Weizmann scientists find the gut's scaffolding 'remembers' inflammation, driving IBD

Researchers at the Weizmann Institute of Science reported in the journal Immunity, announced on 26 August 2026, a new mechanism that may explain why inflammatory bowel diseases such as Crohn's disease and ulcerative colitis keep coming back. The study was led by Dr. Idan Adir in the labs of Prof. Irit Sagi and Dr. Moshe Biton in Weizmann's Department of Immunology and Regenerative Biology, with collaborators at Sheba Medical Center, the Technical University of Denmark and the University of Oulu in Finland.

Most research on these diseases has focused on the immune system. The Weizmann team looked instead at the extracellular matrix, the network of proteins and sugars that surrounds and supports the cells lining the gut. In mice, a single short bout of severe intestinal inflammation left the matrix damaged long after the animals seemed to have recovered: it became less stiff and more porous, and the changes persisted for more than 400 days. When the scientists grew healthy intestinal stem cells as organoids on this damaged scaffold, the cells did not form normal tissue. They became pro-inflammatory cells that release signals attracting immune cells, setting the stage for chronic disease.

The researchers traced the effect to collagen 18, a matrix protein that traps signalling molecules and disrupts the cues that guide stem cells. Silencing it prevented the lasting damage and blocked the development of chronic disease in mice. Analyses of single-cell data and biopsies from people with ulcerative colitis found the same pro-inflammatory cells in inflamed areas. Sagi said it is now clear that the matrix remembers and changes throughout our lives.

Biton noted that current drugs target the immune response, the end result, rather than the root of the problem. The team suggests that matrix changes could help predict where flare-ups will occur, that collagen 18 could become a drug target, and it is developing a blood test that detects matrix protein fragments. The work was done mainly in mice and lab-grown tissue, and how long such tissue memory lasts in humans is not yet known.