Structure of the vasopressin hormone-V2 receptor-β-arrestin1 ternary complex

Abstract

Arrestins interact with G protein-coupled receptors (GPCRs) to stop G protein activation and to initiate key signaling pathways. Recent structural studies shed light on the molecular mechanisms involved in GPCR-arrestin coupling, but whether this process is conserved among GPCRs is poorly understood. Here, we report the cryo-electron microscopy active structure of the wild-type arginine-vasopressin V2 receptor (V2R) in complex with β-arrestin1. It reveals an atypical position of β-arrestin1 compared to previously described GPCR-arrestin assemblies, associated with an original V2R/β-arrestin1 interface involving all receptor intracellular loops. Phosphorylated sites of the V2R C-terminus are clearly identified and interact extensively with the β-arrestin1 N-lobe, in agreement with structural data obtained with chimeric or synthetic systems. Overall, these findings highlight a striking structural variability among GPCR-arrestin signaling complexes.

G protein–coupled receptor interactions with arrestins and GPCR kinases: The unresolved issue of signal bias

“…In the NSTR 1 structure by Huang et al, basic residues in the Arr2 C domain bind the head group of an ordered phosphatidylinositol-4,5-bisphosphate (34). A similar lipid interaction was also proposed for the V 2 R-Arr2 structure but the density is not as convincing (37). Detergent micelles with a highly curved surface could consequently cause more tilting of Arr2 relative to the GPCR versus planar bilayers.…”

“…There are now over 286 structures of GPCRs in complex with G proteins ( 32 ), but only five of a hormone-activated GPCR in complex with arrestin ( Table 1 ): two with neurotensin receptor 1 (NTSR 1 ) ( 33 , 34 ), one with the M 2 muscarinic receptor (M 2 R) ( 35 ), one with the β 1 adrenergic receptor (β 1 AR) ( 36 ), and the most recent with the V 2 R ( 37 ). All these complexes involve Arr2 in a “trimodal mode” interaction, most likely because this arrangement has the highest affinity and exhibits the least dynamics, which typically portends higher resolution data in cryo-EM experiments.…”

“…However, the five structural campaigns used approaches that could bias the conformational space of the particles used for 3D reconstructions, potentially obscuring functionally relevant molecular distinctions ( Table 1 ). For example, four out of the five cryo-EM structures were determined in the presence of Fab30 ( 33 , 35 , 36 ) or its single chain variant (ScFv30) ( 37 ). Fab30 was originally selected based on its ability to bind activated Arr2 in complex with V 2 Rpp and was used to aid the crystal structure determination of the Arr2–V 2 Rpp complex ( 17 ).…”

“…Fab30 was originally selected based on its ability to bind activated Arr2 in complex with V 2 Rpp and was used to aid the crystal structure determination of the Arr2–V 2 Rpp complex ( 17 ). Four structures utilized preactivated forms of Arr2 created by either truncating its C tail ( 34 , 37 ), destabilizing the three-element interaction ( 33 ), or by disrupting the polar core ( 36 ). Two structures used sortase, a protein ligase, to attach a homogenously phosphorylated peptide derived from the V 2 R C tail in place of the native receptor tail ( 35 , 36 ).…”

“…Crosslinking ( 33 , 34 ) or protein fusions ( 33 ) were used in two cases. Both NTSR 1 –Arr2 complexes and the V 2 R–Arr2 complex were reconstituted into detergent micelles ( 33 , 34 , 37 ), whereas the M 2 R–Arr2 and β 1 AR–Arr2 complexes were resolved in the more native bilayer-like environment provided by nanodiscs ( 35 , 36 ). Finally, it should be noted that the established GRK-phosphorylated residues in the M 2 R are in its extended ICL3 ( 39 ), and this cannot be replicated by a C-terminal V 2 Rpp fusion protein.…”