G protein-coupled receptors (GPCRs) activate G proteins and undergo a complex regulation by interaction with GPCR kinases (GRKs) and the formation of receptor–arrestin complexes. However, the impact of individual GRKs on arrestin binding is not clear. We report the creation of eleven combinatorial HEK293 knockout cell clones lacking GRK2/3/5/6, including single, double, triple and the quadruple GRK knockout. Analysis of β-arrestin1/2 interactions for twelve GPCRs in our GRK knockout cells enables the differentiation of two main receptor subsets: GRK2/3-regulated and GRK2/3/5/6-regulated receptors. Furthermore, we identify GPCRs that interact with β-arrestins via the overexpression of specific GRKs even in the absence of agonists. Finally, using GRK knockout cells, PKC inhibitors and β-arrestin mutants, we present evidence for differential receptor–β-arrestin1/2 complex configurations mediated by selective engagement of kinases. We anticipate our GRK knockout platform to facilitate the elucidation of previously unappreciated details of GRK-specific GPCR regulation and β-arrestin complex formation.

E5 is the major transforming oncoprotein of bovine papillomavirus, which activates the platelet-derived growth factor receptor β in a highly specific manner. The short transmembrane protein E5 with only 44 residues interacts directly with the transmembrane segments of the receptor, but structural details are not available. Biophysical investigations are challenging, because the hydrophobic E5 protein tends to aggregate and get cross-linked non-specifically via two Cys residues near its C-terminus. Here, we demonstrate that a truncation by 10 amino acids creates a more manageable protein that can be conveniently used for structure analysis. Synchrotron radiation circular dichroism and solid-state (15)N- and (31)P-nuclear magnetic resonance spectroscopy show that this E5 variant serves as a representative model for the wild-type protein. The helical conformation of the transmembrane segment, its orientation in the lipid bilayer, and the ability to form homodimers in the membrane are not affected by the C-terminal truncation.

The oncogenic E5 protein from bovine papillomavirus is a short (44 amino acids long) integral membrane protein that forms homodimers. It activates platelet-derived growth factor receptor (PDGFR) β in a ligand-independent manner by transmembrane helix-helix interactions. The nature of this recognition event remains elusive, as numerous mutations are tolerated in the E5 transmembrane segment, with the exception of one hydrogen-bonding residue. Here, we examined the conformation, stability, and alignment of the E5 protein in fluid lipid membranes of substantially varying bilayer thickness, in both the absence and presence of the PDGFR transmembrane segment. Quantitative synchrotron radiation circular dichroism analysis revealed a very long transmembrane helix for E5 of ∼26 amino acids. Oriented circular dichroism and solid-state (15)N-NMR showed that the alignment and stability of this unusually long segment depend critically on the membrane thickness. When reconstituted alone in exceptionally thick DNPC lipid bilayers, the E5 helix was found to be inserted almost upright. In moderately thick bilayers (DErPC and DEiPC), it started to tilt and became slightly deformed, and finally it became aggregated in conventional DOPC, POPC, and DMPC membranes due to hydrophobic mismatch. On the other hand, when E5 was co-reconstituted with the transmembrane segment of PDGFR, it was able to tolerate even the most pronounced mismatch and was stabilized by binding to the receptor, which has the same hydrophobic length. As E5 is known to activate PDGFR within the thin membranes of the Golgi compartment, we suggest that the intrinsic hydrophobic mismatch of these two interaction partners drives them together. They seem to recognize each other by forming a closely packed bundle of mutually aligned transmembrane helices, which is further stabilized by a specific pair of hydrogen-bonding residues.

G protein-coupled receptors (GPCRs) comprise the largest family of transmembrane receptors and represent major drug targets. Upon ligand stimulation, GPCRs activate G proteins and undergo a complex regulation by interaction with GPCR kinases (GRKs) and formation of receptor-arrestin complexes. For many GPCRs, this mechanism triggers receptor desensitisation, internalisation, and possibly a second intracellular signalling wave. Here we created eleven different HEK293 knockout cell clones for GRK2, 3, 5, and 6 individually and in combination. These include four single, two double, four triple, and the quadruple GRK knockout. The statistical evaluation of β-arrestin1/2 interactions for twelve different receptors grouped the tested GPCRs into two main subsets: those for which β-arrestin interaction was mediated by either GRK2, 3, 5, or 6 and those that are mediated by GRK2 or 3 only. Interestingly, the overexpression of specific GRKs was found to induce a robust, ligand-independent β-arrestin interaction with the V2R and AT1R. Finally, using GRK knockout cells, PKC inhibitors, and β-arrestin mutants, we present evidence for differential AT1R-β-arrestin2 complex configurations mediated by selective engagement of PKC, GRK2, or GRK6. We anticipate our novel GRK-knockout platform to facilitate the elucidation of previously unappreciated details of GRK-specific GPCR regulation and β-arrestin complex formation.

No abstract