Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Hyperglycemia and Inhibition of Glycogen Synthase in Streptozotocin-treated Mice

Hyperglycemia and Inhibition of Glycogen Synthase in Streptozotocin-treated Mice THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 279, No. 20, Issue of May 14, pp. 20636 –20642, 2004 Printed in U.S.A. Hyperglycemia and Inhibition of Glycogen Synthase in Streptozotocin-treated Mice ROLE OF O-LINKED N-ACETYLGLUCOSAMINE* Received for publication, November 5, 2003, and in revised form, February 6, 2004 Published, JBC Papers in Press, March 10, 2004, DOI 10.1074/jbc.M312139200 Glendon Parker, Rodrick Taylor, Deborah Jones, and Donald McClain‡ From the Veterans Affairs Medical Center and Division of Endocrinology, University of Utah School of Medicine, Salt Lake City, Utah 84132 Glycogen synthase is post-translationally modified by gen synthase is responsive to endocrine factors, including in- both phosphate and O-linked N-acetylglucosamine (O- sulin, glucagon, and catecholamines, as well as to metabolic GlcNAc). In 3T3-L1 adipocytes exposed to high concen- status, such as the concentration of the allosteric activator trations of glucose, O-GlcNAc contributes to insulin re- glucose 6-phosphate (G6P) (3, 4). Glycogen synthase activity is sistance of glycogen synthase. We sought to determine modulated by phosphorylation that directly inhibits the en- whether O-GlcNAc also regulates glycogen synthase in zyme and renders it less sensitive to allosteric activation by vivo. Glycogen synthase activity in fat pad extracts was G6P (5). Insulin stimulation leads to removal of phosphate by inhibited in streptozotocin (STZ)-treated diabetic mice. protein phosphatase 1 (PP1), resulting in increased sensitivity The half-maximal activation concentration for glucose to activation by G6P and a higher level of G6P-independent 6-phosphate (A ) was increased to 830  120 M com- 0.5 activity (6, 7). Glycogen levels, glycogen synthase activity, and pared with 240  20 M in control mice (C, p < 0.01), responsiveness to insulin signaling are all reduced in diabetes while the basal glycogen synthase activity (%I-form) was (8 –11). Both endogenous and exogenous phosphatases are also decreased to 2.4  1.4% compared with 10.1  1.8% in less able to fully activate glycogen synthase in streptozotocin controls (p < 0.01). Glycogen synthase activity remained (STZ)-diabetic rats (12). inhibited after compensatory insulin treatment. After Glycogen synthase activity is also affected by the hexosa- insulin treatment kinetic parameters of glycogen syn- mine biosynthetic pathway, which produces UDP-N-acetylhex- thase were more closely correlated with blood glucose 2 2 osamines (13–17). UDP-N-acetylglucosamine is a substrate for (A , r  0.70; %I-form, r  0.59) than insulin levels 0.5 2 2 O-linked N-acetylglucosaminyltransferase, which transfers the (A , r  0.04; %I-form, r  0.09). Hyperglycemia also 0.5 monosaccharide onto serine and threonine residues of cytosolic resulted in an increase in the level of O-GlcNAc on gly- and nuclear proteins. Data recently published by our labora- cogen synthase (16.1  1.8 compared with 7.0  0.9 arbi- trary intensity units for controls, p < 0.01), even though tory showed that glycogen synthase from extracts of 3T3-L1 the level of phosphorylation was identical in diabetic adipocytes was modified by O-GlcNAc in a glucose-dependent and control mice either with (STZ: 2.9  1.0 and C: 3.2 manner (18). This modification inhibited the enzyme in a man- 0.8) or without (STZ: 12.2  2.8 and C: 13.8  3.0 arbitrary ner analogous to phosphate, and only after enzymatic removal intensity units) insulin treatment. In all mice the per- of O-GlcNAc could the enzyme be fully activated by exogenous cent activation of glycogen synthase that could be PP1 (18). This illustrated a direct link between increased glu- achieved in vitro by recombinant protein phosphatase 1 cose uptake, modification by O-GlcNAc, glycogen synthase in- (230  30%) was significantly greater in the presence of hibition, and resistance of the synthase to activation by insulin -D-N-acetylglucosaminidase (410  60%, p < 0.01). This signaling. We therefore investigated the relative roles of O- synergistic stimulation of glycogen synthase due to co- GlcNAc and phosphate in regulating glycogen synthase in vivo digestion by protein phosphatase 1 and -D-N-acetylglu- in mice made diabetic by low dose STZ treatment. We show cosaminidase was more pronounced in STZ-diabetic that hyperglycemia results in elevated O-GlcNAc on glycogen mice (310  70%) compared with control mice (100  10%, synthase and that removal of O-GlcNAc facilitates activation of p < 0.05). The findings demonstrate that O-GlcNAc has a the enzyme by PP1 especially in diabetes. This confirms that role in the regulation of glycogen synthase both in nor- O-GlcNAc has an important regulatory function in vivo and moglycemia and diabetes. plays a role in mediating the inhibitory effect of hyperglycemia on glycogen synthase in a diabetic animal model. The rate-limiting enzyme in glycogen metabolism, glycogen EXPERIMENTAL PROCEDURES synthase, is a major determinant of overall glucose metabolism Antibodies and Reagents—The following primary antibodies were (1, 2). Because of its central role in glucose metabolism glyco- used in the current study: anti-glycogen synthase (Chemicon Interna- tional, Inc., Temecula, CA), anti-phosphoglycogen synthase (Cell Sig- naling Technology, Inc., Beverly, MA), anti-O-GlcNAc monoclonal IgM antibody (CTD 110.6; a gift of Dr. Gerald Hart, The Johns Hopkins * This work was supported by the research service of the Veterans University, Baltimore, MD) (19). Secondary antibodies used were horse- Administration, National Institutes of Health Grant R01 DK43526, and the Ben B. and Iris M. Margolis Foundation. The costs of publication of radish peroxidase-conjugated anti-rabbit and anti-mouse IgG (Amer- this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. The abbreviations used are: G6P, glucose 6-phosphate; AIU, arbi- ‡ To whom correspondence should be addressed: Division of Endocri- trary intensity units; O-GlcNAc, O-linked N-acetylglucosamine; PP1, nology, University of Utah, 30 North, 2030 East, Salt Lake City, UT protein phosphatase type 1 (-isoform); STZ, streptozotocin; %I-form, 84132. Tel.: 801-581-7755; Fax: 801-585-0956; E-mail: donald.mcclain@ basal glycogen synthase activity; A , half-maximal activation concen- 0.5 hsc.utah.edu. tration for glucose 6-phosphate. This is an Open Access article under the CC BY license. 20636 This paper is available on line at http://www.jbc.org In Vivo Regulation of Glycogen Synthase by O-GlcNAc 20637 sham Biosciences) as well as goat anti-mouse IgM (Calbiochem-Nova- biochem). Succinylated wheat germ agglutinin-agarose was obtained from EY Laboratories (San Mateo, CA). UDP-[6- H]glucose was ob- tained from Amersham Biosciences. The insulin used in this study was recombinant human insulin (NovolinR, NovoNordisk, Bagsvaerd, Denmark). 6-Acetamido-6-deoxycastanospermine was obtained from Industrial Research Ltd. (Wellington, New Zealand). The protease in- hibitors used were the Mini-Complete tablets from Roche Applied Sci- ence. Glucose was measured using a Glucometer Elite glucose meter and test strips from Bayer Corp. (Mishawaka, IN). The saline used was 0.9% sodium chloride (Baxter Health Products, Deerfield, IL). All other enzymes and chemicals were obtained from Sigma. Animals and Treatments—Animals used in this study were male C57BL/6 mice. Mice were maintained in a 12-h light/dark cycle with ad libitum access to food. At 10 –12 weeks of age mice were injected intraperitoneally with either saline or on sequential days with 85, 75, and 50 mg of STZ/kg of body weight to induce hyperglycemia. Blood glucose concentrations were measured daily in random fed animals. The STZ treatment resulted in 72% of the mice developing hyperglyce- FIG.1. Blood glucose and serum insulin concentrations in con- mia (resting blood glucose, 13.9 mM) after 8.7  0.5 days. After at least trol, STZ-treated, and insulin-treated mice. A, blood glucose con- 3 days of hyperglycemia, the mice were weighed, resting blood glucose centrations were measured in control mice (CON, white columns) and levels were measured, and STZ-treated and control mice were treated mice treated with a low dose STZ protocol (STZ, light gray columns). with 0.75 units of insulin/kg of body weight or the equivalent volume of Animals were either left untreated or treated with 0.75 units of insu- saline. The mice were then left without food for 60 min at which time lin/kg of body weight and sacrificed after 60 min (CON  I and STZ the blood glucose levels were measured, serum was collected, and both I, dark gray and black columns, respectively). B, serum insulin concen- groups of animals were sacrificed by cervical dislocation. The animals trations were measured in the same mice. Significant relationships were immediately dissected, and organs were frozen in liquid nitrogen (p  0.05) are indicated. and stored at 80 °C. Insulin levels were measured in serum using the Sensitive Rat Insulin radioimmunoassay kit (Linco Research, Inc., St. ducted in which case it was added after the digestion. The glycosylation Charles, MO). The Institutional Animal Care and Use Committee of the of glycogen synthase and phosphorylation were measured as described University of Utah and Salt Lake City Veterans Affairs approved all previously (18). procedures. Statistical Analysis—Significance was determined by using the Stu- Preparation of Epididymal Fat Pad Extract—Frozen epididymal fat dent’s t test (Microsoft Excel Version X, Microsoft Corp., Redwood, WA). pads were finely diced with a razor and placed in 1 ml of ice-cold buffer The lines of best fit were determined using linear regression and sig- containing 25 mM HEPES, pH 7.4, 100 mM NaCl, 5% glycerol (v/v), and nificance (p  0.05) determined from the degree of freedom and corre- protease inhibitors. They were then immediately homogenized with a lation coefficient. Data are presented as the means  S.E. Polytron PT 2100 homogenizer and PT-DA 2107/2EC probe (setting 26 for 15 s; Kinematica AG, Littau, Switzerland) and centrifuged at RESULTS 20,000  g for 2 min at 4 °C. The infranatant was aspirated, frozen as Characterization of STZ-diabetic Mice—We followed a pro- aliquots in liquid nitrogen, and stored at 80 °C. tocol of low dose STZ treatment to induce moderate hypergly- Analysis of Glycogen Synthase—Glycogen synthase activity was cemia with only mild hypoinsulinemia. STZ-treated mice (Fig. measured as reported previously (18). Basal glycogen synthase activity (%I-form) was defined as the percentage of enzyme activity in the 1A) had an elevation in blood glucose to 20.5  