Claudin-4 Emerges as the Missing Entry Point for a Colorectal Cancer Toxin
For nearly two decades, colorectal cancer researchers have known that a toxin from Bacteroides fragilis can damage the colon and contribute to tumor formation. What remained elusive was the first step: how the toxin gained access to intestinal cells in the first place. A Johns Hopkins-led team has now identified the receptor, and the answer is strikingly specific. The toxin known as BFT binds to claudin-4, a tight-junction protein that helps seal neighboring colon cells together.
The finding closes a major mechanistic gap in a line of investigation that began with earlier evidence that BFT promotes colon tumors and disrupts E-cadherin, a protein central to epithelial integrity. Once the toxin reaches E-cadherin, the colon’s protective lining becomes easier to break down, inflammation can persist, and the tissue environment shifts toward tumor promotion. The new work clarifies the sequence: BFT does not simply strike its target at random. It first docks onto claudin-4, then proceeds to cleave E-cadherin.
That receptor step matters because barrier failure is not just structural damage. In the colon, tight junctions and adhesion complexes preserve a controlled interface between the intestinal lumen and underlying tissue. When those seals are disrupted, bacterial products, inflammatory mediators, and other luminal stressors can amplify local injury. A toxin that exploits a barrier protein to get inside that system is using the colon’s own architecture against it.
A CRISPR Screen Turned a Long Search Into a Clean Hit
The team used genome-wide CRISPR screening in colon epithelial cells to determine which genes were necessary for toxin attachment and injury. Rather than guessing at a candidate receptor, investigators systematically disabled genes and watched for loss of binding. One result stood out immediately: when the gene encoding claudin-4 was turned off, BFT could no longer bind the cells.
That result is unusually direct for a membrane-biology question. It links a bacterial virulence factor to a concrete host receptor, and it does so in a way that fits the observed biology. Without claudin-4, BFT loses its foothold and cannot efficiently reach E-cadherin. The toxin’s damaging chain reaction stops before it starts.
The receptor identity was unexpected. The investigators had anticipated a signaling receptor, such as a G protein-coupled receptor, because many microbial toxins exploit those pathways to enter cells or alter cell behavior. Claudin-4 belongs to a different category altogether. It is a structural protein embedded in tight junctions, the cellular seals that regulate paracellular movement in the intestinal lining. BFT has therefore been revealed as an enzyme with an unusual route of entry: a protease that requires a barrier-forming receptor before it can complete its attack.
Structural Evidence Confirmed a Direct Molecular Pairing
The receptor assignment was not left to genetics alone. Structural biologists collaborated to test whether BFT and claudin-4 physically interact, and laboratory measurements showed a tight complex between the two. Each toxin molecule paired with one receptor molecule, establishing direct biochemical evidence for the binding event.
That confirmation matters because it turns a genetic association into a molecular mechanism. The interaction is not indirect, and it is not merely inferential. BFT appears to recognize claudin-4 as its entry platform, then uses that contact to enable downstream injury to epithelial architecture. For a pathogen-derived toxin, the pathway is elegant and dangerous at the same time.
A single receptor can change the entire therapeutic logic. If the first step is a binding event, then prevention can focus on intercepting the toxin before it ever reaches the cell surface. That is a far cleaner strategy than trying to rescue colon tissue after barrier disruption has already begun.
A Soluble Decoy Protected Mice From Toxin Injury
The same receptor discovery immediately suggested a countermeasure. If BFT needs claudin-4 to attach to colon cells, a soluble mimic might pull the toxin away from its intended target. Investigators designed a molecular decoy containing claudin-4 sequences and tested whether it could intercept BFT before the toxin reached epithelial cells.
In mouse experiments, the decoy worked. Instead of binding to claudin-4 on colon cells, BFT attached to the substitute protein, and the animals were protected from BFT-induced damage. The result does not establish cancer prevention in humans, but it does show that toxin neutralization is biologically feasible in a living intestine. The intervention targets the earliest stage of injury, which is often the most efficient place to intervene when a bacterial product initiates chronic inflammation.
The approach also raises the possibility of future drug development beyond a protein decoy. Smaller molecules or other biologics could potentially be engineered to block the same interaction with better pharmacologic properties. Any such agent would have to compete with the toxin’s affinity for claudin-4 while remaining stable and deliverable in the gut, but the basic therapeutic direction is now visible.
What This Means for Cancer Biology and the Remaining Gaps
The discovery does not mean that every person harboring Bacteroides fragilis faces cancer risk. The bacterium can be present in healthy intestines, and only certain strains produce the toxin. What the new work shows is more specific: when toxin-producing strains are present, BFT has a defined molecular route into the colon barrier, and that route can be blocked in preclinical models.
One limitation remains significant. The exact three-dimensional structure of the BFT–claudin-4 interaction has not yet been captured, and advanced modeling tools have not fully resolved it. That missing structural detail may matter for the design of next-generation inhibitors, because precise shape complementarity often determines whether a compound merely binds or truly blocks an interaction with therapeutic strength.
Another gap is translational. The mouse data show protection from toxin-driven injury, not proof that the same strategy prevents colorectal cancer in people. Long-term human studies will be needed to determine whether intercepting BFT can alter tumor risk, reduce chronic inflammation, or affect related conditions such as diarrhea and bloodstream infection. The path from receptor discovery to prevention is now far shorter, but it is not complete.
For now, the most important advance is mechanistic clarity. A toxin long linked to colon tumor formation now has a confirmed host receptor, a validated route of entry, and a plausible neutralizing strategy. In a field where microbial contributions to cancer often remain diffuse and difficult to pin down, this is a rare case where the first molecular handshake has finally been identified.