Structural Elements in the Gαs and Gαq C Termini That Mediate Selective G Protein-coupled Receptor (GPCR) Signaling

Abstract

Although the importance of the C terminus of the α subunit of the heterotrimeric G protein in G protein-coupled receptor (GPCR)-G protein pairing is well established, the structural basis of selective interactions remains unknown. Here, we combine live cell FRET-based measurements and molecular dynamics simulations of the interaction between the GPCR and a peptide derived from the C terminus of the Gα subunit (Gα peptide) to dissect the molecular mechanisms of G protein selectivity. We observe a direct link between Gα peptide binding and stabilization of the GPCR conformational ensemble. We find that cognate and non-cognate Gα peptides show deep and shallow binding, respectively, and in distinct orientations within the GPCR. Binding of the cognate Gα peptide stabilizes the agonist-bound GPCR conformational ensemble resulting in favorable binding energy and lower flexibility of the agonist-GPCR pair. We identify three hot spot residues (Gαs/Gαq-Gln-384/Leu-349, Gln-390/Glu-355, and Glu-392/Asn-357) that contribute to selective interactions between the β2-adrenergic receptor (β2-AR)-Gαs and V1A receptor (V1AR)-Gαq The Gαs and Gαq peptides adopt different orientations in β2-AR and V1AR, respectively. The β2-AR/Gαs peptide interface is dominated by electrostatic interactions, whereas the V1AR/Gαq peptide interactions are predominantly hydrophobic. Interestingly, our study reveals a role for both favorable and unfavorable interactions in G protein selection. Residue Glu-355 in Gαq prevents this peptide from interacting strongly with β2-AR. Mutagenesis to the Gαs counterpart (E355Q) imparts a cognate-like interaction. Overall, our study highlights the synergy in molecular dynamics and FRET-based approaches to dissect the structural basis of selective G protein interactions.

Keywords: G protein; G protein-coupled receptor (GPCR); Receptor Conformation; cell signaling; fluorescence resonance energy transfer (FRET); molecular dynamics.

FIGURE 1.

Gα C terminus minimally sufficient to detect cognate pathway for six Class A GPCRs. a, schematic of GPCR FRET sensor expressed at plasma membrane in dissociated, low FRET and associated, high FRET state. b–d, G_s-coupled GPCRs (β3-adrenergic receptor; dopamine receptor D1), G_i-coupled GPCRs (α₂-adrenergic receptor; cannabinoid receptor type 1 (CB1)), and G_q-coupled GPCR FRET sensors (α₁-adrenergic receptor; vasopressin 1A receptor) tethered to no-peptide (N), s, i, or q peptide and test for change in FRET upon agonist stimulation. e, GPCR selects for G protein via Gα C terminus. Results are expressed as mean ± S.E. of three independent experiments performed in triplicate (n ≥ 3). Asterisks represent significant differences between the indicated peptide compared with no-peptide using Tukey's multiple comparison test. Full statistical results are in supplemental Tables 1 and 2. *, p ≤ 0.05; ***, p ≤ 0.001; ****, p ≤ 0.0001.

FIGURE 2.

GPCR binds tighter to cognate peptide and is more energetically favored. a, transmembrane (panels i and iv) views of β2-AR·s peptide complex (left) and V_1AR·q peptide complex (right) with peptide movements during simulation; intracellular view of peptide movements during simulation (panels ii and v); intracellular view of receptor movement during simulation (panels iii and vi), with maximum (‡) and minimum (*) positions of TM 6 throughout simulation. b, distances measured (in Å) between TM3 and TM6 for β2-AR·peptide complexes (left) and V_1AR·peptide complexes (right) with s peptide (red), i peptide (blue), and q peptide (green). Population density expressed as a fraction of whole. c, binding energies (−kcal/mol) of β2-AR·peptide and V_1AR·peptide complexes with s, i, and q peptides compared with ΔFRET (30). Binding energy results are expressed as mean ± S.E. of five independent replicates of 100-ns simulations. ΔFRET results are expressed as mean ± S.E. of three independent experiments of at least three repeats per experiments. Table lists the binding energy and ΔFRET values presented in the graph with calculated S.D.

