Effect of acute exercise on glycogen synthase in muscle from obese and diabetic subjects

Abstract

Insulin stimulates glycogen synthase (GS) through dephosphorylation of serine residues, and this effect is impaired in skeletal muscle from insulin-resistant [obese and type 2 diabetic (T2DM)] subjects. Exercise also increases GS activity, yet it is not known whether the ability of exercise to affect GS is impaired in insulin-resistant subjects. The objective of this study was to examine the effect of acute exercise on GS phosphorylation and enzyme kinetic properties in muscle from insulin-resistant individuals. Lean normal glucose-tolerant (NGT), obese NGT, and obese T2DM subjects performed 40 min of moderate-intensity cycle exercise (70% of V̇o_2max). GS kinetic properties and phosphorylation were measured in vastus lateralis muscle before exercise, immediately after exercise, and 3.5 h postexercise. In lean subjects, GS fractional activity increased twofold after 40 min of exercise, and it remained elevated after the 3.5-h rest period. Importantly, exercise also decreased GS K_m for UDP-glucose from ≈0.5 to ≈0.2 mM. In lean subjects, exercise caused significant dephosphorylation of GS by 50–70% (Ser⁶⁴¹, Ser⁶⁴⁵, and Ser^{645,649,653,657}), and phosphorylation of these sites remained decreased after 3.5 h; Ser⁷ phosphorylation was not regulated by exercise. In obese NGT and T2DM subjects, exercise increased GS fractional activity, decreased K_m for UDP-glucose, and decreased GS phosphorylation as effectively as in lean NGT subjects. We conclude that the molecular regulatory process by which exercise promotes glycogen synthesis in muscle is preserved in insulin-resistant subjects.

skeletal muscle is the main site responsible for insulin-stimulated glucose disposal (11). After glucose enters the muscle fibers, it is converted to glucose 6-phosphate (G6P). Upon insulin stimulation during the postprandial period, ∼80% of the glucose that is taken up by the muscle is utilized for glycogen synthesis (37). Insulin stimulates glucose transport/phosphorylation and activates glycogen synthase (GS), which catalyzes incorporation of glucose from uridyl diphosphate (UDP)-glucose into glycogen (7). Glucose transport is normally considered the rate-limiting step for glycogen storage (34), although impaired GS activation by insulin also has been proposed to limit glycogen synthesis in insulin-resistant subjects (13, 14).

GS activity is regulated by covalent phosphorylation and allosteric activation by G6P (7, 16). It is thought that both mechanisms contribute to the stimulation of glycogen synthesis in response to insulin by increasing GS fractional activity and intramuscular G6P concentration, secondary to increased glucose transport (3, 14, 15, 34). The role of G6P-mediated allosteric activation has been demonstrated in muscles from knockin mice expressing a GS mutant insensitive to G6P. These mice displayed normal insulin-stimulated GS dephosphorylation and activation but impaired insulin-stimulated glycogen synthesis (4). In rat muscle, both insulin and exercise increase phosphodependent GS activity by enhancing GS sensitivity for G6P and affinity for UDP-glucose (21, 33). The contribution of phosphorylation-dependent regulation on GS kinetic properties in humans is not known.

Insulin-resistant subjects have impaired insulin-mediated muscle glucose disposal (14, 44, 46). Some studies also have demonstrated impaired insulin signaling through Akt (also known as PKB) (45) and reduced insulin-stimulated GS activation (12, 45). Exercise plays a fundamental role in the treatment and prevention of type 2 diabetes mellitus (T2DM) (18). Although the molecular mechanisms by which exercise improves glucose homeostasis are not fully understood, it is thought that exercise improves glycemia in part by enhancing glucose transport into the contracting muscles fibers (32), where it is oxidized to generate ATP or stored as glycogen. Similarly to insulin, exercise promotes dephosphorylation of GS, leading to its enzymatic activation for glycogen synthesis (21, 25, 30, 35, 36). Notably, it is not entirely clear which phosphorylation sites mediate exercise-induced GS activation in muscle, particularly in humans.

