Shiga toxin-producing *Escherichia coli* (STEC) causes human illness ranging from diarrhea to life-threatening hemolytic uremic syndrome (HUS). In North America, *E. coli* O157:H7 is the predominant serotype associated with human infections and outbreaks, often linked to contaminated beef. While cattle are considered a natural reservoir, the role of swine in STEC O157 infections remains debated. Unlike asymptomatic cattle, pigs can develop edema when infected with *stx₂*-harboring STEC strains. Although outbreaks linked to pork are rare globally, Alberta, Canada, has experienced three such outbreaks in four years (2014, 2016, 2018), prompting this investigation. Previous studies on STEC O157:H7 prevalence in pigs have yielded varying results, with reports ranging from no detection to a prevalence of up to 2%. A recent survey in Alberta found STEC O157:H7 in 1.4% of pigs and 1.8% of pork carcasses. The 2018 outbreak investigation also recovered STEC O157:H7 from pig fecal samples. This study aims to clarify whether pigs are transient carriers or if persistent STEC O157:H7 strains circulate on pig farms, utilizing both conventional and whole-genome sequencing (WGS)-based subtyping methods to analyze 121 isolates from pigs, cattle, and pork-production environments.

Numerous studies have investigated the prevalence of STEC O157:H7 in healthy pigs across various regions. Most studies found no or very low prevalence, except in Japan, Ireland, the UK, Sweden, Norway, the USA, and Canada, where prevalence reached up to 2%. These studies highlighted the need for a deeper understanding of the role of pigs in STEC O157 transmission and persistence. Contrasting results on the genetic relationship between STEC strains from cattle and swine exist. A Korean study found distinct RAPD patterns, while a Norwegian study reported similar PFGE patterns, emphasizing the need for a comprehensive genomic analysis. The existing literature lacks sufficient data on the genetic relationship between cattle and swine STEC strains. Previous studies have also highlighted the importance of different Shiga toxin subtypes in virulence, with *stx2a* being strongly linked to HUS. Therefore, characterizing the *stx* gene profiles in isolates from different sources is crucial.

This study analyzed 121 *E. coli* O157:H7 isolates: 41 from pigs, 51 from cattle, and 30 from pork-production environments in Alberta, Canada. Conventional subtyping methods (Manning clade, lineage-specific polymorphism (LSPA), MLST, Clermont phylo-group) were performed using in silico analysis. WGS-based methods (core SNP analysis, cgMLST, wgMLST, gene content analysis) were also employed. Core SNP analysis used *E. coli* O157:H7 EDL933 as a reference. Phylogenetic trees were constructed using UPGMA and maximum likelihood methods. The pan-genome was analyzed using Roary. Virulence genes (*stx* subtypes, *eae*, *paa*, *toxB*, *ehaA*, *saa*, *lpfA*, *ehxA*, *astA*, *estlA*, *subA*) and antibiotic resistance genes were identified using bioinformatics tools. PlasmidFinder was used to identify plasmid sequences.

Conventional subtyping showed limited strain diversity, with most isolates belonging to Manning clade 2, LSPA lineage I, MLST type 822 (Pasteur Institute's scheme), MLST type 11 (Achtman's scheme), and Clermont phylo-group E. WGS-based subtyping revealed higher discriminatory power, with core SNP analysis, cgMLST, and wgMLST clustering pig and environmental isolates into seven phylogenetic groups. Gene content analysis showed moderate correlation with core genome analyses. Most environmental isolates clustered closer to pig isolates than cattle isolates. A strong association was observed between 2018 outbreak isolates and isolates from farm 9. Pig isolates primarily possessed *stx₂a*, while cattle isolates frequently had both *stx₁a* and *stx₂a*. The *eae* gene was present in nearly all isolates. Only one pig isolate showed antibiotic resistance, while a larger proportion of cattle isolates exhibited resistance. Several plasmid replicons were identified, with some carrying antibiotic resistance genes. The presence of specific replicons correlated with phylogenetic groups.

The findings address the research question by demonstrating a close phylogenetic relationship between *E. coli* O157:H7 isolates from pigs and pork-production environments, suggesting pigs are a significant source of contamination. The limited strain diversity observed using conventional methods was significantly increased by WGS-based methods. The consistent clustering of environmental and pig isolates in multiple phylogenetic analyses strongly supports the link between pig farms and outbreaks. The prevalence of *stx₂a* only in most pig isolates, compared to the *stx₁a* and *stx₂a* profile in cattle isolates, suggests potential differences in virulence. The observed antibiotic resistance patterns suggest possible cross-contamination between cattle and pigs or environmental contamination. The identification of plasmids carrying antibiotic resistance genes highlights the potential for horizontal gene transfer.

This study provides strong evidence that pigs can serve as a significant source of STEC O157:H7 contamination in pork and pork-production environments. The close phylogenetic relationship between isolates from pigs and the environment, combined with the identification of outbreak-associated strains on pig farms, strengthens this conclusion. Future research should focus on investigating the mechanisms of persistence of STEC O157:H7 in pigs and the dynamics of transmission in pork-production systems. Further investigation into the role of different plasmid types in antibiotic resistance and virulence is also warranted.

The study is limited to isolates from Alberta, Canada, and may not be generalizable to other regions. The environmental sampling strategy might not have captured the full diversity of contamination sources. While a strong association was found between isolates from farm 9 and the 2018 outbreak, definitive proof of direct causation requires further epidemiological investigation. The study focused on a specific set of virulence and antibiotic resistance genes and may not capture the complete spectrum of genetic variation.