1, 4-alpha-Glucan phosphorylase from Klebsiella pneumoniae purification, subunit structure and amino acid composition.

1. A 1,4-alpha-glucan phosphorylase from Klebsiella pneumoniae has been purified about 80-fold with an over-all yield greater than 35%. The purified enzyme has been shown to be homogeneous by gel electrophoresis at different pH-values, by isoelectric focusing, by dodecylsulfate electrophoresis and by ultracentrifugation. 2. The molecular weight of the native enzyme has been determined to be 180 000 by ultra-centrifugation studies, in good agreement with the value of 189 000 estimated by gel permeation chromatography. 3. The enzyme dissociates in the presence of 0.1% dodecylsulfate or 5 M guanidine hydrochloride into polypeptide chains. The molecular weight of these polypeptide chains has been found to be 88 000 by dodecylsulfate polyacrylamide gel electrophoresis and 99 000 by sedimentation equilibrium studies, indicating that the native enzyme is composed of two polypeptide chains. 4. The enzyme contains pyridoxalphosphate with a stoichiometry of two moles per 180 000 g protein, confirming that the 1,4-alpha-glucan phosphorylase from Klebsiella pneumoniae is a dimeric enzyme. 5. The amino acid composition of the enzyme has been determined, and its correspondence to that of 1,4-alpha-glucan phosphorylases from other sources is discussed. 6. The pI of the enzyme has been shown to be 5.3 and its pH-optimum to be about pH 5.9. The enzyme is stable in the range from pH 5.9 to 10.5.     

The role of phosphorylase kinase subunits in interaction with glycogen

The interaction of rabbit skeletal muscle phosphorylase kinase with CNBr-activated glycogen results in the formation of a covalent complex. The non-bound kinase was removed by chromatography on DEAE-cellulose and phenyl-Sepharose. The amount of the bound protein increased with an increase in the number of activated groups in the glycogen molecule; the enzyme activity was thereby decreased. The kinase covalently and non-covalently bound to glycogen exhibited a higher affinity for the protein substrate (phosphorylase b) as well as for Mg2+ and Ca2+ than did the kinase in the absence of glycogen. Electrophoresis performed under denaturating conditions showed that the gamma-subunit of phosphorylase kinase is responsible for the enzyme binding to CNBr-glycogen. The effect of cross-linking reagents (glutaric aldehyde, 1.5-difluoro-2.4-dinitrobenzene) on the binding of phosphorylase kinase subunits was studied. Glycogen afforded protection of the gamma-subunit from the cross-linking to other enzyme subunits. An analysis of the subunit composition of phosphorylase kinase covalently bound to CNBr-glycogen and of the enzyme treated with cross-linking reagents in the presence of glycogen-revealed that the gamma-subunit is involved in the specific binding of phosphorylase kinase to glycogen.     

Partial purification and characterization of glycogen phosphorylase from Dictyostelium discoideum

Glycogen phosphorylase was isolated from cells of Dictyostelium discoideum in the culmination stage of development and purified 35-fold. The enzyme had a pH optimum of 6.9 and contained sulfhydryl groups essential for activity. The K(m) values for phosphate and glycogen were 3 mm and 0.06% (w/v), respectively. No dependence on, or stimulation by, any nucleotide was observed and a wide variety of nucleotides and glycolytic intermediates did not inhibit the enzyme. Nucleotide sugars competitively inhibited the enzyme. Guanosine diphosphoglucose and adenosine diphosphoglucose were the most effective, and uridine diphosphoglucose was the least effective of the nucleotide sugars tested. The specific activity of glycogen phosphorylase increased from about 0.004 unit per mg of protein in aggregating cells to about 0.024 unit per mg in culminating cells, and then decreased during sorocarp formation. This increase in enzyme specific activity during the starvation and aging of the system can account for the increased rate of glycogen degradation during this period of development. Amylase specific activity, measured at pH 4.8 and 6.9, varied between 0.005 and 0.013 unit per mg of protein during all stages of development.     

