Genotoxic Escherichia coli Strains Encoding Colibactin, Cytolethal Distending Toxin, and Cytotoxic Necrotizing Factor in Laboratory Rats Original paper

What was studied?

This original research study isolated and characterized genotoxin-encoding Escherichia coli from specific-pathogen–free (SPF) laboratory rats sourced from multiple commercial vendors and academic institutions, then tested whether these strains carried the colibactin (pks) island, cytolethal distending toxin (cdt), and/or cytotoxic necrotizing factor (cnf), and whether those genotypes produced the expected cytopathic effects in vitro. The investigators cultured E. coli from rat feces/rectal swabs, nares, and vaginal swabs, performed biochemical typing, used PCR to assign phylogroups and detect cyclomodulin genes, and then applied targeted phenotyping (HeLa cell assays), serotyping, and whole-genome sequencing on representative isolates to verify pathogenicity islands and related virulence content.

Who was studied?

The study evaluated 52 laboratory rats obtained from 3 vendors and housed across 4 academic institutions, sampled between 2015 and 2017, spanning ages 8 weeks to 2 years, with an approximately even sex distribution and including predominantly Sprague–Dawley animals alongside smaller numbers of Long Evans and transgenic rats. Investigators collected samples from clinically normal rats either immediately on arrival or after housing at the institutions and recovered 69 distinct E. coli isolates across the sampled body sites, enabling comparison of genotoxin prevalence by vendor and institution despite SPF status and the absence of E. coli from vendor surveillance reports.

What were the most important findings?

The central microbiome-relevant signal (MMA) was the high prevalence of genotoxin-positive, phylogroup B2 E. coli colonizing laboratory rats. Across 69 isolates, 65% carried pks, 29% carried cdt, and 6% carried cnf; importantly, cdt and cnf never appeared without pks, and cdt and cnf never co-occurred with each other, creating a consistent pattern in which colibactin served as the foundational genotoxic module and other cyclomodulins layered on top in subsets of strains. All genotoxin-positive isolates belonged to pathogen-associated phylogroup B2, strengthening the inference that these were not incidental commensals but pathobiont-like strains. Functional assays aligned with genotype: live pks+ isolates induced contact-dependent megalocytosis in HeLa cells (consistent with colibactin activity), while sonicates from cdt+ or cnf+ isolates produced the characteristic enlargement phenotype expected from those toxins, and genotoxin-negative isolates behaved like nonpathogenic controls. Serotyping and whole-genome sequencing further supported vendor-linked clustering and confirmed complete pks, cdt, and hemolysin–cnf pathogenicity islands in representative strains.

What are the greatest implications of this study?

This study shows that SPF status does not prevent colonization by genotoxic, B2 phylogroup E. coli, and that vendor source can materially change the burden of pks/cdt/cnf pathotypes entering an animal facility, creating a credible, underrecognized confounder for rat-based studies of inflammation, barrier function, carcinogenesis, neonatal infection, and antimicrobial interventions. Clinicians and translational researchers should interpret these results as a warning that “background microbiota” in rodents can carry functional toxins capable of DNA damage and cell-cycle disruption, meaning experimental outcomes may reflect hidden microbial genotoxicity rather than the intended exposure. The authors also raise a practical biosafety implication: detection of a serotype associated with human uropathogenesis supports vigilance for potential zoonotic transfer and reinforces strict hygiene and protective measures even in low-risk settings.