5) underscores the significant role of MGEs in shaping their antimicrobial resistance profiles. Key AMR regulatory genes, including the RND efflux regulator cpxA and the MFS-associated regulator hns, were found within MGEs. The presence of a mutated gyrA gene within an MGE suggests a potential for rapid dissemination of resistance traits (Figs. 3, 4 and 5). In the genome of E. coli strain EEC_F3, the close association of the bacA gene MGE regions suggests the potential co-transfer of bacitracin resistance alongside other resistance loci. Additional AMR genes, including qacL, ant(3’’)-IIa, cmlA1, aadA3, blaCARB-3 (CARB-3), and dfrA16, are also positioned near MGEs (Fig. 4). Similarly, the genome of E. coli strain EEC_W5 contains multiple resistance and resistance-associated genes, such as qacL, ant(3’’)-IIa, acrE, acrS, parC, rsmA, evgA, pmrF, acrR, acrA, acrB, mphA, mrx, floR, and aadA23, all located in proximity to MGEs (Fig. 5). Fig. 3The coexistence of antibiotic resistance gene and mobile genetic elements found in the genome E. coli strain APEC_AN-25. The antibiotic resistance genes are shown in red, whereas the mobile genetic elements are represented in various colored nodes. The coexistence of antibiotic resistance gene and mobile genetic elements found in the genome E. coli strain APEC_AN-25. The antibiotic resistance genes are shown in red, whereas the mobile genetic elements are represented in various colored nodes. Fig. 4The coexistence of antibiotic resistance genes and mobile genetic elements (MGEs) found in the genome E. coli strain EEC_F3. The antibiotic resistance genes are shown in red, whereas the mobile genetic elements are represented in various colored nodes. The coexistence of antibiotic resistance genes and mobile genetic elements (MGEs) found in the genome E. coli strain EEC_F3. The antibiotic resistance genes are shown in red, whereas the mobile genetic elements are represented in various colored nodes. Fig. 5The coexistence of antibiotic resistance genes and mobile genetic elements found in the genome E. coli strain EEC_W5. The antibiotic resistance genes are shown in red, whereas the mobile genetic elements are represented in various colored nodes. The coexistence of antibiotic resistance genes and mobile genetic elements found in the genome E. coli strain EEC_W5. The antibiotic resistance genes are shown in red, whereas the mobile genetic elements are represented in various colored nodes. Virulence associated gene analysis Comprehensive genome analysis of E. coli strains APEC_AN-25, EEC_F3, and EEC_W5 identified approximately 100 virulence-associated genes. These strains possess the enterobactin biosynthesis (ent) gene cluster (entA, entB, entC, entD, entE, entF, and entS), which regulates enterobactin production and transport. Additionally, the strains contain multiple genes involved in iron uptake and utilization, including fepA, fepB, fepC, fepD, fepE, and fepG (Data S4). The analyzed strains possess a diverse array of fim gene clusters associated with the production and regulation of type 1 fimbriae, including fimA, fimB, fimC, fimD, fimE, fimF, fimG, fimH, and fimI. In APEC AN-25, EEC F3, and EEC W5, additional VAGs such as flgC, flgE, flgG, and flgH were identified. The genomes also contain flh and fli gene clusters, which encode essential regulatory, structural, and assembly components of the bacterial flagellum. Furthermore, motA, ompA, and tssA genes are present, contributing to key aspects of bacterial physiology, survival, and environmental interaction. Notably, the genome of E. coli strain EEC F3 contains the iro gene cluster (iroB, iroC, iroD, iroE, and iroN), which encodes proteins involved in the biosynthesis, modification, transport, and uptake of salmochelin. The genome of EEC W5 harbors the iuc gene cluster (iucA, iucB, iucC, and iucD), which encodes enzymes necessary for the synthesis of aerobactin, a high-affinity hydroxamate siderophore essential for bacterial viability under iron-deficient conditions (Data S4). Metal resistance gene analysis Further comprehensive genome annotation of the strains revealed multiple arsenic resistance (ars) genes, including arsB, arsC, arsH, and arsR (Data S5). Genes associated with copper resistance were also identified, such as copA and the cus gene cluster, both of which are essential for copper efflux and detoxification. The genomes also contained a cor gene cluster, which may provide a mechanism for resistance to divalent cations, potentially including magnesium. Additionally, the strains possessed tehA and tehB, which are genes associated with bacterial resistance to tellurite. Zinc resistance genotypes were also detected, including the ZitB and ZntA efflux systems. Furthermore, the presence of yieF, which may encode a protein reported in E. coli strains to function as both a ferrous iron efflux transporter and a chromate reductase. Pangenome analysis and genome-wide pathogenicity profiling A total of 40 E. coli isolates from avian samples were evaluated, revealing 1559 core genes, 798 soft core genes, 4316 shell genes, and 18,911 cloud genes (Fig. 6a, b). The pangenome analysis of 