Although exercise can enhance insulin-stimulated glucose disposal (5, 6) in insulin-resistant subjects (5, 6), the effect of exercise on GS phosphorylation and affinity for UDP-glucose has not been studied in these individuals. The goal of this study was to examine the effect of exercise on GS kinetic properties and GS dephosphorylation at Ser⁷, Ser⁶⁴¹, and Ser^{645,649,653,657} in normal and insulin-resistant (obese nondiabetic and T2DM) subjects.

METHODS

Subjects.

Seven obese (BMI = 33.6 ± 1.9 kg/m²) normal glucose tolerant (NGT), six obese (BMI = 31.1 ± 1.7) T2DM, and nine lean (BMI = 24 ± 1.0) NGT subjects participated in this study. No subject exercised regularly (0 to 1 exercise sessions/wk). Each subject underwent a medical history, physical examination, and 75-g oral glucose tolerance test (OGTT). Three T2DM subjects took a sulfonylurea, which was stopped 2 days before any experiment to avoid hypoglycemia. Three T2DM subjects were diet treated. In lean and obese subjects normal glucose tolerance was documented with the OGTT (1), and these subjects did not have a family history (first-degree relative) of T2DM. Other than the sulfonylureas, subjects were not taking any medication known to affect glucose or lipid metabolism. The study was approved by the Institutional Review Board of the UTHSCSA, and all subjects gave written voluntary consent.

OGTT.

After an overnight fast, plasma glucose was measured at baseline and 2 h after the ingestion of 75 g of glucose. Plasma insulin and free fatty acid concentrations were measured at baseline.

V̇o_2max testing.

Within 7 days after the OGTT, V̇o_2max was determined using a cycle ergometer and a Metabolic Measurement System (Sensormedics, Savi Park, CA), as described previously (40).

Measurement of insulin sensitivity.

Within 3–7 days after the V̇o_2max test, subjects underwent a 180-min euglycemic hyperinsulinemic (160 mU·m²·min⁻¹) clamp study. Insulin-stimulated glucose metabolism (M) was determined as the mean glucose infusion rate during the last 30 min of the clamp (11).

Acute exercise protocol.

Within 7–10 days after the insulin clamp, subjects underwent an acute exercise experiment with muscle biopsies. After arriving at the Clinical Research Center, subjects rested in bed for 30 min, and a vastus lateralis muscle biopsy was performed under local anesthesia (1% lidocaine) using a Bergström cannula with suction. The muscle was rapidly (within ≈7 s) debrided of adipose and connective tissue and frozen in liquid nitrogen. Subjects then exercised on the cycle ergometer. From the V̇o_2max test, a power output designed to elicit an intensity of 70% of V̇o_2max was calculated for each subject. Subjects cycled at the designated power output for 40 min and were then placed on the bed, and a second muscle biopsy was obtained (at ≈35 min, subjects stopped exercising for ∼20 s for lidocaine application). The subjects then rested in a bed for 210 min, and a third muscle biopsy was performed under local anesthesia. Each biopsy site was separated by ≥5 cm. In two lean and two diabetic subjects the 210-min postexercise biopsies were not available due to technical reasons.

Laboratory analyses.

Plasma insulin was measured by radioimmunoassay (Diagnostic Products, Los Angeles, CA), plasma glucose by the glucose oxidase method (Beckman, Fullerton, CA), and hemoglobin A_1c (Hb A_1c) using a DCA2000 Analyzer (Bayer, Tarrytown, NY). Plasma free fatty acid (FFA) concentration was determined using a colorimetric method (Wako, Richmond, VA).

GS activity.

Muscles were homogenized with a Polytron (Kinematica, Littau-Luzern, Switzerland) in ice-cold buffer (1:400) containing 50 mM Tris·HCl (pH 7.8), 100 mM NaF, and 10 mM EDTA. Homogenates were centrifuged at 3,000 g for 30 min at 4°C. For analysis, 20 μl of supernatant was added to 40 μl of assay buffer containing 25 mM Tris·HCl (pH 7.8), 50 mM NaF, 5 mM EDTA, glycogen (10 mg/ml), different concentrations of UDP-glucose (see below), 0.5 μCi/ml d-[¹⁴C]UDP in the presence of 0.17 mM, and 12 mM G6P. GS %-I form was calculated as activity without G6P in percent of activity at 12 mM G6P; the concentration of UDP-glucose was 1.67 mM. GS affinity for UDP-glucose (K_m) and V_max were analyzed with 0.17 and 12 mM G6P, and assays were conducted with the following concentrations of UDP-glucose: 1.67, 0.8, 0.4, 0.2, 0.1, 0.05, and 0.03 mM. Kinetic data were linearized as Eadie-Hofstee plots, and GS K_m for UDP-glucose was calculated as the reciprocal to the slope and V_max as the intercept with the y-axis. K_m-0.17 and V_max-0.17 refer to the analysis performed with 0.17 mM G6P, and K_m-12 and V_max-12 refer to the analysis performed with 12 mM. Concentrations of UDP-glucose and G6P in stock solutions were determined spectrophotometrically, as described (27). Homogenate protein concentration was measured using a colorimetric assay (DC Protein Assay; Bio-Rad, Hercules, CA).