1.2 mM com- absence of G6P relative to activity in 10 mM G6P. The glycosylated form pared with 7.9  0.6 mM in control mice (p  0.001). Serum of glycogen synthase was quantified by binding the glycosylated pro- insulin levels of the STZ-treated mice (0.8  0.2 ng/ml) tended teins to immobilized succinylated wheat germ agglutinin followed by to be lower than control mice (1.1  0.3 ng/ml), but this differ- SDS-PAGE and staining the blot with anti-glycogen synthase as de- ence was not significant (Fig. 1B). When treated with an intra- scribed previously (18). Phosphorylated glycogen synthase was meas- peritoneal injection of 0.75 units of insulin/kg of body weight ured by staining an immunoblot of fat pad extracts with an antibody specific for phosphorylation at the Ser-640 residue of glycogen synthase. (20), serum insulin values increased after 60 min to similar Densitometry was conducted as described previously (18). levels of 1.8  0.6 ng/ml in STZ-treated animals and 1.5  0.3 Digestion with Jack Bean -D-N-Acetylglucosaminidase and Protein ng/ml in control animals (Fig. 1B; p  0.55). Glucose levels Phosphatase 1—Epididymal fat pad extract (13.5 g) was digested at were reduced with insulin treatment to 2.4  0.3 mM in controls 30 °C in a total volume of 50 lin50mM HEPES, pH 7.4, 5% glycerol, (Fig. 1A; p  0.001). STZ-treated mice were hyperglycemic even 20 mM sodium chloride, 3.6 mM manganese chloride, and protease after insulin treatment with glucose decreasing only to 14.9 inhibitors with or without 1 unit of jack bean -D-N-acetylglucosamini- dase and with or without 0.2 units of rabbit recombinant PP1 (Sigma). 4.3 mM (p  0.07), 6.0-fold more than that in control mice Incubations lacking -D-N-acetylglucosaminidase included 2 mM 6-ac- (p  0.01). etamido-6-deoxycastanospermine, an inhibitor of O-linked -D-N- Glycogen Synthase from Adipose Tissue Is Inhibited in Hy- acetylglucosaminidase. Both -D-N-acetylglucosaminidase and phos- perglycemic Conditions Even after Exogenous Insulin Treat- phatase were either desalted into or diluted in desalting buffer (50 mM ment—We next sought to determine the relative roles of glyce- HEPES, pH 7.4, 5% glycerol, and protease inhibitors). After 30 min 103 mia and insulin levels on regulation of glycogen synthase in l of 20 mg of glycogen/ml, 50 mM sodium fluoride, 2 mM 2-acetamido- 1-amino-1,2-dideoxy--D-glucopyranose, 1 mg/ml bovine serum albumin this mouse model of diabetes. Extracts were made from the fraction V, and 1% (v/v) protein phosphatase 1 inhibitor mixture epididymal fat pads of control and STZ-treated mice and con- (Sigma) were added to stop the digestion. Duplicate glycogen synthase trol and STZ-treated mice that had been treated with insulin assays, with and without 10 mM G6P, were then conducted for 30 min 60 min prior to sacrifice. Glycogen synthase activity was meas- at 37 °C as described previously (18). A stock of PP1 was prepared by ured over a range of 0 –10 mM G6P to measure the half-maxi- resuspension at a concentration of 7 units/l in a 50% (v/v) mixture of mal activation concentration for G6P (A ) and %I-form. Gly- the desalting buffer and glycerol and stored as aliquots at 80 °C. 0.5 To confirm that the conditions described above resulted in deglyco- cogen synthase from the epididymal fat pads of STZ-treated sylation and dephosphorylation of glycogen synthase 400 g of protein mice was inhibited with the A value increasing from 240 0.5 were treated under the same conditions with 30 units of -D-N-acetyl- 20 M G6P in fat pad extracts of control mice to 830  120 M glucosaminidase or 6 units of protein phosphatase 1. The incubation G6P in fat pads of STZ-treated mice (Fig. 2A; p  0.01). A was terminated by addition of 750 l of radioimmune precipitation similar result was observed with basal glycogen synthase ac- assay buffer. Endogenous -D-N-acetylglucosaminidase activity was in- tivity: STZ treatment resulted in a decrease of basal glycogen hibited by prior addition of 1 mM 2-acetamido-1-amino-1,2-dideoxy--D- glucopyranose unless a -D-N-acetylglucosaminidase digestion was con- synthase activity from 10.1  1.8% in control mice to 2.4  20638 In Vivo Regulation of Glycogen Synthase by O-GlcNAc FIG.2. Glycogen synthase activity in epididymal fat pad ex- tracts from STZ- and insulin-treated mice. Epididymal fat pad FIG.3. Degree of phosphorylation of glycogen synthase in ep- extracts (7.5 g of protein) were obtained from control mice (CON), ididymal fat pad extracts from STZ- and insulin-treated mice. STZ-treated mice (STZ), and control and STZ-treated mice also treated Extracts were obtained from the epididymal fat pads of control mice with 0.75 units of insulin/kg of body weight (CON  I and STZ  I, (CON), streptozotocin-treated mice (STZ), and mice treated addition- respectively). Glycogen synthase activity was assayed over a range of 0- ally with 0.75 units of insulin/kg of body weight (CON  I and STZ 10 mM G6P to determine both the A (A) and %I-form (B). Data are 0.5 I). A, immunoblots of the extracts were developed with an antibody represented as the mean  S.E. Significant relationships are indicated specific for phosphorylation at the Ser-640 site. The blot shown is a (p  0.05). single representative exposure in which lanes were arranged to main- tain consistency of presentation. B, the relative levels of phosphorylated glycogen synthase (pGS) were quantified with densitometry. Signifi- 1.4% in STZ-treated mice (Fig. 2B; p  0.01). The inactivation cant relationships (p  0.05) are indicated. Data are represented as the seen with STZ treatment is expected as a result of both dimin- mean  S.E. ished insulin signaling and hyperglycemia. To separate the two factors, we compensated for insulin deficiency by additional harvesting. Basal glycogen synthase activity in mice that were insulin treatment at levels typically used in insulin tolerance not treated with insulin (open circles) correlated equally well tests (21). The treatment significantly activated glycogen syn- with either blood glucose (Fig. 4A; r  0.42, p  0.013) or thase from both STZ-treated and control mice (p  0.01). How- serum insulin levels (Fig. 4B; r  0.38, p  0.020). However, ever, even with compensatory insulin treatment there was still treatment with exogenous insulin (closed circles) eliminated a relative inhibition of A values and basal glycogen synthase the correlation with insulin levels (Fig. 4B; r  0.09, p  0.28), 0.5 activity in STZ-treated mice. Control mice treated with insulin while the correlation with blood glucose levels was even more had an A value of 150  20 M G6P, whereas in STZ-treated significant (Fig. 4A; r  0.59, p  0.005). A values were also 0.5 0.5 mice the A value was 290  40 M G6P (Fig. 2A; p  0.01). better correlated with glycemia than insulinemia. Blood glu- 0.5 The basal glycogen synthase activity showed the same pattern cose concentrations showed significant correlations with the of resistance to stimulation by insulin with 15.1  1.4% for A value of insulin-untreated and -treated mice (Fig. 4C; r 0.5 control mice and 9.1  2.2% for STZ-treated mice (Fig. 2B; 0.70, p  0.005; r  0.61, p  0.005). However, serum insulin p  0.05). concentrations showed no significant correlation with the A 0.5 The Phosphorylation of Glycogen Synthase Is the Same in value in animals treated with or without insulin (Fig. 4D; r Both Hyperglycemic and Control Mice—We next determined 0.23, p  0.08; r  0.04, p  0.51). Both A values and basal 0.5 whether the inhibition of glycogen synthase in the hyperglyce- activities were also significantly correlated with the blood glu- mic mice could be explained by increased phosphorylation of cose levels that were measured prior to insulin treatment (A , 0.5 2 2 the enzyme. Epididymal fat pad extracts from control and r  0.49, p  0.005; %I-form, r  0.27, p  0.05; data not diabetic mice, treated with and without insulin, were resolved shown). by SDS-PAGE, immunostained with an antibody specific for a Glycogen Synthase Is Differentially Modified by O-GlcNAc in key regulatory phosphorylation site (Ser-640) on glycogen syn- STZ-treated Mice—The lack of relationship of phosphorylation thase, and quantified with densitometry (Fig. 3, A and B). The of glycogen synthase to the basal activities and A values 0.5 level of phosphoglycogen synthase was similar in the fat pads of combined with the better correlation of both parameters with control and STZ-treated mice (13.8  3.0 arbitrary intensity blood glucose compared with insulin levels suggests that hy- units (AIU) compared with 12.2  2.8 AIU, respectively). With perglycemia may inhibit glycogen synthase by mechanisms insulin treatment both control and diabetic mice underwent that are independent of insulin action. In cultured 3T3-L1 dephosphorylation at the Ser-640 site equally well with levels adipocytes increased glucose flux results in inhibition of glyco- of phosphorylation of 3.2  0.8 AIU for control and 2.9  1.0 gen synthase due to modification by O-GlcNAc (18). We there- AIU for STZ-treated mice, reductions of 76 and 77%, respec- fore next examined the potential role of O-GlcNAc on glycogen tively. Thus the inhibition of glycogen synthase and its resist- synthase regulation in vivo in control and diabetic mice. ance to activation by insulin in STZ-treated mice could not be To determine whether STZ-induced diabetes increased the explained solely by the phosphorylation state of the enzyme at level of O-GlcNAc on cytosolic and nuclear proteins, we re- Ser-640. solved proteins from epididymal fat pad extracts of STZ-treated Blood Glucose Levels Are a Better Predictor of Glycogen Syn- and untreated mice and stained the resulting immunoblot with thase Kinetics than Serum Insulin Levels—Glycogen synthase an anti-O-GlcNAc antibody to examine global changes in pro- activities in fat pad extracts were next correlated with the tein glycosylation (Fig. 5A). As indicated by the asterisk a blood glucose levels and serum insulin levels at the time of select, and at this stage uncharacterized, population of proteins In Vivo Regulation of Glycogen Synthase by O-GlcNAc 20639 FIG.4. Correlation of glycogen syn- thase kinetic parameters with glu- cose and insulin concentrations in serum. %I-form (A and B) and A (C and 0.5 D) were correlated with blood glucose (A and C) and blood insulin levels (B and D). Data points from epididymal fat pad ex- tracts of insulin-treated (closed circles) and untreated mice (open circles) are shown. The lines of best fit, achieved us- ing linear regression, are shown for both insulin-treated (solid line) and untreated mice (dashed line). Significant (p  0.05 and p  0.005) correlations are indicted by the plus sign and asterisk, respectively. The blood glucose concentrations repre- sent those obtained after insulin treat- ment