21 human isolates revealed 1940 core genes, 550 soft core genes, 4163 shell genes, and 12,364 cloud genes. The comprehensive study of the pangenome revealed that the APEC strain APEC_AN-25 exhibited notable genomic and phylogenetic analogies to the APEC strain Z1323CEC0004 from South Korea. However, the strain EEC_F3 exhibited notable genomic and phylogenetic similarities to the EEC strain E101, as well as to EC4200 from Australia and France, respectively (Fig. 6c). The pangenome-wide study demonstrated that E. coli strains derived from avian sources had significant genomic and phylogenetic resemblance to clinical isolates derived from human origins. The EEC_F3 strain exhibited an intricate phylogenetic relationship with the clinical isolate E. coli strain E-9 from Bangladesh. Conversely, the strain EEC_W5, another strain in the study, showed phylogenetic proximity to the clinical isolate NGE-024. Nonetheless, the strain APEC_AN-25 had no substantial genomic or phylogenetic similarity to clinical isolates in Bangladesh. The results suggest that EEC isolates may have greater zoonotic potential than APEC isolates (Fig. 6d). Fig. 6Pangenome analysis of the E. coli across different geographics and origins. (a) The genome comparison of the 40 different genomes reveled different genomic composition including core (blue), soft (green), shell (yellow) and cloud (purple) genes across the E. coli strains. (b) Genome alignment of the E. coli strains suggests significant genes conservation across the strains (Blue and color denotes the presence and absence of genes, respectively). (c) The phylogenetic analysis of the 40 pangenome depicts the evolutionary relationship of the avian strains of different geographics. (d) The phylogenetic analysis of the 24 pan-genome depicts the evolutionary relationship of the avian strains with the human isolates. Pangenome analysis of the E. coli across different geographics and origins. (a) The genome comparison of the 40 different genomes reveled different genomic composition including core (blue), soft (green), shell (yellow) and cloud (purple) genes across the E. coli strains. (b) Genome alignment of the E. coli strains suggests significant genes conservation across the strains (Blue and color denotes the presence and absence of genes, respectively). (c) The phylogenetic analysis of the 40 pangenome depicts the evolutionary relationship of the avian strains of different geographics. (d) The phylogenetic analysis of the 24 pan-genome depicts the evolutionary relationship of the avian strains with the human isolates. Discussion Monitoring APEC is essential to protect poultry health and farm productivity, and to reduce the risk of zoonotic infections and the spread of drug-resistant strains through the food chain and environmental contamination. This study presents a comprehensive genomic characterization of E. coli strains isolated from poultry anatomical sites B-23 (APEC_AN-25), faeces F-3 (EEC_F3), and water W-5 (EEC_W5) within a single farm environment. Previous research has primarily evaluated E. coli from individual sample types or across multiple farms. In contrast, this study integrates WGS data from diverse ecological niches within a single poultry production system. By analyzing both APEC and EEC strains using Oxford Nanopore long-read sequencing, the study provides deeper insights into strain diversity, genomic architecture, and potential transmission pathways between poultry and their environment. This approach enables comprehensive analysis of the dissemination of AMR genes and VAGs within confined farm environments, an area that remains insufficiently addressed in current literature. High genome completeness is particularly important for accurate gene prediction, virulence profiling, and antimicrobial resistance gene detection, as incomplete assemblies may lead to underrepresentation of key functional loci 38,39. The genome completeness scores were high (99.98% for APEC_AN-25, 98.09% for EEC_F3, and 99.49% for EEC_W5), suggesting that most of the genomic content was captured in the assemblies 40. Species-level assignment was further supported by ANI analysis, which yielded values exceeding 98% compared to the reference strain E. coli K-12 MG1655. As ANI thresholds of 95–96% are widely accepted for prokaryotic species delineation, these results provide strong genomic evidence for accurate taxonomic classification 41. Within the genome, multiple AMR and AMR-regulatory genes were also identified. The identified RND efflux genes may be responsible for reduced susceptibility to a variety of antibiotic classes, such as fluoroquinolones, β-lactams, tetracyclines, macrolides, phenicols, and rifamycins 42–46. Besides, the presence of MFS transporters would be predominantly responsible for resistance to fluoroquinolones, and macrolides 47–51. The presence of other SMR-type genes suggests horizontal gene transfer, thereby broadening the resistance spectra of macrolides, aminoglycosides, and tetracyclines 52,53. However, the presence of pmrF, ugd, and eptA, which each encode pmr-associated phosphoethanolamine transferases, indicates the