Glycogen content.

Glycogen was measured in homogenates prepared for GS analysis. One-hundred fifty micoliters of crude homogenate was hydrolyzed with 300 μl of 1.8 M HCl (100°C, 2.5 h) and glycogen determined fluorometrically as glucose units (27).

Immunoblotting.

Muscles were homogenized (1:25) twice for 15 s in ice-cold buffer containing 50 mM HEPES, 150 mM NaCl, 10 mM Na₄P₂O₇, 30 mM NaF, 1 mM NA₃VO₄, 10 mM EDTA, 2.5 mM benzamidine, and 2 μl/ml protease inhibitor cocktail (P-8340; Sigma), and 1% Triton X-100 was added. Homogenates were rotated for 1 h at 4°C and centrifuged (11,500 g for 10 min at 4°C). Homogenates were diluted to 1.5 μg/μl for immunoblotting, and proteins (≈15 μg) were resolved in 10% SDS gels, transferred to polyvinylidene difluoride membranes (Millipore, Bedford, MA), and probed with the corresponding primary and secondary antibodies. Antibody binding was detected with enhanced chemiluminescence reagents (Millipore, Billerica, MA) and quantified using FUJI LAS-4000 Mini (FujiFilm, Tokyo, Japan).

Antibodies.

Anti-dual phospho-GS Ser^7,10 and anti-phospho-GS Ser⁷ were gifts from D. Grahame Hardie (Dundee, UK). Anti-phospho-GS Ser^641,645 was generated in house, as described previously (3). Anti-total GS was a gift from Oluf Pedersen (Copenhagen, Denmark). Anti-phospho-GS Ser^{645,649,653,657} was obtained from Oncogene (San Diego, CA), and anti-phospho-GS Ser⁶⁴¹ and anti-GSK3α/β Ser^21/9 were obtained from Cell Signaling Technology (Beverly, MA).

Statistics.

Data are presented as means ± SE. Baseline characteristics between groups were compared with one-way ANOVA. The effect of exercise and recovery was analyzed using one-way repeated measures ANOVA followed by Fisher's least significant difference test. A P value <0.05 was considered to be significant.

RESULTS

Subject characteristics.

Table 1 shows the subject characteristics. Obese NGT and T2DM had higher BMI than the lean NGT group. T2DM subjects had significantly higher plasma glucose level (baseline and during the OGTT), Hb A_1c, and FFA concentrations than obese and lean NGT subjects. T2DM and obese subjects were more insulin resistant than the lean subjects based on the lower M value during the insulin clamp. Subjects in the T2DM group had lower V̇o_2max than lean subjects.

Table 1.

Characteristics of subjects

	Lean	Obese	T2DM
Age, yr	43.4 ± 3.6	41.4 ± 3.4	53.3 ± 3.9
Sex (F/M)	6/3	3/4	4/2
Race (H/C/AA)	6/2/1	5/2/0	4/2/0
Weight, kg	63.8 ± 3.6	94.9 ± 7.6	81.5 ± 4.5
BMI, kg/m2	24.0 ± 1.0	33.6 ± 1.9L	31.1 ± 1.7L
V̇o2max, ml·min−1·kg−1	22.3 ± 1.9	19.5 ± 1.2	16.6 ± 2.0L
Hb A1c, %	5.3 ± 0.1	5.1 ± 0.2	9.0 ± 1.4L
Fasting plasma glucose, mmol/l	5.0 ± 0.2	5.4 ± 0.1	10.2 ± 1.7L
OGTT 2-h plasma glucose, mmol/l	6.4 ± 0.4	6.9 ± 0.5	15.9 ± 1.9L
Fasting plasma insulin, μU/l	4.2 ± 1.1	14.8 ± 8.4L	14.1 ± 3.0L
Fasting plasma FFA, μmol/l	531 ± 72	451 ± 57	756 ± 74O
M value, mg·kg−1·min−1	10.5 ± 0.6	8.5 ± 0.9L	6.6 ± 1.3L
Work during exercise, W	79 ± 7	90 ± 8	67 ± 9

