Regulation of Glycogen Synthase Kinase-3 in Human Skeletal Muscle: Effects of Food Intake and Bicycle Exercise

Subjects. Seven healthy men (25 ± 1 years) gave their informed consent to participate in the study, which was approved by the Copenhagen Ethics Committee. Body weight, height, and BMI were 74 ± 2 kg, 178 ± 2 cm, and 23 ± 1 kg/m², respectively. The subjects participated in regular physical activities twice a week on average and were used to biking on a daily basis for local transportation. One to two weeks before the experiments, maximal pulmonary oxygen consumption was determined during an incremental bicycle ergometer test (55 ± 1 ml· kg^-1 · min^-1).

Experimental protocol. The subjects randomly underwent two experimental trials separated by 2-3 weeks. Subjects were instructed to eat a controlled diet (carbohydrate ∼63%, fat ∼22%, and protein ∼15%,average 13.5 MJ/day) for 3 days before each experiment and arrived at the laboratory in the morning after an overnight fast. After 45 min of rest, a needle biopsy from the vastus lateralis was obtained under local anesthesia. A venous catheter was inserted in an antecubital vein for blood sampling. The subjects then performed bicycle exercise for 90 min at 50% Vo_2max(low-intensity trial) or for 55 min at 75% Vo_2max plus 5 min at 90%Vo_2max (high-intensity trial). The total energy expenditure was aimed at being the same and did not differ in the two trials (average 3,800 kJ). Immediately after exercise, another biopsy was obtained from the vastus lateralis. The subjects then rested in the supine or sitting position for 3 h,having free access to water only. Then, a third biopsy was obtained, and the subjects then ingested a carbohydrate-rich meal with an energy content similar to the amount used during the exercise bout plus the expected energy consumption in the 3-h postexercise period. Thus, this meal was identical in the two trials, and consisted of ∼75% carbohydrate, ∼17% fat, and∼8% protein, which amounted to 4,910 kJ in total. The meal was ingested within 15 min. The subjects then rested in the supine or sitting position for 2 h before drinking a carbohydrate beverage, containing 500 kJ. Another hour elapsed before a final biopsy was obtained from the vastus lateralis. The biopsies before and after exercise were always taken in the same leg, and the biopsies taken 3 and 6 h after exercise were always taken in the contra-lateral leg. Incisions for biopsies were spaced 4-5 cm apart. Biopsies were frozen in liquid nitrogen within 20 s. Before and during the last minute of exercise as well as every following hour, blood samples were obtained for measurements of plasma insulin concentrations.

Analytical procedures. Plasma insulin concentrations were measured as previously described (14). For determination of muscle glycogen content, muscle biopsies were freeze-dried and dissected free of blood, fat, and connective tissue before analysis. Glycogen content was determined as glycosyl units after acid hydrolysis (15). Muscle GS activity was measured in a homogenate by a modification of the method of Thomas (16) as described by Richter et al. (17). GS activity was determined in the presence of 0.02, 0.17, and 8 mmol/l glucose-6-phosphate (G6P) and given either as the percent of G6P-independent GS activity (%I-form) (100 × activity in the presence of 0.02 mmol/l G6P divided by the activity at 8 mmol/l G6P [saturated]) or as the percent of fractional velocity (%FV) (100 × activity in the presence of 0.17 mmol/l G6P divided by the activity at 8 mmol/l G6P).

As an indicator of Akt activation, Akt serine⁴⁷³ phosphorylation was measured by Western blotting of muscle lysate using a phosphospecific antibody (New England Biolabs, Beverly, MA) and fluorescent measurements of alkaline phosphatase as a detection system (Promega U.K., Southampton, U.K.),as previously described (14). GSK3α and GSK3β were immunoprecipitated from 100 μg of muscle lysate using an anti-GSK3α (Upstate Biotechnology, Lake Placid, MA) or anti-GSK3β (Transduction Laboratories, San Diego, CA) antibody. A p81-filter paper assay, with a phospho-GS2-peptide (Upstate Biotechnology) as substrate, was used to measure GSK3 activity, as described previously(8,9).

Calculations and statistics. Control samples were added to all immunoblots and activity assays, and assay-to-assay variation was accounted for by expressing the data relative to these samples. Data are expressed as means ± SE. Statistical evaluation was done by one- or two-way analysis of variance with repeated measures, as appropriate. When analysis of variance revealed significant differences, a post hoc test was used to correct for multiple comparisons (Student-Newman-Keuls test). Differences between groups were considered statistically significant if P was <0.05.

