Allosteric communication mechanism in the glucagon receptor

Abstract

Recent drug development suggests agonists may be more promising against glucagon receptor dysregulation in metabolic disorders. Allosteric modulation may pave an alternative way to initiate responses that are required to target these metabolic disorders. Here, we investigated the allosteric communication mechanisms within the glucagon receptor using molecular dynamics simulations on five receptor states. Results highlighted that the extracellular domain is dynamic in the absence of an orthosteric agonist. In the presence of a partial agonist, we observed increased flexibility in the N terminus of the receptor compared with the full agonist-bound receptor. Class B1 G protein-coupled receptor (GPCR) microswitches showed repacking going from the inactive state to the active state, allowing for G protein coupling. In the full agonist- and G protein-bound state, Gα_s showed both translational and rotational movement in the N terminus, core, and α5-helix, thereby forming key interactions between the core of the G protein and the receptor. Finally, the allosteric communication from the extracellular region to the G protein coupling region of the receptor was the strongest in the intracellular negative allosteric modulator-bound state, the full agonist and G protein-bound state, and the full agonist-bound G protein-free state. The residue positions predicted to play a significant role in the allosteric communication mechanism showed overlap with disease-associated mutations. Overall, our study provides insights into the allosteric communication mechanism in a class B1 GPCR, which sets the foundation for future design of allosteric modulators targeting the glucagon receptor.

Keywords: G protein; G protein–coupled receptor; SNP; allosteric communication; class B1; diabetes; glucagon; glucagon receptor; metabolic disorders; molecular dynamics.

Figure 1

Computationally simulated systems employed in the present study. From left to right: intracellular NAM–bound (NNC0640) G protein–free inactive state (Protein Data Bank [PDB] accession number: 5XEZ); ligand– and G protein–free inactive state (PDB accession number: 5XEZ); the partial agonist–bound (NNC1702) G protein–free intermediate state (PDB accession number: 5YQZ); full agonist–bound (glucagon) G protein–free intermediate state (PDB accession number: 6LMK); full agonist– (glucagon) and G protein–bound (Gɑ_s–Gβ₁–Gγ₂) fully active state (PDB accession number: 6LMK). NAM, negative allosteric modulator.

Figure 2

Dynamics of the extracellular domain (ECD) in the glucagon receptor.A, RMSD of the ECD during MD simulations of the intracellular NAM–bound G protein–free, ligand– and G protein–free, partial agonist–bound G protein–free, full agonist–bound (glucagon) G protein–free, and full agonist– (glucagon) and G protein–bound (Gɑ_s–Gβ₁–Gγ₂) states of the glucagon receptor with reference to the ECD of the starting structure of the glucagon receptor. B, root-mean-square fluctuation (RMSF) of individual ECD residues during MD. C, representative snapshots from the most populated conformational cluster extracted from MD simulations of the different glucagon receptor states overlayed with the experimental starting structure. The snapshots were taken from each of the five independent simulation replicates after 500 ns and 1000 ns of simulation time (10 in total per receptor state). For the full agonist–bound (glucagon) G protein–free state, snapshots were taken at 4250 ns and 5000 ns simulation time (eight in total). Data represent the mean ± SD of five (four for the full agonist–bound [glucagon] G protein–free state) independent simulation replicates in A, whereas they represent the mean in B, statistical significance was assessed using a one-way ordinary ANOVA with Dunnett’s multiple comparisons test (ns is not significant; ∗p < 0.05; ∗∗∗p < 0.001). MD, molecular dynamics.

Figure 3

Ligand–receptor interaction differences between a partial and full agonist on the glucagon receptor.A, sequence comparison between NNC1702 and glucagon. B, representative snapshots from MD simulations of the ligand-binding site. The snapshots were taken from the unsupervised clustering (RMSD cutoff: 1.2 Ȧ) of the MD ensemble of snapshots (see Table S1 for more details) and then overlayed with the intermediate X-ray structure (Protein Data Bank [PDB] accession number: 5YQZ) and active state cryo-EM structure (PDB accession number: 6LMK), respectively. Root-mean-square fluctuation of (C) whole ligand or (D) individual residues in each ligand during MD. E, N-terminal contact map of NNC1702 and glucagon with the glucagon receptor. F, total nonbond interaction energies of N-terminal residues of the NNC1702 (residues 2–6) or glucagon (residues 1–6) with the binding site residues listed in E, the interaction energies are comprised of Coulomb and Lennard Jones energies in kJ/mol. G, closeup of N-terminal contact differences between NNC1702 and glucagon from the partial agonist–bound G protein–free state and full agonist– and G protein–bound state, respectively. Data represent the mean ± SD of five (four for the full agonist–bound [glucagon] G protein–free state) independent simulation replicates in C and F, whereas they represent the mean in D, statistical significance was assessed using a one-way ordinary ANOVA with Dunnett's multiple comparisons test (ns is not significant; ∗p < 0.05; ∗∗∗p < 0.001).

