Wortmannin inhibits insulin-stimulated activation of protein phosphatase 1 in rat cardiomyocytes.

A major function of insulin in target tissues is the activation of glycogen synthase. Phosphatidylinositol 3-kinase (PI3K) has been implicated in the insulin-induced activation of glycogen synthase, although the true function of this enzyme remains unclear. Data presented here demonstrate that the PI3K inhibitors wortmannin and LY-294002 block the insulin-stimulated activation of protein phosphatase 1 (PP1) in rat ventricular cardiomyocytes. This loss of phosphatase activation mimics that seen in diabetic cardiomyocytes, in which insulin stimulation fails to activate both PP1 and glycogen synthase. Interestingly, in diabetic cells, insulin stimulated PI3K activity to 300% of that in untreated controls, whereas this activity was increased by only 77% in normal cells. PI3K protein levels, however, were similar in normal and diabetic cells. Our results indicate that PI3K is involved in the stimulation of glycogen synthase activity by insulin through the regulation of PP1. The inability of insulin to stimulate phosphatase activity in diabetic cells, despite a significant increase in PI3K activity, suggests a defect in the insulin signaling pathway that contributes to the pathology of insulin-dependent diabetes.

Nitric oxide inhibits glycogen synthesis in isolated rat hepatocytes.

There is increasing evidence for the existence of intrahepatic regulation of glucose metabolism by Kupffer cell products. Nitric oxide (NO) is known to inhibit gluconeogenic flux through pyruvate carboxylase and phosphoenolpyruvate carboxykinase. However, NO may also influence glucose metabolism at other levels. Using hepatocytes from fasted rats incubated with the NO-donor S-nitroso-N-acetylpenicillamine, we have now found that the synthesis of glycogen from glucose is even more sensitive to inhibition by NO than gluconeogenesis. Inhibition of glycogen production by NO was accompanied by a rise in intracellular glucose 6-phosphate and UDPglucose. Activity of glycogen synthase, as measured in extracts of hepatocytes after the cells had been exposed to NO, was decreased. Experiments with gel-filtered liver extracts revealed that inhibition of glycogen synthase was caused by an inhibitory effect of NO on the conversion of glycogen synthase b into glycogen synthase a.

Effect of glucose phosphorylation on the activation by insulin of skeletal muscle glycogen synthase.

The effect of insulin injection on skeletal muscle glycogen synthase activation was studied in anesthetized, normal, fed rats. Insulin stimulated the conversion of glycogen synthase to the active I form, increased the concentration of glucose 6-phosphate, and activated glycogen synthase phosphatase. A close correlation between glucose 6-phosphate concentrations, per cent glycogen synthase in the active I form, and phosphatase activity was found. When boiled extracts of muscle from control and insulin-injected animals were added to glycogen pellets containing phosphatase 1G, the difference in phosphatase activity between muscle extracts from insulin-injected and control rats was restored, indicating that the phosphatase was activated by heat-stable factors. Deproteinized muscle extracts from control and insulin-injected rats, at concentrations equivalent to those present in muscle, were tested for the activation of glycogen synthase by purified protein phosphatases 1 and 2A. The activation with the insulin extracts was four-fold larger than with the control extracts. When the extracts from insulin-injected rats were treated with glucose 6-phosphatase, the difference in activation with the control rat extracts was canceled. It would appear that, as in other insulin sensitive tissues, in skeletal muscle the increase in glucose 6-phosphate subsequent to the activation of glucose transport by insulin contributes to the activation of glycogen synthase.

Transient ischemia induces regional myocardial glycogen synthase activation and glycogen synthesis in vivo.

Glycogen is consumed during ischemic preconditioning and synthesized during the subsequent period of ischemic tolerance. To better understand this sequence, we examined the effect of brief coronary artery occlusions on regional myocardial glycogen metabolism in intact, anesthetized rats. Sequential 2-min periods of left coronary artery occlusion reduced the glycogen concentration of the anterior left ventricle approximately 30% relative to the posterior region. During subsequent reperfusion, the activity of the physiologically active glycogen synthase I form of glycogen synthase increased threefold in the anterior region (0.58 +/- 0.11 vs. 0.18 +/- 0.08 mumol.g-1.min-1, P < 0.01), stimulating a similar regional increase in glycogen synthesis rate (0.24 +/- 0.04 vs. 0.08 +/- 0.03 mumol.g-1.min-1, P < 0.01). These events were preceded by a rise in regional glucose 6-phosphate concentration, which increased the activity of a myocardial glycogen synthase phosphatase. In diabetic rats glycogen synthase phosphatase activity was significantly lower, and postischemic glycogen synthase activation was significantly impaired. These data suggest the operation of a feedback loop in which transient ischemia leads to a glucose 6-phosphate-mediated increase in the activity of a phosphoprotein phosphatase active toward glycogen synthase. This suggests phospho-protein phosphatase activation may be a feature of the preconditioned myocardium.