Exercise and food intake effects on glycogen metabolism.Figure 1 shows GS activity in the low- and high-intensity exercise trials. Both %I-form and %FV of GS were increased by exercise, and the increase tended to be higher in the high-intensity trial. GS activity was not changed in the first 3 h after exercise. Food intake raised GS activity further only in the high-intensity trial, despite the fact that plasma insulin concentrations were elevated to a similar extent in the two trials (Table 1). The rate of glycogen utilization was higher during high-compared with low-intensity exercise (4.5 ± 0.7 vs. 1.6 ± 0.3 mmol/kg dry wt/min, P < 0.05). In the high-intensity trial,glycogen content decreased further during the first 3 h after exercise, and only in this trial was a significant (P < 0.05) glycogen resynthesis (∼25%) observed after food intake(Fig. 2).

Bicycle exercise in humans does not lead to deactivation of GSK3.Glycogen-depleting exercise, performed at either a high or low intensity, did not decrease the activity of GSK3α or GSK3β(Fig. 3). In fact, exercise increased GSK3α activity in both trials in all subjects except one,giving rise to a modest but significant (P < 0.05) increase in both trials (∼10-20%) (Fig. 3A). In accordance with this increase, Akt serine⁴⁷³ phosphorylation decreased during exercise in the majority of experiments (10 of 14) (Fig. 4), perhaps as a result of the significant (P < 0.05)insulin-lowering effect of exercise (Table 1). However, overall the decrease in Akt phosphorylation did not reach statistical significance.

Increased plasma insulin levels after food intake lead to phosphorylation of Akt and deactivation of GSK3α. Food intake was associated with an increase in plasma insulin concentration, peaking 60 min after food intake in both trials at an average of 85 μU/ml, followed by a slight decrease 2 and 3 h after ingestion (∼70 and ∼60 μU/ml,respectively) (Table 1). Plasma insulin concentrations were similar during the two trials at all time points. Serine phosphorylation on site 473 of Akt increased by ∼2-fold in response to food intake (Fig. 4). The degree of phosphorylation was independent of the intensity of the prior exercise bout. Likewise, GSK3α activity decreased significantly by∼40% in response to food intake during both trials, whereas GSK3βactivity was unchanged (Fig. 3).

The current results demonstrate that low- and high-intensity physical exercise in humans is not associated with a deactivation of GSK3 in skeletal muscle. In fact, the drop in plasma insulin concentration during exercise may cause the minor, but significant, activation of GSK3α. Because GS is highly activated during exercise, our data suggest that this activation in response to exercise is not regulated by a GSK3-dependent mechanism in humans. These findings are in line with our recent human data showing an elevated muscle GS activity 4 h after exercise without any concurrent deactivation of GSK3 (9). The lack of deactivation of GSK3 with exercise in the present study contrasts with observations in muscle from treadmill-exercised rats, in which an exercise-induced decrease in both GSK3α activity and GSK3β activity was demonstrated (8).

For the first time, we show that a physiological stimulus such as intake of food causes deactivation of muscle GSK3α in humans. Because deactivation of GSK3α was recently demonstrated during hyperinsulinemic clamp conditions(9,10),the GSK3α-deactivation observed in the present study is probably caused by the meal-induced hyperinsulinemia, giving rise to increased Akt phosphorylation (Fig. 4). However, deactivation of GSK3β is not observed(Fig. 3B). During euglycemic clamp conditions in healthy subjects and patients with type 2 diabetes, deactivation of GSK3β was also not observed([10] and unpublished observation). In contrast, a marked deactivation of GSK3β was observed in rat skeletal muscle after injections of very large doses of insulin (10 U/kg body wt) (8). The differences between the findings in humans and in rats may relate to the species differences or possibly to the level of hyperinsulinemia achieved, which was clearly supraphysiological in the rats.