FIGURE 3.

Electrostatic, energetic, and structural differences in GPCR/peptide interface identify three hot spots for binding. a, intracellular view of β2-AR·s peptide complex (left) and V_1AR·q peptide complex (right) showing surface representation of the receptor's binding interface coupled with cognate peptide. Charged residues are colored as follows: anionic (red), cationic (blue), non-charged polar (yellow), and hydrophobic (white). b, colored gradation indicating distribution of energetically favored (white to color: unfavorable to favorable) residue positions in s (left) and q (right) peptides based on simulation with cognate receptors. Peptides are rotated 180° about their principal axis to display GPCR contacting interface. Head, neck, and tail regions of C termini are denoted between dashed lines. c, structure-based sequence alignment of Gα_s and Gα_q C termini depicts residues binding to receptor (black) and residues left out of interface (gray). Residue similarity denoted as follows: identical (*); conservative, maintenance of charges (:); semi-conservative, replacement of charges (.); non-synonymous, changed chemical properties (no symbol). d, hot spot residues conferring specificity of Gα_s binding to β2-AR and Gα_q binding to V_1AR.

FIGURE 4.

Single point mutations are sufficient to enhance peptide binding in V_1AR but not β2-AR. a, ΔFRET assay for agonist-stimulated V_1AR testing single-point mutations in q peptide (L349Q and N357E) and s peptide (Q384L and E392N) compared with WT q and s peptide, respectively. b, cAMP assay for single-point mutations (Q384L and E392N) in Gα_s-tethered V_1AR FRET sensor compared with WT or untransfected HEK-293 (basal). c, ΔFRET assay with β2-AR testing single-point mutations in s and q peptides compared with WT. Results are expressed as mean ± S.E. of three independent experiments performed in triplicate. Asterisks represent significance of mutant peptides compared with WT peptide using Student's unpaired t test. *, p ≤ 0.05; **, p ≤ 0.01; ***, p ≤ 0.001; ****, p ≤ 0.0001; n.s., not significant.

FIGURE 5.

Glu-355 hot spot residue in q peptide shows steric clash in β2-AR interface, resolved by mutation to glutamine. a, overlain structures of s (red) and q (blue) peptides in β2-AR interface shows shallow rearranged orientation of q peptide compared with s peptide. b, intermolecular contacts (within 5 Å) made between Gln-390 of s peptide and β2-AR (left) are found outside of suitable binding (>6 Å) radius from Glu-355 in q peptide bound to β2-AR (right). c, ΔFRET assay with β2-AR testing E355Q q peptide mutant in context of single-point mutation, L349Q/E355Q double mutant, and E355Q/N357E double mutant compared with WT q peptide. d, double mutation of Glu-355 and Leu-349 in q peptide to glutamine rearranges mutant q peptide (purple) orientation to s peptide-like (red) orientation within β2-AR interface. e, several intermolecular contacts made between Gln-390 in s peptide and β2-AR are restored in the L349Q/E355Q q peptide interaction with β2-AR (highlighting E355Q residue). f, TM3 to TM6 distance measured in β2-AR bound to s, q, and q double mutant peptides. Results for ΔFRET are expressed as mean ± S.E. of three independent experiments performed in triplicate. Asterisks represent significance of mutant peptides compared with WT peptide using Student unpaired t test. **, p ≤ 0.01; ****, p ≤ 0.0001.

FIGURE 6.

C terminus of cognate G protein stabilizes GPCR conformational flexibility, allowing tighter binding to G protein, deeper insertion of C terminus in receptor interface. Agonist-stimulated GPCRs are able to interact with different G proteins; however, the cognate G protein binds deeper into the receptor based on favorable interactions between the Gα C terminus and the GPCR, whereas the non-cognate interaction is hindered by unfavorable interactions between Gα C terminus and GPCR.