A Harvard-led Colibactin Study Reveals How Gut Microbes Cross-Link DNA to Raise Colorectal Cancer Risk

Colorectal cancer (CRC) is a significant and growing global health crisis. In 2022 alone, the Global Cancer Observatory recorded over 1.92 million new cases and more than 900,000 deaths worldwide, making CRC the second-leading cause of cancer-related mortality. In several high-income countries clinicians have also sounded alarms about a rising incidence among younger adults, a trend that complicates screening strategies and heightens interest in non-dietary drivers of risk such as genetic factors.

A research team led by researchers at Harvard University has for the first time resolved how colibactin, a genotoxic metabolite produced by gut bacteria such as E. coli, cross-links DNA and why that chemistry produces a distinctive mutational footprint associated with colorectal tumors. Published in Science on December 4, 2025, the report synthesizes chemical, structural and biochemical data to deliver a coherent model in which the gut bacterial toxin inserts itself in DNA’s minor groove and creates covalent interstrand cross-links at specific adenine–thymine (AT)-rich sequence motifs.

Methods and the Line of Evidence

To uncover exactly how colibactin damages DNA, the researchers used several complementary laboratory techniques, each answering a different part of the puzzle. Using mass spectrometry, the team pinpointed where colibactin forms stable chemical bonds with DNA. Nuclear magnetic resonance (NMR) spectroscopy then revealed how the toxin wedges into the DNA double helix and links the two strands together. 

In parallel, the Harvard team used carefully designed chemical probes, which are simplified versions of colibactin, to confirm which parts of the molecule are responsible for attaching to DNA. By combining these approaches, they were able to identify both the precise location of the DNA damage and the three-dimensional structure of the cross-link itself. Using multiple, independent methods strengthens confidence that the observed DNA lesions are directly caused by colibactin and explains why they leave a characteristic mutational pattern in the genome.

Sequence Specificity: Why Particular DNA Stretches are Preferred?

A key finding of the study is that colibactin does not damage DNA at random. Instead, the toxin targets specific short DNA sequences that are rich in adenine and thymine, binding within the minor groove of the double helix and attaching an alkyl group to the DNA. These preferred target sites match the “mutational fingerprints” previously seen in certain colorectal cancers. This selectivity helps explain longstanding observations that 5–20% of colorectal tumors show mutation patterns linked to colibactin exposure and establishes a direct mechanistic connection between a gut bacterial toxin and the DNA changes that can drive tumor development.

Chemical Mechanism: How Colibactin Causes DNA Cross-links?

Crucially, the researchers identified an α-amino ketone fragment—a small, chemically unstable core within colibactin—as the key element that guides the toxin to specific DNA sequences. Laboratory experiments that changed this part of the molecule caused colibactin to attack different DNA sequences, showing that it determines both how the toxin positions itself within the DNA double helix and how it links the two strands together. This mechanistic insight reframes colibactin not merely as a genotoxic “warhead” but as a structurally guided cross-linker whose internal chemistry explains why it leaves behind a consistent and recognizable pattern of mutations.

DNA Repair and Mutagenesis: How Lesions Become Mutations?

Interstrand cross-links (ICLs) are highly toxic DNA lesions that block both DNA replication and gene expression (transcription). Cells try to repair these lesions and stabilize stalled replication forks using specialized pathways, including the well-studied Fanconi anemia repair system. However, these repair processes can sometimes make mistakes, introducing DNA mutations. According to Emily Balskus, Professor of Harvard’s Department of Chemistry and Chemical Biology and one of the corresponding authors of the study, “An ICL means that your DNA-damaging agent reacts with both strands of DNA. It links the two strands of DNA together, creating a particularly toxic form of DNA damage to the cell.” 

The study highlights that many of the mutations linked to colibactin likely result from the cell’s attempt to fix the damage. In other words, colibactin starts the problem by linking the DNA strands, and the cell’s repair response can inadvertently turn that damage into permanent mutations that may contribute to tumor formation, especially if the repair machinery is not fully effective or the affected cells are already vulnerable.

Clinical and Translational Implications of Colibactin Toxicity

By identifying the specific DNA damage and the sequences colibactin prefers, the study opens up several potential applications for cancer research and therapeutics. First, the unique pattern of colibactin-induced mutations could help improve computational analyses of tumor genomes and develop tests to detect past exposure in tumor DNA. 

Second, the chemical changes caused by colibactin could serve as biomarkers that might be measured in stool, blood, or tissue to indicate local toxin activity. Third, the findings provide a scientific basis for strategies that either reduce the presence of harmful bacteria (such as targeting pks-positive E. coli, which carry the polyketide synthetase gene island that enables them to produce colibactin) or strengthen DNA repair in people at higher risk. All of these approaches will need thorough testing in clinical studies before being applied in practice.

Insights from Experts and Future Directions

The Harvard-led study supplies a critical piece of the causal puzzle by explaining how a bacterial metabolite can create persistent, mutation-generating DNA lesions. It is a major advance since it moves the link between colibactin and cancer from general correlation to a well-defined chemical and structural mechanism. 

The next important steps include long-term human studies that track markers of colibactin exposure and their connection to colorectal cancer risk, experiments to see if changing the gut microbiome can reduce these DNA lesions, and further research on how differences in people’s DNA repair systems affect their susceptibility. For clinicians and public-health planners, the study reinforces the microbiome as a modifiable domain of cancer risk deserving investment in surveillance and translational science.

Author

Richard Chau