Johns Hopkins Identifies Claudin-4 as the Docking Point for a Colon-Damaging Gut Bacterium Toxin

For nearly twenty years, researchers studying a toxin made by a common gut bacterium could see the damage it did to the colon but not the first step of the attack. They did not know what the toxin grabbed onto when it reached the cells lining the intestine. A team at Johns Hopkins has now identified that docking point, and the finding was published in Nature.

The bacterium is Bacteroides fragilis, which lives in the intestines of up to 20% of healthy people. Carrying it does not mean a person will develop cancer. But certain strains produce a toxin, called BFT, that disrupts the colon's protective lining, causes inflammation, and encourages tumour growth.

The link to cancer was established in a 2009 study in Nature Medicine, in which Cynthia Sears and her colleagues showed the toxin could promote colon tumours. Later experiments showed that BFT breaks down E-cadherin, a protein that holds neighbouring cells together and helps maintain the colon's barrier. When that protein is broken apart, harmful substances and bacteria can interfere with the tissue beneath.

A puzzle remained, though. BFT could degrade E-cadherin, yet it did not appear to attach to it directly. Researchers suspected another molecule was helping the toxin reach its target, but repeated attempts to find it failed. Sears, now the Bloomberg~Kimmel Professor of Cancer Immunotherapy and a professor of medicine at Johns Hopkins, is the study's senior author. "We've made several attempts over time to identify the receptor, so this is an exciting moment," she said.

The solution came through CRISPR, the gene-editing technology that lets scientists switch off genes one at a time and observe the effect. Maxwell White, an M.D./Ph.D. candidate in Sears' laboratory, led a genome-wide screen in collaboration with Matthew Waldor's laboratory at Harvard Medical School. The team disabled genes across colon epithelial cells, the cells forming the colon's inner lining, to find which were needed for the toxin to bind and cause harm.

One gene stood out. It encodes claudin-4, a protein that helps neighbouring cells form tight junctions, the seals that limit what passes between cells. When the researchers disabled that gene, BFT could no longer bind to the colon cells, and without that attachment it could not attack E-cadherin. Claudin-4, in other words, is the receptor: the toxin has to latch onto it first. White said it took a while to get the assay working and validated, but once the screen ran, claudin-4 was "a clear, resounding top hit."

The researchers have already used the discovery to build a molecular decoy that intercepts the toxin before it can damage colon cells. In mice, the strategy protected against the toxin's harmful effects. The work was led by scientists at the Johns Hopkins Kimmel Cancer Center, its Bloomberg~Kimmel Institute for Cancer Immunotherapy, and the Johns Hopkins University School of Medicine, with partial support from the National Institutes of Health.

Sears said understanding how bacterial toxins work could open doors to new approaches for detection and therapy in associated diseases, including diarrhoea, colorectal cancer and bloodstream infections. What to watch is whether the decoy moves beyond mice; no human trials have been reported.