Subthreshold activation of the melanocortin system causes generalized sensitization to anorectic agents in mice
What limits GLP-1 drugs is often tolerability, not efficacy. Releasing an inhibitory melanocortin receptor made mice markedly more sensitive to those drugs — and, notably, no more sensitive to the malaise that usually comes with them.
The same pathway, released: MC3R holds feeding circuits under constant restraint, and removing that restraint changed how strongly they answered.
The limiting factor for GLP-1 drugs is often not whether they work but how much of them a patient can tolerate. Nausea and related malaise shape the dose-escalation schedules, and for some people they set a ceiling below the dose that would deliver the full effect. Any approach that widened that window — more anorectic effect per milligram, without more misery — would matter clinically.
A study from Roger Cone’s laboratory at the University of Michigan, with collaborators at the University of Illinois and Tomi K. Sawyer of Courage Therapeutics, proposes one, and it comes from an unexpected direction: not from the GLP-1 system at all, but from a brake sitting upstream of it.
A receptor that holds things back
The melanocortin-3 receptor is the less-studied sibling of MC4R, the receptor setmelanotide targets. Its role here is inhibitory. MC3R regulates GABA release from the nerve terminals of AgRP neurons, and in doing so tonically suppresses several circuits involved in feeding behaviour and energy balance. It is, in effect, a standing restraint on how strongly those circuits can respond.
Mice lacking MC3R were already known to be unusually sensitive to things that suppress appetite. They lose more weight in response to LPS or IL-1 treatment, to tumour implantation, and to behavioural stressors such as restraint and social isolation. The question this paper asks is whether that sensitivity extends to the drugs people actually take.
Three ways to release it, one result
It does. Careful dose–response curves showed that Mc3r−/− mice had increased sensitivity to the anorectic and weight-loss effects of GLP-1 agonists. Crucially, the same shift could be produced in three independent ways.
Deleting the receptor genetically worked. Blocking it pharmacologically with the MC3R antagonist C11 worked too, in both acute 24-hour and chronic three-day models — which matters, because it argues the effect is not merely a developmental consequence of growing up without the receptor, and could in principle be reproduced with a drug.
And approaching from the other side worked as well. Subthreshold doses of MC4R agonists — setmelanotide, and the compound CTX-1211 — reproduced the phenomenon. The word subthreshold is doing real work: these were doses that on their own had no effect on food intake or body weight, yet they still amplified the response to liraglutide.
The sensitisation was not specific to GLP-1. It extended to the acute satiety factors PYY3-36 and cholecystokinin, and to leptin, the long-term adipostatic signal. Removing the brake made the animals more responsive to anorectic signalling generally, whatever its source.
The part that separates effect from side effect
The most consequential observation is a negative one.
In striking contrast to the anorectic response, Mc3r−/− mice showed no increased sensitivity to the incretin effect of liraglutide, nor to the malaise it induces, as measured by a conditioned taste aversion assay. The two things came apart.
Mapping neuronal activation with Fos immunohistochemistry told the same story anatomically. Low-dose liraglutide produced increased Fos signal across multiple hypothalamic feeding centres in the knockout animals — but no increase in the area postrema, the region associated with emesis.
Our results indicate that the loss or pharmacological inhibition of MC3R can reliably hypersensitize animals to GLP1R agonists without promoting malaise or peripheral incretin effects.
The authors also narrowed the anatomy. Deleting Mc3r specifically in AgRP neurons increased responsiveness to both liraglutide and leptin, placing at least part of the mechanism in those cells.
What it does not establish
The authors are unusually clear about the limits of their own result, and the caveats are substantial enough to state plainly.
Mice do not vomit. They note that it will be important to assess whether MC3R antagonism can blunt GLP-1-induced emesis in a species capable of it, or whether comparable weight loss can be achieved at lower drug concentrations to avoid side effects. Conditioned taste aversion is a proxy for malaise, not a measurement of nausea in a patient.
The pharmacological tool is inadequate. C11 is difficult to synthesise, has limited brain penetrance, and had to be given by intracerebroventricular injection. To the authors’ knowledge, no potent, receptor-subtype-specific, brain-penetrant MC3R antagonist has yet been described — which means the most direct clinical version of this idea cannot currently be tested.
Mechanism remains partly open. Whether MC3R antagonism enhances setmelanotide’s effect through GABA and NPY released onto MC4R neurons, or through MC3R at sites beyond AgRP neurons, remains to be determined. MC3R expression is reported to be sexually dimorphic, and some sex-dependent responses appeared in this study as well; the circuit-level explanation is not yet in hand.
Why it is interesting anyway
Strip it back and the proposition is this: the dose-limiting side effects of GLP-1 drugs and their therapeutic effect are mediated by circuits that can, at least in a mouse, be separated — and the melanocortin system offers a lever on one without the other.
The authors put it no more strongly than the data allow, suggesting MC3R antagonists or MC4R agonists may have value in enhancing the dose–response range of obesity therapeutics, and that more effective MC3R antagonists will be needed to test the two approaches in parallel.
That the MC4R route already has approved chemistry behind it makes the subthreshold finding the more immediately testable of the two. A dose too small to do anything on its own, given alongside a GLP-1 agonist, changed how much that agonist achieved. Whether that survives the journey out of the mouse is the open question.
Source
Naima S. Dahir, Yijun Gui, Yanan Wu, Patrick R. Sweeney, Alix A.J. Rouault, Savannah Y. Williams, Luis E. Gimenez, Tomi K. Sawyer, Stephen T. Joy, Anna K. Mapp and Roger D. Cone. “Subthreshold activation of the melanocortin system causes generalized sensitization to anorectic agents in mice.” The Journal of Clinical Investigation, 2024, volume 134, issue 14, article e178250. DOI: 10.1172/JCI178250.
Affiliations: Life Sciences Institute, Department of Molecular and Integrative Physiology, and Department of Molecular, Cellular and Developmental Biology, University of Michigan; Department of Chemistry, School of Literature, Science, and the Arts, University of Michigan; Department of Molecular and Integrative Physiology, University of Illinois Urbana-Champaign; Courage Therapeutics, Newton, Massachusetts.
The underlying research is the work of the cited authors and their institutions. Maestro Chronicle provides the editorial presentation.