Purification of Glycogen Phosphorylase from Bovine Brain and Immunocytochemical Examination of Rat Glial Primary Cultures Using Monoclonal Antibodies Raised Against This Enzyme

The physiological function in brain of glycogen and the enzyme catalyzing the rate-limiting step in glycogenolysis, glycogen phosphorylase (EC 2.4.1.1), is unknown. As a first step toward elucidating such a function, we have purified bovine brain glycogen phosphorylase isozyme BB 1,700-fold to a specific activity of 24 units/mg protein. When analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and subsequent silver staining, a single major protein band corresponding to an apparent molecular mass of 97 kDa was observed. Mouse monoclonal antibodies raised against the enzyme were purified and shown to be monospecific as indicated by immunoblotting. Immunocytochemical examination of astroglia-rich primary cultures of rat brain cells revealed a colocalization of glycogen phosphorylase with the astroglial marker glial fibrillary acidic protein in many cells. The staining for the enzyme appeared at two levels of intensity. There were other cells in the culture showing no specific staining under the experimental conditions employed. Neurons in neuron-rich primary cultures did not show positive staining. The data suggest that glycogen phosphorylase may be predominantly an astroglial enzyme and that astroglia cells play an important role in the energy metabolism of the brain.     

Procedure for the simultaneous large-scale isolation of pullulanase and 1,4-alpha-glucan phosphorylase from Klebsiella pneumoniae involving liquid-liquid separations.

A procedure for the simultaneous large-scale isolation of pullulanase and 1,4-alpha-glucan phosphorylase from Klebsiella pneumoniae is described. The pullulanase is solubilized from the cell wall by cholate treatment; cells and cell debris are removed by partition in a poly(ethylene glycol) (PEG)-dextran two-phase system and from the upper (PEG) phase of this system the pullulanase is isolated by ultrafiltration and precipitation with N-cetyl,N-,N-,N-trimethyl ammonium bromide to a purity of about 80% with a yield of 70%. The preparations are free of alpha-amylase activity. The cell containing dextran-rich phase is passed through a Manton-Gaulin homogenizer. Then the phosphorylase is separated from the cell debris by partition in a second PEG-dextran system. From the top phase of this system the phosphorylase is isolated by distribution in a PEG-salt two-phase system followed by batch adsorption on carboxymethyl-Sephadex in a yield of 55%, a purity of around 90%, and nearly free of glycosyltransferase activity. All steps in the isolation of the two enzymes can be performed easily in a large scale.     

Identity of the HL-A common portion fragment and human beta2-microglobulin

Glycogen synthetase D from the 17,000 × g supernatant of a homogenate of human polymorphonuclear leukocytes has been purified to a specific activity of 7,4 units/mg protein in a single step, chromatography on Concanavalin A bound to agarose (Con A-Sepharose). The overall recovery of the enzyme was 66% and the entire procedure requires only 3–4 hours. After an in vitro D to I conversion, glycogen synthetase I was purified to a specific activity of 11,5 units/mg protein in a similar procedure.     

Association of gylcogenolysis with cardiac sarcoplasmic reticulum.

Sarcoplasmic reticulum fragments isolated from dog cardiac muscle possess a calcium-accumulating system associated with a series of enzymes linked to glycogenolysis. These enzymes include: adenylate cyclase, cyclic AMP-dependent protein kinase, phosphorylase b kinase, phosphorylase (b/a, 30/1),"debrancher" enzyme, and glycogen (0.3 to 0.7 mg/mg of protein). The sarcoplasmic reticulum preparation produced glucose 1-phosphate and glucose from either endogenous or exogenous glycogen. Both the calcium-accumulating and glycogenolytic enzymes sediment in a single peak at 33% sucrose on a linear continous sucrose density gradient, and the complex remains intact throughout repeated washing. Glycogen particles appear to be associated with the sarcoplasmic reticulum in situ as well as in the isolated microsomal fraction. The sarcoplasmic reticulum-glycogenolytic complex, monitored by a linked enzyme spectrophotometric assay, shows several features: (a) activation of phosphorylase activity to peak rate occurs over a very rapid time course which cannot be duplicated using combinations of purified enzymes; (b) activation is inhibited by protein kinase inhibitor; (c) phosphorylase b functions as in the purified form with respect to AMP (Km, 0.3 mM); (d) in the presence of limiting amounts of glycogen, optimal phosphorylase b activity in the sarcoplasmic reticulum requires the presence of debrancher, and the activity is sensitive to inhibitors of that enzyme such as Tris, which suggests the possiblity that the enzymes bear a specific structual relationship to the glycogen present. Phosphorylase b leads to a activation in the sarcoplasmic reticulum was completely resistant to ethylene glycol bis(beta-aminoethyl either)-N,N'-tetraacetic acid (EGTA). Inhibition of calcium accumulation by or release of bound calcium from sarcoplasmic reticulum by X537A (RO 2-2985) did not alter the EGTA resistance. These results suggest that cardiac sarcoplasmic reticulum is a complex organelle containing functions that may be related to excitation-contraction coupling and intermediary metabolism.     

