Shutoff and agonist-triggered internalization of protease-activated receptor 1 can be separated by mutation of putative phosphorylation sites in the cytoplasmic tail

The thrombin receptor PAR1 becomes rapidly phosphorylated upon activation b y either thrombin or exogenous SFLLRN agonist peptide. Substitution of alan ine for all serine and threonine residues in the receptor's cytoplasmic car boxyl-terminal tail ablated phosphorylation and yielded a receptor defectiv e in both shutoff and agonist-triggered internalization. These observations suggested that activation-dependent phosphorylation of PAR1's cytoplasmic tail is required for both shutoff and agonist-triggered internalization. To identify the phosphorylation site(s) that are necessary for these function s, we generated three mutant receptors in which alanine was substituted for serine and threonine residues in the amino-terminal, middle, and carboxyl- terminal thirds of PAR1's cytoplasmic tail. When stably expressed in fibrob lasts,all three mutated receptors were rapidly phosphorylated in response t o agonist, while a mutant in which all serines and threonines in the cytopl asmic tail were converted to alanines was not. This result suggests that ph osphorylation can occur at multiple sites in PAR1's cytoplasmic tail. Alani ne substitutions in the N-terminal and C-terminal portions of the tail had no effect on either receptor shutoff or agonist-triggered internalization. By contrast, alanine substitutions in the "middle" serine cluster between S er(391) and Ser(406) yielded a receptor with considerably slower shutoff of signaling after thrombin activation than the wild type. Surprisingly, this same mutant was indistinguishable from the wild type in agonist-triggered internalization and degradation. Overexpression of G protein-coupled recept or kinase 2 (GRK2) and GRK3 "suppressed" the shutoff defect of the S --> A (391-406) mutant, consistent with this defect being due to altered receptor phosphorylation. These results suggest that specific phosphorylation sites are required for rapid receptor shutoff, but phosphorylation at multiple a lternative sites is sufficient for agonist-triggered internalization. The o bservation that internalization and acute shutoff were dissociated by mutat ion of PAR1 suggests that there are quantitative or qualitative differences in the requirements or mechanisms for these two processes.