Understanding how toxins work opens doors to new ways of treating disease
For fifteen years, a quiet mystery persisted at the intersection of microbiology and oncology: how does a bacterium carried by one in five healthy people quietly set the stage for colon cancer? Researchers at Johns Hopkins, Harvard, and the Institute of Molecular Biology in Barcelona have now answered that question, tracing the destructive path of Bacteroides fragilis to a single receptor — claudin-4 — that the bacterium's toxin exploits as a doorway into colon tissue. The discovery, published in Nature, does more than explain a mechanism; it illuminates a potential point of intervention, reminding us that the most consequential battles for human health are often fought at scales invisible to the eye.
- A toxin hiding in plain sight for fifteen years has finally been caught in the act — latching onto the claudin-4 receptor to breach colon cells and dismantle their protective barrier.
- Once inside, the toxin destroys E-cadherin, the protein that holds the colon's defenses together, unleashing chronic inflammation that quietly cultivates the conditions for tumor growth.
- Using CRISPR gene-editing, doctoral student Maxwell White systematically stripped away candidate proteins until claudin-4 emerged as the unmistakable key — and crystallographic imaging from Barcelona confirmed the physical bond between toxin and receptor.
- A molecular decoy — a soluble protein engineered to intercept the toxin before it reaches colon cells — has already protected mice from the damage, offering a concrete blueprint for prevention.
- The path to human trials is long, but for the first time there is a precise biological target, transforming a fifteen-year-old mystery into a tractable medical problem.
For fifteen years, scientists knew that Bacteroides fragilis — a microbe found in roughly one in five healthy people — was somehow connected to colon cancer, but the mechanism remained out of reach. Now, a team spanning Johns Hopkins, Harvard, and the Institute of Molecular Biology in Barcelona has finally identified the missing piece: a receptor called claudin-4, which the bacterium's toxin uses as an entry point into colon tissue.
Once the toxin binds to claudin-4, it cleaves E-cadherin, a protein that normally acts as a protective barrier in the colon. With that barrier gone, chronic inflammation sets in and tumors begin to form. Researchers could observe the destruction of E-cadherin for years, but couldn't explain how the toxin was gaining access — until now.
The breakthrough was driven by Maxwell White, a doctoral student in Cynthia Sears's laboratory at Johns Hopkins, who used CRISPR technology to methodically test which proteins the toxin required to cause damage. When claudin-4 was removed from cells, the toxin could no longer attach and E-cadherin remained intact. Colleagues in Barcelona then provided crystallographic proof that the toxin and receptor were physically binding to one another — a mechanism with no known parallel among bacterial toxins.
With the biology confirmed, the team developed a molecular decoy: a soluble protein designed to intercept the toxin before it ever reaches colon cells. Tested in mice, the strategy worked. Sears framed the implications broadly, noting that understanding how bacterial toxins function opens new possibilities for detecting and treating colorectal cancer, diarrhea, and bloodstream infections. Human trials remain years away, but for the first time, researchers have a clear and precise target to aim at.
For fifteen years, scientists knew that a common bacterium was somehow triggering colon cancer, but the mechanism remained locked away. Now, researchers at Johns Hopkins University's Kimmel Cancer Center, working alongside colleagues at Harvard and the Bloomberg-Kimmel Institute for Cancer Immunotherapy, have finally cracked it open. The culprit is Bacteroides fragilis—a microbe present in roughly one in five healthy people—and the key to its destructive power is a receptor called claudin-4.
The bacteria produce a toxin that, once it latches onto claudin-4 on the surface of colon cells, sets off a cascade of damage. The toxin cleaves a protein called E-cadherin, which normally acts as a protective barrier in the colon. When that barrier breaks down, chronic inflammation takes hold, and tumors begin to form. For years, researchers could see that the toxin was destroying E-cadherin, but they couldn't explain how it was getting in the door. The missing piece was claudin-4—the receptor that acts as the toxin's entry point.
The breakthrough came through a combination of genetic tools and international collaboration. Maxwell White, a doctoral student in Cynthia Sears's laboratory at Johns Hopkins, used CRISPR gene-editing technology to systematically test which proteins the toxin needed to bind to in order to cause damage. Working with Matthew Waldor at Harvard, White discovered that when claudin-4 was removed from cells, the toxin could no longer attach, and E-cadherin remained intact. "It took time to get the experiment working," White recalled, "but once we did, it was clear that claudin-4 was the target." The finding was surprising enough that Sears and her team reviewed the existing literature on bacterial toxins and found no comparable mechanism—most toxins don't require an additional receptor to breach cells.
To confirm the toxin and receptor were physically binding to each other, the team brought in Xavier Gomis and Ulrich Eckhard from the Institute of Molecular Biology in Barcelona, who provided crystallographic evidence of the interaction. With that proof in hand, the researchers moved to living systems. At Harvard, they tested their findings in laboratory mice, creating a soluble protein that acted as a molecular decoy—a trap that intercepted the toxin before it could reach colon cells. The strategy worked. Mice treated with this decoy protein were protected from the tissue damage that Bacteroides fragilis toxin normally causes.
Sears, who leads the immunotherapy research at Johns Hopkins, framed the implications broadly. "Understanding how bacterial toxins function opens doors to new ways of detecting and treating associated diseases, including diarrhea, colorectal cancer, and bloodstream infections." The discovery, published in Nature, suggests that blocking claudin-4 or intercepting the toxin before it binds could become a preventive strategy for people who carry the bacteria. The next phase will be moving from mouse models to human trials, a process that typically takes years but now has a clear biological target to aim at.
Citações Notáveis
Understanding how bacterial toxins function opens doors to new ways of detecting and treating associated diseases, including diarrhea, colorectal cancer, and bloodstream infections.— Cynthia Sears, Johns Hopkins immunotherapy researcher
Once we got the experiment working, it was clear that claudin-4 was the target.— Maxwell White, doctoral researcher at Johns Hopkins
A Conversa do Hearth Outra perspectiva sobre a história
Why did it take fifteen years to find this receptor?
Because they were looking at the damage without understanding the mechanism. They could see the toxin breaking down E-cadherin, but not how it was getting access to the cells in the first place. It's like watching a thief steal from a house without knowing which door they're using.
And claudin-4 was the door?
Exactly. Once they used CRISPR to systematically test which proteins mattered, claudin-4 stood out. When they removed it, the toxin couldn't attach at all.
The bacteria are in one in five healthy people. Does that mean most people with the bacteria won't get cancer?
Right. The bacteria alone isn't sufficient. You need the bacteria, the toxin, the receptor, and probably other factors—genetics, diet, immune status. The bacteria is necessary but not sufficient.
So the molecular decoy they created—how does that actually work?
It's a trap. It's a protein that looks like the real target on colon cells, so the toxin binds to it instead. The toxin gets stuck to the decoy and never reaches the actual cells.
And it worked in mice?
Completely. The mice treated with the decoy were protected from tissue damage. That's why this matters—it's not just understanding the problem anymore. It's a potential solution.
What comes next?
Human trials, eventually. But first, more animal work to make sure the approach is safe and effective. Then the long road of clinical testing.