Comparative purification and characterization of invertebrate muscle glycogen synthase from the porcine parasite Ascaris suum

Glycogen synthase has been purified from the obliquely striated muscle of the swine parasite Ascaris suum. The muscle contains a concentration of glycogen synthase and glycogen which is 20-fold and 15-fold, respectively, greater than rabbit skeletal muscle. The enzyme could not be solubilized with salivary amylase, but partial solubilization was achieved by activation of endogenous phosphorylase. The enzyme was purified to 85-90% homogeneity (specific activity = 4.3 units/mg) by DEAE-cellulose, Sepharose 4B, and glucosamine 6-phosphate chromatography. The purified glycogen synthase was substantially similar to rabbit skeletal muscle enzyme with respect to Mr (gel electrophoresis and gel filtration), pH dependence, aggregation properties, temperature dependence, and kinetic constants for substrates and activators. Glycogen synthase I was converted to glycogen synthase D by the cyclic AMP-dependent protein kinase. The cyclic AMP-dependent protein kinase catalyzed the incorporation of 1.3 mol of phosphate into each glycogen synthase I subunit and the concomitant interconversion to glycogen synthase D. Since glycogen is the sole fuel utilized by this organism during nonfeeding periods of the host, the characterization of this enzyme provides further insight into the regulatory mechanisms which determine glycogen turnover.     

Regulatory properties of phosphorylase from chicken skeletal muscle

The main kinetic parameters for purified phosphorylase kinase from chicken skeletal muscle were determined at pH 8.2: Vm = 18 micromol/min/mg; apparent Km values for ATP and phosphorylase b from rabbit muscle were 0.20 and 0.02 mM, respectively. The activity ratio at pH 6.8/8.2 was 0.1-0.4 for different preparations of phosphorylase kinase. Similar to the rabbit enzyme, chicken phosphorylase kinase had an absolute requirement for Ca2+ as demonstrated by complete inhibition in the presence of EGTA. Half-maximal activation occurred at [Ca2+] = 0.4 microM at pH 7.0. In the presence of Ca2+, the chicken enzyme from white and red muscles was activated 2-4-fold by saturating concentrations of calmodulin and troponin C. The C0.5 value for calmodulin and troponin C at pH 6.8 was 2 and 100 nM, respectively. Similar to rabbit phosphorylase kinase, the chicken enzyme was stimulated about 3-6-fold by glycogen at pH 6.8 and 8.2 with half-maximal stimulation occurring at about 0.15% glycogen. Protamine caused 60% inhibition of chicken phosphorylase kinase at 0.8 mg/ml. ADP (3 mM) at 0.05 mM ATP caused 85% inhibition with Ki = 0.2 mM. Unlike rabbit phosphorylase kinase, no phosphorylation of the chicken enzyme occurred in the presence of the catalytic subunit of cAMP-dependent protein kinase. Incubation with trypsin caused 2-fold activation of the chicken enzyme.     

Isolation and characterization of rabbit skeletal muscle protein phosphatases C-I and C-II

Previous studies have shown that phosphorylase phosphatase can be isolated from rabbit liver and bovine heart as a form of Mr approximately 35,000 after an ethanol treatment of tissue extracts. This enzyme form was designated as protein phosphatase C. In the present study, reproducible methods for the isolation of two forms of protein phosphatase C from rabbit skeletal muscle to apparent homogeneity are described. Protein phosphatase C-I was obtained in yields of up to 20%, with specific activities toward phosphorylase a of 8,000-16,000 units/mg of protein. This enzyme represents the major phosphorylase phosphatase activity present in the ethanol-treated muscle extracts. The second enzyme, protein phosphatase C-II, had a much lower specific activity toward phosphorylase a (250-900 units/mg). Phosphatase C-I and phosphatase C-II had Mr = 32,000 and 33,500, respectively, as determined by sodium dodecyl sulfate disc gel electrophoresis. The two enzymes displayed distinct enzymatic properties. Phosphatase C-II was associated with a more active alkaline phosphatase activity toward p-nitrophenyl phosphate than was phosphatase C-I. Phosphatase C-II activities were activated by Mn2+, whereas phosphatase C-I was inhibited. Phosphatase C-I was inhibited by rabbit skeletal muscle inhibitor 2 while phosphatase C-II was not inhibited. Both enzymes dephosphorylated glycogen synthase and phosphorylase kinase, but displayed different specificities toward the alpha- and beta-subunit phosphates of phosphorylase kinase (Ganapathi, M. K., Silberman, S. R., Paris, H., and Lee, E. Y. C. (1980) J. Biol. Chem. 246, 3213-3217). The amino acid compositions of the two proteins were similar. Peptide mapping of the two proteins showed that they are distinct proteins and do not have a precursor-proteolytic product relationship.     