Another possibility is that insulin injections in rats result in hypoglycemia giving rise to epinephrine secretion(18). β-adrenergic stimulation of primary adipocytes decreased GSK3 activity immunoprecipitable with an antibody recognizing both α and β isoform of GSK3(19). If β-adrenergic stimulation by epinephrine in response to hypoglycemia in the insulin-injected rats also deactivates GSK3 in muscle, this could be the indirect mechanism by which supraphysiological insulin concentrations cause deactivation of GSK3β. However, if so, we would expect similar effects of epinephrine during exercise in humans because plasma epinephrine levels, especially during the high-intensity exercise trial, are known to be elevated four- to sixfold compared with rest (20). Thus,more studies are needed to clarify the role of epinephrine in the regulation of GSK3 activity in skeletal muscle.

Hyperinsulinemia as a result of food intake activates GS to a greater extent after high-intensity exercise compared with low-intensity exercise, and is accompanied by a significant glycogen resynthesis in the period after food intake. Interestingly, this enhanced insulin action is not mediated through changes in insulin signaling, as both Akt and GSK3 were affected similarly in the two trials. Recently, we reported that enhanced insulin action in exercised versus rested human skeletal muscle was also not a consequence of enhanced intracellular insulin signaling(9,14). In fact, in those studies, a very strong correlation existed between the degree of glycogen breakdown during the exercise bout and the subsequent rate of muscle glucose uptake after insulin stimulation (r² =0.52, n = 14, P < 0.01 [unpublished observations]). In the present study, a significant glycogen resynthesis occurred in response to food intake in the high-intensity trial, in which a lower glycogen content was also evident after exercise. In fact, good negative correlations exist between the glycogen level before food intake and the following: 1) the increase in glycogen content after food intake (r² = 0.43, n = 13, P < 0.01) and 2) the increase in GS activity after food intake (e.g., for %I-form r² = 0.31, n = 13, P < 0.05)(Fig. 5). These observations are in line with other recent observations indicating that muscle glycogen content is important for the ability of both insulin and contractions to stimulate muscle glucose uptake in rats(21,22,23).

In contrast to the present study in humans, exercise in rats leads to deactivation of GSK3α and GSK3β in muscle(8). The explanation of this difference is not easily found, but it cannot be excluded that even more severe exercise than in the present study, leading to more severe glycogen depletion, could cause deactivation of GSK3 also in human muscle. Nevertheless, our findings that two rather strenuous exercise bouts consuming close to 4 MJ, leading to marked glycogen depletion and GS activation, do not lead to GSK3 deactivation suggests that GSK3 deactivation during exercise is not physiologically important for activation of GS in human skeletal muscle. A putative alternative mechanism for dephosphorylation of GS in response to exercise could be an activation of protein phosphatases (PPs), especially PP1(24). PP1 is targeted to glycogen through binding to regulatory binding proteins. For example, protein targeting to glycogen (PTG) has been shown to form complexes to the catalytic subunit of PP1 and to glycogen. In addition, there may be an association with other enzymes involved in glycogen metabolism(25). Thus, regulation of the binding properties of these binding proteins may be important for the regulation of GS activity in response to exercise. In fact, a very recent but preliminary report indicates that in the absence of the regulatory glycogen binding subunit G_M (or R_GL) of PP1, a marked reduction in GS activation in response to contractile activity is observed in mouse skeletal muscle (26). To our knowledge, only one study has measured muscular PP1 activity in response to exercise. Thus, in humans PP1 activity actually decreases during maximal isometric muscle contractions and returns to pre-exercise levels early in the recovery period (27). Whereas these findings do not offer any explanation for the postexercise-increased GS activity after isometric contractions, they do not exclude a regulatory role of PP1 during more dynamic glycogen-depleting exercise as in the present study.

In conclusion, food intake is associated with muscular deactivation of GSK3α, supporting a physiological role for GSK3α in the regulation of GS activity in human skeletal muscle. However, the regulation of GS in response to food intake (elevated plasma insulin) is not solely dependent on GSK3α deactivation because muscle glycogen content clearly has a signaling-independent influence on the effects of food intake on insulin action. Finally, in contrast to rodents, moderate and intense exercise in humans does not lead to GSK3 deactivation in skeletal muscle, excluding the kinase as an important regulator of GS activity during physical activity in humans.

This study was supported by grant no. 504-14 from the Danish National Research Foundation. J.F.P.W. was supported by a postdoctoral fellowship from the Danish Medical Research Council.

We are grateful to Betina Bolmgren, Winnie Taagerup, and Irene B. Nielsen for superior technical assistance.