Data are means ± SE. Blood samples were taken after an overnight fast. BMI, body mass index; FFA, free fatty acids; H, Hispanic; C, Caucasian; AA, African-American; F, female; M, male. ^LP < 0.05 vs. lean; ^OP < 0.05 vs. obese.

Glycogen content.

Glycogen content in muscle was not significantly different in muscle from lean, obese, or diabetic subjects prior to exercise (Fig. 1A). Exercise robustly decreased muscle glycogen content in lean, obese, and T2DM subjects, and the decrease in glycogen content caused by exercise (delta from baseline to exercise time point) was similar between groups (185 ± 25, 159 ± 32, and 177 ± 39 mmol/kg dry wt in lean, obese, and T2DM, respectively, P < 0.05 in all groups vs. baseline). No significant increase in muscle glycogen content was observed between the preexercise and the 3.5-h postexercise time points.

Fig. 1.

Effect of exercise on glycogen content and glycogen synthase (GS) fractional activity. Muscle biopsies were taken prior to, immediately after, and 3.5 h after 40 min of cycling at 70% V̇o_2max and analyzed for glycogen content (A), GS fractional activity with a physiological concentration of uridyl diphosphate (UDP)-glucose (B), GS fractional activity with 1.67 mM UDP-glucose in assay buffer (C), and GS in the absence of glucose 6-phosphate (%I-form) with 1.67 mM UDP-glucose (D). *P < 0.05 compared with basal (prior to exercise). Open bars, lean subjects; filled bars, obese subjects; hatched bars, T2DM subjects.

GS activity.

Baseline GS fractional activity, measured with a physiological concentration of UDP-glucose (FV_0.03), was not significantly different in the lean, obese, and T2DM groups, although there was a tendency for a lower FV_0.03 in T2DM compared with lean subjects (Fig. 1B). Exercise increased the FV_0.03 in lean, obese, and T2DM subjects by 1.9-, 2.3-, and 3.2-fold, respectively (P < 0.05 vs. baseline in all groups); GS activity measured immediately after 40 min of exercise was similar between groups (Fig. 1B). After 3.5 h of rest, GS activity (FV_0.03) remained elevated in the three groups (P < 0.05 vs. basal in all groups); there were no differences in FV_0.03 between groups at the 3.5-h postexercise time point (Fig. 1B). GS fractional activity, measured with a high concentration of UDP-glucose (FV_1.67), was increased significantly in lean, obese, and diabetic subjects after 40 min of exercise (P < 0.05 vs. baseline in all groups), and FV_1.67 remained elevated during the 3.5-h postexercise period (Fig. 1C). There were no statistically significant differences in FV_1.67 between groups after 40 min of exercise or at the 3.5-h postexercise time point (Fig. 1C). At baseline, GS %I-form was similar in lean, obese, and T2DM groups (Fig. 1D). Exercise caused a time-dependent increase in GS %I-form that in all of the groups was highest at the 3.5-h postexercise time point (Fig. 1D). There were no differences between groups in maximal GS activity calculated from Eadie-Hofstee plots (Table 2). GS activity measured with 1.67 and 12 mM UDP-glucose (often called total activity) was also similar in all groups (data not shown). Exercise did not affect maximal GS activity (Table 3). Figure 2A shows that there was a strong, direct correlation between muscle glycogen content and maximal GS activity at baseline when pooled data from the lean, obese, and diabetic groups were analyzed. GS fractional activity (FV_0.03) measured in pooled samples from the three groups before, during, and after exercise correlated inversely with muscle glycogen content (Fig. 2B).

Table 2.

