U32 - In Situ Neutralization of Colibactin-Mediated Genotoxicity via Engineered Gut Commensals

Introduction: : Colorectal cancer (CRC) remains one of the leading causes of cancer-related mortality worldwide, with mounting evidence implicating gut microbial dysbiosis in its pathogenesis. Among the microbial contributors, Escherichia coli strains harboring the polyketide synthase (pks) island produce colibactin, a potent genotoxin that induces DNA double-strand breaks and promotes tumorigenesis. Notably, pks⁺ bacteria are present in over half of CRC patients and are enriched in individuals consuming Western-style diets. Despite increasing recognition of colibactin’s role in CRC initiation, no FDA-approved therapeutics currently exist to neutralize its genotoxic effects.
This project aims to fill that gap by developing a novel microbial-based therapeutic approach using engineered commensal E. coli to deliver ClbS, a colibactin-inactivating enzyme. Naturally produced by pks⁺ strains to prevent self-inflicted DNA damage, ClbS irreversibly hydrolyzes the reactive cyclopropane moiety of colibactin. We hypothesize that ClbS-expressing E. coli can effectively neutralize colibactin in situ, thereby reducing DNA damage and tumorigenesis in the gut.
Our approach integrates synthetic biology, organoid co-cultures, and mouse models to evaluate the efficacy and safety of surface-displayed ClbS. This work will provide mechanistic insights into colibactin-driven carcinogenesis, establish a platform for microbiome-based CRC prevention, and contribute broadly to the field of engineered microbial therapeutics. By targeting microbial genotoxins at their source, we propose a transformative strategy to mitigate CRC risk in individuals colonized by pks⁺ bacteria.

Materials and

Methods: : We engineered E. coli strains to express ClbS, a colibactin-inactivating enzyme, on the bacterial surface using an OmpA-based display system. Plasmid-based and genome-integrated constructs were generated and validated by Western blot and immunofluorescence.

To assess colibactin neutralization in vitro, human colon epithelial cell lines and primary colon organoids were co-cultured with pks⁺ E. coli (NC101) and ClbS-expressing E. coli. DNA damage was quantified by γH2AX staining and comet assay. For mutation profiling, organoids were co-cultured long-term, followed by whole-genome sequencing to detect colibactin-associated signatures. Therapeutic efficacy was assessed in C57BL/6 Apc min mouse model, with tumor burden, γH2AX staining, and histopathology as readouts. Safety was evaluated via weight tracking and colonic inflammation profiling.

Results, Conclusions, and Discussions:: Colibactin, a genotoxin produced by pks⁺ gut bacteria, has been implicated in colorectal cancer (CRC) development through DNA alkylation and strand breaks. Inspired by the natural self-protection mechanism of pks⁺ bacteria via the ClbS enzyme, we engineered non-pathogenic E. coli to express ClbS on their surface, enabling in situ neutralization of colibactin. By anchoring ClbS using an optimized outer membrane protein scaffold, we significantly enhanced its neutralization efficacy compared to secreted ClbS.
Surface-displayed ClbS effectively prevented colibactin-induced DNA damage in human and murine epithelial cell lines and colon organoids. In vivo, oral administration of ClbS-displaying E. coli reduced tumor burden in a colitis-associated CRC mouse model without inducing adverse effects. Furthermore, this strategy selectively inhibited the colonization of pks⁺ E. coli while sparing isogenic pks-deleted strains, indicating specificity toward colibactin-mediated ecological advantages.
Mechanistically, our findings suggest that ClbS-display may neutralize colibactin-mediated fitness advantages, altering bacterial competition dynamics. Future studies using germ-free or synthetic microbiota models will help clarify these interactions. Broadly, this work introduces a new therapeutic paradigm—targeting bacterial genotoxins with engineered microbes—offering a path forward for CRC prevention and deeper insights into microbiome-host interactions.

Acknowledgements and/or References (Optional)::