Claudin-4 identified as the docking receptor for a colon-damaging gut bacterial toxin

Researchers at Johns Hopkins have identified how a toxin from the gut bacterium Bacteroides fragilis first attaches to cells lining the colon. The toxin, known as BFT, binds the tight-junction protein claudin-4, and a decoy protein built on this finding protected mice from toxin-induced damage. The work, led by Cynthia Sears, was published in Nature.

A long-standing gap in the mechanism

B. fragilis is found in the intestines of up to 20% of healthy people, and carrying it does not in itself signal cancer risk. Certain strains, however, produce BFT, which can disrupt the colon's protective lining, drive inflammation and encourage tumour growth. Sears and colleagues showed in a 2009 Nature Medicine study that the toxin promotes colon tumours in mice.

Earlier work established that BFT is a protease that cleaves E-cadherin, a protein that holds neighbouring epithelial cells together. Yet the toxin did not seem to bind E-cadherin directly. A separate receptor was suspected, but previous attempts to find it had not succeeded.

A genome-wide screen points to claudin-4

Maxwell White, an M.D./Ph.D. candidate in Sears' laboratory, led a genome-wide CRISPR screen with Matthew Waldor's group at Harvard Medical School. The team switched off genes one by one in colon epithelial cells and looked for those needed for the toxin to bind and cause harm.

The gene for claudin-4 emerged as the clear top hit. Claudin-4 helps form tight junctions, the seals that restrict the passage of substances between cells. When the gene was disabled, BFT could no longer bind the cells and E-cadherin was not cleaved.

Working with structural biologists F. Xavier Gomis-Rüth and Ulrich Eckhard at the Molecular Biology Institute of Barcelona, the team also showed in laboratory experiments that BFT and claudin-4 form a tight complex, with one toxin molecule paired with one receptor molecule.

An unexpected kind of receptor

The result was a surprise. The researchers had anticipated a signalling receptor, such as a G protein-coupled receptor. Claudin-4 instead serves mainly a structural role in linking cells. The team also noted that most proteases act directly on their substrates, whereas BFT appears to need a separate receptor first. A review of earlier literature turned up no other toxin known to work this way.

A decoy protects mice

If BFT requires claudin-4, a substitute might draw it away. The researchers designed a soluble decoy protein carrying claudin-4 sequences. In experiments with Min Dong's laboratory at Harvard Medical School, including Kang Wang, BFT bound the decoy rather than the receptors on colon cells, and the mice were protected from the toxin's damaging effects.

White said the approach could be developed further with small molecules or other biologics that have better pharmacological properties. The team is now exploring which options might work best.

Limitations and next steps

Several questions remain open:

• No experimental three-dimensional structure of the BFT and claudin-4 complex has yet been obtained, and AlphaFold has not fully resolved the interaction. A structure would help in designing compounds that block binding.
• The mouse experiments showed protection against toxin-induced tissue damage. They did not show that the strategy prevents colorectal cancer.
• The safety and efficacy of a decoy in people have not been tested.

For researchers, the study adds a mechanistic link between a toxin-producing gut bacterium and the breakdown of epithelial barrier proteins. It also offers a defined target for work on bacterial contributions to colorectal cancer, and potentially to other toxin-associated conditions such as diarrhoea and bloodstream infections, as the authors suggest.

Funding came from the Bloomberg~Kimmel Institute for Cancer Immunotherapy, Janssen Research and Development, Cancer Research UK, the National Institutes of Health and the Howard Hughes Medical Institute.

Source: Scientists finally discover how a common gut bacterium can trigger colon cancer (ScienceDaily: health and medicine, 31 October 2026).

Image: AI-generated illustration.