This page describes the structural basis for how the glucagon-like peptide-1 receptor (GLP-1R), is activated by both peptide ligands (GLP-1) and recently developed small-molecule antagonists (PF-06882961 and CHU-128). High-resolution cryo-EM structures reveal distinct ligand binding modes and explain why PF-06882961 mimics GLP-1 signalling more closely than CHU-128. This could be through water-mediated networks, TM6/TM7 rearrangements, and biased agonism. These findings offer important insights for the design of oral drugs.
You may include any references to papers as in: the use of JSmol in Proteopedia [1] or to the article describing Jmol [2] to the rescue.
Background
The GLP-1R is a class B1 G-protein-coupled receptor (GPCRs) essential for maintaining glucose homeostasis, regulating appetite, and supporting metabolic health. Activation of GLP-1R by the endogenous hormone GLP-1 enhances insulin secretion, slows gastric emptying, and promotes satiety.
Current GLP-1R therapeutics are peptide-based drugs, which are effective but require injection and often cause gastrointestinal side effects. These limitations have driven interest in developing orally available small-molecule agonists with similar beneficial effects.
Recent structural and pharmacological studies show that small molecules can bind and activate GLP-1R in ways that differ from peptide ligands. Some compounds, such as PF-06882961, closely mimic GLP-1–like signalling, while others exhibit strong pathway selectivity (biased agonism). Understanding these structural differences is crucial for designing next-generation oral GLP-1R agonists with improved efficacy, safety, and tolerability.
Pharmacalogical Profile of PF-06882961 and CHU-128
There are multiple GLP-1R pathways: cAMP, pERK1/2, intracellular calcium, β-arrestin recruitment, Gs conformational change, and receptor internalization. The small peptide agonists displayed different signalling behaviours across these pathways. Both non-peptides were potent and full agonists for cAMP production, showing ~30-fold lower potency than GLP-1 and similar ability to induce Gs activation. However, CHU-128 was inactive in every other pathway, even at high concentrations, indicating extreme Gs/cAMP bias.
In contrast, PF-06882961 activated all pathways, although with reduced potency. PF-06882961 also had only subtle biased agonism compared to GLP-1. Despite CHU-128 having ~10-fold higher binding affinity, PF-06882961 showed higher efficacy in cAMP signalling.
The data suggested that PF-06882961 behaves similarly to GLP-1 in overall signalling and receptor regulation, whereas CHU-128 displays narrow, cAMP-restricted profiles.
Structural Overview of GLP-1R Conformations Induced by Peptide and Small-Molecule Agonists
High-resolution cryo-EM structures of human GLP-1R in complex with GLP-1, PF-06882961, or CHU-128 gave an understanding of how these agonists engage and activate the receptor.
The GLP-1 bound receptor showed extensive peptide interactions across the transmembrane pocket, extracellular loops, and a deep water-mediated hydrogen-bond network critical for receptor activation.
Binding of the small molecule agonists seemed more superficial but was still distinct.CHU-128 occupied a planar orientation with limited overlap to the peptide and failed to engage key TM7 and water-network interactions, whereas PF-06882961 adopted a deeper, elongated structure that overlapped the GLP-1 N-terminal binding region and stabilised a rich structural water network similar to that observed for the peptide. This differential binding might explain why PF-06882961 has a broad signalling profile similar to GLP-1
Structural Basis for PF-06882961’s GLP-1-Like Efficacy
PF-06882961 drives the GLP-1 receptor into a GLP-1–like active state by pushing TM7 sharply inward, sealing the deep binding pocket and recreating the polar interaction network normally formed by the peptide’s N-terminal residues. Its carboxylate group mimics GLP-1’s D15, engaging R380 and stabilising a chain of interactions through E373, R310, and W306 that support efficient Gs coupling.
The ligand also recruits a structured water network and positions F385 into the same hydrophobic site used by GLP-1’s F12, allowing key TM1–TM2 residues to adopt their active conformations. These combined structural features explain how a small molecule can reproduce the signalling efficacy and receptor activation pattern of the native peptide.
CHU-128 Shows Biased Agonism
CHU-128 fails to pull TM7 inward or seal the deep binding pocket, thus the TM6–ECL3–TM7 region remains in a more open, flexible conformation. This prevents the formation of the continuous polar and water-mediated network that normally stabilises the fully active GLP-1R conformation.
As a result, CHU-128 can partially stabilise the deep polar core, which is sufficient for Gs coupling and cAMP production, but it cannot support the conformational features required for the other pathways described before. This produces a strongly biased signaling profile in which only the cAMP pathway is robustly activated, explaining its highly selective Gs-biased agonism.
Conclusion
Diabetes and obesity are major modern health concerns, and efforts have been made to develop orally delivered non-peptide GLP-1R drugs with little success. Only three compound series have progressed into clinical development. This work highlights how different small molecules engage and activate the receptor in distinct ways. The high-resolution cryo-EM structures reveal multiple ligand-specific binding pockets and active-state conformations, offering a detailed blueprint for structure-based drug design. Among the compounds studied, PF-06882961 most closely replicates the pharmacological actions of GLP-1, and the structural data explain the basis of this similarity. Overall, these findings provide new opportunities for developing more effective small-molecule GLP-1R agonists and could guide drug discovery efforts across other class B1 GPCRs.
Created for course BI3323-Aug2025
