Figure 1.

Structural overview of the inactive and active states of the GLP-1 receptor and the list of the Tc5b-modified exenatide sequences used in this study. (a) Crystal structure of the ligand-free and thus inactive state of the GLP-1R, including the 7TM region and the ECD domain. The extracellular domain (ECD, green), the transmembrane helices (TM1–7, shades of blue), the intracellular (ICL1–3, shades of orange), and the extracellular (ECL1–3, shades of turquoise) segments are highlighted. Cryo-EM-derived structures of the active states of GLP-1R with its orthosteric ligands (red) (b) GLP-1 and (c) exendin-4/exenatide (Ex4). Their flexible stalk (depicted in black) and some ICL and ECL were not built into the models (due to the absence of corresponding densities), indicating the flexibility of these regions even in the activated state. Interaction partners on the cytoplasmic side (Ras-like domain of Gαs in complex with Gβ, Gγ, and Nb35) docked to the activated TM domain are not shown. (d) NMR ensemble of structures of Ex4 determined in 30% v/v trifluoroethanol-containing water. Ex4 adopts a Trp-cage tertiary structure, formed by the C-proximal helical part, a 3₁₀ helix, and a polyproline II helix, encompassing the central W25 residue. The α-helical part is also recognized by the ECD domain, whereas the latter two segments are absent in the cryo-EM structure of GLP-1R-bound Ex4 (c). In contrast, the N-terminal part of the ligand is highly dynamic in the solution (d) but becomes well-defined within the 7TM binding pocket of the receptor–ligand complex (c). (e) Single-letter amino acid sequences of Tc5b-modified Ex4 derivatives in this study. The superscript on the left indicates the number of residues of the truncated variants, while the subscript on the right indicates the applied amino acid mutation compared to the Ex4-Tc5b base sequence. Amino acid residues within the sequence at positions highlighted in red may contribute to either salt bridge formation or disulfide cross-linking.