The Colibactin Hypothesis: What Gut Bacteria May Have to Do with It

When people talk about the microbiome and cancer, the conversation often stays at the level of generality: 'gut health matters,' 'bacteria affect everything.' The colibactin story is different. It is specific. A specific type of E. coli carries a specific gene cluster that produces a specific toxin that causes a specific pattern of DNA damage, and that exact pattern of DNA damage shows up in a subset of human colorectal cancers. This is not hand-waving. It is molecular epidemiology. But it is also not a settled question. The gap between 'this bacterium can damage DNA in a lab model' and 'this bacterium causes cancer in people' is significant, and this article will walk through exactly where the evidence stands as of early 2026.

What is colibactin, and which bacteria produce it?

Colibactin is a genotoxin, meaning it directly damages DNA. It is produced by strains of Escherichia coli (and some other Enterobacteriaceae) that carry a genomic island called the pks (polyketide synthase) locus. These strains are referred to as pks+ E. coli. The pks island encodes a series of enzymes that synthesize colibactin through a complex biosynthetic pathway. The final product is unstable and difficult to isolate, which delayed its structural characterization for years after its biological effects were first described.

The genotoxic activity of pks+ E. coli was first reported by Nougayrede et al. in 2006 in Science. They showed that infection of mammalian cells with pks+ E. coli caused DNA double-strand breaks, cell cycle arrest, and cellular swelling (a phenomenon called megalocytosis). Subsequent work confirmed that colibactin causes interstrand DNA crosslinks, which are among the most damaging types of DNA lesions and the same type of damage caused by some chemotherapy drugs (Wilson et al., Science, 2019).

pks+ E. coli are not rare exotic organisms. They colonize the gut of an estimated 20 to 40% of healthy adults in Western populations, with some studies reporting even higher prevalence. They belong to the phylogenetic group B2 of E. coli, which also includes many strains associated with urinary tract infections and neonatal meningitis. The pks island is transmitted via horizontal gene transfer between bacteria, meaning it can spread within the gut microbial community.

What did the 2020 Nature organoid study show?

The pivotal study linking colibactin to a cancer-relevant mutational signature was published by Pleguezuelos-Manzano et al. in Nature in 2020. The researchers exposed human intestinal organoids (miniature, lab-grown structures that replicate the architecture and cell types of the human intestinal lining) to pks+ E. coli over a period of 5 months.

After exposure, they sequenced the genomes of the organoids and found two distinct mutational signatures that were not present in control organoids exposed to pks-negative E. coli. These signatures, now cataloged as SBS88 (a single-base substitution pattern dominated by T>N mutations at specific sequence contexts) and ID18 (a small insertion/deletion pattern), are specific to colibactin exposure. The patterns are distinctive enough that they can be identified in whole-genome sequencing data as a fingerprint of prior colibactin activity.

The researchers then searched for these signatures in a large database of human colorectal cancer genomes. They found that approximately 5 to 10% of colorectal cancer samples contained the colibactin-associated mutational signature, with some studies reporting higher proportions depending on the cohort and detection methodology. The presence of this signature in human tumors means that colibactin had been actively damaging DNA in the tissue that eventually became cancerous. It does not prove that colibactin caused the cancer, but it places colibactin at the scene of the crime.

Is there a link between colibactin and young-onset colorectal cancer specifically?

This is one of the most active areas of investigation, and the data is preliminary. Several research groups have reported that the colibactin mutational signature may be enriched in young-onset CRC tumors compared to those from older patients, which would be consistent with the hypothesis that early-life microbiome exposures contribute to the rising incidence in younger adults.

A 2022 study by Lee-Six et al. published in Nature examined normal (non-cancerous) colon tissue from healthy donors and found the colibactin signature present in stem cell clones, suggesting that colibactin-mediated DNA damage is occurring in normal intestinal tissue throughout life, not just in the context of established tumors. The accumulation of these mutations over time could theoretically contribute to cancer initiation, particularly if colibactin exposure begins early in life when the microbiome is being established.

However, the evidence linking colibactin specifically to young-onset CRC (as opposed to CRC generally) is still circumstantial. The studies are small, and confounding factors (diet, geography, other microbiome differences between younger and older patients) have not been fully controlled for. This is an area where larger, well-designed studies are needed before drawing conclusions.

If colibactin-producing E. coli are common, why does not everyone get cancer?

This is the right question, and it is the main reason colibactin cannot be called a proven cause of CRC. If 20 to 40% of adults carry pks+ E. coli and the lifetime risk of colorectal cancer is roughly 4 to 5%, then the vast majority of people exposed to colibactin never develop the disease. Several factors likely explain this.

First, cancer is a multi-hit process. The prevailing model of colorectal carcinogenesis (the adenoma-carcinoma sequence) requires the accumulation of multiple driver mutations in specific genes (APC, KRAS, TP53, SMAD4, among others) over years to decades. Colibactin may contribute some of these mutations, but its damage alone is probably not sufficient. Other genetic and environmental factors are needed.