at the time of sacrifice and tissue harvesting. fied by O-GlcNAc was elevated in hyperglycemic mice (Fig. 5B). Densitometry of this population of glycogen synthase showed a glycosylation level of 16.1  0.9 AIU relative to 7.0  1.8 AIU in control mice (Fig. 5C; p  0.005). Both bands shown in Fig. 5B represent differently modified populations of glycogen syn- thase that were resolved by SDS-PAGE. Removal of O-GlcNAc Improves Activation of Glycogen Syn- thase by Protein Phosphatase 1—To provide direct evidence for a role of O-GlcNAc in the regulation of glycogen synthase we conducted a series of digestions of epididymal fat pad extracts with jack bean -D-N-acetylglucosaminidase and submaximal levels of recombinant rabbit PP1 or a combination of both treatments (Fig. 6). Averaged across the combined populations of all mice, phosphatase digestion alone resulted in a 230 30% increase (p  0.005), and hexosaminidase digestion alone resulted in an activation of 30  10% (p  0.67) (Fig. 6A). The combination of both hexosaminidase and phosphatase treat- ment resulted in a 410  60% increase (p  0.005), a significant increase compared with phosphatase alone (p  0.005). This synergistic effect on activation of glycogen synthase by removal of both phosphate and O-GlcNAc is more pronounced in dia- betic mice. In control mice, the increase seen with both treat- ments is 100  10% greater than with phosphatase alone FIG.5. Modification of glycogen synthase with O-GlcNAc in compared with 310  70% greater in STZ-treated hyperglyce- epididymal fat pad extracts from STZ-treated mice. A, extracts mic mice (p  0.05). from the epididymal fat pads of STZ-treated () and control () mice The glycosylation and phosphorylation of glycogen synthase were resolved by SDS-PAGE, and the resulting immunoblot was was changed as a result of digestion with either -D-N-acetyl- stained with the anti-O-GlcNAc antibody CTD110.6. An example of an unidentified protein differentially glycosylated in STZ-treated mice is glucosaminidase or protein phosphatase 1. Digestion with -D- indicated by an asterisk. B, extracts from the fat pads of STZ-treated () N-acetylglucosaminidase resulted in a 45  4% (p  0.01) and control () mice were incubated with immobilized succinylated decrease in binding to succinylated wheat germ agglutinin. wheat germ agglutinin. An immunoblot of the isolated O-GlcNAc-mod- Digestion with recombinant protein phosphatase 1 decreased ified proteins was stained with an anti-glycogen synthase antibody. The two bands present in the immunoblot indicate differences in the post- phosphorylation by 82  7% (p  0.01). Interestingly -D-N- translational modification and migration of glycogen synthase. C, levels acetylglucosaminidase digestion increased the level of glycogen of glycosylated glycogen synthase (gGS) were quantified by densitom- synthase phosphorylation by 165  7% (p  0.01). This indi- etry. Data are represented as the mean  S.E., and significance (p cates that removal of O-GlcNAc facilitates rephosphorylation of 0.05) is indicated by a line. CON, control. the enzyme. This would help explain the lower degree of gly- in STZ-treated mice exhibited an increase in glycosylation. To cogen synthase activation with only -D-N-acetylglucosamini- dase (Fig. 6A, open bars; control: 40  10% and STZ-treated: determine whether glycogen synthase (84 kDa) was one of this group, we incubated extracts with immobilized succinylated 30  10%) compared with -D-N-acetylglucosaminidase- wheat germ agglutinin to isolate proteins that were hypergly- dependent activation in combined digests (Fig. 6A, closed bars; cosylated with O-GlcNAc. Resolution of these proteins and control: 100  10% and STZ-treated: 310  70%). Phosphatase subsequent probing with an anti-glycogen synthase antibody digestion did not change the glycosylation status of the enzyme. indicated that the relative amount of glycogen synthase modi- In mice that were not treated with insulin the absolute 20640 In Vivo Regulation of Glycogen Synthase by O-GlcNAc synthase is dynamically regulated by multiple mechanisms, including substrate availability, hormone signaling, subcellu- lar localization, targeting of phosphatase, and allosteric acti- vation (4, 22, 23, 25). In addition we recently reported that glycogen synthase is regulated by modification with O-GlcNAc (18). Treatment of 3T3-L1 adipocytes with high concentrations of glucose resulted in an increase in glycosylation of glycogen synthase, which inactivated the enzyme and contributed to a reduced response to insulin signaling. Removal of O-GlcNAc was able to correct the resistance to PP1-mediated activation of glycogen synthase (18). We therefore sought to extend the findings in cultured cells and determine whether O-GlcNAc exerts a biologically signifi- cant effect on glycogen synthase activity in the intact organism where glycogen synthase is subject to the full range of physio- logic regulation by hormones and metabolic factors. Results reported here verify that O-GlcNAc modification does occur and does affect glycogen synthase activity in vivo. Glycogen syn- thase from STZ-diabetic mice was inhibited even after insulin treatment. This inhibition of glycogen synthase was more tightly correlated with blood glucose than insulin levels. Addi- tionally the inhibition of glycogen synthase in these diabetic animals could not be explained by increased phosphorylation at the key regulatory 3a site (Ser-640) either before or after insu- lin treatment. This normal dephosphorylation after insulin treatment in vivo demonstrates that dephosphorylation may be necessary but not sufficient for complete activation of glycogen synthase. Full activation of glycogen synthase in vitro required both -D-N-acetylglucosaminidase and PP1. The increase in the efficiency of activation by -D-N-acetylglucosaminidase was much greater in extracts from STZ-treated mice, which also have increased levels of glycosylation. This illustrates a rela- FIG.6. Activation of glycogen synthase with -D-N-acetylglu- tionship between hyperglycemia, glycosylation of glycogen syn- cosaminidase and protein phosphatase 1. A, epididymal fat pad thase, and resistance to activation by PP1. In this model of extracts (13.5 g of protein) from control and STZ-treated mice were digested with 1 unit of -D-N-acetylglucosaminidase (open columns), 0.2 moderate and short term hyperglycemia the ability of insulin to units of protein phosphatase 1 (shaded columns), or both together for 30 lead to dephosphorylation was unimpeded, implying no critical min at 30 °C(hatched columns). Glycogen synthase activity was as- effects on insulin signal transduction upstream of glycogen sayed with 0 and 10 mM glucose 6-phosphate in duplicate. The percent- synthase. However, other studies using different model sys- age of increase in %I-form (average  S.E.) was determined for all mice (COMBINED), control mice only (CONTROL), and STZ-treated diabetic tems have demonstrated inhibition of insulin signal trans- mice (STZ-TREATED). The -D-N-acetylglucosaminidase-dependent duction pathways mediated by the hexosamine biosynthetic component of the glycogen synthase activation observed in the com- pathway (26 –30). bined digests is represented as black columns. Significant increases The level of O-GlcNAc on nuclear and cytosolic proteins is a (p  0.05) in STZ-diabetic compared with control mice are indicated by asterisks. B, epididymal fat pad extracts (400 g of protein) from control function of cellular nutritional status. Flux through the hexos- mice were digested with either 30 units of -D-N-acetylglucosaminidase amine biosynthetic pathway is rate-limited by the concentra- or 6 units of protein phosphatase 1 under the same conditions used in tion of fructose 6-phosphate, and hexosamine flux increases A and C. The resulting mixtures were incubated overnight with O- with increased glucose or fatty acid uptake (14, 15, 31). The GlcNAc-specific succinylated wheat germ agglutinin-agarose, and im- mobilized proteins were resolved on an immunoblot. O-GlcNAc-modi- terminal metabolite of the hexosamine biosynthetic path- fied glycogen synthase (gGS) was measured by staining with an anti- way, UDP-N-acetylglucosamine, is a substrate for O-linked glycogen synthase antibody. The degree of phosphorylation at the N-acetylglucosaminyltransferase, which covalently modifies regulatory Ser-640 site (pGS) was measured by staining an immunoblot serine and threonine residues with O-GlcNAc (32). The K of each digest with a phosphospecific glycogen synthase antibody. C, epididymal fat pad extracts from insulin-untreated control and STZ- value of O-linked N-acetylglucosaminyltransferase for UDP-N- treated mice were digested with 1 unit of -D-N-acetylglucosaminidase acetylglucosamine is relatively high (33). Thus, high levels of and 0.2 units of protein phosphatase 1 compared with protein phospha- O-GlcNAc on nuclear and cytoplasmic proteins reflect a high tase 1 alone. The resulting difference in %I-form that is attributable to level of nutrient uptake (32, 34). This makes O-GlcNAc an -D-N-acetylglucosaminidase is plotted against blood glucose concentra- tions from each animal (control mice, n  2; STZ-treated mice, n  3). attractive candidate for a nutrient sensing mechanism and for The line of best fit was achieved using linear regression. mediating the cellular consequences of excess nutrient flux (34 –37). Consistent with the proposed nutrient sensing role of O- increase in basal glycogen synthase activity seen after diges- GlcNAc, insulin resistance develops in mice with transgenic tion with PP1 plus -D-N-acetylglucosaminidase, compared expression of O-linked N-acetylglucosaminyltransferase or the with PP1 treatment alone, correlated significantly with blood rate-limiting enzyme for the hexosamine biosynthetic pathway: glucose concentrations (r  0.98, p  0.005; Fig. 6C). The glutamine:fructose-6-phosphate amidotransferase (38, 39). In correlation, however, was eliminated with insulin treatment. addition to modulation of glycogen synthase activity, a direct DISCUSSION role for O-GlcNAc in metabolic signaling has been demon- strated by several studies. Glycosylation of the Akt site on Metabolism of glycogen is dependent on the metabolic de- mands of the cell and the total organism (3, 22–24). Glycogen endothelial nitric-oxide synthase prevents activation by insulin In Vivo Regulation of Glycogen Synthase by O-GlcNAc 20641 signaling (40). Modification by O-GlcNAc of transcription fac- glycogen synthase and its partners is necessary for resolving tors involved in metabolism, such as Sp1, YY1, and c-myc, the exact function of O-GlcNAc in modulating glycogen metab- affects protein-protein interactions, transcriptional activation, olism. Recently new methodologies have been developed to and protein stability (41– 47). Pharmacological inhibition of the facilitate the identification of O-GlcNAc sites using mass spec- troscopy (58), and applying these techniques to identify O- enzyme that removes O-GlcNAc from proteins results in down- regulation of elements of the insulin signaling pathway and GlcNAc sites on glycogen synthase should further clarify the mechanisms of inhibition. reduced glucose uptake by glucose transporter 4 (30). Glycogen synthase has been previously regarded as being REFERENCES primarily regulated by phosphorylation-dephosphorylation and 1. Park, K. S., Ciaraldi, T. P., Carter, L., Mudaliar, S., Nikoulina, S. E., Webster, allosteric activation by G6P. These factors are interdependent: N. J., and Henry, R. R. (2000) Metabolism 49, 962–968 increased phosphorylation reduces sensitivity to G6P activa- 2. Crosson, S. M., Khan, A., Printen, J., Pessin, J. E., and Saltiel, A. R. (2003) J. Clin. Investig. 