Common activation mechanism of class A GPCRs

Abstract

Class A G-protein-coupled receptors (GPCRs) influence virtually every aspect of human physiology. Understanding receptor activation mechanism is critical for discovering novel therapeutics since about one-third of all marketed drugs target members of this family. GPCR activation is an allosteric process that couples agonist binding to G-protein recruitment, with the hallmark outward movement of transmembrane helix 6 (TM6). However, what leads to TM6 movement and the key residue level changes of this movement remain less well understood. Here, we report a framework to quantify conformational changes. By analyzing the conformational changes in 234 structures from 45 class A GPCRs, we discovered a common GPCR activation pathway comprising of 34 residue pairs and 35 residues. The pathway unifies previous findings into a common activation mechanism and strings together the scattered key motifs such as CWxP, DRY, Na⁺ pocket, NPxxY and PIF, thereby directly linking the bottom of ligand-binding pocket with G-protein coupling region. Site-directed mutagenesis experiments support this proposition and reveal that rational mutations of residues in this pathway can be used to obtain receptors that are constitutively active or inactive. The common activation pathway provides the mechanistic interpretation of constitutively activating, inactivating and disease mutations. As a module responsible for activation, the common pathway allows for decoupling of the evolution of the ligand binding site and G-protein-binding region. Such an architecture might have facilitated GPCRs to emerge as a highly successful family of proteins for signal transduction in nature.

Keywords: GPCR; activation mechanism; allostery; computational biology; drug discovery; genetic diseases; human; molecular biophysics; signal transduction; structural biology; systems biology.

Figure 1.. An increasing number of reported class A GPCR structures facilitates studies on common activation mechanism.

(a) Distribution of structures in different states in the non-olfactory class A GPCR tree as of October 1, 2018. (b) Common GPCR activation mechanism and the residue-level triggers are not well understood.

Figure 1—figure supplement 1.. The pattern of conservation of residues and the map of number of disease-associated mutations in human class A GPCRs.

The alignment of 286 non-olfactory, class A human GPCRs were obtained from the GPCRdb (Hauser et al., 2018; Isberg et al., 2016; Pándy-Szekeres et al., 2018; Isberg et al., 2015) and sent for the sequence conservation score calculation for all residue positions by the Protein Residue Conservation Prediction (Capra and Singh, 2007) tool with scoring method ‘Property Entropy” (Mirny and Shakhnovich, 1999). To obtain disease-associated mutations, we performed database integration and literature investigation for all 286 non-olfactory class A GPCRs. Four commonly used databases (UniProt [The UniProt Consortium, 2017], OMIM [Amberger et al., 2011], Ensembl [Zerbino et al., 2018] and GPCRdb) were first filtered by disease mutations and then merged. Finally, we collected 435 disease mutations from 61 class A GPCRs (Figure 1—source data 2).

Figure 2.. Understanding GPCR activation mechanism by RRCS and ∆RRCS.

(a) Comparison of residue contact (RC) (Venkatakrishnan et al., 2016) and residue residue contact score (RRCS) calculations. RRCS can describe the strength of residue-residue contact quantitatively in a much more accurate manner than the Boolean descriptor RC. (b) RRCS and ΔRRCS calculation for a pair of active and inactive structures can capture receptor conformational change upon activation. (c) Two types of conformational changes (i.e. switching and repacking contacts) can be defined by RRCS to quantify the global, local, major and subtle conformational changes in a systematic way. (d) Two criteria of identifying conserved residue rearrangements upon receptor activation by RRCS and ΔRRCS. Thirty-four residues pairs were identified based on the criteria (see Materials and methods, Figure 2—source datas 1 and 2 for details), only six of them were discovered before (Venkatakrishnan et al., 2016).

Figure 2—figure supplement 1.. Calculation of RRCS and ΔRRCS.

(a) Workflow of RRCS calculation. (b) Examples of RRCS and ΔRRCS calculation for two residues pairs. (c) Statistics of residue contacts and contact types for six receptors (bRho, β₂AR, M2R, µOR, A_2AR and κ-OR) in their inactive and active states. Contact type describes physicochemical properties of two interacted amino acids that form a pair. The amino acids with hydrophobic side chains (one-letter code: A, V, I, L, M, P, F, Y, W) contribute to the majority of residue contacts, either within themselves (50.1%) or with other amino acids (37.8%).

Figure 3.. Common activation pathway of class A GPCRs.