Glycogen synthase activity in muscles from lean, obese, and T2DM subjects

	Maximal Glycogen Synthase Activity, mmol·min−1·kg dry wt−1
	Lean	Obese	T2DM
Basal	16.6 ± 1.4	14.7 ± 1.0	13.9 ± 0.8
After exercise	15.2 ± 1.3	14.0 ± 1.1	12.8 ± 1.9
3 h Postexercise	13.8 ± 1.1	15.3 ± 1.6	16.5 ± 4.1

Data are means ± SE. T2DM, type 2 diabetes mellitus. Maximal glycogen synthase activity was calculated from Eadie-Hofstee plots with 12 mM glucose 6-phosphate.

Table 3.

GSK-3β Ser⁹ phosphorylation in lean, obese, and T2DM subjects before, immediately after, and 3.5 h after 40 min of exercise

	GSK-3β Ser9 Phosphorylation (Arbitrary Units)
	Lean	Obese	T2DM
Preexercise	21.1 ± 3.1	26.1 ± 3.1	22.3 ± 1.2
Postexercise	24.4 ± 4.0	21.7 ± 2.2	20.0 ± 3.5
3.5 h Postexercise	21.3 ± 2.8	25.7 ± 3.7	25.1 ± 3.3

Data are means ± SE.

Fig. 2.

Correlation between glycogen content and GS activity. A: correlation between maximal GS activity with 1.67 mM UDP-glucose (V_max-12) and glycogen content at baseline. B: correlation between glycogen content and GS fractional activity (FV_0.03) measured before, during, and after exercise. Open symbols, before exercise; filled symbols, immediately after exercise; filled symbols with cross, 3.5 h after exercise.

GS affinity to UDP-glucose was measured with a physiological (0.17 mM) and a pharmacological concentration (12 mM) of G6P. K_m-0.17 in baseline muscle samples was ∼0.5 mM, which is in agreement with our previous data in rat muscle (21, 23). Interestingly, baseline K_m-0.17 was 47% higher in muscles from diabetic subjects compared with lean individuals (P < 0.05; Fig. 3A). Exercise decreased K_m-0.17 to ∼0.2 mM in all groups, and no differences were observed between groups. GS affinity for UDP-glucose remained elevated 3.5 h after exercise in all groups (Fig. 3A). In all three groups, 12 mM G6P increased the affinity of GS to UDP-glucose, as evidenced by a decrease in K_m-12 to 0.15 mM (Fig. 3B). Exercise did not significantly affect K_m-12 in any of the groups (Fig. 3B).

Fig. 3.

Effect of exercise on GS affinity for UDP-glucose. Muscle biopsies were taken prior to, immediately after, and 3.5 h after 40 min of cycling at 70% V̇o_2max and GS affinity for UDP-glucose analyzed with a physiological (0.17 mM) concentration of glucose 6-phosphate (A) and with 12 mM glucose 6-phosphate (B). *Significantly different from basal (prior to exercise); †significantly different from lean. Open bars, lean subjects; filled bars, obese subjects; hatched bars, T2DM subjects.

GS protein content and phosphorylation.

At baseline, GS protein content was decreased significantly in T2DM subjects (P < 0.05 vs. lean; Fig. 4A) and was accompanied by a trend for a reduction (17%, P = not significant) in GS total activity (Table 2). At baseline, GS Ser⁶⁴¹ phosphorylation (corrected for protein content) was similar in the three groups (Fig. 4B). Exercise significantly decreased GS Ser⁶⁴¹ phosphorylation to a similar level in all groups (P < 0.05 vs. baseline in the 3 groups), and it remained decreased 3.5 h after exercise (Fig. 4B). Prior to exercise, baseline GS Ser^{645,649,653,657} phosphorylation (corrected to GS protein content) was similar in all groups (Fig. 4C). Exercise caused GS Ser^{645,649,653,657} dephosphorylation to a similar level in the three groups (P < 0.05 vs. rest). GS Ser^{645,649,653,657} phosphorylation remained significantly decreased 3.5 h after exercise in lean and obese subjects (P < 0.05 vs. baseline in both groups; Fig. 4C). GS Ser⁷ phosphorylation was similar between groups after correcting for GS protein content and was not influenced by exercise (Fig. 4D). In line with this finding, there was no difference in baseline GS Ser^{7 + 10} phosphorylation between groups, and it did not change with exercise (data not shown). Analysis of pooled data from lean, obese, and T2DM subjects before and after exercise revealed that GS Ser⁶⁴¹ and Ser^{645,649,653,657} phosphorylation correlated inversely with GS FV_0.03 (r = −0.580 and −0.602, respectively, P < 0.001 for both) and directly with K_m-0.17 (r = 0.546 and 0.524 respectively, P < 0.001 for both). Notably, there was a strong curvilinear relationship between GS FV_0.03 and K_m-0.17 (Fig. 5A). GS fractional activity, measured with 0.03 mM (FV_0.03), correlated closely with GS fractional activity measured with 1.67 mM UDP-glucose (FV_1.67) in the assay buffer (Fig. 5B).