The modulator protein dissociates the catalytic subunit of hepatic protein phosphatase G from glycogen.

1. The phosphorylase phosphatase and glycogen-synthase phosphatase activities associated with the glycogen particles from rat liver were progressively inhibited by incubation with modulator protein. However, the phosphorylase phosphatase activity of the catalytic subunit was entirely recovered after destruction of the modulator and the regulatory subunit(s) by trypsin. 2. Inhibition of protein phosphatase G by modulator was associated with a translocation of the phosphorylase phosphatase activity (measured after incubation with trypsin) from glycogen to the soluble fraction. The degree of inhibition of phosphatase G corresponded closely to the extent to which the phosphorylase phosphatase activity was released from the glycogen particles. Incubation of glycogen-free protein phosphatase G with modulator did not change the affinity of the enzyme for added glycogen, but decreased the amount of phosphatase that could be bound to glycogen. 3. The phosphorylase phosphatase activity that was released from the glycogen particles by modulator migrated on gel filtration as a complex (Mr 106,000) of the catalytic subunit with modulator. Phosphorylase phosphatase activity could be transferred from glycogen-bound protein phosphatase G to modulator that was covalently bound to Sepharose. After elution from the column, the enzyme was identified as the free catalytic subunit (Mr 37,000).     

Binary affinity chromatography for the purification of glycogen phosphorylase

A binary affinity chromatography medium was prepared and found to be useful for the purification and quantitative isolation of glycogen phosphorylase from rabbit skeletal muscle and liver. Glycogen is used as the binary ligand as it has affinity toward both the column matrix and the enzyme. Agarose beads derivatized with concanavalin A bound glycogen to the level of 35 mg/ml. The glycogen-impregnated beads were able to bind 9 mg/ml of phosphorylase a or b. The phosphorylase is tightly bound so that the column can be washed free of contaminants before quantitative elution of the phosphorylase by 2 M glucose, which releases the glycogen-phosphorylase complex. It appears that binary affinity chromatography may have general utility for the isolation and purification of enzymes and other specific binding agents.     

Purification and properties of phosphorylase from baker's yeast

A rapid, reliable method for purification of phosphorylase, yielding 200-400 mg pure phosphorylase from 8 kg of pressed baker's yeast, is described. The enzyme is free of phosphorylase kinase activity but contains traces of phosphorylase phosphatase activity. Phosphorylase constitutes 0.5-0.8% of soluble protein in various strains of yeast assayed immunochemically. The subunit molecular weight (Mr) of yeast phosphorylase is around 100,000. The enzyme is composed of two subunits in various ratios, differing slightly in molecular weight and N-terminal sequence. Both are active. Only the enzyme species containing the larger subunit can form tetramers and higher oligomers. The activated enzyme is dimeric. Correlated with specific activity (1 to 110 U/mg), phosphorylase contained between less than 0.1 to 0.74 covalently bound phosphate per subunit. Inactive forms of phosphorylase could be activated by phosphorylase kinase and [gamma-32P]ATP with concomitant phosphorylation of a single threonine residue in the aminoterminal region of the large subunit. The small subunit was not labeled. The incorporated phosphate could be removed by yeast phosphorylase phosphatase, resulting in loss of activity of phosphorylase, which could be restored by ATP and phosphorylase kinase.     

Metabolism of the reserve polysaccharide of Streptococcus mitior (mitis): is there a second alpha-1,4-glucan phosphorylase?