Impaired skeletal muscle glycogen synthase activation by insulin in the Goto-Kakizaki (G/K) rat.

The Goto-Kakizaki (G/K) rat is an animal model of non-insulin-dependent diabetes mellitus, with early hyperglycaemia, hyperinsulinaemia, and insulin resistance. We have studied the effect of insulin on the activation of glycogen synthase in the G/K rat and in the original parent strain, the Wistar rat. After insulin injection, glycogen synthase I activity, glycogen synthase phosphatase activity and glucose 6-phosphate content in skeletal muscle were significantly increased in the Wistar rats. In the G/K rats, insulin injection resulted in a reduced activation of skeletal muscle glycogen synthase, which was not significant when compared with the control rats without insulin, and no increases in glycogen synthase phosphatase and glucose 6-phosphate were seen. In adipose tissue the activation of glycogen synthase by insulin was normal in the G/K rats. Previous investigations have shown that glucose disappearance rates are low in the G/K rat. However, stimulation of glucose transport was reported to be normal in the G/K rat. A defective activation of glucose accumulation into glycogen by skeletal muscle may contribute to explain the hyperglycaemia in the G/K rat.

Opposite effects of hyperglycemia and insulin deficiency on liver glycogen synthase phosphatase activity in the diabetic rat.

The specific effect of hyperglycemia on the reported decrease in liver glycogen synthase phosphatase activity was studied in STZ-induced diabetic rats with normal fasting insulinemia. Four groups of animals were investigated: control (nondiabetic), diabetic hyperglycemic (STZ), diabetic normoglycemic (STZ followed by 3-day phloridzin treatment), and a diabetic normoglycemic group injected with glucose to reinstate hyperglycemia. None of the treatments significantly altered fasting plasma insulin and glucagon concentrations. We found that hepatic synthase phosphatase activity decreased in STZ-induced diabetic rats and was further markedly reduced when glycemia was normalized in the diabetic animals. This additional decrease in phosphatase activity was almost fully reversed when hyperglycemia was restored by acute glucose infusion of the normoglycemic diabetic rats. In parallel, the levels of liver G6P and F6P were markedly reduced in the diabetic normoglycemic rats and restored with reinstatement of hyperglycemia. In contrast, liver microsomal glucose-6-phosphatase activity was enhanced and glucokinase activity was lowered in all diabetic groups, regardless of glycemia. Our results indicate that hyperglycemia per se counteracts part of the loss of hepatic synthase phosphatase in diabetic animals and provokes the stable conversion of synthase phosphatase from a less active to a more active form.

Regulation of glycogen synthase in human muscle during isometric contraction and recovery.

Six subjects performed isometric contraction (66% maximal force) to fatigue with the knee extensor muscles. Biopsies were taken from the quadriceps femoris muscle at rest, at fatigue and 1 min after termination of contraction. In three of the subjects recovery from contraction occurred in the presence of an intact circulation (non-occluded, NON) to the thigh, whereas in the other three the circulation during recovery was occluded (OCC). Glycogen synthase fractional activity (GSF) decreased in all subjects from (mean +/- SE) 0.53 +/- 0.06 at rest to 0.37 +/- 0.04 at fatigue (P < 0.001). In the OCC group GSF returned to the pre-exercise value within 1 min after termination of contraction (0.59 +/- 0.07 at rest vs. 0.57 +/- 0.04 at 1 min post-exercise), whereas in the NON group GSF increased to a higher extent (0.48 +/- 0.09 at rest vs. 0.70 +/- 0.06 at 1 min post-exercise). The increase in GSF during the 1-min recovery was almost three-fold higher in the NON group (0.15 +/- 0.02 vs. 0.38 +/- 0.03). Cyclic AMP-dependent protein kinase (cAMP-PK) (assayed at 0/100 microM and 0.2/100 microM cAMP) did not change at fatigue or during recovery in either group. Glycogen synthase phosphatase (GSP) increased at fatigue by approximately 30% (P < 0.05 vs. rest). It is concluded that isometric contraction mediated inactivation of GS (i.e. phosphorylation of GS) is due to activation of a protein kinase(s) but not cAMP-PK. The rapid activation of GS in the NON group demonstrates that a humoral factor(s), possibly insulin and/or oxygen, is responsible for this phenomenon.

