Antimicrobial resistance (AMR) poses an imminent threat to global public health, driven in part by the widespread use of antimicrobials in both humans and animals. Within the dairy cattle industry, Gram-negative coliforms such as Escherichia coli and Klebsiella pneumoniae stand out as major causative agents of clinical mastitis. These same bacterial species are frequently associated with severe infections in humans, including bloodstream and urinary tract infections, and contribute significantly to the alarming surge in antimicrobial-resistant bacterial infections worldwide. Additionally, mastitis-causing coliforms often carry AMR genes akin to those found in hospital-acquired strains, notably the extended-spectrum beta-lactamase genes. This raises concerns regarding the potential transmission of resistant bacteria and AMR from mastitis cases in dairy cattle to humans. In this narrative review, we explore the distinctive characteristics of antimicrobial-resistant E. coli and Klebsiella spp. strains implicated in clinical mastitis and human infections. We focus on the molecular mechanisms underlying AMR in these bacterial populations and critically evaluate the potential for interspecies transmission. Despite some degree of similarity observed in sequence types and mobile genetic elements between strains found in humans and cows, the existing literature does not provide conclusive evidence to assert that coliforms responsible for mastitis in cows pose a direct threat to human health. Finally, we also scrutinize the existing literature, identifying gaps and limitations, and propose avenues for future research to address these pressing challenges comprehensively.

Colistin resistance in Escherichia coli is of public health significance for its use to treat multidrug-resistant Gram-negative infections. Amino acid variations in PmrB have been implicated in colistin resistance in E. coli . In this cross-sectional study, 288 generic E. coli isolates from surveillance of broiler chicken and feedlot cattle feces, retail meat, wastewater, and well water were whole-genome sequenced. Phylogroup designation and screening for two amino acid substitutions in PmrB putatively linked to colistin resistance (Y358N, E123D) were performed in silico . Three additional data sets of publicly available E. coli assemblies were similarly scrutinized: (i) E. coli isolates from studies identifying the Y358N or E123D substitutions, (ii) colistin-susceptible E. coli isolates reported in the literature, and (iii) a random sampling of 14,700 E. coli assemblies available in the National Center for Biotechnology Information public database. Within all data sets, ≥95% of phylogroup B1 and C isolates have the PmrB Y358N variation. The PmrB E123D amino acid substitution was only identified in phylogroup B2 isolates, of which 94%–100% demonstrate the substitution. Both PmrB amino acid variations were infrequent in other phylogroups. Among published colistin susceptible isolates, colistin minimum inhibitory concentrations (MICs) were not higher in isolates bearing the E123D and Y358N amino acid variations than in isolates without these PmrB substitutions. The E123D and Y358N PmrB amino acid substitutions in E. coli appear strongly associated with phylogroup. The previously observed associations between Y358N and E123D amino acid substitutions in PmrB and colistin resistance in E. coli may be spurious. IMPORTANCE: Colistin is a critical last-resort treatment for extensively drug-resistant Gram-negative infections in humans. Therefore, accurate identification of the genetic mechanisms of resistance to this antimicrobial is crucial to effectively monitor and mitigate the spread of resistance. Examining over 16,000 whole-genome sequenced Escherichia coli isolates, this study identifies that PmrB E123D and Y358N amino acid substitutions previously associated with colistin resistance in E. coli are strongly associated with phylogroup and are alone not sufficient to confer a colistin-resistant phenotype. This is a critical clarification, as both substitutions are identified as putative mechanisms of colistin resistance in many publications and a common bioinformatic tool. Given the potential spurious nature of initial associations of these substitutions with colistin resistance, this study's findings emphasize the importance of appropriate experimental design and consideration of relevant biological factors such as phylogroup when ascribing causal mechanisms of resistance to chromosomal variations.

Background/Objectives: Due to the numerical dominance of environmental and commensal strains, understanding antimicrobial resistance (AMR) transmission in Escherichia coli requires consideration of non-clinical as well as pathogenic isolates. In this cross-sectional study, associations between the genetic context of non-clinical E. coli and AMR carriage are examined in isolates sampled from different niches within a One Health continuum. Methods: Two hundred eighty-eight E. coli isolates collected in Alberta, Canada (2018–2019) from wastewater, well water, feces of broiler chickens and feedlot cattle, and retail beef and chicken meat were selected from existing surveillance collections using a stratified random sampling structure. Using short-read whole genome assemblies, phylogenetic relationships were inferred from pan-genome multiple sequence alignments. Principal coordinate analysis and permutational analysis of variance (PERMANOVA) of a Jaccard dissimilarity matrix derived from gene presence/absence data were used to investigate contributions of source and AMR strata to observe genetic dissimilarity. Population clustering and gene under- or over-representation by source and cluster were also explored. Results: Minimal phylogenetic segregation of isolates was noted based on source or AMR strata, and both contributed significant but small proportions of observed genetic dissimilarity, with the largest proportion attributed to phylogroup. There was notable diversity of E. coli within and between sources; however, in some larger clusters, differential gene presence/absence was potentially linked to ecological niche rather than source of isolation. Conclusions: This study highlights the ecological complexity of AMR in E. coli in non-clinical contexts, offering a novel lens on how niche-specific factors can influence population structure and AMR carriage. It also provides insight into apparent discrepancies in the literature regarding clustering of E. coli by source. These findings support a more integrative One Health approach to AMR surveillance, emphasizing the need to account for microbial diversity and niche-specific adaptation across interconnected systems.