Colibactin: Biosynthesis, Mechanisms, and Detection Methods

Colibactin is a genotoxin produced by certain gut bacteria that damages host DNA, and its connection to colorectal cancer has made it one of the most intensely studied microbial metabolites of the past two decades. What makes colibactin particularly tricky to work with is that it is chemically unstable and produced in tiny quantities, which kept its full structure a mystery for years after its biological effects were first documented. The compound is built by a large cluster of bacterial genes, activated through an unusual prodrug mechanism, and has driven the development of creative detection strategies ranging from mass spectrometry to glowing chemical probes.

Which Bacteria Carry the Colibactin Machinery

Colibactin production is encoded by a stretch of DNA called the pks genomic island, roughly 54 kilobases long, that contains around 19 genes directing the molecule’s synthesis and export. The island was first characterized in certain strains of Escherichia coli, but screening of over 1,500 isolates across the Enterobacteriaceae family revealed it in other species as well, including Klebsiella pneumoniae, Enterobacter aerogenes, and Citrobacter koseri.¹ Further genomic comparisons showed the island is also conserved in Serratia marcescens and even the plant pathogen Erwinia oleae.²

Among E. coli, the pks island clusters heavily within phylogenetic group B2, a lineage known for causing infections outside the gut. Epidemiological work on hospitalized patients found that the colibactin-associated genes clbB and clbN identified a subset of B2 strains with extremely high virulence scores and a disproportionate representation among bloodstream isolates.³ Interestingly, colibactin output is not uniform across species. Klebsiella and Citrobacter koseri strains tend to produce colibactin more consistently and at higher levels than most E. coli isolates do.⁴

How Colibactin Is Built

The pks island encodes an assembly line of enzymes that belongs to a hybrid class, combining the logic of two major natural-product-building systems. One type, called nonribosomal peptide synthetases, stitches together amino acid building blocks. The other, polyketide synthases, extends a carbon chain using small organic acid units. In colibactin biosynthesis, both types work together as a multi-protein megacomplex, passing intermediates from one enzyme module to the next like items on a conveyor belt.⁵ One of the unusual polyketide building blocks used in this assembly line is aminomalonate, which is not a standard substrate for these kinds of enzymes.⁶

One of the more surprising biochemical twists in the pathway involves a cofactor called S-adenosylmethionine, or SAM. SAM is best known for donating methyl groups in countless cellular reactions, but a specialized module in the colibactin assembly line uses it as a structural building block to form an amide bond and then construct a cyclopropane ring, a strained three-membered carbon ring that turns out to be central to colibactin’s ability to damage DNA.⁷ That cyclopropane ring is effectively the warhead of the molecule, the chemical feature that allows it to react with DNA once activated.

The Prodrug Strategy

Colibactin is not made in its final, active form inside the bacterial cell. Instead, the assembly line produces a precursor called precolibactin, which is linked to a protective chemical tag, a scaffold built around D-asparagine. This tag keeps the molecule inactive while it travels through the cell’s interior, preventing it from attacking the bacterium’s own DNA. Once precolibactin is exported to the periplasm, the space between the inner and outer bacterial membranes, a peptidase enzyme called ClbP cleaves off the protective tag and releases the mature, DNA-damaging form of colibactin.

Structural studies of ClbP have revealed how it recognizes its target. The D-asparagine sidechain of precolibactin slots into a binding pocket where it forms hydrogen bonds with specific residues, including S188, H257, and N331. The orientation of N331 is itself locked in place by a chain of interactions involving E92 and K235. N331 is strictly conserved among similar prodrug-activating peptidases in other bacterial systems but is absent from the wider enzyme family, suggesting it evolved specifically to recognize D-asparagine-tagged prodrugs.⁸ This prodrug design is not unique to colibactin; other bacterial natural products use a similar strategy, but colibactin’s version has drawn particular attention because of the molecule’s relevance to human disease.

How Bacteria Protect Themselves from Their Own Toxin

Producing a compound that attacks DNA presents an obvious problem: how does the bacterium avoid poisoning itself? The pks island addresses this with a gene called clbS, which encodes a self-resistance protein. When clbS was deleted from a colibactin-producing E. coli strain, the bacteria could still damage host cells normally, but they activated their own SOS DNA-repair response and stopped replicating. That self-toxicity worsened dramatically when the bacteria also lacked a standard DNA repair pathway, confirming that ClbS is genuinely protecting the bacterium from its own product.⁹

ClbS works through at least two mechanisms. It can hydrolyze colibactin directly, breaking the molecule down before it reaches DNA, and it can also bind DNA, presumably shielding it from attack.¹⁰ This dual strategy gives the producing bacterium a robust safety net. The existence of ClbS also has evolutionary implications: bacteria that do not produce colibactin but share an ecological niche with producers face selective pressure to acquire resistance of their own, and evidence suggests this dynamic has shaped the distribution of drug-resistant versus colibactin-producing lineages in different regions.¹¹