Mechanism of activation of liver glycogen synthase by swelling.

The mechanism linking the stimulation of liver glycogen synthesis to swelling induced either by amino acids or hypotonicity was studied in hepatocytes, in gel-filtered liver extracts, and in purified preparations of particulate glycogen to which glycogen-metabolizing enzymes are bound. High concentrations of KCl, but not of potassium glutamate, were found to inhibit glycogen synthesis in permeabilized hepatocytes. Similarly, physiological concentrations (30-50 mM) of Cl- ions were also found to inhibit synthase phosphatase in vitro, whereas 10-20 mM Cl- ions, a concentration found in swollen hepatocytes, did not inhibit synthase phosphatase. Synthase phosphatase activity was more sensitive to inhibition by Cl- ions at low (0.1%) than at high (1%) concentrations of glycogen. By contrast, 10 mM glutamate and aspartate, a concentration observed in hepatocytes incubated with glutamine or proline, stimulated synthase phosphatase in vitro. Therefore, it is proposed that the fall in intracellular Cl- concentration as well as the increase in intracellular glutamate and aspartate concentrations, that are observed in swollen hepatocytes in the presence of amino acids, are responsible, at least in part, for the stimulation of synthase phosphatase and, hence, of glycogen synthesis.

Increased synthase phosphatase activity is responsible for the super-activation of glycogen synthase in hepatocytes from fasted obese Zucker rats.

Addition of 60 mM glucose caused a similar partial activation of glycogen synthase in hepatocytes isolated from overnight fasted Wistar rats and from fasted lean Zucker (Fa/fa?) rats. In contrast, the activation went rapidly to completion in cells from fasted obese (fa/fa) rats. Subsequent addition of 4 microM microcystin, a potent inhibitor of type 1 and type 2A protein phosphatases, induced a rapid inactivation of glycogen synthase, which occurred at a similar rate in all three types of hepatocytes. This suggests that the super-activation of glycogen synthase in hepatocytes from fasted obese rats is not due to a lower synthase kinase activity. Glycogen synthase phosphatase was quantitatively assayed in broken-cell preparations from the same livers, with exogenous synthase b as substrate. The synthase phosphatase activity in the fa/fa livers was 3-fold higher than that in the livers from both lean Zucker rats and Wistar rats. This difference has to be attributed to an increased synthase phosphatase activity of the glycogen-bound protein phosphatase-1 in livers of fasted obese rats. The results suggest that in the latter animals the available insulin exceeds the insulin resistance of the liver. The resulting overexpression of the insulin-dependent synthase-phosphatase-1G activity may explain the super-activation of glycogen synthase in response to a glucose challenge.

Regulation of glycogen synthesis in human skeletal muscle: does cellular glycogen control glycogen synthase phosphatase activity?

Contrary to the accepted feedback control mechanism of glycogen biosynthesis in skeletal muscle, evidence is presented here leading to the conclusion that glycogen does not control the activity of glycogen synthase phosphatase in intact human skeletal muscle tissue.

Identification of a glycogen synthase phosphatase from yeast Saccharomyces cerevisiae as protein phosphatase 2A.