Node represents structurally equivalent residue with the GPCRdb numbering (Isberg et al., 2016) while the width of edge is proportional to the average ∆RRCS among six receptors (bRho, β₂AR, M2R, µOR, A_2AR and κ-OR). Four layers were qualitatively defined based on the topology of the pathway and their roles in activation: signal initiation (layer 1), signal propagation (layer 2), microswitches rewiring (layer 3) and G-protein coupling (layer 4).

Figure 3—figure supplement 1.. Rearrangements of ligand-residue contacts in ligand-binding pocket are not conserved, reflecting diverse ligand recognition modes.

(a) Sphere representation of antagonist- and agonist-bound receptor crystal structures. (b) Diverse LRCS and ∆LRCS reveal the repertoire of ligand recognition across class A GPCRs. The agonist or antagonist was treated as a single residue when calculating LRCS and ∆LRCS. As shown by the calculated ∆RRCS, no ligand-residue pair exhibits conserved rearrangements upon activation. (c) Conserved conformational changes were only observed at the very bottom of ligand-binding pocket (6×48, 3×40 and 6×44).

Figure 4.. The common activation mechanism is the shared portion of various downstream pathways of different class A GPCRs.

(a) Intracellular binding partners used in the active state structures. (b) Comparison of RRCS for active (green) and inactive (orange) states of eight receptors with different intracellular binding partners, including four recently solved cryo-EM structures of G_i/o-bound receptors (5-HT_1B receptor, rhodopsin, A₁R and µOR) (Tsai et al., 2018; García-Nafría et al., 2018; Kang et al., 2018; Koehl et al., 2018; Draper-Joyce et al., 2018) whose resolutions were low (usually ≥3.8 Å for the GPCR part). Nevertheless, almost all conserved residue rearrangements in the pathway can be observed from them. Three of 34 residues pairs were shown here, see Figure 4—figure supplements 1 and 2 for the remaining 31 residue pairs.

Figure 4—figure supplement 1.. The switching conformation change is conserved upon receptor activation.

Comparison of RRCS for active (green) and inactive (orange) states of eight receptors with different intracellular binding partners, including four recently solved cryo-EM structures of G_i/o-bound receptors (5-HT_1B receptor, rhodopsin, A₁R and µOR) whose resolutions were low (usually ≥3.8 Å for the GPCR part). Nevertheless, almost all conserved residue rearrangements in the pathway can be observed from them. Nineteen of 34 residues pairs were shown here, see Figure 4 and Figure 4—figure supplement 2 for the remaining residue pairs.

Figure 4—figure supplement 2.. The repacking conformation change is conserved upon receptor activation.

Comparison of RRCS for active (green) and inactive (orange) states of eight receptors with different intracellular binding partners, including four recently solved cryo-EM structures of G_i/o-bound receptors (5-HT_1B receptor, rhodopsin, A₁R and µOR) whose resolutions were low (usually ≥3.8 Å for the GPCR part). Nevertheless, almost all conserved residue rearrangements in the pathway can be observed from them. Twelve of 34 residues pairs were shown here, see Figure 4 and Figure 4—figure supplement 1 for the remaining residue pairs.

Figure 5.. Common activation model of class A GPCRs reveals major changes upon GPCR activation.

(a) Active and inactive state structures form compact clusters in the 2D inter-helical contact space: RRCS_TM3-TM7 (X-axis) and RRCS_TM3-TM6 (Y-axis). GPCR activation is best described by the outward movement of TM6 and inward movement of TM7, resulting in switch in the contacts of TM3 from TM6 to TM7. (b) Common activation model for class A GPCRs. Residues are shown in circles, conserved contact rearrangements of residue pairs upon activation are denoted by lines.

Figure 5—figure supplement 1.. Global conformational change upon activation.

(a) Distinct clustering of inactive- and active-state structures in two-dimentional inter-helical contact space RRCS_TM5-TM6 vs. RRCS_TM3-TM6. (b) The inter-helical contacts comparison between inactive- and active-state structures. (c) Receptor-specific inter-helical contacts for all class A GPCR structures (inactive, intermediate and active states are colored in orange, cyan and green, respectively). These results demonstrate that receptor activation involves the elimination of TM3-TM6 contacts, formation of TM3-TM7 and TM5-TM6 contacts, reflecting the outward movement of the cytoplasmic end of TM6 away from TM3, the inward movement of TM7 towards TM3 and the repacking of TM5 and TM6.

