Sudha Shenoy

Arrestins and their yeast homologs, arrestin-related trafficking adaptors (ARTs), share a stretch of 29 amino acids called the ART motif. However, the functionality of that motif is unknown. We now report that deleting this motif prevents agonist-induced ubiquitination of β-arrestin2 (β-arr2) and blocks its association with activated G protein-coupled receptors (GPCRs). Within the ART motif, we have identified a conserved phenylalanine residue, Phe116, that is critical for the formation of β-arr2-GPCR complexes. β-arr2 Phe116Ala mutant has negligible effect on blunting β2-adrenergic receptor-induced cAMP generation unlike β-arr2, which promotes rapid desensitization. Furthermore, available structures for inactive and inositol hexakisphosphate 6-activated forms of bovine β-arr2 revealed that Phe116 is ensconced in a hydrophobic pocket, whereas the adjacent Phe117 and Phe118 residues are not. Mutagenesis of Phe117 and Phe118, but not Phe116, preserves GPCR interaction of β-arr2. Surprisingly, Phe116 is dispensable for the association of β-arr2 with its non-GPCR partners. β-arr2 Phe116Ala mutant presents a significantly reduced protein half-life compared with β-arr2 and undergoes constitutive Lys-48-linked polyubiquitination, which tags proteins for proteasomal degradation. We also found that Phe116 is critical for agonist-dependent β-arr2 ubiquitination with Lys-63-polyubiquitin linkages that are known mediators of protein scaffolding and signal transduction. Finally, we have shown that β-arr2 Phe116Ala interaction with activated β2-adrenergic receptor can be rescued with an in-frame fusion of ubiquitin. Taken together, we conclude that Phe116 preserves structural stability of β-arr2, regulates the formation of β-arr2-GPCR complexes that inhibit G protein signaling, and promotes subsequent ubiquitin-dependent β-arr2 localization and trafficking.

The glucagon receptor (GCGR) activated by the peptide hormone glucagon is a seven-transmembrane G protein-coupled receptor (GPCR) that regulates blood glucose levels. Ubiquitination influences trafficking and signaling of many GPCRs, but its characterization for the GCGR is lacking. Using endocytic colocalization and ubiquitination assays, we have identified a correlation between the ubiquitination profile and recycling of the GCGR. Our experiments revealed that GCGRs are constitutively ubiquitinated at the cell surface. Glucagon stimulation not only promoted GCGR endocytic trafficking through Rab5a early endosomes and Rab4a recycling endosomes, but also induced rapid deubiquitination of GCGRs. Inhibiting GCGR internalization or disrupting endocytic trafficking prevented agonist-induced deubiquitination of the GCGR. Furthermore, a Rab4a dominant negative (DN) that blocks trafficking at recycling endosomes enabled GCGR deubiquitination, whereas a Rab5a DN that blocks trafficking at early endosomes eliminated agonist-induced GCGR deubiquitination. By down-regulating candidate deubiquitinases that are either linked with GPCR trafficking or localized on endosomes, we identified signal-transducing adaptor molecule-binding protein (STAMBP) and ubiquitin-specific protease 33 (USP33) as cognate deubiquitinases for the GCGR. Our data suggest that USP33 constitutively deubiquitinates the GCGR, whereas both STAMBP and USP33 deubiquitinate agonist-activated GCGRs at early endosomes. A mutant GCGR with all five intracellular lysines altered to arginines remains deubiquitinated and shows augmented trafficking to Rab4a recycling endosomes compared with the WT, thus affirming the role of deubiquitination in GCGR recycling. We conclude that the GCGRs are rapidly deubiquitinated after agonist-activation to facilitate Rab4a-dependent recycling and that USP33 and STAMBP activities are critical for the endocytic recycling of the GCGR.

The glucagon receptor (GCGR) activated by the peptide hormone glucagon is a seven-transmembrane G protein-coupled receptor (GPCR) that regulates blood glucose levels. Ubiquitination influences trafficking and signaling of many GPCRs, but its characterization for the GCGR is lacking. Using endocytic colocalization and ubiquitination assays, we have identified a correlation between the ubiquitination profile and recycling of the GCGR. Our experiments revealed that GCGRs are constitutively ubiquitinated at the cell surface. Glucagon stimulation not only promoted GCGR endocytic trafficking through Rab5a early endosomes and Rab4a recycling endosomes, but also induced rapid deubiquitination of GCGRs. Inhibiting GCGR internalization or disrupting endocytic trafficking prevented agonist-induced deubiquitination of the GCGR. Furthermore, a Rab4a dominant negative (DN) that blocks trafficking at recycling endosomes enabled GCGR deubiquitination, whereas a Rab5a DN that blocks trafficking at early endosomes eliminated agonist-induced GCGR deubiquitination. By down-regulating candidate deubiquitinases that are either linked with GPCR trafficking or localized on endosomes, we identified signal-transducing adaptor molecule-binding protein (STAMBP) and ubiquitin-specific protease 33 (USP33) as cognate deubiquitinases for the GCGR. Our data suggest that USP33 constitutively deubiquitinates the GCGR, whereas both STAMBP and USP33 deubiquitinate agonist-activated GCGRs at early endosomes. A mutant GCGR with all five intracellular lysines altered to arginines remains deubiquitinated and shows augmented trafficking to Rab4a recycling endosomes compared with the WT, thus affirming the role of deubiquitination in GCGR recycling. We conclude that the GCGRs are rapidly deubiquitinated after agonist-activation to facilitate Rab4a-dependent recycling and that USP33 and STAMBP activities are critical for the endocytic recycling of the GCGR.

Small ubiquitin like modifier (SUMO) conjugation or SUMOylation of βarrestin2 promotes its association with the clathrin adaptor protein AP2 and facilitates rapid β2 adrenergic receptor (β2AR) internalization. However, disruption of the consensus SUMOylation site in βarrestin2, did not prevent βarrestin2's association with activated β2ARs, dopamine D2 receptors (D2Rs), angiotensin type 1a receptors (AT1aRs) and V2 vasopressin receptors (V2Rs). To address the role of SUMOylation in the trafficking of βarrestin and GPCR complexes, we generated and characterized a yellow fluorescent protein (YFP) tagged βarrestin2-SUMO1 chimeric protein, which is resistant to de-SUMOylation. In HEK-293 cells, YFP-SUMO1 predominantly localized in the nucleus, whereas YFP-βarrestin2 is cytoplasmic. YFP-βarrestin2-SUMO1 in addition to being cytoplasmic, is localized at the nuclear membrane. Nonetheless, βarrestin2-SUMO1 associated robustly with agonist-activated β2ARs as evaluated by co-immunoprecipitation, confocal microscopy and bioluminescence resonance energy transfer (BRET). βarrestin2-SUMO1 associated strongly with the D2R, which forms transient complexes with βarrestin2. But, βarrestin2-SUMO1 and βarrestin2 showed equivalent binding with the V2R, which forms stable complexes with βarrestin2. βarrestin2 expression level directly correlated with the steady state levels of the unmodified form of RanGAP1, which upon SUMOylation associates with nuclear membrane. On the other hand, βarrestin2-SUMO1 not only localized at the nuclear membrane, but also formed a macromolecular complex with RanGAP1. Taken together, our data suggest that SUMOylation of βarrestin2 promotes its protein interactions at both cell and nuclear membranes. Furthermore, βarrestin2-SUMO1 presents as a useful tool to characterize βarrestin2 recruitment to GPCRs, which form transient and unstable complex with βarrestin2.