111, 1423–1432 tion, but G6P changes the conformation of the protein to simul- 3. Lawrence, J. C., Jr., and Roach, P. J. (1997) Diabetes 46, 541–547 taneously activate the enzyme and make it a better substrate 4. Ferrer, J. C., Favre, C., Gomis, R. R., Fernandez-Novell, J. M., Garcia-Rocha, M., de la Iglesia, N., Cid, E., and Guinovart, J. J. (2003) FEBS Lett. 546, for dephosphorylation (4, 23, 48, 49). In diabetes this regulation 127–132 breaks down, and glycogen synthase is inhibited and less re- 5. Skurat, A. V., Dietrich, A. D., and Roach, P. J. (2000) Diabetes 49, 1096 –1100 sponsive to insulin activation even though G6P levels are un- 6. Newgard, C. B., Brady, M. J., O’Doherty, R. M., and Saltiel, A. R. (2000) Diabetes 49, 1967–1977 changed (9, 10, 22, 50 –52). In some situations this may be 7. Dent, P., Lavoinne, A., Nakielny, S., Caudwell, F. B., Watt, P., and Cohen, P. partly attributable to upstream effects of the hexosamine path- (1990) Nature 348, 302–308 8. Chang, A. Y. (1972) Can. J. Biochem. 50, 714 –717 way on signal transduction elements such as phosphatidyli- 9. Groop, L. C., Bonadonna, R. C., DelPrato, S., Ratheiser, K., Zyck, K., nositol 3-kinase, insulin receptor substrate-1, or -catenin (26, Ferrannini, E., and DeFronzo, R. A. (1989) J. Clin. Investig. 84, 205–213 10. Thorburn, A. W., Gumbiner, B., Bulacan, F., Wallace, P., and Henry, R. R. 27, 30, 53–55). However, in studies on glycogen synthase from (1990) J. Clin. Investig. 85, 522–529 the livers of STZ-treated rats, diabetes also directly inhibited 11. Kaslow, H. R., Eichner, R. D., and Mayer, S. E. (1979) J. Biol. Chem. 254, the ability of phosphatases to activate the enzyme (12). The 4674 – 4677 12. Akatsuka, A., Singh, T. J., and Huang, K. P. (1983) Arch. Biochem. Biophys. current findings regarding glycosylation of the enzyme may 220, 426 – 434 explain these results. 13. McClain, D. A., and Crook, E. D. (1996) Diabetes 45, 1003–1009 14. Marshall, S., Bacote, V., and Traxinger, R. R. (1991) J. Biol. Chem. 266, The simplest model to explain these findings is that O- 4706 – 4712 GlcNAc is directly inhibitory in a manner analogous to phos- 15. Hawkins, M., Barzilai, N., Liu, R., Hu, M., Chen, W., and Rossetti, L. (1997) J. Clin. Investig. 99, 2173–2182 phate and that removal is required for full activation of the 16. Marshall, S. (2002) Curr. Opin. Endocrinol. Diabetes 9, 160 –167 enzyme. In diabetic conditions modification of glycogen syn- 17. McClain, D. A. (2001) Curr. Opin. Endocrinol. Diabetes 8, 186 –191 thase by O-GlcNAc is increased. Removal of either O-GlcNAc or 18. Parker, G. J., Lund, K. C., Taylor, R. P., and McClain, D. A. (2003) J. Biol. Chem. 278, 10022–10027 phosphate alone is not sufficient for full activation of the en- 19. Comer, F. I., Vosseller, K., Wells, L., Accavitti, M. A., and Hart, G. W. (2001) zyme, but removal of O-GlcNAc facilitates the activation of the Anal. Biochem. 293, 169 –177 20. Roduit, R., Masiello, P., Wang, S. P., Li, H., Mitchell, G. A., and Prentki, M. synthase by dephosphorylation. This hexosaminidase-depend- (2001) Diabetes 50, 1970 –1975 ent increase in activation of the synthase by PP1 is greater in 21. Lauro, D., Kido, Y., Castle, A. L., Zarnowski, M. J., Hayashi, H., Ebina, Y., and diabetes, consistent with an increased level of O-GlcNAc on the Accili, D. (1998) Nat. Genet. 20, 294 –298 22. Shulman, R. G., Bloch, G., and Rothman, D. L. (1995) Proc. Natl. Acad. Sci. enzyme. There are two sites of phosphorylation on glycogen U. S. A. 92, 8535– 8542 synthase that are sufficient for inhibition of the enzyme: Ser-7 23. Villar-Palasi, C., and Guinovart, J. J. (1997) FASEB J. 11, 544 –558 24. Fernandez-Novell, J. M., Roca, A., Bellido, D., Vilaro, S., and Guinovart, J. J. and Ser-640 (5, 56). If either of these were modified by O- (1996) Eur. J. Biochem. 238, 570 –575 GlcNAc, then dephosphorylation alone would not be sufficient 25. Saltiel, A. R., and Kahn, C. R. (2001) Nature 414, 799 – 806 to fully activate the total enzyme population, especially in 26. Patti, M. E., Virkamaki, A., Landaker, E. J., Kahn, C. R., and Yki-Jarvinen, H. (1999) Diabetes 48, 1562–1571 diabetic conditions. Deglycosylation by hexosaminidase would 27. Kim, Y. B., Zhu, J. S., Zierath, J. R., Shen, H. Q., Baron, A. D., and Kahn, B. B. lower the threshold for phosphatase-mediated activation by (1999) Diabetes 48, 310 –320 28. Hawkins, M., Barzilai, N., Chen, W., Angelov, I., Hu, M., Cohen, P., and prior removal of an inhibitory group on the other serine. The Rossetti, L. (1996) Diabetes 45, 1734 –1743 inability of hexosaminidase digestion alone to activate glycogen 29. Federici, M., Menghini, R., Mauriello, A., Hribal, M. L., Ferrelli, F., Lauro, D., Sbraccia, P., Spagnoli, L. G., Sesti, G., and Lauro, R. (2002) Circulation 106, synthase was explained in part by the phosphorylation status 466 – 472 of glycogen synthase. In the digestion conditions used in this 30. Vosseller, K., Wells, L., Lane, M. D., and Hart, G. W. (2002) Proc. Natl. Acad. study, removal of O-GlcNAc by hexosaminidase was followed Sci. U. S. A. 99, 5313–5318 31. Yki-Jarvinen, H., Virkamaki, A., Daniels, M. C., McClain, D., and Gottschalk, by subsequent rephosphorylation of the enzyme by endogenous W. K. (1998) Metabolism 47, 449 – 455 kinases. This would have blunted the expected activation of 32. Wells, L., Vosseller, K., and Hart, G. W. (2001) Science 291, 2376 –2378 33. Kreppel, L. K., and Hart, G. W. (1999) J. Biol. Chem. 274, 32015–32022 glycogen synthase. Alternatively glycosylation may sterically 34. McClain, D. A. (2002) J. Diabetes Complicat. 16, 72– 80 hinder access by phosphatases or result in a conformation of 35. Obici, S., Wang, J., Chowdury, R., Feng, Z., Siddhanta, U., Morgan, K., and Rossetti, L. (2002) J. Clin. Investig. 109, 1599 –1605 the synthase that is a poorer substrate for phosphatase. G6P is 36. Rumberger, J. M., Wu, T., Hering, M. A., and Marshall, S. (2003) J. Biol. well known to change the conformation of glycogen synthase Chem. 278, 28547–28552 and make it a better substrate for phosphatases (57). O-GlcNAc 37. Patti, M. E., Butte, A. J., Crunkhorn, S., Cusi, K., Berria, R., Kashyap, S., Miyazaki, Y., Kohane, I., Costello, M., Saccone, R., Landaker, E. J., may block the conformational changes induced by G6P or even Goldfine, A. B., Mun, E., DeFronzo, R., Finlayson, J., Kahn, C. R., and limit access of G6P itself, impeding the ability of phosphatase Mandarino, L. J. (2003) Proc. Natl. Acad. Sci. U. S. A. 100, 8466 – 8471 38. Hebert, L. F., Jr., Daniels, M. C., Zhou, J., Crook, E. D., Turner, R. L., to activate the enzyme. The role of O-GlcNAc could also be an Simmons, S. T., Neidigh, J. L., Zhu, J. S., Baron, A. D., and McClain, D. A. indirect one. O-GlcNAc is now documented to interfere with (1996) J. Clin. Investig. 98, 930 –936 39. McClain, D. A., Lubas, W. A., Cooksey, R. C., Hazel, M., Parker, G. J., Love, some protein-protein interactions (44). Thus, the association of D. C., and Hanover, J. A. (2002) Proc. Natl. Acad. Sci. U. S. A. 99, PP1 with glycogen-targeting regulatory subunits and glycogen 10695–10699 synthase may well be inhibited by glycosylation of the synthase 40. Du, X. L., Edelstein, D., Dimmeler, S., Ju, Q., Sui, C., and Brownlee, M. (2001) J. Clin. Investig. 108, 1341–1348 or other components of the glycogen synthetic machinery. 41. Yang, X., Su, K., Roos, M. D., Chang, Q., Paterson, A. J., and Kudlow, J. E. The current results are significant in that they demonstrate (2001) Proc. Natl. Acad. Sci. U. S. A. 98, 6611– 6616 42. Kamemura, K., Hayes, B. K., Comer, F. I., and Hart, G. W. (2002) J. Biol. a role and provide a mechanism by which excess nutrients lead Chem. 277, 19229 –19235 directly to insulin resistance and altered protein function in 43. Roos, M. D., Su, K., Baker, J. R., and Kudlow, J. E. (1997) Mol. Cell. Biol. 17, diabetes. The identification of the sites of glycosylation on 6472– 6480 20642 In Vivo Regulation of Glycogen Synthase by O-GlcNAc 44. Hiromura, M., Choi, C. H., Sabourin, N. A., Jones, H., Bachvarov, D., and 51. Kaslow, H. R., and Eichner, R. D. (1984) Am. J. Physiol. 247, E581–E584 Usheva, A. (2003) J. Biol. Chem. 278, 14046 –14052 52. Rossetti, L., Giaccari, A., and DeFronzo, R. A. (1990) Diabetes Care 13, 45. Majumdar, G., Harmon, A., Candelaria, R., Martinez-Hernandez, A., Raghow, 610 – 630 R., and Solomon, S. S. (2003) Am. J. Physiol. 285, E584 –E591 53. Singh, L. P., and Crook, E. D. (2000) J. Investig. Med. 48, 251–258 46. Clark, R. J., McDonough, P. M., Swanson, E., Trost, S. U., Suzuki, M., Fukuda, 54. Virkamaki, A., Daniels, M. C., Hamalainen, S., Utriainen, T., McClain, D., and M., and Dillmann, W. H. (2003) J. Biol. Chem. 278, 44230 – 44237 Yki-Jarvinen, H. (1997) Endocrinology 138, 2501–2507 47. Kang, H. T., Ju, J. W., Cho, J. W., and Hwang, E. S. (2003) J. Biol. Chem. 278, 55. Singh, L. P., Gennerette, D., Simmons, S., and Crook, E. D. (2001) J. Diabetes 51223–51231 Complicat.15, 88 –96 48. Roach, P. J., Takeda, Y., and Larner, J. (1976) J. Biol. Chem. 251, 1913–1919 56. Skurat, A. V., Wang, Y., and Roach, P. J. (1994) J. Biol. Chem. 269, 49. Camici, M., DePaoli-Roach, A. A., and Roach, P. J. (1984) J. Biol. Chem. 259, 25534 –25542 3429 –3434 57. Villar-Palasi, C. (1991) Biochim. Biophys. Acta 1095, 261–267 50. Yki-Jarvinen, H., Mott, D., Young, A. A., Stone, K., and Bogardus, C. (1987) 58. Wells, L., Vosseller, K., Cole, R. N., Cronshaw, J. M., Matunis, M. J., and Hart, J. Clin. Investig. 80, 95–100 G. W. (2002) Mol. Cell. Proteomics 1, 791– 804 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Biological Chemistry Unpaywall

Hyperglycemia and Inhibition of Glycogen Synthase in Streptozotocin-treated Mice

Parker, Glendon; Taylor, Rodrick; Jones, Deborah; McClain, Donald
Journal of Biological Chemistry – May 1, 2004
Download PDF
7 pages

Loading next page...
