Semaglutide and the GLP-1 Receptor: What Binding and Receptor-Engagement Assays Report

Structural work has described how semaglutide occupies the glucagon-like peptide-1 receptor (GLP-1R). Cryo-electron microscopy structures and 3D variability analysis of semaglutide- and taspoglutide-bound GLP-1R-Gs protein complexes, generated in vitro from purified receptor-G protein complexes, reported peptide interactions broadly similar to those of GLP-1 but different motions within the receptor and within the bound peptides [1]. The authors framed these differences as molecular determinants of how distinct agonists engage the GLP-1R [1]. Competition radioligand binding is the assay format most often used to place semaglutide alongside other GLP-1R ligands. In one in vitro characterisation, competition binding experiments using either [125I]GLP-1(7-36)NH2 or [3H]orforglipron identified the nonpeptide agonist orforglipron as a high-affinity (inhibition constant Ki = 1 nM) and selective ligand of the human GLP-1R, while signal transduction assays in the same report showed low intrinsic efficacy for effector activation and negligible beta-arrestin recruitment [2]. The same work used the measured receptor Ki together with unbound in vivo concentrations in mice expressing the human GLP-1R to calculate predicted receptor occupancy, an animal experiment that reported low calculated occupancy accompanying a full biological response [2]. In CRISPR-Cas9 gene-edited rats carrying a sensitised receptor (Glp1rS33W), target engagement in pancreas and brain was compared with peptide-based GLP-1R agonists, and diet-induced obese animals of that line were used for comparisons between orally administered orforglipron and subcutaneously injected semaglutide [2]. Functional receptor assays have also been used to separate binding affinity from signalling bias. A study developing G protein-biased peptide GLP-1R agonists reported that the ligand modifications needed to reduce beta-arrestin recruitment usually also reduce GLP-1R affinity, and selected a lead compound with acute signalling potency comparable to semaglutide in vitro [3]. That work further reported biased agonist-specific GLP-1R internalisation profiles at clinically relevant pharmacological concentrations in vitro, cAMP signalling that was differentially modulated by single and double GLP1R coding variants observed in human populations when tested in cell assays, and blood glucose effects assessed in mice [3]. Semaglutide frequently appears as the in vitro benchmark in analog discovery programmes. A receptor activation assay in human GLP-1R-expressing cell lines, combined with a human serum albumin binding assay, reported that the long-acting analog GZR18 showed binding affinity for albumin and for the GLP-1 receptor similar to semaglutide and liraglutide, with pharmacokinetic profiling carried out separately in rats and cynomolgus monkeys [4]. In another medicinal chemistry series, in vitro functional assays indicated that GLP-1R/cholecystokinin-1 receptor co-agonists retained full agonism potency at both receptors, with subsequent comparisons against semaglutide performed in diet-induced obese and db/db mice [5]. Surface plasmon resonance measurements and plasma stability testing were used in vitro to rank albumin-binding fusion peptides before comparison with liraglutide and semaglutide in db/db mice and rhesus monkeys [6]. Assay development itself has been a subject of published work. Two cell-based assays for neutralising anti-drug antibodies, one against semaglutide and one against endogenous GLP-1, were optimised in vitro across three validation iterations; reported sensitivity for the semaglutide assay moved from 3,400 ng/mL to 98 ng/mL of antibody, and drug tolerance moved from 2.5 nM semaglutide when detecting 3,400 ng/mL antibody to 4.8-5.6 nM semaglutide when detecting 1,000 ng/mL antibody [7]. The authors identified the concentration of drug standard used for receptor activation, sample pre-treatment, and the binding properties of the control antibody as the factors governing those assay characteristics [7]. Receptor expression in the test system determines what a binding or proliferation assay can show. Quantitative PCR, immunofluorescence, and Western blotting across six neuroendocrine neoplasm cell lines reported that BON, NEC1452, and NEC1583 expressed significantly lower GLP-1R transcript and protein than GOT1, NT-3, and NEC913, and in vitro viability assays showed 19% and 22% growth increases in GOT1 and NT-3 after semaglutide exposure; in mice bearing GOT1 xenografts, tumour volume increased by 72% with semaglutide treatment [8]. Separately, a computational molecular docking analysis examined semaglutide binding affinity to GLP-1R protein, and transcriptome datasets from human liver samples reported GLP-1R downregulation in non-alcoholic fatty liver disease, particularly in advanced stages [9]. Binding assays relevant to semaglutide are not limited to the receptor. In a randomised, double-blind, three-period crossover study in 31 people with type 2 diabetes, total semaglutide exposure was unaffected by combination with insulin icodec, whereas maximum concentration was higher and occurred earlier for the combination; in vitro albumin binding studies and animal pharmacokinetic studies were cited by the authors as supporting competition for albumin binding locally at the injection site [10]. A review of the compound's development describes the addition of a C18 fatty-diacid side chain and specific amino-acid substitutions as the molecular modifications intended to enhance albumin binding, associated with a reported peptide half-life of approximately 160 hours in the clinical programme [11].