The brain has its own cleaning crew capable of removing harmful deposits. In Alzheimer's disease, this system functions poorly. Researchers at Cold Spring Harbor Laboratory have now identified a molecular switch that reactivates these cleaning cells: the protein PTP1B. When it is inhibited, the amyloid plaques characteristic of Alzheimer's shrink and memory improves significantly. The study was published on February 2, 2026 in the Proceedings of the National Academy of Sciences (PNAS).
Microglia: The Brain's Janitors
Central to this approach are microglia, the immune cells of the brain. Their job is to clear away damaged cells, foreign bodies and harmful protein deposits. In Alzheimer's, the protein amyloid-beta accumulates into plaques that damage neurons. Microglia are supposed to clear these plaques but do so too little in Alzheimer's patients.
PTP1B is an enzyme that regulates microglia activity. More precisely, it brakes the SYK signaling pathway, which drives microglia to engulf and remove plaques. When PTP1B is active, the brakes are on. The team led by Yuxin Cen and Nicholas Tonks at Cold Spring Harbor Laboratory investigated what happens when those brakes are released.
What the Mouse Experiments Show
The researchers worked with mice genetically engineered to develop Alzheimer's symptoms, including amyloid plaques and memory impairment. In a first approach, PTP1B was genetically switched off. The microglia of these mice were more active and cleared more plaques. In one-year-old Alzheimer's mice raised without PTP1B, plaque burden was significantly lower than in untreated control animals.
In a second step, the researchers tested a pharmacological PTP1B inhibitor called DPM-1003, developed by the company DepYmed, in which the research group has a financial interest. Mice treated with DPM-1003 for five weeks had around 25 percent fewer amyloid plaques than untreated Alzheimer's mice. More strikingly, their performance in memory tests matched that of healthy control mice rather than untreated Alzheimer's animals.
The 25 percent reduction sounds modest but needs to be put in context: other approaches that achieved similar reductions in plaque burden did not produce this degree of memory improvement. The researchers suspect that PTP1B influences not only plaque clearance but also neuronal communication directly.
Multiple Mechanisms, Multiple Diseases
PTP1B is not an unknown protein. It has been studied as a potential target for type 2 diabetes and obesity for decades, because it plays a role in insulin and leptin signaling pathways. Previous attempts to develop PTP1B inhibitors for diabetes ran into problems with selectivity and side effects, without disproving the concept.
For Alzheimer's therapy, this dual function could be an advantage: a drug that inhibits PTP1B might simultaneously benefit dementia, type 2 diabetes and obesity. Whether that holds in humans remains entirely unknown. But pharmaceutical companies have strong interest in compounds with multiple mechanisms of action, which increases the chances of investment and development.
Where the Findings Fall Short
The key caveat: mice are not humans. Alzheimer's mouse models replicate important features of the disease, but human Alzheimer's is more complex and develops over decades. Many compounds that show impressive results in mice fail in clinical trials. Bridging the gap from animal to human remains one of the hardest challenges in neurology.
DPM-1003 has not yet undergone any human studies. Side effect profile, optimal dosage and long-term effects are unknown. DepYmed says it is working on next-generation compounds with better selectivity. Phase 1 clinical trials are planned but have not yet begun.
Where This Fits in Alzheimer Research
Several Alzheimer's drugs have been approved in recent years that target amyloid plaques directly, including lecanemab (Leqembi) and donanemab. These antibody therapies show moderate effects in early disease stages but are expensive, must be administered intravenously, and carry serious side effect risks including brain bleeding and brain swelling.
The PTP1B approach is complementary: rather than attacking plaques directly, it activates the body's own clearing system. The Tonks lab has already begun discussions with other research groups to assess whether combining this approach with existing Alzheimer's drugs could be beneficial. When first results from human studies might be available remains open. The research group has set a target of Phase 1 clinical trials within the next two to three years.