Fig. 4.

Effect of exercise on GS phosphorylation. Muscle biopsies were taken prior to, immediately after, and 3.5 h after 40 min of cycling at 70% V̇o_2max, and GS expression and phosphorylation were assessed. A: total GS expression in biopsies from lean, obese, and T2DM subjects taken before, immediately after, and 3.5 h after exercise. B: GS Ser⁶⁴¹ phosphorylation corrected for GS expression. C: GS Ser^{645,649,653,657} phosphorylation corrected for GS expression. D: GS Ser⁷ phosphorylation corrected for GS expression. E: representative blots showing GS phosphorylation at Ser⁶⁴¹, Ser^{645,649,653,657}, Ser⁷, and GS expression. *Significantly different from basal (prior to exercise); †significantly different from lean. Open bars, lean subjects; filled bars, obese subjects; hatched bars, T2DM subjects. L, lean; O, obese; D, diabetic.

Fig. 5.

Correlation between GS phosphorylation and fractional activity. A: relationship between GS affinity (K_m-0.17) for UDP-glucose and GS fractional activity (FV_0.03). B: relationship between GS fractional activity measured with 1.67 mM (FV_1.67) and 0.03 mM (FV_0.03) UDP-glucose. Inverted triangles, lean; squares, obese; circles, T2DM. Open symbols, before exercise; filled symbols, immediately after exercise; filled symbols with cross, 3.5 h after exercise.

GSK-3 phosphorylation.

GSK3β Ser⁹ phosphorylation was similar in lean, obese, and diabetic subjects (Table 3). Exercise did not influence GSK-3β Ser⁹ phosphorylation in any group of subjects.

DISCUSSION

GS activity is regulated by covalent phosphorylation and the allosteric activator G6P (7, 16). Recent studies performed in rodents suggest that allosteric activation by G6P plays a more prominent role than phosphorylation in the regulation of muscle GS activity and glycogen synthesis in response to insulin (3, 4). Nonetheless, the phosphorylation state of GS influences its enzymatic activity by altering its affinity for G6P and the K_m for its substrate UDP-glucose (16). In addition, G6P increases GS affinity for UDP-glucose (21, 23, 24, 33). These interactions allow for a marked increase in the rate of glycogen synthesis in response to relatively small changes in the level of the allosteric activator G6P and the substrate UDP-glucose when GS phosphorylation is reduced.

The originality of this study lies in that we demonstrate that exercise promotes an increase in GS affinity for UDP-glucose in human skeletal muscle and that exercise increases GS affinity for UDP-glucose with similar efficiency in muscles from lean, obese, and T2DM subjects. Previous studies in rodents have shown that strenuous contractions induced via electrical stimulation increase GS affinity for UDP-glucose (21, 22). The present study is novel in that it demonstrates that a more physiological mode of exercise (40 min of moderate-intensity cycling) is able to reduce significantly the K_m of GS for UDP-glucose in human subjects that are both insulin sensitive and insulin resistant (obese and type 2 diabetic). The GS K_m of ≈0.2 mM UDP-glucose after exercise at a physiological concentration of G6P (0.17 mM) is far above the physiological concentration of UDP-glucose (0.03 mM) in resting muscles (28, 31), and the increased affinity for UDP-glucose would be expected to increase the rate of glycogen synthesis after exercise. Our data suggest that increased GS affinity for UDP-glucose represents an important physiological mechanism by which contraction promotes glycogen synthesis.

