Injected straight into the fluid around the spinal cord, semaglutide ↗ eased pain in mice with fresh burns. The relief did not come from the drug acting on the injury. It came from the spinal cord's own support cells switching on the body's homemade opioids.
That is the finding of a mouse study published August 30 in Neuropharmacology ↗ by Qiang Fang and colleagues in the School of Basic Medical Sciences at Lanzhou University, in Gansu, China. Semaglutide is the molecule Novo Nordisk sells as Ozempic and Wegovy, a GLP-1 receptor ↗ drug that people inject under the skin for diabetes and weight loss. Here it was doing something unrelated to metabolism, and by an unusual route: a needle into the spinal fluid, not a shot in the belly.
What the mice showed
The researchers gave mice a second-degree burn, the kind of injury that leaves the skin painfully oversensitive. Two signs of that oversensitivity are mechanical allodynia, where a light touch that should not hurt does, and thermal hyperalgesia, an exaggerated flinch from mild heat. Both appeared, and at the height of the pain the spinal cord ramped up its production of the GLP-1 receptor, the same dock semaglutide grabs. The team tracked that rise with immunoblotting, antibody staining, and RNAscope in situ hybridization, three ways of counting a protein and the gene behind it inside tissue.
Intrathecal semaglutide, meaning the drug delivered into the fluid bathing the spinal cord, blunted both the touch pain and the heat pain. When the team added a compound that blocks the GLP-1 receptor, the relief faded. That is the control that matters, because it ties the effect to the receptor rather than to the injection itself. The pain relief also showed up in female mice, not just males, a detail that is easy to skip and often skipped.
The part that is new
Where the GLP-1 receptor sits is the twist. Most of it was on astrocytes, the star-shaped support cells that outnumber neurons and were long treated as scaffolding rather than signaling partners. Semaglutide raised spinal levels of Penk, the gene for proenkephalin, and of enkephalin itself, one of the opioids the body makes on its own. In dishes of spinal astrocytes, the drug pushed the same gene up and drove the cells to release more enkephalin into the surrounding fluid. In tissue, the enkephalin tracked with astrocytes more than with neurons or with microglia, the cord's resident immune cells.
Then the closing move. When the researchers neutralized enkephalin or blocked the delta-opioid receptor ↗, the dock those homemade opioids act on, semaglutide stopped working. The drug was not numbing the pain directly. It was telling astrocytes to hand over the body's own opioids, which then did the analgesic work at a receptor semaglutide never touches.
Why it is worth reading, and where to stop
This lands next to a growing pile of human data hinting that people on GLP-1 drugs report less pain. A cross-sectional study last week found GLP-1 users with diabetes had less painful nerve damage ↗, while conceding a single snapshot cannot say why. A spinal enkephalin route is exactly the kind of mechanism those associations have been missing. It is a candidate answer, not a confirmed one.
The reasons to hold back are the usual ones, and they are large. This is a mouse, one pain model, from a single lab. The abstract reports directions, not effect sizes, so the size of the relief is not on the table. Most of all, the drug went into the spinal fluid. Nobody takes semaglutide that way, and whether a subcutaneous dose reaches the spinal GLP-1 receptor in useful amounts is a separate, unanswered question. The enkephalin and receptor-blocking experiments are the strongest part, because they show the pathway is carrying the effect rather than merely being present.
What the study does is turn a loose clinical hunch into a testable circuit: a metabolic drug, a support cell, and the body's own opioids, wired together in the spinal cord. That is a lead worth chasing, not a treatment worth prescribing.