Semaglutide in the Spinal Cord: Enkephalin Does the Work
A group at Lanzhou University published work in Neuropharmacology on 30 August 2026 that does something the GLP-1 literature has mostly not done this year. It follows an effect all the way down to the molecule that produces it, and then removes that molecule to check.
The setting is a mouse model of second-degree burn injury. The intervention is semaglutide delivered intrathecally, directly into the spinal space. The finding is that the pain relief runs through enkephalin released by astrocytes, acting at the delta-opioid receptor.
A framing note. This is an animal study using a route of delivery that has nothing to do with how the compound is used in people, and semaglutide is not in the Peptra Labs catalogue. Nothing below is guidance about human use or about pain management.
What the semaglutide experiment established
The authors began by characterising the model rather than by treating it. They report that burn injury increased spinal GLP-1 receptor expression during the peak phase of pain hypersensitivity, using immunoblotting, immunofluorescence and RNAscope in situ hybridisation.
Where that receptor sat matters for everything after. Receptor immunoreactivity and Glp1r transcripts showed a substantial association with GFAP-positive astrocytic profiles, with detectable signal also present in microglia and in neurons.
Astrocytes, in other words, rather than neurons as the principal location.
Intrathecal semaglutide then attenuated mechanical allodynia and thermal hyperalgesia, and the authors report that those effects were reduced by pharmacological antagonism of the GLP-1 receptor. Acute responsiveness was also observed in female mice, which is worth stating because a great deal of preclinical pain work is run in males only and does not check.
The chain, and the two places it was cut
This is the part that separates the paper from most of what we have covered this month.
The authors report that semaglutide increased spinal Penk messenger RNA and enkephalin immunoreactivity in living animals, and that semaglutide increased Penk expression and extracellular enkephalin in primary spinal astrocyte-enriched cultures. Penk is the gene encoding proenkephalin, the precursor of the endogenous opioid peptides.
In spinal tissue, enkephalin immunoreactivity was more frequently associated with GFAP-positive profiles than with Iba1-positive or NeuN-positive profiles, which is to say more with astrocytes than with microglia or neurons.
| Step in the proposed chain | How it was supported |
|---|---|
| Receptor rises after injury | immunoblotting, immunofluorescence, RNAscope |
| Receptor sits mainly on astrocytes | GFAP colocalisation, signal also in microglia and neurons |
| Semaglutide reduces pain behaviours | mechanical allodynia and thermal hyperalgesia |
| Effect needs the receptor | pharmacological GLP-1 receptor antagonism |
| Semaglutide raises enkephalin | Penk mRNA and protein, in vivo and in culture |
| Effect needs enkephalin | enkephalin neutralisation |
| Effect needs the delta receptor | delta-opioid receptor antagonism |
Three separate blocking experiments, each removing one link. Antagonising the receptor cut the effect at the top. Neutralising enkephalin cut it in the middle. Antagonising the delta-opioid receptor cut it at the bottom.
That structure is what a mechanism looks like when it is demonstrated rather than proposed. Showing that something rises alongside an effect is correlation. Showing that removing it abolishes the effect is a necessity claim, and this paper makes three of them.
What changed since the earlier work
Spinal GLP-1 receptors and pain are not a new pairing. A 2014 paper in the Journal of Neuroscience reported that activating spinal GLP-1 receptors specifically suppresses pain hypersensitivity, and a 2015 paper in the British Journal of Pharmacology attributed the analgesia of a non-peptide agonist to beta-endorphin released from spinal microglia.
Set that against the new work and the difference is specific. The earlier account put the source in microglia and the mediator as beta-endorphin. This one puts the principal source in astrocytes and the mediator as enkephalin acting at the delta-opioid receptor.
The new paper does report signal in microglia as well, so these are not flatly incompatible. Different agonists, different injury models and different timepoints can genuinely engage different cells. But anyone citing a single sentence about how spinal GLP-1 receptor activation relieves pain is now citing two different mechanisms, and should say which.
What the semaglutide result does not show
The delivery route is the constraint, and it is a large one.
Intrathecal administration places semaglutide directly in the cerebrospinal fluid around the spinal cord, bypassing everything that determines where a subcutaneous injection ends up. Semaglutide reaches the central nervous system poorly by ordinary routes, which is the premise of a separate literature about why its central trials have disappointed. Nothing here says that systemically administered semaglutide does any of this.
The model is also acute and the readouts are reflexive withdrawal thresholds in mice. Mechanical allodynia and thermal hyperalgesia measure what they measure.
And there is no lasting claim about duration, tolerance or repeated administration in the abstract, so none should be inferred.
Why this matters to the rest of the month
We have written repeatedly since early August that the GLP-1 literature is built from outcomes and markers with no direct mechanistic measurement underneath it.
That has been a fair description of the clinical work, and it is worth saying plainly that this paper is the counterexample. It is in mice, by a route nobody uses, and semaglutide is not a compound we supply. Within those limits, the semaglutide pathway here is fully specified, with necessity tests at three points, and it is the sort of thing the clinical literature has been reasoning about without.
We covered a paper on the mitochondrial axis earlier in the month that pushed in a similar direction on a different system.
What this has to do with research compounds
The transferable content is the shape of the evidence rather than the compound.
The peptide with the largest pain literature in our catalogue is BPC-157, and that literature is almost entirely of the first kind: administration, then a measured improvement, then a proposed mechanism. We covered a Harvard review of the pain literature earlier this month and a rat tendon healing study, and our own summary of mechanisms and preclinical evidence is honest about how much of it is proposal.
The difference between a proposed pathway and one with a neutralisation experiment behind it, as in the semaglutide work above, is the single most useful thing to look for when reading a preclinical claim about any compound. Tirzepatide has a large clinical literature and a comparatively thin mechanistic one, which is the opposite imbalance and just as worth noticing.
Everything we supply is on a research use only basis, and our ranking of the most-studied research peptides counts primary studies for each compound.
References
- He Y, Gao J, Liu Y, Wu S, Kuang J, Chen D, et al. Intrathecal semaglutide attenuates burn injury-induced pain in mice through a spinal GLP-1R-linked enkephalin/delta-opioid receptor pathway. Neuropharmacology, 30 August 2026, 111163. doi 10.1016/j.neuropharm.2026.111163
- Gong N, Xiao Q, Zhu B, Zhang CY, Wang YC, Fan H, et al. Activation of spinal glucagon-like peptide-1 receptors specifically suppresses pain hypersensitivity. Journal of Neuroscience, 9 April 2014, 34(15):5322-5334
- Fan H, Gong N, Li TF, Ma AN, Wu XY, Wang MW, et al. The non-peptide GLP-1 receptor agonist WB4-24 blocks inflammatory nociception by stimulating beta-endorphin release from spinal microglia. British Journal of Pharmacology, January 2015, 172(1):64-79
The products referenced on this site are supplied for laboratory research use only. They are not medicines and are not intended for human or veterinary use. This article summarises published research for informational purposes and is not medical advice. Statements about third-party studies belong to their authors.