 
/lp/unpaywall/hyperglycemia-and-inhibition-of-glycogen-synthase-in-streptozotocin-O53tth71ce

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Unpaywall
ISSN
0021-9258
DOI
10.1074/jbc.m312139200
Publisher site
See Article on Publisher Site

Abstract

THE JOURNAL OF BIOLOGICAL CHEMISTRY Vol. 279, No. 20, Issue of May 14, pp. 20636 –20642, 2004 Printed in U.S.A. Hyperglycemia and Inhibition of Glycogen Synthase in Streptozotocin-treated Mice ROLE OF O-LINKED N-ACETYLGLUCOSAMINE* Received for publication, November 5, 2003, and in revised form, February 6, 2004 Published, JBC Papers in Press, March 10, 2004, DOI 10.1074/jbc.M312139200 Glendon Parker, Rodrick Taylor, Deborah Jones, and Donald McClain‡ From the Veterans Affairs Medical Center and Division of Endocrinology, University of Utah School of Medicine, Salt Lake City, Utah 84132 Glycogen synthase is post-translationally modified by gen synthase is responsive to endocrine factors, including in- both phosphate and O-linked N-acetylglucosamine (O- sulin, glucagon, and catecholamines, as well as to metabolic GlcNAc). In 3T3-L1 adipocytes exposed to high concen- status, such as the concentration of the allosteric activator trations of glucose, O-GlcNAc contributes to insulin re- glucose 6-phosphate (G6P) (3, 4). Glycogen synthase activity is sistance of glycogen synthase. We sought to determine modulated by phosphorylation that directly inhibits the en- whether O-GlcNAc also regulates glycogen synthase in zyme and renders it less sensitive to allosteric activation by vivo. Glycogen synthase activity in fat pad extracts was G6P (5). Insulin stimulation leads to removal of phosphate by inhibited in streptozotocin (STZ)-treated diabetic mice. protein phosphatase 1 (PP1), resulting in increased sensitivity The half-maximal activation concentration for glucose to activation by G6P and a higher level of G6P-independent 6-phosphate (A ) was increased to 830  120 M com- 0.5 activity (6, 7). Glycogen levels, glycogen synthase activity, and pared with 240  20 M in control mice (C, p < 0.01), responsiveness to insulin signaling are all reduced in diabetes while the basal glycogen synthase activity (%I-form) was (8 –11). Both endogenous and exogenous phosphatases are also decreased to 2.4  1.4% compared with 10.1  1.8% in less able to fully activate glycogen synthase in streptozotocin controls (p < 0.01). Glycogen synthase activity remained (STZ)-diabetic rats (12). inhibited after compensatory insulin treatment. After Glycogen synthase activity is also affected by the hexosa- insulin treatment kinetic parameters of glycogen syn- mine biosynthetic pathway, which produces UDP-N-acetylhex- thase were more closely correlated with blood glucose 2 2 osamines (13–17). UDP-N-acetylglucosamine is a substrate for (A , r  0.70; %I-form, r  0.59) than insulin levels 0.5 2 2 O-linked N-acetylglucosaminyltransferase, which transfers the (A , r  0.04; %I-form, r  0.09). Hyperglycemia also 0.5 monosaccharide onto serine and threonine residues of cytosolic resulted in an increase in the level of O-GlcNAc on gly- and nuclear proteins. Data recently published by our labora- cogen synthase (16.1  1.8 compared with 7.0  0.9 arbi- trary intensity units for controls, p < 0.01), even though tory showed that glycogen synthase from extracts of 3T3-L1 the level of phosphorylation was identical in diabetic adipocytes was modified by O-GlcNAc in a glucose-dependent and control mice either with (STZ: 2.9  1.0 and C: 3.2 manner (18). This modification inhibited the enzyme in a man- 0.8) or without (STZ: 12.2  2.8 and C: 13.8  3.0 arbitrary ner analogous to phosphate, and only after enzymatic removal intensity units) insulin treatment. In all mice the per- of O-GlcNAc could the enzyme be fully activated by exogenous cent activation of glycogen synthase that could be PP1 (18). This illustrated a direct link between increased glu- achieved in vitro by recombinant protein phosphatase 1 cose uptake, modification by O-GlcNAc, glycogen synthase in- (230  30%) was significantly greater in the presence of hibition, and resistance of the synthase to activation by insulin -D-N-acetylglucosaminidase (410  60%, p < 0.01). This signaling. We therefore investigated the relative roles of O- synergistic stimulation of glycogen synthase due to co- GlcNAc and phosphate in regulating glycogen synthase in vivo digestion by protein phosphatase 1 and -D-N-acetylglu- in mice made diabetic by low dose STZ treatment. We show cosaminidase was more pronounced in STZ-diabetic that hyperglycemia results in elevated O-GlcNAc on glycogen mice (310  70%) compared with control mice (100  10%, synthase and that removal of O-GlcNAc facilitates activation of p < 0.05). The findings demonstrate that O-GlcNAc has a the enzyme by PP1 especially in diabetes. This confirms that role in the regulation of glycogen synthase both in nor- O-GlcNAc has an important regulatory function in vivo and moglycemia and diabetes. plays a role in mediating the inhibitory effect of hyperglycemia on glycogen synthase in a diabetic animal model. The rate-limiting enzyme in glycogen metabolism, glycogen EXPERIMENTAL PROCEDURES synthase, is a major determinant of overall glucose metabolism Antibodies and Reagents—The following primary antibodies were (1, 2). Because of its central role in glucose metabolism glyco- used in the current study: anti-glycogen synthase (Chemicon Interna- tional, Inc., Temecula, CA), anti-phosphoglycogen synthase (Cell Sig- naling Technology, Inc., Beverly, MA), anti-O-GlcNAc monoclonal IgM antibody (CTD 110.6; a gift of Dr. Gerald Hart, The Johns Hopkins * This work was supported by the research service of the Veterans University, Baltimore, MD) (19). Secondary antibodies used were horse- Administration, National Institutes of Health Grant R01 DK43526, and the Ben B. and Iris M. Margolis Foundation. The costs of publication of radish peroxidase-conjugated anti-rabbit and anti-mouse IgG (Amer- this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. The abbreviations used are: G6P, glucose 6-phosphate; AIU, arbi- ‡ To whom correspondence should be addressed: Division of Endocri- trary intensity units; O-GlcNAc, O-linked N-acetylglucosamine; PP1, nology, University of Utah, 30 North, 2030 East, Salt Lake City, UT protein phosphatase type 1 (-isoform); STZ, streptozotocin; %I-form, 84132. Tel.: 801-581-7755; Fax: 801-585-0956; E-mail: donald.mcclain@ basal glycogen synthase activity; A , half-maximal activation concen- 0.5 hsc.utah.edu. tration for glucose 6-phosphate. This is an Open Access article under the CC BY license. 20636 This paper is available on line at http://www.jbc.org In Vivo Regulation of Glycogen Synthase by O-GlcNAc 20637 sham Biosciences) as well as goat anti-mouse IgM (Calbiochem-Nova- biochem). Succinylated wheat germ agglutinin-agarose was obtained from EY Laboratories (San Mateo, CA). UDP-[6- H]glucose was ob- tained from Amersham Biosciences. The insulin used in this study was recombinant human insulin (NovolinR, NovoNordisk, Bagsvaerd, Denmark). 6-Acetamido-6-deoxycastanospermine was obtained from Industrial Research Ltd. (Wellington, New Zealand). The protease in- hibitors used were the Mini-Complete tablets from Roche Applied Sci- ence. Glucose was measured using a Glucometer Elite glucose meter and test strips from Bayer Corp. (Mishawaka, IN). The saline used was 0.9% sodium chloride (Baxter Health Products, Deerfield, IL). All other enzymes and chemicals were obtained from Sigma. Animals and Treatments—Animals used in this study were male C57BL/6 mice. Mice were maintained in a 12-h light/dark cycle with ad libitum access to food. At 10 –12 weeks of age mice were injected intraperitoneally with either saline or on sequential days with 85, 75, and 50 mg of STZ/kg of body weight to induce hyperglycemia. Blood glucose concentrations were measured daily in random fed animals. The STZ treatment resulted in 72% of the mice developing hyperglyce- FIG.1. Blood glucose and serum insulin concentrations in con- mia (resting blood glucose, 13.9 mM) after 8.7  0.5 days. After at least trol, STZ-treated, and insulin-treated mice. A, blood glucose con- 3 days of hyperglycemia, the mice were weighed, resting blood glucose centrations were measured in control mice (CON, white columns) and levels were measured, and STZ-treated and control mice were treated mice treated with a low dose STZ protocol (STZ, light gray columns). with 0.75 units of insulin/kg of body weight or the equivalent volume of Animals were either left untreated or treated with 0.75 units of insu- saline. The mice were then left without food for 60 min at which time lin/kg of body weight and sacrificed after 60 min (CON  I and STZ the blood glucose levels were measured, serum was collected, and both I, dark gray and black columns, respectively). B, serum insulin concen- groups of animals were sacrificed by cervical dislocation. The animals trations were measured in the same mice. Significant relationships were immediately dissected, and organs were frozen in liquid nitrogen (p  0.05) are indicated. and stored at 80 °C. Insulin levels were measured in serum using the Sensitive Rat Insulin radioimmunoassay kit (Linco Research, Inc., St. ducted in which case it was added after the digestion. The glycosylation Charles, MO). The Institutional Animal Care and Use Committee of the of glycogen synthase and phosphorylation were measured as described University of Utah and Salt Lake City Veterans Affairs approved all previously (18). procedures. Statistical Analysis—Significance was determined by using the Stu- Preparation of Epididymal Fat Pad Extract—Frozen epididymal fat dent’s t test (Microsoft Excel Version X, Microsoft Corp., Redwood, WA). pads were finely diced with a razor and placed in 1 ml of ice-cold buffer The lines of best fit were determined using linear regression and sig- containing 25 mM HEPES, pH 7.4, 100 mM NaCl, 5% glycerol (v/v), and nificance (p  0.05) determined from the degree of freedom and corre- protease inhibitors. They were then immediately homogenized with a lation coefficient. Data are presented as the means  S.E. Polytron PT 2100 homogenizer and PT-DA 2107/2EC probe (setting 26 for 15 s; Kinematica AG, Littau, Switzerland) and centrifuged at RESULTS 20,000  g for 2 min at 4 °C. The infranatant was aspirated, frozen as Characterization of STZ-diabetic Mice—We followed a pro- aliquots in liquid nitrogen, and stored at 80 °C. tocol of low dose STZ treatment to induce moderate hypergly- Analysis of Glycogen Synthase—Glycogen synthase activity was cemia with only mild hypoinsulinemia. STZ-treated mice (Fig. measured as reported previously (18). Basal glycogen synthase activity (%I-form) was defined as the percentage of