Fig 1 semaglutide · GLP-1R mechanism hover or focus a part · space: pause semaglutide · cAMP high · insulin high · glucose-dependent
Schematic, not to scale. Mechanism only: no doses, no outcomes, no comparisons between people or drugs. Educational, not medical advice.
Hover or focus a labelled part for a one-line explanation. Try low glucose with semaglutide: the receptor signal is high but insulin release is not triggered.
Fig 1b the molecule
native GLP-1(7–37) His7 Gly37 Lys26 Lys34 DPP-4 cuts after Ala8 semaglutide Aib8 (Ala8 replaced): DPP-4 resistant Arg34 (native Lys34) γGlu + 2×OEG linker C18 fatty diacid albumin at the receptor over time (schematic) more none native: DPP-4 cleaves it, fades semaglutide: persists time, compressed (no units) Fig 1c appetite and reward
GLP-1 receptors in the brain (rodent focus) front hypothalamus (arcuate) VTA nucleus accumbens area postrema / NTS engaged by GLP-1RA in rodent studies GLP-1R present; rodent local-activation data only Addiction-related effects in people: still emerging. Only small phase 2 trials (n = 48, 50); mixed.
Sources, and what is established
Educational only, not medical advice. This figure explains a mechanism. It gives no dosing, no schedule, no expected outcomes and no advice about who should or should not use any medicine. Talk to a clinician about medicines.
Numbers shown, and where they come from. 94% sequence homology to human GLP-1 (the label’s wording; consistent with 2 substitutions in the 31 residues 7–37, Lau et al., J Med Chem 2015 ); elimination half-life of approximately 1 week and more than 99% bound to plasma albumin , both in patients with type 2 diabetes (FDA label for semaglutide injection, DailyMed set id adec4fd2-…, sections 11, 12.1, 12.3 , read version 20, June 2026). The positions 8, 26, 34 are residue numbers in the GLP-1(7–37) convention. n = 48 and n = 50 are trial sizes (below). No other number is shown.The molecule. The label states Ala8 is modified for DPP-4 stability, Lys26 carries a hydrophilic spacer and a C18 fatty diacid for albumin binding (the main protraction mechanism), and a minor change at 34 ensures only one fatty diacid attaches. Lau et al. report the two substitutions as Aib8 and Arg34 with acylation at Lys26. PubChem’s name for semaglutide (CID 56843331 , molecular formula C187H291N45O59) lists a 2-amino-2-methylpropanoyl (Aib) residue second in the chain, an Arg at the Arg34 position, and the γGlu plus two 2-[2-(2-aminoethoxy)ethoxy]acetyl (OEG) linker leading to octadecanedioic acid on the Lys26 side chain, which is how the linker is drawn here. Native GLP-1 is cleaved at the N-terminus by DPP-4 (Deacon et al., Diabetes 1995 : intact GLP-1 was extensively degraded to the N-terminally truncated (9–36)amide after subcutaneous or intravenous administration in people). The bead chain is a schematic of the sequence and is not a structure-based model.Receptor and signalling. GLP-1R is a class B GPCR that signals mainly through Gs; the peptide is clasped between the extracellular domain and the transmembrane bundle (Zhang et al., Nature 2017, cryo-EM of GLP-1R–Gs ). GLP-1 raises cAMP in beta cells, which acts through PKA and through Epac2, and Epac couples cAMP to fast Ca²⁺-dependent exocytosis (Holz, Diabetes 2004 ). Glucose triggers secretion via metabolism, KATP channel closure, depolarisation and Ca²⁺ entry (Rorsman and Ashcroft, Physiol Rev 2018 , which also notes mouse versus human beta-cell differences). The label states that semaglutide stimulates insulin and lowers glucagon secretion in a glucose-dependent manner: when glucose is high insulin secretion is stimulated. Reviewed for GLP-1 physiology in Drucker, Cell Metab 2018 .Brain: appetite circuits (preclinical). In rodents, labelled liraglutide bound arcuate-nucleus neurons and needed GLP-1R (Secher et al., J Clin Invest 2014 ); semaglutide accessed brainstem, septal nucleus and hypothalamus through circumventricular organs without crossing the blood-brain barrier and activated several areas (Gabery et al., JCI Insight 2020 ).Brain: reward circuits (preclinical). In rats, GLP-1-producing neurons of the nucleus of the solitary tract project to the VTA and nucleus accumbens, and GLP-1R activation there reduced food intake, especially highly palatable food (Alhadeff et al., Endocrinology 2012 ). Those experiments did not test systemic semaglutide at these sites.Brain: people (early, mixed). One phase 2 randomised trial of once-weekly semaglutide in 48 non-treatment-seeking adults with alcohol use disorder reported lower alcohol intake in a laboratory task, with no change in average drinks per calendar day (Hendershot et al., JAMA Psychiatry 2025 ). A phase 2 randomised trial of oral semaglutide in 50 treatment-seeking adults reported no significant change in laboratory craving or drinks per day but fewer heavy drinking days (Schacht et al., Am J Psychiatry 2026 ). A systematic review describes the evidence for substance use disorders as promising but early, with preclinical and early clinical studies (Völker et al., Front Pharmacol 2025 ). Treat addiction-related effects as an open research question.
What was simplified. One beta cell and one capillary stand for the islet; mouse and human beta cells differ in channels and transporters (Rorsman and Ashcroft). Receptor numbers, sizes, positions and speeds are illustrative, and time is compressed. The glucose slider and the cAMP/insulin levels come from a qualitative model, not measurements: they show only that GLP-1R signalling potentiates a glucose-triggered response rather than starting it. Omitted: glucagon suppression, slowed gastric emptying, receptor internalisation and biased signalling, other incretin pathways, the vagus nerve, endothelial transport of the drug, all clinical risks and effects.
Preclinical means animal or cell work; clinical means studies in people. Do not read the brain panel as evidence about any effect in a person.