Blocking one protein regrew knee cartilage in mice and in donated human tissue

Cartilage in an old mouse’s knee is thinner than in a young one, and the difference tracks a single enzyme: 15-PGDH, which roughly doubles in the aged animals. Stanford Medicine’s Helen Blau and Nidhi Bhutani, senior authors of a paper in Science published Nov. 27, 2025, report that a small-molecule drug blocking that enzyme brought the cartilage back to normal thickness, and that cartilage taken from people’s replaced knees began regenerating after a week in the same inhibitor.

The result stays at the level of mice and laboratory dishes. No human trial of the approach for cartilage has been run, and the paper’s own senior author says the work is far from a prescription.

Aged mice and the injured knee

The enzyme belongs to a class Blau’s laboratory calls gerozymes, proteins that accumulate with age and degrade tissue. In Stanford Medicine’s account of the work, first authors Mamta Singla and Yu Xin Wang treated old mice with the inhibitor and saw knee cartilage restored to normal thickness. The animals also moved more normally and put more weight on the treated limb, a behavioral sign that the structural change translated into a working joint.

A second model mimicked an ACL-type tear, the kind of ligament injury that leads many people to osteoarthritis years later. Untreated mice developed osteoarthritis within four weeks. Mice that received the inhibitor were strongly protected, according to the journal citation and summary for the paper, titled “Inhibition of 15-hydroxy prostaglandin dehydrogenase promotes cartilage regeneration” and listed in Science volume 391, issue 6789.

The author list runs to 15 names, from Singla and Wang through Elena Monti, Yudhishtar Bedi and Pranay Agarwal to Stuart Goodman, with Blau and Bhutani last. Funding came from the National Institutes of Health, the Baxter Foundation and the Li Ka Shing Foundation, among others. The date matters: Stanford’s release is dated Nov. 27, 2025, roughly ten months before this week, so the work is not fresh data.

Knee-replacement tissue in the dish

The human evidence is smaller in scale and different in kind. Cartilage was taken from patients undergoing knee replacement, tissue that is by definition badly worn, and exposed to the inhibitor for one week. Stanford reports that the samples showed lower 15-PGDH and began generating functional articular cartilage.

One week in a dish says nothing about what happens in a living knee over months, how deep the effect reaches or whether the new tissue can bear load. The paper establishes that human cartilage cells carry the same switch. A patient’s joint, with its synovial fluid, inflammation and decades of mechanical wear, is a harder test that has not been attempted.

Cells that switch from scar to gristle

The cellular change is what makes the result more than a thickness measurement. In old chondrocytes, the cartilage-making cells, the share expressing 15-PGDH together with cartilage-degradation genes fell from 8 percent to 3 percent after treatment. The share expressing fibrocartilage genes, the stiffer repair tissue that forms after injury, dropped from 16 percent to 8 percent. Cells expressing the genes for hyaline cartilage, the smooth glassy kind that lines healthy joints, rose from 22 percent to 42 percent.

Blau describes the strategy as empowering cells already sitting in the joint to work better, as she explained in an interview with KTVU. That framing sidesteps stem cell transplants and implanted scaffolds: the inhibitor can be taken orally or injected locally into the joint, and it reprograms cartilage cells toward making collagen and the lubricating glycosaminoglycans.

Muscle drug turned joint drug

A related inhibitor is further along than the cartilage work. Epirium Bio’s oral compound MF-300 targets the same enzyme for sarcopenia, age-related muscle weakness, and in its Phase 1 results reported no serious adverse events among 70 healthy adults across single and multiple ascending doses. That safety record in healthy volunteers is the reason Stanford’s team expresses optimism about launching cartilage trials.

The need is large. The National Institute of Arthritis and Musculoskeletal and Skin Diseases describes osteoarthritis as the most common form of arthritis, with the knee among the joints most often affected, and treatment today runs from exercise, weight management and braces to medication and, when those fail, surgery. Stanford puts osteoarthritis at about 20 percent of the U.S. population and about $65 billion a year in direct medical costs, with no drug that slows or reverses it. KTVU’s report adds that over one million knee and hip replacements are performed each year in the United States.

Blau was blunt about timing. “This is not ready for prime time,” she told KTVU, and she estimated two to four years if development goes smoothly while stressing the uncertainty. As of Stanford’s account, no clinical trial has tested a 15-PGDH inhibitor on cartilage in people, so the first human data on the joint question remain unwritten.

This article was produced with the assistance of AI and reviewed by Morning Overview editors prior to publication.

More from Morning Overview