Structural Insights into the Process of GPCR-G Protein Complex Formation

Abstract

The crystal structure of the β2-adrenergic receptor (β2AR) bound to the G protein adenylyl cyclase stimulatory G protein (Gs) captured the complex in a nucleotide-free state (β2AR-Gs^empty). Unfortunately, the β2AR-Gs^empty complex does not provide a clear explanation for G protein coupling specificity. Evidence from several sources suggests the existence of a transient complex between the β2AR and GDP-bound Gs protein (β2AR-Gs^GDP) that may represent an intermediate on the way to the formation of β2AR-Gs^empty and may contribute to coupling specificity. Here we present a structure of the β2AR in complex with the carboxyl terminal 14 amino acids from Gαs along with the structure of the GDP-bound Gs heterotrimer. These structures provide evidence for an alternate interaction between the β2AR and Gs that may represent an intermediate that contributes to Gs coupling specificity.

Keywords: G protein; G protein-coupled receptor; coupling specificity; intermediate state; protein engineering.

Figure 1:. Design of the β2AR-T4L-GsCT-CC construct and overall β2AR-T4L-GsCT-CC structure.

(A) Topology map of β2AR-T4L-GsCT-CC construct. (B) Agonist competition studies show β2AR-T4L-GsCT-CC and β2AR-T4L-GsCT exhibit higher affinity for the agonist isoproterenol (ISO) than do β2AR-T4L or β2AR (p < 0.0001) (n=3). The differences between β2AR-T4L-GsCT-CC and β2AR-T4L-GsCT are not statistically significant (p = 0.13). Statistics comparing were performed by extra sum-of-squares F test. Data are represented as mean ± SEM. (C) Overall structure of β2AR-T4L-GsCT-CC. (D) Structure alignment of β2AR-T4L-GsCT-CC structure with active β2AR structure (Protein Data Bank accession number 3SN6). (E) Alignment of β2AR-T4L-GsCT-CC structure with inactive β2AR structure (Protein Data Bank accession number 2RH1). TM6 in β2AR-T4L-GsCT-CC is almost identical to its position in the active structure, and is ~14Å away compared to inactive structure. See also Figure S1, S2, S3 and Table S1.

Figure 2:. Identification of an alternate interaction between the α5 helix and β2AR

(A and B) Two different views comparing the orientation of GsCT from β2AR-T4L-GsCT-CC structure (yellow) with GsCT from β2AR-Gs complex structure (cyan). (C) E392^Gαs interact with R131^3.50, R389^Gαs interacts with Y141^ICL2 and T68^2.39 in the β2AR-T4L-GsCT-CC structure. (D) Y391^Gαs forms cation-π interaction with R131^3.50, H387^Gαs interacts with Y141^ICL2 and Q142^ICL2 in β2AR-Gs complex structure (Protein Data Bank accession number 3SN6). See also Figure S4 and S5.

Figure 3:. The structure of the α5 helix in Gs^GDP

(A) Alignment of Gs^GDP with Gi₁^GDP (Protein Data Bank accession number 1GP2). (B) Alignment of Gs^GDP with Gq^GDP (Protein Data Bank accession number 3AH8). The overall structures are similar, but the C terminal end of the α5 helix was not resolved in Gi₁^GDP or Gq^GDP. (C) Y391^Gαs and H387^Gαs on the α5 helix interact with R38^Gαs, H41^Gαs and R42^Gαs on the αN-β1 loop. Cation-pi interaction is shown with a green arrow, hydrogen bonds are shown with green dashes. (D) Conformational changes in the GsCT in going from the Gs^GDP structure (orange) to the β2AR-Gs^empty complex structure (cyan, Protein Data Bank accession number 3SN6) involve disrupting an extensive network of hydrophobic and polar interactions between the α5 helix and the rest of the Ras domain. See also Figure S2 and Table S1.

Figure 4:. R389^Gαs and E392^Gαs are essential for efficient complex formation

(A) Fluorescence spectra of monobromobimane–labeled β2AR in presence of 1 µM WT Gs, 1 µM R389A/E392A Gs or 1 µM H387A/Y391A Gs. Each spectrum represents an average of 3 independent spectra. (B) β2AR triggered 3H-GDP release of WT Gs, R389A/E392A Gs or H387A/Y391A Gs (n=3). The statistically significant differences (p < 0.05) between the wild type and mutant are labeled with an asterisk (*). (C) Isoproterenol competition curves of β2AR reconstituted into high-density lipoprotein (HDL) particles, either as receptor alone or in presence of WT Gs, R389A/E392A Gs or H387A/Y391A Gs. The values and statistics of the agonist competition curves are shown in the inset table. Data are given as mean±SEM from 4–6 independent experiments performed in duplicate or triplicate. Statistics comparing WT and mutant Gs were performed by extra sum-of-squares F test. The differences between WT Gs and R389A/E392A Gs are statistically significant (p < 0.0001), so are the differences between WT Gs and H387A/Y391A Gs (p < 0.0001). However, the differences between R389A/E392A Gs and H387A/Y391A Gs are not statistically significant (p = 0.13). See also Figure S6.

Figure 5:. Interaction between the β2AR and R389^Gαs and E392^Gαs may initiate GDP release

(A) Interaction between the β2AR and R389^Gαs and E392^Gαs would affect interactions between H387^Gαs, Y391^Gαs and the αN-β1 loop. These changes could be propagated to the P loop and β6-α5 loop through β1 strand and α5 helix. (B) H387^Gαs interacts with H41^Gαs from β1 strand; Q390^Gαs interacts with T242^Gαs at α2-β4 loop. (C) Effect of Gαs H387A/Q390A mutation on basal GDP release of Gαs. The wild type and mutant are statistically different (p < 0.0001). Statistics comparing the H387A/Q390A with WT Gs was performed by the extra sum-of-squares F test where a model of individually best-fitted values of Koff of the two different datasets was compared to a model where the parameters were shared among the datasets. The data are given as mean±SD of three independent experiments.

Figure 6:. Proposed process of GPCR – G protein complex formation

(A) Before the formation of β2AR-Gs protein complex, α5 helix of Gs^GDP is bent with H387^Gαs and Y391^Gαs interacting with the RAS domain while E392^Gαs and R389^Gαs are surface exposed. (B and C) An early interaction with the β2AR involves E392^Gαs and R389^Gαs that requires straightening of the α5 helix, which may trigger GDP release. The β2AR-T4L-GsCT-CC structure may represent this intermediate state. (D) Relative large conformation changes between GPCR and Gs are required to change from the intermediate state shown in panel C to nucleotide-free complex (Protein Data Bank accession number 3SN6). H387^Gαs and Y391^Gαs interact with β2AR in this state (zoom in window on top of panel D). See also Figure S5.

Figure 7:. The proposed intermediate complex may contribute to G protein coupling specificity.

Interactions between R389^Gαs, E392^Gαs and D^3.49, R^3.50 suggest the interaction pattern may be conserved among most Family A GPCRs. The orientation of the Gs peptide (yellow) requires larger outward movement of TM6 (Green, β2AR). While the smaller movement of TM6 observed in Gi coupled GPCRs is not compatible with this binding mode (Grey, µOR).