Other proteins besides ß-arrestins and GRKs contribute to GPCR desensitization and couple receptors directly to novel signalling proteins and pathways. The regulators of G protein signalling (RGS proteins) are a family (>30 mammalian isoforms) of multifunctional signalling proteins that bind Ga subunits to modulate linked signalling events (De Vries et al. 2000; Ross and Wilkie 2000; Zhong and Neubig 2001). RGS proteins contain a conserved 130 amino acid core domain (RGS domain) that directly binds activated Ga-GTP subunits and serves as a GTPase-activating protein (GAP) and/or effector antagonists for Ga. In reconstitution assays using purified proteins, most RGS proteins block the function of Gia and Gqa family members whereas certain family members are selective regulators of G12/13a; no
RGS proteins have yet been reported that directly regulate Gsa functions. As a class, proteins with RGS domains negatively regulate receptor-directed G protein signalling.
Aside from their shared capacity to block G protein signalling, RGS proteins differ widely in their overall size and amino acid identity, and possess a remarkable variety of structural domains and motifs (Hepler 1999; Siderovski et al. 1999). Based on RGS domain amino acid identities and comparative overall structural and functional similarities, RGS proteins have been classified into six distinct subfamilies (RZ, R4, R7, R12, RA, and RL) (Ross and Wilkie 2000). Members of the RZ and R4 subfamilies, with a few exceptions, are small 20-30 kDa proteins that contain short amino (N) and C-terminal regions flanking the RGS domain. In contrast, the R7, R12, RA, and RL subfamily members, with a few exceptions, are much larger proteins (up to 160 kDa) that possess longer N- and C termini flanking the RGS domain that contain various binding domains and motifs for other proteins. Among the reported non-Ga binding partners for various RGS proteins are GP5 subunits, activated forms of the small G proteins rho and rap1/2,14:3:3, adenomous polyposis coli, P-catenin, glycogen synthase kinase-3p, coatomer proteins, as well as others (for review see De Vries etal. 2000; Ross and Wilkie 2000). Thus RGS proteins serve both as inhibitors of Ga-directed signalling events and as scaffold/effector proteins that link activated Ga subunits to diverse signalling proteins and pathways.
Although RGS proteins bind activated Ga-GTP subunits, they may interact directly with GPCRs to modulate their functions (Fig. 7.1b). RGS12 is alternatively spliced (Chatterjee and Fisher 2000) and longer variant forms contain an N terminal PDZ domain that recognizes specific binding motifs on target proteins. Many GPCRs contain PDZ binding motifs at their C termini, and a screen of receptor C termini revealed that the PDZ domain of RGS12 binds to a specific PDZ binding motif of the interleukin CXCR2 receptor, but not other receptors (Snow et al. 1998a). This observation suggests that certain GPCRs serve active roles in recruiting RGS proteins to regulate function of linked G proteins.
RGS protein binding to both GPCRs and G proteins may contribute to signalling selectivity. Accumulated evidence indicates that most, but not all, RGS proteins exhibit a surprising lack of selectivity for the target Ga in vitro. With few exceptions, most RGS proteins are GAPs for most Gia family members and/or Gq/11a. How then does an RGS protein decide which Ga to selectively regulate in a cellular context? Recent studies provide compelling evidence that RGS/G protein interactions are dictated by linked GPCR in cells. Wilkie and co-workers demonstrated that purified RGS1, RGS4, and RGS16, when introduced directly into cells, selectively inhibited inositol lipid/Ca++ signalling by the m3 muscarinic cholinergic receptor (m3ChoR) compared with the CCK receptor (Xu et al. 1999); in stark contrast, RGS2 displayed no preference. Receptor selectivity of RGS4 is conferred by its N-terminal domain since truncated RGS4 lacking this domain exhibited no receptor-selectivity. However, RGS4 selectivity for m3ChoR was restored by combined addition of the N terminus and the RGS core domain (Zeng et al. 1998).
Additional evidence for direct RGS/GPCR interactions has come from genetic studies in Caenorhabditis elegans. The worm RGS proteins Eat-16 and Egl-10 and members of the mammalian R7 family of RGS proteins (RGS7, RGS6, RGS9, RGS11) each contain an RGS domain, a poorly understood DEP (dishevelled, egl-10, pleckstrin) domain, and a highly unusual G protein gamma-like (GGL) domain. GGL domains specifically bind GP5 subunits in mammals or the equivalent (GPB-2) in worms (Cabrera et al. 1998; Snow et al. 1998b; Hajdu-Cronin et al. 1999). GP5 is unusual among identified mammalian GP subunits in that it is the only isoform localized to the cytosol and it shares only 53% amino acid identity with the other homologous G|3 subunits (p1—4) (Watson etal. 1996). By forming a complex, RGS/G^5 resemble conventional G protein G^y subunits with an RGS domain attached. In Caenorhabditis elegans, serotonin receptors linked to worm Goa at the neuromuscular junction stimulate egg-laying behaviour, and muscarinic receptors linked to Gq oppose these actions. Muscarinic agonists and Egl-10/GPB-2 selectively block Goa actions, whereas serotonin and Eat-16/GPB-2 selectively block Gq actions (Hajdu-Cronin etal. 1999). Speculative models (Sierra et al. 2000) propose that serotonin receptors and Goa complex with Eat-16/GPB-2 and that serotonin-directed receptor activation releases Eat-16 to block Gq actions. In reciprocal fashion, muscarinic-directed dissociation of Egl-10/GPB-2 from Gqa blocks Goa actions. Consistent with this idea, mammalian RGS9/G|35 complex substitutes for conventional G^y in supporting Ga/|35/RGS coupling to receptors in reconstitution systems using purified proteins (T.K. Harden and A.G. Gilman, personal communication).
Collectively, these findings predict that certain RGS proteins serve as multifunctional modulators of GPCR signalling. Direct RGS coupling to GPCRs may dictate which G proteins and linked signalling pathways are regulated. Formation of a stable GPCR/Ga/RGS complex could, in fact, facilitate more efficient and finely tuned signalling since the GAP activity of RGS would allow rapid exchange and rebinding of GTP on Ga to sustain, rather than inhibit the signalling event (Ross and Wilkie 2000). Finally, RGS proteins may serve as scaffolding proteins that assemble multiple signalling proteins in a complex to switch GPCR and G protein signalling between classical second messenger pathways and novel signalling pathways.
Was this article helpful?