A glycogen synthase phosphatase was purified from the yeast Saccharomyces cerevisiae. The purified yeast phosphatase displayed one major protein band which coincided with phosphatase activity on nondenaturing polyacrylamide gel electrophoresis. This phosphatase had a molecular mass of about 160,000 Da determined by gel filtration and was comprised of three subunits, termed A, B, and C. The subunit molecular weights estimated by sodium dodecyl sulfate-polyacrylamide gel electrophoresis were 60,000 (A), 53,000 (B), and 37,000 (C), indicating that this yeast glycogen synthase phosphatase is a heterotrimer. On ethanol treatment, the enzyme was dissociated to an active species with a molecular weight of 37,000 estimated by gel filtration. The yeast phosphatase dephosphorylated yeast glycogen synthase, rabbit muscle glycogen phosphorylase, casein, and the alpha subunit of rabbit muscle phosphorylase kinase, was not sensitive to heat-stable protein phosphatase inhibitor 2, and was inhibited 90% by 1 nM okadaic acid. Dephosphorylation of glycogen synthase, phosphorylase, and phosphorylase kinase by this yeast enzyme could be stimulated by histone H1 and polylysines. Divalent cations (Mg2+ and Ca2+) and chelators (EDTA and EGTA) had no effect on dephosphorylation of glycogen synthase or phosphorylase while Mn2+ stimulated enzyme activity by approximately 50%. The specific activity and kinetics for phosphorylase resembled those of mammalian phosphatase 2A. An antibody against a synthetic peptide corresponding to the carboxyl terminus of the catalytic subunit of rabbit skeletal muscle protein phosphatase 2A reacted with subunit C of purified yeast phosphatase on immunoblots, whereas the analogous peptide antibody against phosphatase 1 did not. These data show that this yeast glycogen synthase phosphatase has structural and catalytic similarity to protein phosphatase 2A found in mammalian tissues.

Purification and characterization of the glycogen-bound protein phosphatase from rat liver.

Glycogen-bound protein phosphatase G from rat liver was transferred from glycogen to beta-cyclodextrin (cycloheptaamylose) linked to Sepharose 6B. After removal of the catalytic subunit and of contaminating proteins with 2 M NaCl, elution with beta-cyclodextrin yielded a single protein on native polyacrylamide gel electrophoresis and two polypeptides (161 and 54 kDa) on sodium dodecyl sulfate-polyacrylamide gel electrophoresis. Several lines of evidence indicate that the latter polypeptides are subunits of the protein phosphatase G holoenzyme. First, these polypeptides were also present, together with the catalytic subunit, in the extensively purified holoenzyme. Also, polyclonal antibodies against these polypeptides were able to bind the holoenzyme. Further, while bound to cyclodextrin-Sepharose, the polypeptides were able to recombine with separately purified type-1 (AMD) catalytic subunit, but not with type-2A (PCS) catalytic subunit. The characteristics of the reconstituted enzyme resembled those of the nonpurified protein phosphatase G. At low dilutions, the spontaneous phosphorylase phosphatase activity of the reconstituted enzyme was about 10 times lower than that of the catalytic subunit, but it was about 1000-fold more resistant to inhibition by the modulator protein (inhibitor-2). In contrast with the free catalytic subunit, the reconstituted enzyme co-sedimented with glycogen, and it was able to activate purified liver glycogen synthase b. Also, the synthase phosphatase activity was synergistically increased by a cytosolic phosphatase and inhibited by physiological concentrations of phosphorylase alpha and of Ca2+.

Insulinlike effects of vanadate on hepatic glycogen metabolism in nondiabetic and streptozocin-induced diabetic rats.

The effect of oral administration of sodium orthovanadate for 5 wk on hepatic glycogen metabolism was studied in control and streptozocin-induced diabetic rats. Diabetes caused hyperglycemia (5-fold increase), hypoinsulinemia (85% decrease), and hyperglucagonemia (4-fold increase). There were also marked decreases in liver glycogen and activities of glycogen-metabolizing enzymes in liver. Although vanadate administration in control animals showed no significant effect on the various parameters measured except for a 70% decrease in plasma insulin, this treatment in diabetic rats restored these parameters to near control values. In diabetic rats, glycogen synthase a and the activity ratio (activity of glycogen synthase a divided by activity of total glycogen synthase) decreased to 30% of control levels and were restored to approximately 70-80% of control values after vanadate administration. A similar pattern was observed for the activity of synthase phosphatase. The activities of glycogenolytic enzymes, i.e., phosphorylase (activity of phosphorylase a and activity of total phosphorylase), phosphorylase kinase, and protein kinase (in presence or absence of cAMP), were significantly decreased by 40-70% in diabetic rats. These enzyme activities were recovered to 70-100% of control values after vanadate treatment. Phosphorylase phosphatase was not altered by diabetes, but the vanadate treatment of both groups, i.e., control and diabetic rats, showed a 25% increase in its activity (P less than 0.01). In conclusion, these results show insulinlike in vivo action of vanadate on various parameters related to hepatic glycogen metabolism.

Oral administration of vanadate to streptozotocin-diabetic rats restores the glucose-induced activation of liver glycogen synthase.

