Substrate recognition determinants of human eIF2α phosphatases

Phosphorylation of the translation initiation factor eIF2α is a rapid and vital cellular defence against many forms of stress. In mammals, the levels of eIF2α phosphorylation are set through the antagonistic action of four protein kinases and two heterodimeric protein phosphatases. The phosphatases are composed of the catalytic subunit PP1 and one of two related non-catalytic subunits, PPP1R15A or PPP1R15B (R15A or R15B). Here, we generated a series of R15 truncation mutants and tested their properties in mammalian cells. We show that substrate recruitment is encoded by an evolutionary conserved region in R15s, R15A(325–554) and R15B(340–639). G-actin, which has been proposed to confer selectivity to R15 phosphatases, does not bind these regions, indicating that it is not required for substrate binding. Fragments containing the substrate-binding regions but lacking the PP1-binding motif trapped the phospho-substrate and caused accumulation of phosphorylated eIF2α in unstressed cells. Activity assays in cells showed that R15A(325–674) and R15B(340–713), encompassing the substrate-binding region and the PP1-binding region, exhibit wild-type activity. This work identifies the substrate-binding region in R15s, that functions as a phospho-substrate trapping mutant, thereby defining a key region of R15s for follow up studies.