enzyme activity in the 1A) had an elevation in blood glucose to 20.5  1.2 mM com- absence of G6P relative to activity in 10 mM G6P. The glycosylated form pared with 7.9  0.6 mM in control mice (p  0.001). Serum of glycogen synthase was quantified by binding the glycosylated pro- insulin levels of the STZ-treated mice (0.8  0.2 ng/ml) tended teins to immobilized succinylated wheat germ agglutinin followed by to be lower than control mice (1.1  0.3 ng/ml), but this differ- SDS-PAGE and staining the blot with anti-glycogen synthase as de- ence was not significant (Fig. 1B). When treated with an intra- scribed previously (18). Phosphorylated glycogen synthase was meas- peritoneal injection of 0.75 units of insulin/kg of body weight ured by staining an immunoblot of fat pad extracts with an antibody specific for phosphorylation at the Ser-640 residue of glycogen synthase. (20), serum insulin values increased after 60 min to similar Densitometry was conducted as described previously (18). levels of 1.8  0.6 ng/ml in STZ-treated animals and 1.5  0.3 Digestion with Jack Bean -D-N-Acetylglucosaminidase and Protein ng/ml in control animals (Fig. 1B; p  0.55). Glucose levels Phosphatase 1—Epididymal fat pad extract (13.5 g) was digested at were reduced with insulin treatment to 2.4  0.3 mM in controls 30 °C in a total volume of 50 lin50mM HEPES, pH 7.4, 5% glycerol, (Fig. 1A; p  0.001). STZ-treated mice were hyperglycemic even 20 mM sodium chloride, 3.6 mM manganese chloride, and protease after insulin treatment with glucose decreasing only to 14.9 inhibitors with or without 1 unit of jack bean -D-N-acetylglucosamini- dase and with or without 0.2 units of rabbit recombinant PP1 (Sigma). 4.3 mM (p  0.07), 6.0-fold more than that in control mice Incubations lacking -D-N-acetylglucosaminidase included 2 mM 6-ac- (p  0.01). etamido-6-deoxycastanospermine, an inhibitor of O-linked -D-N- Glycogen Synthase from Adipose Tissue Is Inhibited in Hy- acetylglucosaminidase. Both -D-N-acetylglucosaminidase and phos- perglycemic Conditions Even after Exogenous Insulin Treat- phatase were either desalted into or diluted in desalting buffer (50 mM ment—We next sought to determine the relative roles of glyce- HEPES, pH 7.4, 5% glycerol, and protease inhibitors). After 30 min 103 mia and insulin levels on regulation of glycogen synthase in l of 20 mg of glycogen/ml, 50 mM sodium fluoride, 2 mM 2-acetamido- 1-amino-1,2-dideoxy--D-glucopyranose, 1 mg/ml bovine serum albumin this mouse model of diabetes. Extracts were made from the fraction V, and 1% (v/v) protein phosphatase 1 inhibitor mixture epididymal fat pads of control and STZ-treated mice and con- (Sigma) were added to stop the digestion. Duplicate glycogen synthase trol and STZ-treated mice that had been treated with insulin assays, with and without 10 mM G6P, were then conducted for 30 min 60 min prior to sacrifice. Glycogen synthase activity was meas- at 37 °C as described previously (18). A stock of PP1 was prepared by ured over a range of 0 –10 mM G6P to measure the half-maxi- resuspension at a concentration of 7 units/l in a 50% (v/v) mixture of mal activation concentration for G6P (A ) and %I-form. Gly- the desalting buffer and glycerol and stored as aliquots at 80 °C. 0.5 To confirm that the conditions described above resulted in deglyco- cogen synthase from the epididymal fat pads of STZ-treated sylation and dephosphorylation of glycogen synthase 400 g of protein mice was inhibited with the A value increasing from 240 0.5 were treated under the same conditions with 30 units of -D-N-acetyl- 20 M G6P in fat pad extracts of control mice to 830  120 M glucosaminidase or 6 units of protein phosphatase 1. The incubation G6P in fat pads of STZ-treated mice (Fig. 2A; p  0.01). A was terminated by addition of 750 l of radioimmune precipitation similar result was observed with basal glycogen synthase ac- assay buffer. Endogenous -D-N-acetylglucosaminidase activity was in- tivity: STZ treatment resulted in a decrease of basal glycogen hibited by prior addition of 1 mM 2-acetamido-1-amino-1,2-dideoxy--D- glucopyranose unless a -D-N-acetylglucosaminidase digestion was con- synthase activity from 10.1  1.8% in control mice to 2.4  20638 In Vivo Regulation of Glycogen Synthase by O-GlcNAc FIG.2. Glycogen synthase activity in epididymal fat pad ex- tracts from STZ- and insulin-treated mice. Epididymal fat pad FIG.3. Degree of phosphorylation of glycogen synthase in ep- extracts (7.5 g of protein) were obtained from control mice (CON), ididymal fat pad extracts from STZ- and insulin-treated mice. STZ-treated mice (STZ), and control and STZ-treated mice also treated Extracts were obtained from the epididymal fat pads of control mice with 0.75 units of insulin/kg of body weight (CON  I and STZ  I, (CON), streptozotocin-treated mice (STZ), and mice treated addition- respectively). Glycogen synthase activity was assayed over a range of 0- ally with 0.75 units of insulin/kg of body weight (CON  I and STZ 10 mM G6P to determine both the A (A) and %I-form (B). Data are 0.5 I). A, immunoblots of the extracts were developed with an antibody represented as the mean  S.E. Significant relationships are indicated specific for phosphorylation at the Ser-640 site. The blot shown is a (p  0.05). single representative exposure in which lanes were arranged to main- tain consistency of presentation. B, the relative levels of phosphorylated glycogen synthase (pGS) were quantified with densitometry. Signifi- 1.4% in STZ-treated mice (Fig. 2B; p  0.01). The inactivation cant relationships (p  0.05) are indicated. Data are represented as the seen with STZ treatment is expected as a result of both dimin- mean  S.E. ished insulin signaling and hyperglycemia. To separate the two factors, we compensated for insulin deficiency by additional harvesting. Basal glycogen synthase activity in mice that were insulin treatment at levels typically used in insulin tolerance not treated with insulin (open circles) correlated equally well tests (21). The treatment significantly activated glycogen syn- with either blood glucose (Fig. 4A; r  0.42, p  0.013) or thase from both STZ-treated and control mice (p  0.01). How- serum insulin levels (Fig. 4B; r  0.38, p  0.020). However, ever, even with compensatory insulin treatment there was still treatment with exogenous insulin (closed circles) eliminated a relative inhibition of A values and basal glycogen synthase the correlation with insulin levels (Fig. 4B; r  0.09, p  0.28), 0.5 activity in STZ-treated mice. Control mice treated with insulin while the correlation with blood glucose levels was even more had an A value of 150  20 M G6P, whereas in STZ-treated significant (Fig. 4A; r  0.59, p  0.005). A values were also 0.5 0.5 mice the A value was 290  40 M G6P (Fig. 2A; p  0.01). better correlated with glycemia than insulinemia. Blood glu- 0.5 The basal glycogen synthase activity showed the same pattern cose concentrations showed significant correlations with the of resistance to stimulation by insulin with 15.1  1.4% for A value of insulin-untreated and -treated mice (Fig. 4C; r 0.5 control mice and 9.1  2.2% for STZ-treated mice (Fig. 2B; 0.70, p  0.005; r  0.61, p  0.005). However, serum insulin p  0.05). concentrations showed no significant correlation with the A 0.5 The Phosphorylation of Glycogen Synthase Is the Same in value in animals treated with or without insulin (Fig. 4D; r Both Hyperglycemic and Control Mice—We next determined 0.23, p  0.08; r  0.04, p  0.51). Both A values and basal 0.5 whether the inhibition of glycogen synthase in the hyperglyce- activities were also significantly correlated with the blood glu- mic mice could be explained by increased phosphorylation of cose levels that were measured prior to insulin treatment (A , 0.5 2 2 the enzyme. Epididymal fat pad extracts from control and r  0.49, p  0.005; %I-form, r  0.27, p  0.05; data not diabetic mice, treated with and without insulin, were resolved shown). by SDS-PAGE, immunostained with an antibody specific for a Glycogen Synthase Is Differentially Modified by O-GlcNAc in key regulatory phosphorylation site (Ser-640) on glycogen syn- STZ-treated Mice—The lack of relationship of phosphorylation thase, and quantified with densitometry (Fig. 3, A and B). The of glycogen synthase to the basal activities and A values 0.5 level of phosphoglycogen synthase was similar in the fat pads of combined with the better correlation of both parameters with control and STZ-treated mice (13.8  3.0 arbitrary intensity blood glucose compared with insulin levels suggests that hy- units (AIU) compared with 12.2  2.8 AIU, respectively). With perglycemia may inhibit glycogen synthase by mechanisms insulin treatment both control and diabetic mice underwent that are independent of insulin action. In cultured 3T3-L1 dephosphorylation at the Ser-640 site equally well with levels adipocytes increased glucose flux results in inhibition of glyco- of phosphorylation of 3.2  0.8 AIU for control and 2.9  1.0 gen synthase due to modification by O-GlcNAc (18). We there- AIU for STZ-treated mice, reductions of 76 and 77%, respec- fore next examined the potential role of O-GlcNAc on glycogen tively. Thus the inhibition of glycogen synthase and its resist- synthase regulation in vivo in control and diabetic mice. ance to activation by insulin in STZ-treated mice could not be To determine whether STZ-induced diabetes increased the explained solely by the phosphorylation state of the enzyme at level of O-GlcNAc on cytosolic and nuclear proteins, we re- Ser-640. solved proteins from epididymal fat pad extracts of STZ-treated Blood Glucose Levels Are a Better Predictor of Glycogen Syn- and untreated mice and stained the resulting immunoblot with thase Kinetics than Serum Insulin Levels—Glycogen synthase an anti-O-GlcNAc antibody to examine global changes in pro- activities in fat pad extracts were next correlated with the tein glycosylation (Fig. 5A). As indicated by the asterisk a blood glucose levels and serum insulin levels at the time of select, and at this stage uncharacterized, population of proteins In Vivo Regulation of Glycogen Synthase by O-GlcNAc 20639 FIG.4. Correlation of glycogen syn- thase kinetic parameters with glu- cose and insulin concentrations in serum. %I-form (A and B) and A (C and 0.5 D) were correlated with blood glucose (A and C) and blood insulin levels (B and D). Data points from epididymal fat pad ex- tracts of insulin-treated (closed circles) and untreated mice (open circles) are shown. The lines of best fit, achieved us- ing linear regression, are shown for both insulin-treated (solid line) and untreated mice (dashed line). Significant (p  0.05 and p  0.005) correlations are indicted by the plus sign