The GLP-1–based obesity medication semaglutide reduces food intake and bodyweight through multiple behavioural mechanisms, notably by increasing satiation and satiety, suppressing food reward, and commonly producing nausea. The brainstem dorsal vagal complex (DVC) has been identified as a critical site mediating these phenotypic components of semaglutide′s anorectic action. This study combined metabolic and behavioural phenotyping with activity-dependent genetic labelling and chemogenetic manipulations to define which GLP-1 receptor (GLP-1R) populations within the DVC are engaged by semaglutide and to test whether distinct circuits mediate separable behavioural effects.
Semaglutide activated a substantial proportion of GLP-1R-expressing neurons in the area postrema (AP). Activation of GLP-1RAP neurons was functionally linked to downstream recruitment of neurons in the nucleus tractus solitarius (NTS), and this AP-to-NTS recruitment underlies several of semaglutide′s behavioural effects.
Mapping of semaglutide-activated neurons across the DVC revealed an unexpected dissociation: while the AP contained a large fraction of semaglutide-activated GLP-1R-expressing neurons, most semaglutide-activated neurons in the NTS did not express GLP-1R. Thus, semaglutide recruits a mixed neural ensemble in the DVC in which GLP-1R-expressing AP neurons appear to engage downstream, predominantly non-GLP-1R NTS neurons.
This anatomical and receptor-expression dissociation suggested that distinct DVC circuits could mediate separate behavioural components of semaglutide′s action, and raised the possibility that targeting downstream NTS populations might permit modulation of therapeutic versus side-effect profiles.
The investigators used an activity-dependent genetic labelling approach referred to as Sema-TRAP to identify neurons recruited by semaglutide in the brainstem. Following Sema-TRAP labelling, chemogenetic reactivation of the labelled neuronal populations allowed causal testing of whether reactivation alone could reproduce semaglutide′s effects. The study combined these circuit-mapping and manipulation techniques with metabolic and behavioural phenotyping assays to assess effects on satiation, satiety, behavioural proxies of nausea, motivation for palatable Western diet, and bodyweight. Reported findings were largely sex-independent.
Specific quantitative measures, exact sample sizes, and detailed protocol parameters for behavioural assays and molecular manipulations were not provided in the source summary.
Chemogenetic reactivation of Sema-TRAP–labelled neurons in the NTS alone was sufficient to recapitulate the acute phenotypic effects observed with systemic semaglutide administration. Reactivation produced increased satiation, induced behavioural proxies of nausea, suppressed motivation for a Western diet (interpreted as reduced food reward), and lowered bodyweight. These causal data indicate that NTS neurons engaged by semaglutide are capable of driving multiple anorectic and aversive components of the drug response.
Given that most semaglutide-activated NTS neurons did not express GLP-1R, the findings imply that semaglutide engages GLP-1RAP neurons that, in turn, recruit non-GLP-1R NTS populations to effect behaviour.
When GLP-1R expression in the AP was knocked down prior to Sema-TRAP labelling, recruitment of Sema-TRAP–labelled NTS neurons was abolished. Correspondingly, activation of the Sema-TRAPNTS ensemble that otherwise elicited increases in satiation, nausea, suppression of food reward, and bodyweight reduction no longer occurred following AP GLP-1R knockdown.
Notably, despite this blockade of AP-dependent recruitment of NTS neurons and the associated behavioural effects, semaglutide′s effects on satiety and bodyweight remained intact after AP GLP-1R knockdown. This dissociation indicates that semaglutide recruits multiple, separable circuits within the brainstem: some effects depend on GLP-1R signaling in the AP and subsequent NTS recruitment, while other effects, including satiety and bodyweight regulation, can be mediated independently of GLP-1RAP-dependent NTS recruitment.
The demonstration that semaglutide recruits dissociable anorectic circuits—GLP-1RAP neurons that engage downstream non-GLP-1R NTS neurons to drive satiation, nausea, and food reward suppression—suggests potential strategies to refine GLP-1–based therapies. Specifically, selectively targeting or modulating downstream NTS populations that mediate desirable metabolic effects while avoiding recruitment of pathways that produce nausea might enable better-tolerated GLP-1–based obesity drugs.
The authors propose that non-GLP-1R NTS neurons downstream of GLP-1RAP represent candidate therapeutic targets to tune the balance between efficacy and adverse effects. Details such as the precise molecular identity of these NTS neurons, translational validation in humans, and specific therapeutic interventions were not reported in the source summary and remain to be determined.
In sum, the work maps semaglutide-recruited circuits in the brainstem and provides causal evidence that GLP-1R signaling in the area postrema recruits downstream NTS neurons to mediate several acute behavioural effects of semaglutide, including increased satiation, nausea, and reduced food reward. Some semaglutide effects on satiety and bodyweight are preserved despite AP GLP-1R knockdown, indicating separable mechanisms and potential opportunities to refine GLP-1–based obesity treatments for improved tolerability.