Dynamic Role of the G Protein in Stabilizing the Active State of the Adenosine A2A Receptor

Abstract

Agonist binding in the extracellular region of the G protein-coupled adenosine A2A receptor increases its affinity to the G proteins in the intracellular region, and vice versa. The structural basis for this effect is not evident from the crystal structures of A_2AR in various conformational states since it stems from the receptor dynamics. Using atomistic molecular dynamics simulations on four different conformational states of the adenosine A_2A receptor, we observed that the agonists show decreased ligand mobility, lower entropy of the extracellular loops in the active-intermediate state compared with the inactive state. In contrast, the entropy of the intracellular region increases to prime the receptor for coupling the G protein. Coupling of the G protein to A_2AR shrinks the agonist binding site, making tighter receptor agonist contacts with an increase in the strength of allosteric communication compared with the active-intermediate state. These insights provide a strong basis for structure-based ligand design studies.

Keywords: G protein-coupled receptors (GPCRs); adenosine receptor; allosteric communication; entropy; ligand mobility; molecular dynamics simulations.

Graphical abstract

Figure 1

Conformational Sampling of A_2AR Bound to NECA in Four Different States (A) Conformational ensembles from the MD simulations clustered by comparisons of the distances between TM3-TM6 and TM3-TM7. MD ensembles for A_2AR bound to NECA in four different states were projected on to these two distances and contour maps plotted for the C_α-C_α distances of R102^3.50-E228^6.30 and R102^3.50-Y288^7.53. The numbers 1, 2, and 3 in the figures correspond to the C_α-C_α distances in the crystal structures of inactive (PDB: 3PWH, number 1), active-intermediate (PDB: 2YDV, number 2), and the mini-Gs-bound fully active state of A_2AR (PDB: 5G53, number 3). (B) Representative structures extracted from the most populated cluster of A_2AR bound to the agonist NECA in the inactive state (R_NECA), the active-intermediate state (R′_NECA), and the fully active G protein-bound state (R^∗·G_NECA). The R^∗·G⁻_NECA state is a metastable state observed upon MD simulation of the receptor after removal of the G protein. The color scheme ranges from red to blue, with blue indicating low flexibility and red high flexibility. The flexibility is quantified by the B factor calculated from root-mean-square fluctuation in Å.

Figure 2

Free Energy of Agonist Binding to Conformational States of A_2AR The binding free energies were calculated using the Bennett Acceptance Ratio free energy perturbation method (see the STAR Methods); NECA, colored bars; adenosine, open bars. The error bars are the SD.

Figure 3

Mobility and Binding of NECA in Different Conformational States (A) Spatial distribution function of the agonist NECA calculated centering on the nitrogen atom from the primary amine group and the oxygen atom of the hydroxyl group in the sugar ring of NECA, blue and red arrows in (B) (see Figure S3 for data on adenosine). (B) The protein-ligand contacts for NECA binding in R_NECA, R′_NECA, R^∗·G_NECA, and R^∗·G⁻_NECA states of A_2AR. The protein-ligand contacts that are polar are marked in red and hydrophobic residue contacts are shown in blue. The percentage of snapshots within the MD simulations for each of these protein-ligand contacts is given (aggregated trajectory of 1 μs, 50,000 snapshots per calculation). N253 makes hydrogen bonds with two different N atoms on the adenine ring and the percentage shown is the sum of both.

Figure 4

Entropic Effects of EC and IC Regions in Various Conformational States (A) Torsional entropy of the residues in the extracellular (EC) and intracellular (IC) regions of A_2AR in various conformational states when bound to agonists NECA (solid colored bars) and ADO (white bars). The torsional entropy is shown in units of the Boltzmann constant k_B. (B) The thermal B factor calculated from the root-mean-square fluctuations (RMSF) using the formula B factor = (8π²/3) (RMSF)² of the C_α atoms in the EC regions are shown as a heatmap for the four conformational states of A_2AR.

Figure 5

G-Protein Coupling Leads to Increase in the Strength of the Allosteric Coupling Allosteric communication pipelines from the EC region of the receptor to the G protein coupling region in the NECA-bound A_2AR in various conformational states. The thickness of the pipelines shown is proportional to the strength of correlation in torsional angle motion of residues involved in this pipeline of communication.

Figure 6

Effect of Na⁺ Ions in Diverse Conformational States Effect of Na⁺ ion in the MD simulations of the inverse agonist ZM241385-bound inactive state R (R_ZMA241385/Na), agonist NECA-bound inactive state R (R_NECA/Na), active-intermediate state R′ (R′_NECA/Na), and G protein-bound fully active state R^∗·G (R^∗·G_NECA/Na). The Na⁺ ion retains the hydrogen bonds with residues D52^2.50 and S91^3.39 and a water-mediated hydrogen bond with W246^6.48 during the MD simulations of ZM241385-bound R state and as seen in the crystal structure of ZM241385-bound inactive state of A_2AR (PDB: 4EIY). Agonist NECA binding in the R′ and R^∗·G states disrupts the interaction of the Na⁺ with S91^3.39 but retains the interaction with D52^2.50. The waters present in the ZM241385-bound R state rearrange in the agonist-bound simulations.

Figure 7

The Effect of G Protein Coupling in Increasing the Ligand Affinity Going from the Active-Intermediate R′ State to G Protein-Bound Fully Active (R∗·G) State. (A) The ligand-receptor contacts that showing over 20% increase in population between R′_NECA and R^∗·G_NECA are shown. The numbers shown near each contact is the contraction in the average distance in each of these contacts going from R′_NECA to R^∗·G_NECA. (B) Representative structures of NECA binding site in R′_NECA (pink) and R^∗·G_NECA (green) states with the residues that show significant contraction of ligand-residue distances in (A). (C) The non-bond interaction energy (kcal/mol) between agonist NECA and the residues in the ligand binding site A_2AR in the inactive state (R, red), active-intermediate state (R′, orange) and fully active state (R^∗·G, black). (D) The residues shown by their Ballesteros-Weinstein numbering scheme located in the allosteric communication pipeline from the EC region connecting the G protein-coupling residues via the ligand binding site. The size of the sphere is proportional to their strength contribution to the allosteric pipeline. Residues shown in gray spheres show reduced affinity for agonist when mutated to alanine, and those shown in blue spheres have an increased affinity for agonist, and the maroon sphere residues show less than 10% change in ligand binding upon mutation to alanine compared with the wild-type. The allosteric communication residues to the nucleotide (shown as outline in the figure) binding site in the G protein are shown in green spheres. The G protein numbering is taken from the PDB structure of A_2AR with mini-G_s bound (PDB: 5G53). The A_2AR receptor is shown in green and mini-G_s is in light blue.