potential for resistance to polymyxins. These enzymes modify lipid A by adding phosphoethanolamine, a process that reduces bacterial susceptibility to cationic antimicrobial peptides such as polymyxins 54–56. These genes either neutralize antimicrobials or modify bacterial structures to facilitate multifactorial drug resistance of the strains. The AMR determinants identified in this study are consistent with previous findings from Bangladesh 57,58. The analyzed isolates harbored a diverse array of antimicrobial resistance determinants, including aadA1, blaTEM, cmlA1, mdf(A), sul3, qnrS1, and tet(A), which aligns with the findings reported in the present study. The presence of these resistance genes in both APEC and commensal E. coli populations suggests that poultry production systems may serve as important reservoirs of antimicrobial resistance. In settings where antibiotic use is widespread and insufficiently regulated, strong selective pressure can promote the persistence and spread of bacteria carrying transferable resistance genes, facilitating their dissemination within microbial communities and potentially beyond the poultry environment 17,59. The strains harbored various MGEs, which may facilitate further antimicrobial resistance profile 60. Notably, the RND efflux regulator cpxA and the MFS-associated regulator hns were identified within MGEs, suggesting horizontal gene transfer 61. The identification of a mutated gyrA gene within an MGE supports the potential for rapid dissemination of fluoroquinolone resistance traits 61. Within the genome of E. coli strain EEC_F3, the proximity of the bacA gene to MGE regions indicates a possible co-transfer of bacitracin resistance with other resistance loci. In all strains, the presence of AMR genes (qacL, ant(3’’)-IIa, cmlA1, aadA3, blaCARB-3, and dfrA16) adjacent to MGEs further supports the possibilities of horizontal resistance gene acquisition 61. These findings suggest that the MGEs in the E. coli strains may enhance the strain’s adaptability and increase the risk of resistance spread within the bacterial communities 61,62. The genome of the strains harbored key VAGs, including the ent gene cluster, which may act as a siderophore that facilitates iron acquisition in nutrient-limited environments 63. The presence of additional iron uptake and utilization genes (fep cluster) may enhance the strains’ ability to acquire and metabolize iron in nutrient-limited conditions 64. The identified fim gene cluster in the strains also suggests their capacity for adhesion, colonization, and host interaction 65. This gene cluster encodes the structural and regulatory elements required for the assembly and function of type 1 fimbriae, a crucial adhesin system in E. coli that facilitates attachment to host tissues and surfaces 66. The VAG determinants identified in the present study are largely consistent with findings in previous studies conducted in Bangladesh 57,58. However, an important distinction was observed in the detection of the ompA gene, a notable VAG determinant not reported in earlier studies. In all three strains, the identified ompA gene encodes a major outer membrane protein that contributes to cell envelope integrity, mediates attachment to host cells, and supports immune evasion 67. The identification of this gene in the current analysis highlights potential variability in the virulence profiles of E. coli strains circulating in poultry. Additionally, the iro gene cluster in the genomes of EEC_F3 and EEC_W5 encodes essential proteins responsible for the biosynthesis, modification, transport, and uptake of salmochelin 68. This C-glucosylated siderophore, derived from enterobactin, plays a crucial role in bacterial evasion of the host’s immune system 68. However, the absence of the iro gene cluster in the genome of APEC_AN-25 suggests that this deficiency reflects ecological specialization and genomic diversity, rather than reduced pathogenic potential 69. The genome of the strain EEC_W5 harbors an iuc gene cluster, which encodes enzymes required for the synthesis of aerobactin, a high-affinity hydroxamate siderophore required for bacterial viability under iron-deficient conditions 70. However, the gene cluster was not detected in the APEC_AN-25 and EEC_F3 isolates. The absence of this gene cluster in these faecal and anatomical isolates likely reflects niche-specific selective pressures. In these environments, alternative iron-acquisition systems may be sufficient to support survival and colonization 70. This distribution pattern highlights differential genomic adaptation driven by distinct ecological contexts 71,72. Unlike earlier research, the current study found a number of metal and heavy-metal resistance genes that had not been previously documented in similar studies from the region 57,58. These genes are associated with resistance to toxic metals such as arsenic, copper, tellurite, and manganese. Detailed genome annotation profiles of the strains identified several ars genes, suggesting the potential of the strains for detoxification and efflux of arsenic compounds 73. Therefore, the presence of ars suggests