Isolated hepatocytes from streptozotocin-diabetic rats failed to respond to a glucose load with an activation of glycogen synthase. This lesion was associated with severely decreased activities of glycogen-synthase phosphatase and of glucokinase. All these defects were abolished after consumption for 13-18 days of drinking water containing Na3VO4 (0.7 mg/ml), and they were partially restored after 3.5 days, when the blood glucose concentration was already normalized. In all conditions the maximal extent of activation of glycogen synthase in cells closely parallelled the activity of glycogen-synthase phosphatase.

Particulate-associated protein phosphatases of rat hepatomas as compared with the enzymes of rat liver.

In the course of investigating the neoplastic alterations of protein phosphatases, the particulate fractions of rat liver and AH-13, a strain of rat ascites hepatoma, were chromatographed on DEAE-cellulose and assayed for protein phosphatase using glycogen synthase D and phosphorylase a as substrates. The synthase phosphatase activity of rapidly growing AH-13 was due almost entirely to a divalent cation-inhibited protein phosphatase, tentatively designated phosphatase N, the level of which was elevated remarkably in the hepatoma as compared with liver. Other hepatomas including primary hepatoma induced with 3'-methyl-4-dimethylaminoazobenzene also exhibited high levels of this phosphatase. Phosphatase N exhibited Mr = 49,000 (gel filtration) and has been partially purified with little alteration in properties. Partially purified phosphatase N was inhibited by divalent cations, rabbit skeletal muscle polypeptide inhibitor-2 and heparin, and released the catalytic subunit of type-1 protein phosphatase upon tryptic digestion. It is therefore apparent that phosphatase N is a type-1 protein phosphatase. There is some evidence to suggest that the high levels of phosphatase N in neoplastic cells are due primarily to enhanced synthesis of its non-catalytic (regulatory) subunit.

Insulin resistance is associated with reduced fasting and insulin-stimulated glycogen synthase phosphatase activity in human skeletal muscle.

Insulin-stimulated glycogen synthase activity in human skeletal muscle correlates with insulin-mediated glucose disposal rate (M) and is reduced in insulin-resistant subjects. We have previously reported reduced insulin-stimulated glycogen synthase activity associated with reduced fasting glycogen synthase phosphatase activity in skeletal muscle of insulin-resistant Pima Indians. In this study we investigated the time course for insulin stimulation of glycogen synthase and synthase phosphatase during a 2-h high-dose insulin infusion (600 mU/min per m2) in six insulin-sensitive caucasians (group S) and in five insulin-resistant Pima Indians (group R). Percutaneous muscle biopsies were obtained from the quadriceps femoris muscle after insulin infusion for 0, 10, 20, 40, and 120 min. In group S, insulin-stimulated glycogen synthase activity increased with time and was significantly higher than in group R. In group S, synthase phosphatase activity increased significantly by 25% at 10 min and then decreased gradually. No significant change in synthase phosphatase was seen in group R and activity was lower than group S at 0 to 20 min. These data suggest that a low basal synthase phosphatase activity and a defect in its response to insulin explain, at least in part, reduced insulin stimulation of skeletal muscle glycogen synthase associated with insulin resistance.

ADP and glucose as possible synergistic partners in the stimulation of liver glycogen synthase activation.

Glucose administered either intravenously or orally causes liver glycogen synthase activation independent of a rise in circulating insulin. In vitro, physiological concentrations of glucose stimulate synthase phosphatase activity but only in the presence of a second effector which reduced the A0.5 for glucose. Caffeine and certain methylxanthines have been in vitro models for a putative natural effector. The present study demonstrates that, in vitro, ADP also reduced the A0.5 for glucose comparable to the effect of caffeine. The maximum stimulation by glucose in the presence of caffeine or ADP was comparable. The effect of ADP was specific among the major nucleoside diphosphates. However, the A0.5 for ADP was greater than the normal liver concentration which does not change in response to either glucose or insulin administration. The effect of ADP appeared distinct from that of the methylxanthines since it was observed that at near saturating concentrations of ADP and of glucose, stimulation was increased by addition of theophylline. Similarly, addition of adenosine, a natural cell constituent, caused increased stimulation. Subsequently, it was shown that adenosine reduced the A0.5 for ADP to a nearly physiological concentration. Thus, while ADP is not the inducible putative effector which has been predicted it may be part of an intracellular amplification system for glycogen synthase activation which increases the sensitivity to an induced effector. The present work suggests that the effective concentration of the natural ligand may be less than originally anticipated. This work also suggests that the putative effector could be structurally related to adenosine. Phosphorylase phosphatase activity known to be stimulated by ADP and glucose is further stimulated by the combination which may be acting in synergy.