and asterisk, respectively. The blood glucose concentrations repre- sent those obtained after insulin treat- ment at the time of sacrifice and tissue harvesting. fied by O-GlcNAc was elevated in hyperglycemic mice (Fig. 5B). Densitometry of this population of glycogen synthase showed a glycosylation level of 16.1  0.9 AIU relative to 7.0  1.8 AIU in control mice (Fig. 5C; p  0.005). Both bands shown in Fig. 5B represent differently modified populations of glycogen syn- thase that were resolved by SDS-PAGE. Removal of O-GlcNAc Improves Activation of Glycogen Syn- thase by Protein Phosphatase 1—To provide direct evidence for a role of O-GlcNAc in the regulation of glycogen synthase we conducted a series of digestions of epididymal fat pad extracts with jack bean -D-N-acetylglucosaminidase and submaximal levels of recombinant rabbit PP1 or a combination of both treatments (Fig. 6). Averaged across the combined populations of all mice, phosphatase digestion alone resulted in a 230 30% increase (p  0.005), and hexosaminidase digestion alone resulted in an activation of 30  10% (p  0.67) (Fig. 6A). The combination of both hexosaminidase and phosphatase treat- ment resulted in a 410  60% increase (p  0.005), a significant increase compared with phosphatase alone (p  0.005). This synergistic effect on activation of glycogen synthase by removal of both phosphate and O-GlcNAc is more pronounced in dia- betic mice. In control mice, the increase seen with both treat- ments is 100  10% greater than with phosphatase alone FIG.5. Modification of glycogen synthase with O-GlcNAc in compared with 310  70% greater in STZ-treated hyperglyce- epididymal fat pad extracts from STZ-treated mice. A, extracts mic mice (p  0.05). from the epididymal fat pads of STZ-treated () and control () mice The glycosylation and phosphorylation of glycogen synthase were resolved by SDS-PAGE, and the resulting immunoblot was was changed as a result of digestion with either -D-N-acetyl- stained with the anti-O-GlcNAc antibody CTD110.6. An example of an unidentified protein differentially glycosylated in STZ-treated mice is glucosaminidase or protein phosphatase 1. Digestion with -D- indicated by an asterisk. B, extracts from the fat pads of STZ-treated () N-acetylglucosaminidase resulted in a 45  4% (p  0.01) and control () mice were incubated with immobilized succinylated decrease in binding to succinylated wheat germ agglutinin. wheat germ agglutinin. An immunoblot of the isolated O-GlcNAc-mod- Digestion with recombinant protein phosphatase 1 decreased ified proteins was stained with an anti-glycogen synthase antibody. The two bands present in the immunoblot indicate differences in the post- phosphorylation by 82  7% (p  0.01). Interestingly -D-N- translational modification and migration of glycogen synthase. C, levels acetylglucosaminidase digestion increased the level of glycogen of glycosylated glycogen synthase (gGS) were quantified by densitom- synthase phosphorylation by 165  7% (p  0.01). This indi- etry. Data are represented as the mean  S.E., and significance (p cates that removal of O-GlcNAc facilitates rephosphorylation of 0.05) is indicated by a line. CON, control. the enzyme. This would help explain the lower degree of gly- in STZ-treated mice exhibited an increase in glycosylation. To cogen synthase activation with only -D-N-acetylglucosamini- dase (Fig. 6A, open bars; control: 40  10% and STZ-treated: determine whether glycogen synthase (84 kDa) was one of this group, we incubated extracts with immobilized succinylated 30  10%) compared with -D-N-acetylglucosaminidase- wheat germ agglutinin to isolate proteins that were hypergly- dependent activation in combined digests (Fig. 6A, closed bars; cosylated with O-GlcNAc. Resolution of these proteins and control: 100  10% and STZ-treated: 310  70%). Phosphatase subsequent probing with an anti-glycogen synthase antibody digestion did not change the glycosylation status of the enzyme. indicated that the relative amount of glycogen synthase modi- In mice that were not treated with insulin the absolute 20640 In Vivo Regulation of Glycogen Synthase by O-GlcNAc synthase is dynamically regulated by multiple mechanisms, including substrate availability, hormone signaling, subcellu- lar localization, targeting of phosphatase, and allosteric acti- vation (4, 22, 23, 25). In addition we recently reported that glycogen synthase is regulated by modification with O-GlcNAc (18). Treatment of 3T3-L1 adipocytes with high concentrations of glucose resulted in an increase in glycosylation of glycogen synthase, which inactivated the enzyme and contributed to a reduced response to insulin signaling. Removal of O-GlcNAc was able to correct the resistance to PP1-mediated activation of glycogen synthase (18). We therefore sought to extend the findings in cultured cells and determine whether O-GlcNAc exerts a biologically signifi- cant effect on glycogen synthase activity in the intact organism where glycogen synthase is subject to the full range of physio- logic regulation by hormones and metabolic factors. Results reported here verify that O-GlcNAc modification does occur and does affect glycogen synthase activity in vivo. Glycogen syn- thase from STZ-diabetic mice was inhibited even after insulin treatment. This inhibition of glycogen synthase was more tightly correlated with blood glucose than insulin levels. Addi- tionally the inhibition of glycogen synthase in these diabetic animals could not be explained by increased phosphorylation at the key regulatory 3a site (Ser-640) either before or after insu- lin treatment. This normal dephosphorylation after insulin treatment in vivo demonstrates that dephosphorylation may be necessary but not sufficient for complete activation of glycogen synthase. Full activation of glycogen synthase in vitro required both -D-N-acetylglucosaminidase and PP1. The increase in the efficiency of activation by -D-N-acetylglucosaminidase was much greater in extracts from STZ-treated mice, which also have increased levels of glycosylation. This illustrates a rela- FIG.6. Activation of glycogen synthase with -D-N-acetylglu- tionship between hyperglycemia, glycosylation of glycogen syn- cosaminidase and protein phosphatase 1. A, epididymal fat pad thase, and resistance to activation by PP1. In this model of extracts (13.5 g of protein) from control and STZ-treated mice were digested with 1 unit of -D-N-acetylglucosaminidase (open columns), 0.2 moderate and short term hyperglycemia the ability of insulin to units of protein phosphatase 1 (shaded columns), or both together for 30 lead to dephosphorylation was unimpeded, implying no critical min at 30 °C(hatched columns). Glycogen synthase activity was as- effects on insulin signal transduction upstream of glycogen sayed with 0 and 10 mM glucose 6-phosphate in duplicate. The percent- synthase. However, other studies using different model sys- age of increase in %I-form (average  S.E.) was determined for all mice (COMBINED), control mice only (CONTROL), and STZ-treated diabetic tems have demonstrated inhibition of insulin signal trans- mice (STZ-TREATED). The -D-N-acetylglucosaminidase-dependent duction pathways mediated by the hexosamine biosynthetic component of the glycogen synthase activation observed in the com- pathway (26 –30). bined digests is represented as black columns. Significant increases The level of O-GlcNAc on nuclear and cytosolic proteins is a (p  0.05) in STZ-diabetic compared with control mice are indicated by asterisks. B, epididymal fat pad extracts (400 g of protein) from control function of cellular nutritional status. Flux through the hexos- mice were digested with either 30 units of -D-N-acetylglucosaminidase amine biosynthetic pathway is rate-limited by the concentra- or 6 units of protein phosphatase 1 under the same conditions used in tion of fructose 6-phosphate, and hexosamine flux increases A and C. The resulting mixtures were incubated overnight with O- with increased glucose or fatty acid uptake (14, 15, 31). The GlcNAc-specific succinylated wheat germ agglutinin-agarose, and im- mobilized proteins were resolved on an immunoblot. O-GlcNAc-modi- terminal metabolite of the hexosamine biosynthetic path- fied glycogen synthase (gGS) was measured by staining with an anti- way, UDP-N-acetylglucosamine, is a substrate for O-linked glycogen synthase antibody. The degree of phosphorylation at the N-acetylglucosaminyltransferase, which covalently modifies regulatory Ser-640 site (pGS) was measured by staining an immunoblot serine and threonine residues with O-GlcNAc (32). The K of each digest with a phosphospecific glycogen synthase antibody. C, epididymal fat pad extracts from insulin-untreated control and STZ- value of O-linked N-acetylglucosaminyltransferase for UDP-N- treated mice were digested with 1 unit of -D-N-acetylglucosaminidase acetylglucosamine is relatively high (33). Thus, high levels of and 0.2 units of protein phosphatase 1 compared with protein phospha- O-GlcNAc on nuclear and cytoplasmic proteins reflect a high tase 1 alone. The resulting difference in %I-form that is attributable to level of nutrient uptake (32, 34). This makes O-GlcNAc an -D-N-acetylglucosaminidase is plotted against blood glucose concentra- tions from each animal (control mice, n  2; STZ-treated mice, n  3). attractive candidate for a nutrient sensing mechanism and for The line of best fit was achieved using linear regression. mediating the cellular consequences of excess nutrient flux (34 –37). Consistent with the proposed nutrient sensing role of O- increase in basal glycogen synthase activity seen after diges- GlcNAc, insulin resistance develops in mice with transgenic tion with PP1 plus -D-N-acetylglucosaminidase, compared expression of O-linked N-acetylglucosaminyltransferase or the with PP1 treatment alone, correlated significantly with blood rate-limiting enzyme for the hexosamine biosynthetic pathway: glucose concentrations (r  0.98, p  0.005; Fig. 6C). The glutamine:fructose-6-phosphate amidotransferase (38, 39). In correlation, however, was eliminated with insulin treatment. addition to modulation of glycogen synthase activity, a direct DISCUSSION role for O-GlcNAc in metabolic signaling has been demon- strated by several studies. Glycosylation of the Akt site on Metabolism of glycogen is dependent on the metabolic de- mands of the cell and the total organism (3, 22–24). Glycogen endothelial nitric-oxide synthase prevents activation by insulin In Vivo Regulation of Glycogen Synthase by O-GlcNAc 20641 signaling (40). Modification by O-GlcNAc of transcription fac- glycogen synthase and its partners is necessary for resolving tors involved in metabolism, such as Sp1, YY1, and c-myc, the exact function of O-GlcNAc in modulating glycogen metab- affects protein-protein interactions, transcriptional activation, olism. Recently new methodologies have been developed to and protein stability (41– 47). Pharmacological inhibition of the facilitate the identification of O-GlcNAc sites using mass spec- troscopy (58), and applying these techniques to identify O- enzyme that removes O-GlcNAc from proteins results in down- regulation of elements of the insulin signaling pathway and GlcNAc sites on glycogen synthase should further clarify the mechanisms of inhibition. reduced glucose uptake by glucose transporter 4 (30). Glycogen synthase has been previously regarded as being REFERENCES primarily regulated by phosphorylation-dephosphorylation and 1. Park, K. S., Ciaraldi, T. P., Carter, L., Mudaliar, S., Nikoulina, S. E., Webster, allosteric activation by G6P. These factors are interdependent: N. J., and Henry, R. R. (2000) Metabolism 49, 962–968 increased phosphorylation reduces sensitivity to G6P activa- 2. Crosson, S. M., Khan, A., Printen, J., Pessin, J. E., and Saltiel, A. R. (2003) J. Clin. Investig. 