additional resistance mechanisms, possibly aiding improved survival of the E. coli strains in arsenic-polluted settings 73. The presence of cus genes in the strains’ genomes supports additional copper resistance. The cus cluster constitutes a tripartite efflux system, which includes cusA as the inner membrane transporter, cusB as the periplasmic adaptor protein, and cusC as the outer membrane channel 74. Additionally, cusF serves as the periplasmic copper-binding chaperone, while cusR functions as a two-component response regulator, and cusS acts as the sensor kinase 74,75. ogether, these components facilitate copper efflux under high copper stress, potentially conferring copper resistance in E. coli strains. The strains also had a possible mechanism of magnesium resistance, as suggested by the presence of the cor gene cluster. This gene cluster is involved in the transport of magnesium, required for maintaining the balance of essential divalent cations, and supports metal tolerance under stress conditions 76. Additionally, the presence of the mntP gene suggests possible manganese resistance in the strains, which encode a manganese efflux protein that facilitates the export of excess intracellular Mn²⁺, thereby preventing toxicity and preserving metal ion homeostasis 77. The identified tellurite resistance genes (tehA and tehB) further support resistance to tellurite 78. The coexistence of AMR markers and metal resistance genes in the genomes of the studied strains indicates the potential for co-selection between metal tolerance and antibiotic resistance. In settings where heavy metals and antibiotics coexist—such as poultry production systems that use feed additives, disinfectants, and environmental pollutants—bacteria may encounter concurrent selective pressures that promote the retention of genes conferring resistance to both stressors 79. The persistence of metal resistance determinants may indirectly facilitate the preservation and transmission of AMR genes throughout bacterial populations 80. The interaction between metal exposure and antimicrobial resistance highlights the need to recognize environmental pollutants as supplementary factors in the emergence and maintenance of AMR within agricultural ecosystems 81. The pangenome of avian strains revealed 1559 core genes, 798 soft-core genes, 4316 shell genes, and 18,911 cloud genes, suggesting continuous evolution and substantial genetic heterogeneity among strains 82,83. However, the pangenome of clinical and avian strains showed that E. coli strains from avian sources had significant genomic and phylogenetic similarity to clinical isolates from human sources. The EEC_F3 strain exhibited an intricate phylogenetic relationship with the clinical isolate E. coli strain E-9 from Bangladesh. Conversely, the strain EEC_W5, another strain in the study, showed strong phylogenetic proximity to the clinical isolate NGE-024. Nonetheless, the strain APEC_AN-25 had no substantial genomic or phylogenetic similarity to clinical isolates in Bangladesh. The results suggest that EEC isolates may have greater zoonotic potential than APEC isolates. Conclusion This study highlights APEC’s importance in the poultry industry and provides a comprehensive genomic characterization of three avian-associated E. coli strains, revealing significant antimicrobial resistance, metal tolerance, and potential zoonotic risks. The identification of several AMR genes and multidrug efflux systems indicates the presence of multiple resistance mechanisms that may reduce susceptibility to several important antibiotic classes. The detection of key AMR-associated genes, such as cpxA and hns, within mobile genetic elements may indicate horizontal gene transfer and contribute to the emergence or dissemination of MDR phenotypes. Notably, the identification of the ompA gene represents a distinct finding of this study compared with previous reports from Bangladesh, highlighting possible variation in the virulence profiles of poultry-associated E. coli strains. Another notable observation was the presence of multiple metal and heavy-metal resistance genes conferring tolerance to arsenic, copper, manganese, and tellurite. The coexistence of these resistance determinants with AMR genes suggests a co-selection mechanism where environmental exposure to heavy metals may indirectly promote the persistence and spread of AMR. Pangenome and phylogenetic analyses indicated that the EEC strains shared greater genomic similarity with human clinical isolates than the APEC strain, suggesting a potential zoonotic association. These observations support the value of ongoing surveillance and monitoring efforts to better understand the emergence and distribution of pathogenic and resistant E. coli strains within the poultry sector. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1 Supplementary Material 2 Supplementary Material 2 Supplementary Material 3 Supplementary Material 3 Supplementary Material 4 Supplementary Material 4 Supplementary Material 5 Supplementary Material 5 Supplementary Material 6 Supplementary Material 6 Supplementary Material 7 Supplementary Material 7