Decreased activity and impaired hormonal control of protein phosphatases in rat livers with a deficiency of phosphorylase kinase.

1. Livers from gsd/gsd rats, which do not express phosphorylase kinase activity, also contain much less particulate type-1 protein phosphatases. In comparison with normal Wistar rats, the glycogen/microsomal fraction contained 75% less glycogen-synthase phosphatase and 60% less phosphorylase phosphatase activity. This was largely due to a lower amount of the type-1 catalytic subunit in the particulate fraction. In the cytosol, the synthase phosphatase activity was also 50% lower, but the phosphorylase phosphatase activity was equal. 2. Both Wistar rats and gsd/gsd rats responded to an intravenous injection of insulin plus glucose with an acute increase (by 30-40%) in the phosphorylase phosphatase activity in the liver cytosol. In contrast, administration of glucagon or vasopressin provoked a rapid fall (by about 25%) in the cytosolic phosphorylase phosphatase activity in Wistar rats, but no change occurred in gsd/gsd rats. 3. Phosphorylase kinase was partially purified from liver and subsequently activated. Addition of a physiological amount of the activated enzyme to a liver cytosol from Wistar rats decreased the V of the phosphorylase phosphatase reaction by half, whereas the non-activated kinase had no effect. The kinase preparations did not change the activity of glycogen-synthase phosphatase, which does not respond to glucagon or vasopressin. Furthermore, the phosphorylase phosphatase activity was not affected by addition of physiological concentrations of homogeneous phosphorylase kinase from skeletal muscle (activated or non-activated). 4. It appears therefore that phosphorylase kinase plays an essential role in the transduction of the effect of glucagon and vasopressin to phosphorylase phosphatase. However, this inhibitory effect either is specific for the hepatic phosphorylase kinase, or is mediated by an unidentified protein that is a specific substrate of phosphorylase kinase.

Regulation of protein phosphatase-1G from rabbit skeletal muscle. 1. Phosphorylation by cAMP-dependent protein kinase at site 2 releases catalytic subunit from the glycogen-bound holoenzyme.

The glycogen-associated form of protein phosphatase-1 (PP-1G) is a heterodimer comprising a 37-kDa catalytic (C) subunit and a 161-kDa glycogen-binding (G) subunit, the latter being phosphorylated by cAMP-dependent protein kinase at two serine residues (site 1 and site 2). Here the amino acid sequence surrounding site 2 has been determined and this phosphoserine shown to lie 19 residues C-terminal to site 1 in the primary structure. The sequence in this region is: (sequence; see text) At physiological ionic strength, phosphorylation of glycogen-bound PP-1G was found to release all the phosphatase activity from glycogen. The released activity was free C subunit, and not PP-1G, while the phospho-G subunit remained bound to glycogen. Dissociation reflected a greater than or equal to 4000-fold decrease in affinity of C subunit for G subunit and was readily reversed by dephosphorylation. Phosphorylation and dephosphorylation of site 2 was rate-limiting for dissociation and reassociation of C subunit. Release of C subunit was also induced by the binding of anti-site-1 Fab fragments to glycogen-bound PP-1G. At near physiological ionic strength, PP-1G and glycogen concentration, site 2 was autodephosphorylated by PP-1G with a t0.5 of 2.6 min at 30 degrees C, approximately 100-fold slower than the t0.5 for dephosphorylation of glycogen phosphorylase under the same conditions. Site 2 was a good substrate for all three type-2 phosphatases (2A, 2B and 2C) with t0.5 values less than those toward the alpha subunit of phosphorylase kinase. At the levels present in skeletal muscle, the type-2A and type-2B phosphatases are potentially capable of dephosphorylating site 2 in vivo within seconds. Site 1 was at least 10-fold less effective than site 2 as a substrate for all four phosphatases. In conjunction with information presented in the following paper in this issue of this journal, the results substantiate the hypothesis that PP-1 activity towards the glycogen-metabolising enzymes is regulated in vivo by reversible phosphorylation of a targetting subunit (G) that directs the C subunit to glycogen--protein particles. The efficient dephosphorylation of site 2 by the Ca2+/calmodulin-stimulated protein phosphatase (2B) provides a potential mechanism for regulating PP-1 activity in response to Ca2+, and represents an example of a protein phosphatase cascade.