In agreement with previous studies, exercise increased GS fractional activity similarly in normal, obese, and diabetic subjects (8, 30), indicating elevated sensitivity for G6P. Our study also demonstrated a much greater increase in GS fractional activity after exercise with a physiological UDP-glucose concentration in the assay buffer (≈30 μM) compared with a high UDP-glucose concentration (16, 19). Glycogen content is a powerful regulator of GS fractional activity (9, 15, 21). In this study, 40 min of exercise at 70% V̇o_2max decreased glycogen content similarly in lean, obese, and T2DM subjects and caused a comparable increase in GS fractional activity. These findings support the concept that glycogen breakdown (i.e., decreased muscle glycogen concentration) contributes to the activation of GS (17), and the regulation of GS activity by glycogen is preserved in insulin-resistant subjects.

In this study, we performed a detailed kinetic analysis of GS activity at rest, during exercise, and after exercise in muscle from lean insulin-sensitive and insulin-resistant (obese nondiabetic and T2DM) subjects. At rest, GS affinity for UDP-glucose was lower (higher K_m-017) in obese and T2DM subjects compared with lean individuals, whereas GS fractional activity was comparable between the three groups. This suggests that GS affinity for UDP-glucose is a more sensitive parameter of changes in enzymatic activity than measurements of GS fractional activity. It should be noted that the function of GS may also be affected by other mechanisms aside from covalent phosphorylation and allosteric modification, including subcellular compartmentalization (30) and posttranslational modification (42). Although one study found that the cellular localization of glycogen particles was similar between lean and T2DM subjects at baseline and after 10 wk of endurance training (26), in the future it will be important to evaluate the effect that acute exercise has on GS localization and posttranslational modification in muscle from healthy and T2DM individuals.

Another goal of this study was to examine whether exercise stimulated dephosphorylation of GS in insulin-resistant subjects to the same degree as in lean individuals. Interestingly, exercise decreased GS Ser⁶⁴¹ and Ser^{645,649,653,657} phosphorylation with similar efficiency in lean, obese, and T2DM subjects. These sites, which have been shown to regulate GS activity (39), are dephosphorylated by exercise in normal (insulin-sensitive) muscle (21, 25, 30). GSK-3 phosphorylates the COOH-terminal sites of GS (Ser⁶⁴¹, Ser⁶⁴⁵, Ser⁶⁴⁹, and Ser⁶⁵³) (7). Nonetheless, we did not observe changes in GSK-3β Ser⁹ phosphorylation during exercise despite the decreases in GS Ser⁶⁴¹ and Ser^{645,649,653,657} phosphorylation. These findings indicate that, in human skeletal muscle, exercise-mediated GS dephosphorylation and activation occur independently of changes in GSK-3 inactivation. This is consistent with a prior study performed in mice that showed that insulin, but not exercise, requires GSK-3 inactivation to promote dephosphorylation and activation of GS (3, 25). Experiments in knockout mice demonstrated that exercise, unlike insulin, requires the presence of protein phosphatase-1 (PP1) regulatory subunit R_GL/GM for the activation of GS (2). In this study, we did not have sufficient muscle tissue to evaluate the role of PP1 R_GL/GM on the regulation of GS activity during exercise, but this would be an important topic for investigation in the future.

Mutagenesis analysis of overexpressed muscle GS in COS cells (38) as well as analysis of muscle biopsies from human subjects have suggested that GS Ser⁷ and/or GS Ser⁷⁺¹⁰ phosphorylation are important regulatory sites in controlling GS activity and glycogen synthesis. Previously, we found that contraction increases GS Ser⁷ phosphorylation in rat muscles, but phosphorylation of this site was not associated with GS FV (21, 23). In the present study, we did not observe increases in GS Ser⁷ or GS Ser^{7 + 10} phosphorylation in rested and exercised muscles in any group. These data agree with results from Prats et al. (30), who reported that acute exercise did not increase GS Ser⁷ or GS Ser^{7 + 10} phosphorylation in muscle from young healthy humans, although we cannot rule out transient phosphorylation of these sites at earlier time points. Overall, the present data suggest that exercise promotes GS activity via dephosphorylation of GS Ser⁶⁴¹ and Ser^{645,649,653,657}, and this is supported further by the observed direct correlation between phosphorylation of these sites and GS activity (increased affinity for UDP-glucose and G6P). These findings agree with those reported in previous studies (20, 23, 30) and suggest that exercise regulates GS via dephosphorylation of Ser⁶⁴¹ and Ser^{645,649,653,657} in muscles from both insulin-sensitive and insulin-resistant subjects.