Figure 4

Repacking of microswitches in the glucagon receptor. A, 3D representation of the P^6.47Y(E)^6.53Q^7.49, P^6.47x(L)^6.48x(L)^6.49G^6.50, and H^2.50E^3.50T^6.42Y^7.57 microswitches in the glucagon-bound glucagon receptor in complex with Gɑ_s EM structure (Protein Data Bank accession number: 6LMK). B, population density plot of minimal distance in any side-chain atoms between the residues E362^6.53 and Q392^7.49 (1), population density plot of χ1 angle in Q392^7.49 (2), and closeup view of P^6.47Y(E)^6.53Q^7.49 microswitch in the glucagon receptor, highlighting dynamics among the different states of the receptor (3). C, population density plot of ɸ angle in P356^6.47 (1), closeup view of P^6.47x(L)^6.48x(L)^6.49G^6.50 microswitch in the glucagon receptor, highlighting dynamics among the different states of the receptor (2), population density plot of minimal distance in Cɑ atoms between the residues E245^3.50 and Y343^6.34 measuring TM3–TM6 distance transition (3), and closeup view of TM6 outward movement between inactive, intermediate, and active states of the receptor (4). D, population density plot of minimal distance in any side-chain atoms between the residues E245^3.50 and L357^6.48 (1), population density plot of minimal distance in any side-chain atoms between the residues E245^3.50 and T351^6.42 (2), and closeup view of H^2.50E^3.50T^6.42Y^7.57 microswitch in the glucagon receptor, highlighting dynamics among the different states of the receptor (3). Representative snapshots were taken from the top cluster after performing unsupervised clustering (RMSD cutoff: 0.85 Ȧ) of the MD ensemble of snapshots (see Table S3 for more details). Data represent five (four for the full agonist–bound [glucagon] G protein–free state) independent simulation replicates.

Figure 5

The G protein interface of the glucagon receptor is dynamic.A, principal component (PC) analysis of Gɑ_s dynamics consisting of the first (60%) and second (11%) PCs performed on Gɑ_s's Cɑ atoms. (B) PC1 and (C) PC2 representation of Gɑ_s portrayed onto the glucagon receptor experimental structure. The PC movement is illustrated using 30 representative snapshots taken from GROMACS function gmx anaeig. D, ɑ5-helix movement shown from PCs. E, segment-based receptor–Gɑ_s pairs identified through contact analysis. F, G protein interface of the glucagon receptor overlayed with experimentally validated positions that affect G protein signaling. Color gradients indicate low (gray) to high (orange) contact frequency of receptor residue with G protein. Data represent five independent simulation replicates.

Figure 6

Allosteric communication in the glucagon receptor.A, allosteric communication from the Gα_s coupling interface residues to various structural regions in the glucagon receptor displayed on the MD starting structure for each state. The lines represent the allosteric communication pipelines with the darkest color having the highest pathway population. B, pathway population for top five allosteric pipelines (see Fig. S6 for more details). C, distribution of hub scores along the allosteric communication pathways for the different glucagon receptor states. Statistical significance was assessed using a two-way ANOVA with Dunnett's multiple comparisons test (ns is not significant; ∗p < 0.05; ∗∗∗p < 0.001; between the intracellular NAM–bound G protein–free state and others). Data represent five (four for the full agonist–bound [glucagon] G protein–free state) independent simulation replicates.

Figure 7

SNP positions are part of the allosteric communication in the glucagon receptor. SNP locations and the allostery are portrayed on the full agonist–(glucagon) and G protein–bound receptor state, with the G protein removed from the final figure rendering. Red spheres indicate that these SNP positions have been associated with disease traits. SNP data were taken from the study by van der Velden et al. (25). Data represent five independent simulation replicates.