The alpha-1,4-glucan phosphorylase (alpha-1,4-glucan: orthophosphate glucosyltransferase; EC 2.4.1.1) associated with the particulate cell fraction of Streptococcus mitior strain S3 was compared with the soluble maltodextrin phosphorylase that had been previously isolated from the same organism (Walker et al., 1969). The particulate enzyme was more sensitive to the glycogen content of the cell than the soluble euzyme; its activity was highest when the cells were grown under conditions favoring high glycogen storage. Substrate specificities of the two high activity towards endogenous glycogen, whereas low-molecular-weight maltodextrins were the preferred substrates for the soluble phosphorylase. The purification of the particulate phosphorylase included incubation of the particulate fraction in 160 mM sodium phosphate-10 mM sodium citrate-0.1% (wt/vol) Triton X-100 buffer (pH 6.7) and ion-exchange chromatography on diethylamino-ethyl- Sephadex A-50. The purified enzyme was fully soluble. The value for the purification factor was variable and depended on (i) the substrate used and (ii) whether the synthetic or the degradative reaction was being measured. The solubilization resulted in considerable changes in the properties of the phosphorylase: the pH optimum for activity was raised from 6.0 to 7.0-7.5 and the substrate specificity was altered. Consequently, the purified enzyme bore greater similarity to the soluble maltodextrin phosphorylase. The reported results are best explained in terms of a single phosphorylase, the specificity which is determind by its binding state in the cell. The enzyme acts as a glycogen phosphorylase in the particulate state and as a maltodextrin phosphorylase when soluble. The equilibrium between the two forms is related to the glycogen content of the cells.     

Allosteric effectors and trehalose protect larval Manduca sexta fat body glycogen phosphorylase B against thermal denaturation

In this paper we assessed the ability of modulators of the activity of glycogen phosphorylase b from the fat body of larval Manduca sexta to stabilize the enzyme against thermal denaturation. This approach has allowed us to distinguish between modulators that stabilize the enzyme, presumably through some conformational effect, from those that do not affect thermal stability. For example, 5'-AMP and 5'-IMP are both positive modulators of the enzyme and the K(m)s for AMP and IMP were similar, 0.71 and 1.09 mM, respectively. However, the V(max) for AMP (123 nmol/mg/min) was 10 times higher than the value found for IMP (12.5 nmol/mg/min) and AMP increased the thermal stability of glycogen phosphorylase b, however IMP did not increase the enzyme's thermal stability. Indeed, IMP decreased both the allosteric activation of the enzyme by AMP and the thermal protection conferred by AMP. The allosteric inhibitors ADP and ATP, which in vertebrate phosphorylase bind to the same site as AMP, both increased the thermal stability of the enzyme, however with less efficiency than AMP. Inorganic phosphate increased thermal stability, but glycogen and amylose did not. Glycerol, at 600 mM, protected the enzyme against thermal inactivation, whereas sorbitol at the same concentration did not show any effect. Among the polyols tested, trehalose was the most effective in conferring thermal stability. In fact, in the presence of 20 mM AMP and 600 mM trehalose, 90% of the enzyme activity remained after 20 min at 60 degrees C.     

Quantitation of glycogen synthase and phosphorylase protein in mouse liver: correlation between enzymatic protein and enzyme activity

Phosphorylase and glycogen synthase protein were measured in normal and genetically diabetic (C57BL/KsJ db/db) mice liver extracts using rocket immunoelectrophoresis, and these data correlated with measurements of total phosphorylase and total glycogen synthase activities, respectively. Phosphorylase protein in 5-week-old normal mice was about 5 micrograms/mg protein and reached 8 micrograms/mg protein by 9 weeks. In comparison, the diabetic mice had elevated levels of phosphorylase protein (11-13 micrograms/mg protein) which correlated with an increased total phosphorylase activity compared to normals. The correlation coefficient for the phosphorylase activity vs protein plot was highly significant (r = 0.73, P less than 0.001). The molar concentration of phosphorylase subunit in normal mouse liver was calculated to be 11 microM and up to 23 microM in the diabetic mice. The liver concentration of glycogen synthase was relatively constant in normal mice at 400 ng/mg protein (corresponding to approximately 1.4 microM) but varied from 230 to 441 ng/mg protein (0.9 to 1.8 microM) in diabetic mice. There was little correlation between glycogen synthase activity and enzymatic protein (r = 0.15). These results indicate (1) that phosphorylase is present at concentrations approximately 10 times that of glycogen synthase, and (2) that glycogen synthase activity is relatively more dependent upon factors other than the amount of enzymatic protein.     

Inhibition of muscle glycogen phosphorylase b by vitamin B2 and its coenzyme forms

Kinetic studies have demonstrated that vitamin B2 and its coenzyme forms FMN and FAD are potent inhibitors of glycogen phosphorylase b from rabbit skeletal muscle. The inhibition of the enzyme by flavins has a co-operative character (Hill coefficients exceed unity). Glycogen phosphorylase b bound to FMN or FAD does not reveal catalytic activity, whereas the enzyme bound to riboflavin retains about 16% of the initial catalytic activity.     