111, 1423–1432 tion, but G6P changes the conformation of the protein to simul- 3. Lawrence, J. C., Jr., and Roach, P. J. (1997) Diabetes 46, 541–547 taneously activate the enzyme and make it a better substrate 4. Ferrer, J. C., Favre, C., Gomis, R. R., Fernandez-Novell, J. M., Garcia-Rocha, M., de la Iglesia, N., Cid, E., and Guinovart, J. J. (2003) FEBS Lett. 546, for dephosphorylation (4, 23, 48, 49). In diabetes this regulation 127–132 breaks down, and glycogen synthase is inhibited and less re- 5. Skurat, A. V., Dietrich, A. D., and Roach, P. J. (2000) Diabetes 49, 1096 –1100 sponsive to insulin activation even though G6P levels are un- 6. Newgard, C. B., Brady, M. J., O’Doherty, R. M., and Saltiel, A. R. (2000) Diabetes 49, 1967–1977 changed (9, 10, 22, 50 –52). In some situations this may be 7. Dent, P., Lavoinne, A., Nakielny, S., Caudwell, F. B., Watt, P., and Cohen, P. partly attributable to upstream effects of the hexosamine path- (1990) Nature 348, 302–308 8. Chang, A. Y. (1972) Can. J. Biochem. 50, 714 –717 way on signal transduction elements such as phosphatidyli- 9. Groop, L. C., Bonadonna, R. C., DelPrato, S., Ratheiser, K., Zyck, K., nositol 3-kinase, insulin receptor substrate-1, or -catenin (26, Ferrannini, E., and DeFronzo, R. A. (1989) J. Clin. Investig. 84, 205–213 10. Thorburn, A. W., Gumbiner, B., Bulacan, F., Wallace, P., and Henry, R. R. 27, 30, 53–55). However, in studies on glycogen synthase from (1990) J. Clin. Investig. 85, 522–529 the livers of STZ-treated rats, diabetes also directly inhibited 11. Kaslow, H. R., Eichner, R. D., and Mayer, S. E. (1979) J. Biol. Chem. 254, the ability of phosphatases to activate the enzyme (12). The 4674 – 4677 12. Akatsuka, A., Singh, T. J., and Huang, K. P. (1983) Arch. Biochem. Biophys. current findings regarding glycosylation of the enzyme may 220, 426 – 434 explain these results. 13. McClain, D. A., and Crook, E. D. (1996) Diabetes 45, 1003–1009 14. Marshall, S., Bacote, V., and Traxinger, R. R. (1991) J. Biol. Chem. 266, The simplest model to explain these findings is that O- 4706 – 4712 GlcNAc is directly inhibitory in a manner analogous to phos- 15. Hawkins, M., Barzilai, N., Liu, R., Hu, M., Chen, W., and Rossetti, L. (1997) J. Clin. Investig. 99, 2173–2182 phate and that removal is required for full activation of the 16. Marshall, S. (2002) Curr. Opin. Endocrinol. Diabetes 9, 160 –167 enzyme. In diabetic conditions modification of glycogen syn- 17. McClain, D. A. (2001) Curr. Opin. Endocrinol. Diabetes 8, 186 –191 thase by O-GlcNAc is increased. Removal of either O-GlcNAc or 18. Parker, G. J., Lund, K. C., Taylor, R. P., and McClain, D. A. (2003) J. Biol. Chem. 278, 10022–10027 phosphate alone is not sufficient for full activation of the en- 19. Comer, F. I., Vosseller, K., Wells, L., Accavitti, M. A., and Hart, G. W. (2001) zyme, but removal of O-GlcNAc facilitates the activation of the Anal. Biochem. 293, 169 –177 20. Roduit, R., Masiello, P., Wang, S. P., Li, H., Mitchell, G. A., and Prentki, M. synthase by dephosphorylation. This hexosaminidase-depend- (2001) Diabetes 50, 1970 –1975 ent increase in activation of the synthase by PP1 is greater in 21. Lauro, D., Kido, Y., Castle, A. L., Zarnowski, M. J., Hayashi, H., Ebina, Y., and diabetes, consistent with an increased level of O-GlcNAc on the Accili, D. (1998) Nat. Genet. 20, 294 –298 22. Shulman, R. G., Bloch, G., and Rothman, D. L. (1995) Proc. Natl. Acad. Sci. enzyme. There are two sites of phosphorylation on glycogen U. S. A. 92, 8535– 8542 synthase that are sufficient for inhibition of the enzyme: Ser-7 23. Villar-Palasi, C., and Guinovart, J. J. (1997) FASEB J. 11, 544 –558 24. Fernandez-Novell, J. M., Roca, A., Bellido, D., Vilaro, S., and Guinovart, J. J. and Ser-640 (5, 56). If either of these were modified by O- (1996) Eur. J. Biochem. 238, 570 –575 GlcNAc, then dephosphorylation alone would not be sufficient 25. Saltiel, A. R., and Kahn, C. R. (2001) Nature 414, 799 – 806 to fully activate the total enzyme population, especially in 26. Patti, M. E., Virkamaki, A., Landaker, E. J., Kahn, C. R., and Yki-Jarvinen, H. (1999) Diabetes 48, 1562–1571 diabetic conditions. Deglycosylation by hexosaminidase would 27. Kim, Y. B., Zhu, J. S., Zierath, J. R., Shen, H. Q., Baron, A. D., and Kahn, B. B. lower the threshold for phosphatase-mediated activation by (1999) Diabetes 48, 310 –320 28. Hawkins, M., Barzilai, N., Chen, W., Angelov, I., Hu, M., Cohen, P., and prior removal of an inhibitory group on the other serine. The Rossetti, L. (1996) Diabetes 45, 1734 –1743 inability of hexosaminidase digestion alone to activate glycogen 29. Federici, M., Menghini, R., Mauriello, A., Hribal, M. L., Ferrelli, F., Lauro, D., Sbraccia, P., Spagnoli, L. G., Sesti, G., and Lauro, R. (2002) Circulation 106, synthase was explained in part by the phosphorylation status 466 – 472 of glycogen synthase. In the digestion conditions used in this 30. Vosseller, K., Wells, L., Lane, M. D., and Hart, G. W. (2002) Proc. Natl. Acad. study, removal of O-GlcNAc by hexosaminidase was followed Sci. U. S. A. 99, 5313–5318 31. Yki-Jarvinen, H., Virkamaki, A., Daniels, M. C., McClain, D., and Gottschalk, by subsequent rephosphorylation of the enzyme by endogenous W. K. (1998) Metabolism 47, 449 – 455 kinases. This would have blunted the expected activation of 32. Wells, L., Vosseller, K., and Hart, G. W. (2001) Science 291, 2376 –2378 33. Kreppel, L. K., and Hart, G. W. (1999) J. Biol. Chem. 274, 32015–32022 glycogen synthase. Alternatively glycosylation may sterically 34. McClain, D. A. (2002) J. Diabetes Complicat. 16, 72– 80 hinder access by phosphatases or result in a conformation of 35. Obici, S., Wang, J., Chowdury, R., Feng, Z., Siddhanta, U., Morgan, K., and Rossetti, L. (2002) J. Clin. Investig. 109, 1599 –1605 the synthase that is a poorer substrate for phosphatase. G6P is 36. Rumberger, J. M., Wu, T., Hering, M. A., and Marshall, S. (2003) J. Biol. well known to change the conformation of glycogen synthase Chem. 278, 28547–28552 and make it a better substrate for phosphatases (57). O-GlcNAc 37. Patti, M. E., Butte, A. J., Crunkhorn, S., Cusi, K., Berria, R., Kashyap, S., Miyazaki, Y., Kohane, I., Costello, M., Saccone, R., Landaker, E. J., may block the conformational changes induced by G6P or even Goldfine, A. B., Mun, E., DeFronzo, R., Finlayson, J., Kahn, C. R., and limit access of G6P itself, impeding the ability of phosphatase Mandarino, L. J. (2003) Proc. Natl. Acad. Sci. U. S. A. 100, 8466 – 8471 38. Hebert, L. F., Jr., Daniels, M. C., Zhou, J., Crook, E. D., Turner, R. L., to activate the enzyme. The role of O-GlcNAc could also be an Simmons, S. T., Neidigh, J. L., Zhu, J. S., Baron, A. D., and McClain, D. A. indirect one. O-GlcNAc is now documented to interfere with (1996) J. Clin. Investig. 98, 930 –936 39. McClain, D. A., Lubas, W. A., Cooksey, R. C., Hazel, M., Parker, G. J., Love, some protein-protein interactions (44). Thus, the association of D. C., and Hanover, J. A. (2002) Proc. Natl. Acad. Sci. U. S. A. 99, PP1 with glycogen-targeting regulatory subunits and glycogen 10695–10699 synthase may well be inhibited by glycosylation of the synthase 40. Du, X. L., Edelstein, D., Dimmeler, S., Ju, Q., Sui, C., and Brownlee, M. (2001) J. Clin. Investig. 108, 1341–1348 or other components of the glycogen synthetic machinery. 41. Yang, X., Su, K., Roos, M. D., Chang, Q., Paterson, A. J., and Kudlow, J. E. The current results are significant in that they demonstrate (2001) Proc. Natl. Acad. Sci. U. S. A. 98, 6611– 6616 42. Kamemura, K., Hayes, B. K., Comer, F. I., and Hart, G. W. (2002) J. Biol. a role and provide a mechanism by which excess nutrients lead Chem. 277, 19229 –19235 directly to insulin resistance and altered protein function in 43. Roos, M. D., Su, K., Baker, J. R., and Kudlow, J. E. (1997) Mol. Cell. Biol. 17, diabetes. The identification of the sites of glycosylation on 6472– 6480 20642 In Vivo Regulation of Glycogen Synthase by O-GlcNAc 44. Hiromura, M., Choi, C. H., Sabourin, N. A., Jones, H., Bachvarov, D., and 51. Kaslow, H. R., and Eichner, R. D. (1984) Am. J. Physiol. 247, E581–E584 Usheva, A. (2003) J. Biol. Chem. 278, 14046 –14052 52. Rossetti, L., Giaccari, A., and DeFronzo, R. A. (1990) Diabetes Care 13, 45. Majumdar, G., Harmon, A., Candelaria, R., Martinez-Hernandez, A., Raghow, 610 – 630 R., and Solomon, S. S. (2003) Am. J. Physiol. 285, E584 –E591 53. Singh, L. P., and Crook, E. D. (2000) J. Investig. Med. 48, 251–258 46. Clark, R. J., McDonough, P. M., Swanson, E., Trost, S. U., Suzuki, M., Fukuda, 54. Virkamaki, A., Daniels, M. C., Hamalainen, S., Utriainen, T., McClain, D., and M., and Dillmann, W. H. (2003) J. Biol. Chem. 278, 44230 – 44237 Yki-Jarvinen, H. (1997) Endocrinology 138, 2501–2507 47. Kang, H. T., Ju, J. W., Cho, J. W., and Hwang, E. S. (2003) J. Biol. Chem. 278, 55. Singh, L. P., Gennerette, D., Simmons, S., and Crook, E. D. (2001) J. Diabetes 51223–51231 Complicat.15, 88 –96 48. Roach, P. J., Takeda, Y., and Larner, J. (1976) J. Biol. Chem. 251, 1913–1919 56. Skurat, A. V., Wang, Y., and Roach, P. J. (1994) J. Biol. Chem. 269, 49. Camici, M., DePaoli-Roach, A. A., and Roach, P. J. (1984) J. Biol. Chem. 259, 25534 –25542 3429 –3434 57. Villar-Palasi, C. (1991) Biochim. Biophys. Acta 1095, 261–267 50. Yki-Jarvinen, H., Mott, D., Young, A. A., Stone, K., and Bogardus, C. (1987) 58. Wells, L., Vosseller, K., Cole, R. N., Cronshaw, J. M., Matunis, M. J., and Hart, J. Clin. Investig. 80, 95–100 G. W. (2002) Mol. Cell. Proteomics 1, 791– 804

Journal

Journal of Biological Chemistry – Unpaywall

Published: May 1, 2004

There are no references for this article.