Every day our bodies quietly erase billions of expired cells. Now researchers in Kyoto have learned to redirect that natural cleanup system against cancer and autoimmune disease.
A team at Kyoto University’s Institute for Integrated Cell-Material Sciences (iCeMS) has developed a synthetic protein called Crunch, short for Connector for Removal of Unwanted Cell Habitat, that tricks the immune system into engulfing harmful cells. The breakthrough, published in Nature Biomedical Engineering, could offer a simpler, more adaptable treatment platform than current cell-based therapies.
Hijacking A Housekeeping System
Normally, immune cells known as phagocytes detect dying cells by their surface “eat me” signals and swiftly digest them, a process called efferocytosis. Crunch reprograms that process. By modifying a natural protein called Protein S, the researchers built a molecule that no longer binds only to dead cells but instead can be tailored to recognize specific live cells, such as tumor cells or rogue immune cells driving lupus. Once attached, Crunch links those cells to phagocytes, which finish the job of engulfment and disposal.
“What we’ve done is take that natural cleaning system and reprogram it to target living cells that shouldn’t be there,” said first author Yuki Yamato of Kyoto University.
In lab experiments and mouse models, Crunch cleared cancer cells engineered with identifiable markers. It also depleted overactive B cells in a lupus model, reducing tissue damage and disease signs. Crucially, Crunch does not kill cells directly. Instead, it labels them for removal, leveraging an ancient housekeeping pathway that already functions safely in every organ.
From Tumors To Autoimmunity
The team tested Crunch in multiple scenarios. In melanoma-bearing mice, tumors shrank significantly when Crunch was administered. In a lupus-prone strain, B cells were removed from blood and spleen, kidney inflammation subsided, and harmful autoantibodies declined. These results suggest the method can be tuned across very different diseases by swapping in new targeting sensors.
“We think this could become a new kind of therapy that can be adapted to many conditions,” said senior author Jun Suzuki. “It’s the ecosystem for the various therapeutic tools.”
Unlike chimeric antigen receptor T cells (CAR-T), which require harvesting, genetic modification, and reinfusion of patient cells, Crunch is a protein that could, in principle, be injected directly. That simplicity could make it more accessible and adaptable. The targeting component can be drawn from nanobodies, single-chain antibodies, or other recognition modules already in use for antibody drugs and CAR-T therapies.
Promise And Questions Ahead
There are hurdles before Crunch reaches the clinic. The protein must be manufactured at scale, optimized for long circulation in the body, and carefully tested for off-target effects. Because it depends on the presence of phagocytes, its effectiveness may vary across tissues. But the underlying concept, using the immune system’s existing waste disposal machinery rather than forcing cells to kill, offers a strikingly elegant strategy.
The metaphor is simple: instead of inventing new weapons, Crunch sticks labels on problem cells and lets the body’s janitors do what they already do best. If further work confirms its safety and versatility, this approach could open a new chapter in treating cancers, autoimmune diseases, and perhaps other conditions driven by destructive cells.
Explainer: How Crunch Works
Our immune system removes dying cells every day using phagocytes, cells that act like microscopic garbage collectors. These phagocytes recognize a chemical “eat me” tag on the surface of dead cells, swallow them, and recycle their parts. Crunch hijacks this system. Built from a natural protein called Protein S, it replaces the usual tag-recognition part with a sensor for a disease-causing cell. Once Crunch binds to its chosen target, it connects that cell to phagocytes. The phagocytes then engulf and digest the living cell, treating it as if it were already marked for clearance. Because the targeting module is customizable, Crunch can be adapted to label many different kinds of harmful cells.
Journal: Nature Biomedical Engineering
DOI: 10.1038/s41551-025-01483-9