Second, the level of exposure matters. Carrying pks+ E. coli does not mean continuous, high-level colibactin production. The density of pks+ bacteria, their contact with the mucosal surface, the host's immune response, and the local colonic environment (pH, mucus layer integrity, co-colonizing bacteria) all modulate actual genotoxic exposure. A person with low-level colonization may experience minimal DNA damage.

Third, DNA repair mechanisms are continuously active. The body has multiple pathways for detecting and repairing DNA damage, including the crosslinks caused by colibactin. In most cases, the damage is repaired before it leads to permanent mutations. Individuals with impaired DNA repair (due to inherited mutations in mismatch repair genes, for example, as in Lynch syndrome) may be more vulnerable to colibactin-mediated damage, though this specific interaction has not been studied extensively.

What other gut bacteria are linked to colorectal cancer?

Colibactin-producing E. coli is not the only bacterium implicated in CRC biology. The two other most studied species are Fusobacterium nucleatum and toxigenic Bacteroides fragilis. Each operates through different mechanisms, and the field is increasingly recognizing that the microbiome's role in CRC likely involves multiple species acting through multiple pathways.

Fusobacterium nucleatum is an oral bacterium that colonizes colorectal tumors at high density. It has been found enriched in CRC tissue compared to adjacent normal tissue in dozens of independent studies. Mechanistically, F. nucleatum promotes tumor cell proliferation through its FadA adhesin protein, inhibits natural killer cell activity (suppressing anti-tumor immunity), and promotes a pro-inflammatory microenvironment. A 2022 paper in Nature showed that specific subtypes of F. nucleatum (Fna C2) are particularly enriched in CRC and possess specific genetic adaptations for colonizing the lower GI tract (Zepeda-Rivera et al., 2024).

Bacteroides fragilis toxin (BFT), produced by enterotoxigenic B. fragilis (ETBF) strains, triggers inflammatory signaling cascades in colon cells through cleavage of E-cadherin, activation of NF-kB, and induction of STAT3 signaling. Animal models have shown that ETBF colonization can promote colon tumor formation, particularly in the context of concurrent inflammation. In humans, ETBF has been found more frequently in mucosal biopsies from CRC patients than healthy controls (Dejea et al., mBio, 2015).

An important concept emerging from this research is that these bacteria may not act in isolation. Dejea et al. reported in Science in 2018 that pks+ E. coli and ETBF frequently co-occur in bacterial biofilms on the colonic mucosa of familial adenomatous polyposis (FAP) patients. This suggests a model in which multiple pathogenic bacteria cooperate within biofilm structures to create a pro-carcinogenic environment.

Can you test for colibactin-producing bacteria or do anything about them?

As of early 2026, there is no clinical test available to consumers or clinicians for assessing colibactin exposure or pks+ E. coli colonization. Standard stool microbiome tests (16S rRNA sequencing or shotgun metagenomics) can detect E. coli, but most do not specifically identify pks+ strains. Research-grade assays for the pks gene cluster exist, but they are not validated for clinical use and would not be actionable even if they were, because there is no proven intervention.

There is no evidence that probiotics, dietary changes, or any commercial product can selectively reduce pks+ E. coli colonization without affecting the broader gut ecosystem. Antibiotic treatment targeting E. coli would be non-selective and could cause more harm than benefit by disrupting the microbiome broadly. Some research groups are investigating bacteriophage therapy (viruses that target specific bacteria) and competitive exclusion strategies, but these are in preclinical stages.

This is an important point to emphasize: the colibactin research is currently informing our understanding of CRC biology, not guiding clinical practice. It is a research finding, not a consumer health finding. Anyone selling a product or test that claims to 'address' colibactin risk is ahead of the science.

What does this mean for tracking symptoms and staying informed?

The colibactin hypothesis does not change what you should do about colorectal cancer risk in any immediate, practical way. It does not mean your gut bacteria are giving you cancer. It does not mean you should test your microbiome or take supplements. What it does mean is that the relationship between gut bacteria and colorectal cancer is more specific and mechanistic than 'gut health matters,' and that research in this area is progressing rapidly.

From a practical standpoint, the same advice applies regardless of the colibactin research: follow screening guidelines appropriate for your age and risk factors, pay attention to symptoms that warrant evaluation (rectal bleeding, persistent bowel changes, unexplained weight loss), and bring any persistent concerns to your doctor. Tools like GLP1Gut can help you track symptoms and bring data to your doctor, which is relevant for any GI concern. But symptom tracking is about documentation, not cancer prevention.

Staying informed about this research is worthwhile because it may eventually lead to new screening strategies, new risk stratification tools, or even new prevention approaches. But we are not there yet, and anyone who tells you otherwise is overstating the evidence.