Short-term hormonal control of protein phosphatases involved in hepatic glycogen metabolism

Abstract

The prominent protein phosphatases involved in liver glycogen metabolism are the AMD (ATP, Mg-dependent, type-1) and PCS (polycation-stimulated, type-2A) phosphatases. The glycogen synthase phosphatase activity, measured from the rate of activation of liver glycogen synthase, is virtually accounted for by AMD phosphatases; the bulk of the activity belongs to the glycogen-bound protein phosphatase G and a small part is present in the cytosol. The major part of the phosphorylase phosphatase activity present in the post-mitochondrial supernatant is shared by protein phosphatase G and cytosolic enzymes, and a minor part belongs to a microsomal AMD phosphatase. In the liver cytosol, the phosphorylase phosphatase activity is about equally distributed between AMD and PCS phosphatases. Studies in vivo as well as on isolated, perfused livers have shown that glucagon (which raises the level of cyclic AMP) as well as vasopressin (which increases the cytosolic Ca 2+ concentration) decrease the phosphorylase phosphatase activity in liver extract or cytosol (filtered through Sephadex G-25) by about 25% within a few minutes. These effects were not additive, and the activity of glycogen synthase phosphatase was not affected. Conversely, insulin as well as glucose increased both phosphatase activities by about 25%, and these effects were additive. Vanadate mimicked the effect of insulin on the perfused liver. All the activity changes were only observed when the assays were performed at high tissue concentration. Upon subcellular fractionation all the effects were well expressed in the cytosol, but not in the particulate fraction (glycogen and microsomes). However, quantitatively the hormonal responses were largely lost during the fractionation procedure; they could be restored by recombination of the liver cytosol from a hormone-treated rat with the particulate fraction from either a treated or an untreated animal. It appears that the effects of glucagon, insulin and glucose are mediated by cytosolic, transferable effectors of the V max of protein phosphatases. These effectors are eluted in the void volume of a Sephadex G-25 column. Rats of the gsd/gsd strain, which have a genetic deficiency of hepatic phosphorylase kinase, responded to an injection of insulin plus glucose with a normal increase in the cytosolic phosphorylase phosphatase activity. In contrast, they failed to respond to glucagon as well as vasopressin. A transient 80% inhibition of the phosphorylase phosphatase activity could be induced in vitro in a concentrate liver cytosol from Wistar rats upon addition of MgATP. No change in the phosphorylase phosphatase activity occurred when the same experiment was performed with cytosol from gsd/gsd rats, except after addition of phosphorylase kinase (either partially purified from rat liver, or pure enzyme from rabbit skeletal muscle). Additional data indicate that phosphorylase kinase, when covalently activated by glucagon treatment or stimulated by vasopressin-released Ca 2+, triggers the inhibition of phosphorylase phosphatase by direct phosphorylation of a hitherto unidentified protein substrate, which is distinct from inhibitor-1.