Although both GS fractional activity and affinity for UDP-glucose correlated with GS Ser⁶⁴¹ and Ser^{645,649,653,657} phosphorylation, there was no linear relationship between GS FV_0.03 and K_m-0.17. Rather, there was a curvilinear relationship, as seen in rat muscles when GS activity was manipulated by altering glycogen content and by exercise (16). Also, we found that GS FV_0.03 declined below 20% before a substantial increase in K_m-0.17 was noted. Although we do not know the reason for this early decline in GS FV_0.03, this finding could explain why GS affinity for UDP-glucose may be a more sensitive parameter for changes in enzymatic activity than measurements of fractional activity. It would be important to investigate in future studies whether the higher K_m-0.17 observed in T2DM occurs in conjunction with impaired insulin-induced GS activation.

Insulin-mediated GS activation and dephosphorylation are impaired in insulin-resistant subjects (14, 29, 43). It is well documented that exercise increases nonoxidative glucose disposal (8, 10). Christ-Roberts et al. (6) showed that nonoxidative glucose storage was enhanced immediately after exercise in the diabetic subjects, and glucose disposal has been reported to correlate with GS fractional activity (5). Therefore, our finding that exercise normally dephosphorylates and activates GS may help to explain why exercise stimulates glycogen synthesis and enhances insulin sensitivity in insulin-resistant subjects.

This study evaluated the GS response to a single bout of aerobic exercise. However, it is possible that differences in the GS response could be identified in insulin-resistant subjects by testing the effect of a different exercise challenge. For example, St-Onge et al. (41) demonstrated that GS protein content in muscle increases in response to 6 wk of electrical stimulation in subjects that do not carry the XbaI GS polymorphism, whereas muscle contraction does not increase GS content in carriers of this mutation. In the future, it will be interesting to investigate whether abnormalities in GS function can be detected by testing the response to other exercise/contraction protocols.

In summary, exercise increased GS activity during exercise, and its activity remained elevated in the postexercise period. The sustained effect of exercise on GS activity may contribute to the increases in insulin-mediated glucose disposal after exercise in human skeletal muscle. Overall, we did not observe major differences in exercise-stimulated GS dephosphorylation and activation between lean, obese, and T2DM individuals, suggesting that the molecular mechanism by which exercise promotes glycogen synthesis in muscle is preserved in insulin-resistant subjects. Future studies will be important to further examine how exercise works to activate GS in muscle from healthy and insulin-resistant individuals.

GRANTS

This study was supported by grants from the American Diabetes Association (N. Musi and R. A. DeFronzo), the National Institutes of Health (AG-030979, DK-80157, and DK-089229 to N. Musi and DK-24092 to R. A. DeFronzo), the San Antonio Nathan Shock Center (N. Musi), the South Texas Health Research Center (N. Musi), the US Department of Veterans Affairs (R. A. DeFronzo), the Novo Nordisk Foundation (J. Jensen), the United Kingdom (UK) Medical Research Council (K. Sakamoto), Diabetes UK (K. Sakamoto), and the Dundee and District of Diabetes UK Volunteer Group (K. Sakamoto).

DISCLOSURES

No conflicts of interest, financial or otherwise, are declared by the authors.

AUTHOR CONTRIBUTIONS

J.J., P.T., J.T.S., M.M.-C., and N.M. performed the experiments; J.J., P.T., J.T.S., M.M.-C., R.A.D., and N.M. analyzed the data; J.J., P.T., J.T.S., M.M.-C., R.A.D., K.S., and N.M. interpreted the results of the experiments; J.J., R.A.D., and N.M. prepared the figures; J.J., K.S., and N.M. drafted the manuscript; J.J., P.T., M.M.-C., R.A.D., K.S., and N.M. edited and revised the manuscript; J.J., P.T., J.T.S., M.M.-C., R.A.D., K.S., and N.M. approved the final version of the manuscript; N.M. did the conception and design of the research.