Why Colibactin Took So Long to Characterize

Colibactin’s biological effects were described years before anyone could isolate and determine its structure, which is unusual for a metabolite attracting this much research attention. The core issue is that colibactin is chemically unstable. Its active form degrades rapidly, and the amounts produced by bacteria are minuscule.¹² When researchers finally managed to trap a mature colibactin derivative from wild-type E. coli, they did so by chemically converting its unstable central portion into a more stable form in situ. Even then, they estimated they were capturing only a few percent of the colibactin the bacteria actually produced.¹³

The eventual structural determination relied on a combination of genetics, isotope labeling, tandem mass spectrometry, and chemical synthesis.¹⁴ The picture that emerged was of a symmetrical molecule with two reactive cyclopropane warheads, each capable of alkylating a strand of DNA. When both warheads engage opposite strands, the result is an interstrand cross-link, one of the most dangerous forms of DNA damage a cell can sustain.

How Colibactin Damages DNA and What Happens to Cells

Recent structural work has clarified that colibactin alkylates DNA within the minor groove, preferring regions rich in adenine and thymine bases. This sequence preference explains the specific mutational patterns found in cells and tumors exposed to colibactin-producing bacteria. Unexpectedly, the chemically unstable central motif of the molecule is what mediates the sequence specificity of cross-linking, meaning the very feature that makes colibactin hard to study is the same feature responsible for its targeted damage.¹⁵

At the cellular level, the consequences of colibactin exposure go beyond simple cell death. Mammalian cells that survive an initial encounter with colibactin-producing bacteria display hallmarks of cellular senescence: chronic DNA double-strand breaks, prolonged cell-cycle arrest, and increased activity of senescence markers. These senescent cells do not just sit quietly. They secrete growth factors and other signaling molecules that can push neighboring cells toward proliferation, creating what researchers describe as bystander genotoxic and oncogenic effects.¹⁶ Even a brief, transient contact between cancer cells and colibactin-producing E. coli was sufficient to increase tumor growth in mouse models, sustained by this senescence-driven secretion of growth factors.¹⁷

Colibactin’s Fingerprints in Colorectal Tumors

Whole-genome sequencing of colorectal cancers has revealed specific mutational signatures tied to colibactin exposure. Two signatures in particular, known as SBS88 and ID18, are now considered characteristic markers. In a large Japanese cohort, these signatures were detected as early clonal events in about 45% of non-hypermutated colorectal cancer cases, meaning the damage was present from the earliest stages of tumor development.¹⁸ The fact that these signatures appear early suggests colibactin-mediated DNA damage may be an initiating event in a substantial fraction of colorectal cancers, rather than something that merely accelerates existing tumors.

Beyond DNA damage alone, spatial metabolomic profiling of right-sided colorectal tumors has shown that colibactin-producing E. coli can reshape the local tumor environment. Infected cancer cells accumulate lipid droplets and undergo phospholipid remodeling, creating an immunosuppressive microenvironment with fewer infiltrating CD8+ T cells. This lipid-driven survival mechanism also appeared to help cancer cells resist chemotherapy, and patients at advanced stages who were colonized by colibactin-producing E. coli had lower overall survival compared to those at earlier stages.¹⁹

Pathology Beyond the Gut

Colibactin’s role is not confined to colorectal cancer. In neonatal infection models, colibactin proved to be a virulence factor for E. coli K1, a serotype that causes bloodstream infections and meningitis in newborns. Inactivating the colibactin pathway genes clbA and clbP in a virulent K1 strain significantly reduced its ability to colonize the gut, cause DNA double-strand breaks, and produce lethal invasive disease in neonatal rats.²⁰ Similar findings emerged in avian pathogenic E. coli, where deleting the clbH gene, needed for colibactin synthesis, prevented the development of meningitis in a mouse model, with infected animals showing fewer clinical symptoms and less tissue damage.²¹

Colibactin also appears to contribute to chronic intestinal inflammation. In a mouse model where the mucosal barrier was transiently disrupted, colonization with colibactin-producing E. coli triggered chronic inflammation resembling ulcerative colitis, with significantly higher immune cell infiltration compared to mice colonized by strains lacking a functional colibactin pathway.²²

Detecting Colibactin-Producing Bacteria

Because colibactin itself is too unstable and scarce to measure directly in clinical samples, most detection strategies target either the genes encoding its production or the enzymes that activate it.