Figure 5—figure supplement 2.. An inverse-agonism of class A GPCRs by preventing the collapse of Na⁺ pocket.

(a) Comparison of binding poses of antagonist (bosentan [Shihoya et al., 2017]) and inverse agonists (IRL2500 [Nagiri et al., 2019], BIIL260 [Hori et al., 2018] and ritanserin [Peng et al., 2018]). Inverse agonists diffuse deeper in the ligand-binding pocket to touch the Na⁺ pocket. These key residues around the Na⁺ pocket are shown. (b) Comparison of the sum of contact scores of the conserved residue pairs around the Na⁺ pocket (RRCS_{sodium_pocket}) between inactive- and active-state structures. The collapse of the Na⁺ pocket leads to a denser repacking of six residues (five residue pairs), reflected by higher RRCS_{sodium_pocket} scores compared to that of the inactive state structures. (c) Distribution of three inverse agonist bound structures in the 2D inter-helical contact space: RRCS_TM3-TM7 (X-axis) and RRCS_TM3-TM6 (Y-axis). These inverse agonists bound structures (5×33, 6BQG and 6K1Q) located in the inactive state region with zero RRCS_TM3-TM7 but high RRCS_TM3-TM6 scores, despite deeper binding modes.

Figure 6.. Experimental validation of the common activation mechanism.

(a) cAMP accumulation assay and (b) radioligand binding assay: both validated the common activation pathway-guided design of CAMs/CIMs for A_2AR. Wildtype (WT), CAMs and CIMs are shown in black, green and orange, respectively. (c) Mechanistic interpretation of common activation pathway-guided CAMs/CIMs design. N.D.: basal activity was too high to determine an accurate EC₅₀ value.

Figure 6—figure supplement 1.. Experimental validation of common activation pathway-guided CAM/CIM design for A 2A R.

(a) Cell surface expression of the WT A _2AR and its mutants. WT, CAMs and CIMs are colored by black, orange and green, respectively. (b) Mapping of validated CAMs/CIMs to the common activation pathway. (c) The mechanisms of CAM/CIM design. CAMs and CIMs are in green and orange, respectively.

Figure 6—figure supplement 2.. Experimental validation of common activation pathway-guided CAM/CIM design for G_s-coupled 5-HT₇ and G_i-coupled 5-HT_1B receptors.

(a) cAMP assay of G_s-coupled 5-HT₇ receptor used an EPAC-based BRET cAMP sensor (n = 6). WT, CAMs and CIMs are colored by black, orange and green, respectively. (b) Trafficking assay measured WT and mutant 5-HT₇ receptors in the plasma membrane (PM). (c) cAMP measurement of 5-HT_1B receptor. CAMs and CIMs are shown in green and orange, respectively. (d) Cell surface expression of mutant 5-HT_1B receptors relative to that of WT. DHE, dihydroergotamine.

Figure 7.. Importance of the common activation pathway in pathophysiological and biological contexts.

(a) Comparison of disease-associated mutations in the common activation pathway (further decomposed into layers 1–4), ligand-binding pocket, G-protein-coupling region and other regions. Red line denotes the mean value. (b) Mapping of disease-associated mutations in class A GPCRs to the common activation pathway. (c) Key roles of the residues constituting the common activation pathway have been reported in numerous experimental studies on class A GPCRs. Two hundred seventy two (272) CAMs/CIMs from 41 receptors were mined from the literature for the 14 hub residues (i.e. residues that have more than one edges in the pathway).

Figure 7—figure supplement 1.. The common activation pathway can be used to mechanistically interpret disease-associated mutations and CAMs/CIMs.

(a) Pathway-guided mechanistic interpretations of two disease mutations. (b) Pathway-guided mechanistic interpretations of four CAMs/CIMs.

Figure 7—figure supplement 2.. Residues in the common activation pathway are more conserved than other functional regions of GPCR.

(a) Illustration of different functional regions of GPCR. (b-d) Sequence pattern of the G protein-coupling region (b), ligand-binding pocket (c) and the common activation pathway (d). (e) Distribution of sequence identity (left) and similarity (right) for functional regions across 286 non-olfactory class A receptors.

Author response image 1.. Receptor activation process via intermediate structures (i.e., only agonist but not G protein bound structures) are in receptor- and ligand-specific manner, thereby highlighting the modular nature and complexity of the activation pathways.

Author response image 2.. Calculated RRCS for the HETX motif in class B GPCRs.