Identification of protein-phosphatase-1-binding domains on the glycogen and myofibrillar targetting subunits

The specificity of the catalytic subunit of protein phosphatase-1 (PP1(c)) is modified by regulatory subunits that target it to particular subcellular locations. Here, we identify PP1(c)-binding domains on G(L) and G(M), the subunits that target PP1(c) to hepatic and muscle glycogen, respectively, and on M(110), the subunit that targets PP1(c) to smooth muscle myosin. G(M)-(G63-T93) interacted with PP1(c) and prevented G(L) from suppressing the dephosphorylation of glycogen phosphorylase, but it did not dissociate G(L) from PP1(c) or affect other characteristic properties of the PP1G(L) complex. These results indicate that G(L) contains two PP1(c)-binding sites, the region which suppresses the dephosphorylation of glycogen phosphorylase being distinct from that which enhances the dephosphorylation of glycogen synthase. At higher concentrations, G(M)-(G63-N75) had the same effect as G(M)-(G63-T93), but not if Ser67 was phosphorylated by cyclic-AMP-dependent protein kinase. Thus, phosphorylation of Ser67 dissociates G(M) from PP1(c) because phosphate is inserted into the PP1(c)-binding domain of G(M). M(110)-(M1-E309) and M(110)-(M1-F38), but not M(110)-(D39-E309), mimicked the M(110) subunit in stimulating dephosphorylation of the smooth muscle myosin P-light chain and heavy meromyosin in vitro. However, in contrast to the M(110) subunit and M(110)-(M1-E309), neither M(110)-(M1-F38) nor M(110)-(D39-E309) suppressed the PP1(c)-catalysed dephosphorylation of glycogen phosphorylase. These observations suggest that the region which stimulates the dephosphorylation of myosin is situated within the N-terminal 38 residues of the M(110) subunit, while the region which suppresses the dephosphorylation of glycogen phosphorylase requires the presence of at least part of the region 39-309 which contains seven ankyrin repeats. M(110)-(M1-F38) displaced G(L) from PP1(c), while G(M)-(G63-T93) displaced M(110) from PP1(c) in vitro. These observations indicate that the region(s) of PP1(c) that interact with G(M)/G(L) and M(110) overlap, explaining why different forms of PP1(c) contain just a single targetting subunit.