Gene-Based Approaches

The most straightforward method is PCR or quantitative PCR targeting genes in the pks island, such as clbA, clbB, clbN, or clbQ. This can be done on DNA extracted directly from stool samples without needing to culture individual bacterial strains. In one study, about 20% of stool samples tested positive for pks island genes, with copy numbers reaching up to 1.3 million per milligram of feces in some individuals.²³ A larger screening study detected E. coli in about 90% of participants and found that roughly a quarter carried pks-positive strains, though a one-time stool measurement did not distinguish individuals who went on to develop advanced colorectal lesions from those who did not.²⁴ That last finding is a critical caveat: carrying pks-positive bacteria in your gut at a single time point does not, by itself, predict cancer risk in a screening setting.

Enzyme Activity Probes

A complementary strategy sidesteps gene detection altogether and instead asks whether the colibactin-activating enzyme ClbP is actually functioning. Researchers developed fluorogenic probes, synthetic molecules mimicking precolibactin’s structure, that light up when ClbP cleaves them. In vitro characterization of ClbP’s substrate preferences guided the design of these probes, which can be used in high-throughput formats to identify colibactin-producing bacteria.²⁵ Refined versions of these probes were then adapted for clinical screening, allowing rapid identification of high-colibactin-producing strains directly from patient isolates.²⁶

More recently, chemiluminescent probes based on a different chemical scaffold have been developed. These phenoxy-dioxetane probes react selectively with ClbP and offer improved sensitivity and speed compared to the earlier fluorescent versions, with the added advantage of working directly in complex stool suspensions without purification steps.²⁷ The practical significance is that these probes can tell you not just whether pks genes are present, as PCR does, but whether the bacteria are actively producing and activating colibactin.

Blocking Colibactin at the Source

If ClbP is the gatekeeper that converts inactive precolibactin into the DNA-damaging mature form, then inhibiting ClbP should neutralize colibactin without killing the bacteria themselves. Two independent research groups have pursued this idea using boron-based small molecules designed to fit into ClbP’s active site. Crystallographic studies confirmed that these boronic acid compounds form a covalent bond with the catalytic serine of ClbP, effectively locking the enzyme in an inactive state.²⁸

In mouse experiments, the most effective compound suppressed colibactin-induced DNA damage by about 98% and prevented the tumor-promoting effects of colibactin-producing bacteria. In a colorectal cancer mouse model, animals receiving the compound in their drinking water developed roughly 3.5 times fewer tumors than untreated controls colonized by the same bacteria.²⁹ These inhibitors are still in preclinical stages, but the strategy is appealing because it targets the toxin rather than the bacterium, potentially avoiding the broad disruption of gut microbial communities caused by antibiotics.

Diet, Inflammation, and Colibactin’s Reach

The relationship between colibactin-producing bacteria and their host is not static; it is modulated by environmental factors, particularly diet and inflammation. In mice fed a low-fiber diet, mucosal inflammation increased, leading to diminished signaling through a receptor called PPAR-γ and elevated luminal nitrate levels. That nitrate boost promoted both the growth of pks-positive E. coli and colibactin-induced DNA damage. Supplementing the diet with soluble fiber in the form of inulin reversed both the inflammatory and polyposis phenotypes.³⁰ This work suggests that fiber may be protective in part because it reduces the conditions that allow colibactin-producing strains to thrive and do damage.

A separate line of investigation looked at putrescine, a polyamine naturally produced by certain probiotic E. coli strains. In vitro, putrescine inhibited the growth of a colibactin-producing strain in a concentration-dependent manner. In a mouse colorectal cancer model, putrescine supplementation reduced both tumor counts and tumor size in animals colonized by colibactin-producing bacteria.³¹ These findings are from animal models and it is too early to translate them into dietary recommendations, but they point toward the broader theme that colibactin’s impact depends heavily on context: what else is in the gut, what the host is eating, and how inflamed the intestinal lining is.

The Fitness Cost Puzzle

Producing a complex genotoxin is not free. The pks island is large, the biosynthetic machinery is elaborate, and the bacterium has to invest in self-resistance on top of that. A preprint analyzing the interplay between colibactin production and antibiotic resistance across E. coli populations suggests that carrying both capabilities may simply be too costly. Multi-drug-resistant lineages and colibactin-producing lineages appear to be largely incompatible: colibactin producers remain susceptible to most antibiotics and tend to dominate in regions with low antimicrobial use, while resistant lineages are more common where antibiotic use is high. In high-antibiotic-use environments, resistant strains may drive colibactin producers toward extinction, while in low-use settings, non-producing resistant strains face pressure to at least acquire immunity to colibactin so they can compete during gut colonization.¹¹ If this co-evolutionary dynamic holds up, it means antibiotic stewardship practices could indirectly influence the prevalence of a cancer-associated genotoxin in human gut communities, an unexpected connection between antimicrobial policy and oncology.