Role of the synthase domain of Ags1p in cell wall alpha-glucan biosynthesis in fission yeast

The cell wall is important for maintenance of the structural integrity and morphology of fungal cells. Besides beta-glucan and chitin, alpha-glucan is a major polysaccharide in the cell wall of many fungi. In the fission yeast Schizosaccharomyces pombe, cell wall alpha-glucan is an essential component, consisting mainly of (1,3)-alpha-glucan with approximately 10% (1,4)-linked alpha-glucose residues. The multidomain protein Ags1p is required for alpha-glucan biosynthesis and is conserved among cell wall alpha-glucan-containing fungi. One of its domains shares amino acid sequence motifs with (1,4)-alpha-glucan synthases such as bacterial glycogen synthases and plant starch synthases. Whether Ags1p is involved in the synthesis of the (1,4)-alpha-glucan constituent of cell wall alpha-glucan had remained unclear. Here, we show that overexpression of Ags1p in S. pombe cells results in accumulation of (1,4)-alpha-glucan. To determine whether the synthase domain of Ags1p is responsible for this activity, we overexpressed Ags1p-E1526A, which carries a mutation in a putative catalytic residue of the synthase domain, but observed no accumulation of (1,4)-alpha-glucan. Compared with wild-type Ags1p, this mutant Ags1p showed a markedly reduced ability to complement the cell lysis phenotype of the temperature-sensitive ags1-1 mutant. Therefore, we conclude that, in S. pombe, the production of (1,4)-alpha-glucan by the synthase domain of Ags1p is important for the biosynthesis of cell wall alpha-glucan.     

Purification and partial characterization of bovine heart phosphorylase kinase

Bovine heart phosphorylase kinase has been isolated by a procedure involving precipitation with polyethylene glycol, DEAE-Sephacel chromatography and calmodulin-Sepharose affinity chromatography. The isolated enzyme had a specific activity of 8.3 IU/mg of protein at pH 8.2 at 30 degrees C in the presence of 1% glycogen. The native enzyme had a sedimentation coefficient of 23 S and the Mr of the alpha', beta, gamma, and delta subunits, were 140,000, 130,000, 46,000, and 18,000, respectively. Activation of the phosphorylase kinase by the catalytic subunit of bovine heart cAMP-dependent protein kinase increases the pH 6.8/8.2 activity ratio from 0.01 to 0.32-0.38. Glycogen (1%) decreased the Km of the activated phosphorylase kinase at pH 6.8 for phosphorylase b from 5.5 to 1.25 mg/ml. Trypsin treatment increased the pH 6.8 activity but decreased the pH 8.2 activity. During this process the alpha' subunit was converted to a Mr 110,000 polypeptide and the enzyme activity was converted essentially to a 5.9 S species having an apparent Mr of 100,000 as determined by gel filtration. On extended trypsin treatment only one major polypeptide corresponding to the beta subunit remained. The same polypeptide was present in the active fractions following gel filtration of the trypsinized kinase.     

Isolation and characterization of cyclic AMP-independent glycogen synthase kinase from rat skeletal muscle

Glycogen synthase kinase was isolated from rat skeletal muscle. This kinase, which is cyclic nucleotide-independent and calcium-independent, was separated from phosphorylase kinase, cyclic AMP-dependent protein kinase and phosvitin kinase by phosphocellulose chromatography. Gel filtration on Sephadex G-100 resolved the glycogen synthase kinase into two fractions with apparent molecular weights of 68 000 (peak I) and 52 000 (peak II). This step also separated glycogen synthase kinase from the catalytic subunit of the cyclic AMP-dependent protein kinase, which had an apparent molecular weight of 39 000. Peak II glycogen synthase kinase activity was not affected by the addition of calcium, EGTA or a number of cyclic nucleotides. In addition to ATP, dATP would serve as the phosphate donor. Other trinucleotides tested were either poor or ineffective substrates. Activity was about 5-fold greater with Mg2+ than with Mn2+. Glycogen stimulated activity about 25%. Modifications of the methods of Soderling et al. ((1970) J. Biol. Chem. 245, 6317--6328) and Nimmo et al. ((1976) Eur. J. Biochem. 68, 21--30) were developed for purification of glycogen synthease (UDPglucose:glycogen 4-alpha D-glucosyltransferase, EC 2.4.1.11) to specific activity of 35 units/mg of protein. Using this preparation of glycogen synthase as substrate, the phosphorylation and inactivation catalyzed by glycogen synthase kinase was compared to that catalyzed by cyclic AMP-dependent protein kinase or phosphorylase kinase. Each of the kinases had different specificities for phosphorylation sites on glycogen synthase.