- Top of page
There are many factors that can influence glucose uptake by contracting skeletal muscle during exercise and although one may be intramuscular glycogen content, this relationship is at present not fully elucidated. To test the hypothesis that muscle glycogen concentration influences glucose uptake during exercise, 13 healthy men were studied during two series of experiments. Seven men completed 4 h of two-legged knee extensor exercise 16 h after reducing of muscle glycogen by completing 60 min of single-legged cycling (Series 1). A further six men completed 3 h of two-legged knee extensor exercise on two occasions: one after 60 min of two-legged cycling (16 h prior to the experimental trial) followed by a high carbohydrate diet (HCHO) and the other after the same exercise followed by a low carbohydrate diet (LCHO) (Series 2). Muscle glycogen was decreased by 40 % when comparing the pre-exercised leg (EL) with the control leg (CL) prior to exercise in Series 1. In addition, muscle glycogen was decreased by the same magnitude when comparing LCHO with HCHO in Series 2. In Series 1, glucose uptake was 3-fold higher in the first 60 min of exercise, in the presence of unchanged pre-exercise GLUT4 protein in EL compared with CL, suggesting that the lower glycogen, and not the exercise the day before, might have provided the stimulus for increased glucose uptake. Despite the same magnitude of difference in pre-exercise glycogen concentration when comparing Series 1 with Series 2, neither direct-nor isotopic tracer-determined glucose uptake was higher in LCHO compared with HCHO in Series 2. However, arterial concentrations of insulin and glucose were lower, while free fatty acids and adrenaline were higher in LCHO compared with HCHO. These data suggest that pre-exercise glycogen content may influence glucose uptake during subsequent exercise. However, this is only the case when delivery of substrates and hormones remains constant. When delivery of substrates and hormones is altered, the potential effect of glycogen on glucose uptake is negated.
It is well established that the rate of muscle glycogen utilization during prolonged exercise is influenced, in part, by initial glycogen concentration (Gollnick et al. 1972; Gollnick et al. 1981; Hargreaves et al. 1995, van Hall et al. 1995; Weltan et al. 1998; Blomstrand & Saltin, 1999). This is probably due to the fact that glycogen can bind to glycogen phosphorylase, the enzyme responsible for glycogen breakdown, to increase its activity (Johnson, 1992). It has also been suggested that glucose uptake during exercise is influenced by glycogen content (Richter et al. 2001a), although the effect of pre-exercise glycogen content on glucose uptake is unclear (Gollnick et al. 1981; Hargreaves et al. 1995; Blomstrand & Saltin, 1999, Richter et al. 2001b). The confusion within the literature may be related to the methodology utilized to deplete muscle glycogen prior to exercise. One experimental approach to determine the effect of glycogen content on glucose uptake during contraction is to deplete one leg of glycogen by single-legged exercise prior to performing two-legged exercise. Such an approach resulted in higher glucose uptake in a leg that commenced exercise with lower glycogen content when compared with the contralateral leg, which contained a normal glycogen concentration (Gollnick et al. 1981; Blomstrand & Saltin, 1999). While this experimental approach has the advantage of ensuring that the delivery of hormones and substrates to the contracting muscles is identical, the effect of the prior exercise in one limb may affect glucose transport per se. Indeed, since GLUT4 total protein was not measured in the previous studies that adopted this model (Gollnick et al. 1981; Blomstrand & Saltin, 1999) the authors could not rule out the possibility that the stimulus for the augmented glucose uptake was increased GLUT4 protein and not glycogen content per se.
An alternative experimental approach is to perform two experimental trials following a standardized exercise bout and the ingestion of either a high fat or high carbohydrate diet in the intervening period. While this method has the advantage of normalizing pre-exercise activity levels, it may result in an altered delivery of substrates and/or hormones to the contracting muscles. Such fluctuations may influence glucose uptake independent of glycogen availability. In the study by Hargreaves et al. (1995), while a direct measure of glucose flux across the contracting limb was not made, isotopic tracer-determined glucose uptake was not different during exercise when comparing two trials with differing pre-exercise glycogen content. However, the authors conceded that the maintenance of glucose rate of disappearance (Rd) in the low glycogen trial might have been influenced by the lower plasma glucose and insulin and higher plasma free fatty acids (FFA) during exercise, a situation where glucose uptake would be expected to decrease (Galbo et al. 1979; Hargreaves et al. 1991). In addition, dietary manipulation such as that adopted by Hargreaves et al. (1995) also results in altered adrenaline secretion (Galbo et al. 1979), which has been demonstrated to influence glucose uptake during exercise (Watt et al. 2001).
Clearly, the effect of glycogen content on glucose uptake requires further clarification and we aimed to do so by performing two series of experiments. In Series 1, we had humans perform 4 h of two-legged knee extensor exercise, 16 h after performing one-legged cycling exercise. Along with a measurement of glucose uptake and glycogen content, we also measured GLUT4 gene and protein expression. We hypothesized that glucose uptake would be augmented in the previously exercised leg and that this would be due to a difference in glycogen content per se and not due to any alteration in GLUT4 protein content. In Series 2, we had subjects perform two trials of two-legged exercise for 3 h following a standardized exercise bout performed 16 h prior to the trials and the ingestion of either a high fat or high carbohydrate diet in the intervening period. We attempted to alter pre-exercise glycogen content by a similar magnitude as we altered glycogen content between legs in Series 1. We hypothesized that despite a difference in muscle glycogen, glucose uptake would be similar when comparing trials because of the counter-regulatory effects of altered arterial delivery of substrates and/or hormones.
- Top of page
The results from this study demonstrate that, despite a similar magnitude of difference in pre-exercise glycogen content when comparing Series 1 with Series 2, initial glucose uptake was only different when comparing conditions in Series 1. Therefore, our data suggest that glucose uptake is affected by pre-exercise glycogen content, but altering the delivery of substrates and/or hormones can override the effect of glycogen in the control of glucose uptake by contracting skeletal muscle.
In Series 1, pre-exercise glycogen content was reduced by approximately 40 % in one leg by prior exercise, the day before the experimental trial. As a result, glucose uptake was increased approximately 3-fold in the first 60 min of exercise when comparing the glycogen depleted with the control leg. It was not surprising that the augmented increase in glucose uptake was not observed late in exercise because neither glycogen (Table 1) nor the measurement of other intramuscular metabolites (Table 2) were different when comparing the depleted with the control leg at the end of exercise. It is of note that the difference in glucose uptake was not apparent after 120 min of exercise in Series 1. Since muscle samples were not obtained throughout exercise in Series 1, we are uncertain whether glycogen concentration was different when comparing EL with CL at this time. However, since glycogen binds to glycogen phosphorylase to increase its activity (Johnson, 1992), it is possible that the difference in glycogen between EL and CL was abolished by this time.
It could be argued that because the difference in glucose uptake was transient (Fig. 2), this was mediated not by the difference in pre-exercise glycogen content per se, but by the residual effects of prior exercise in the depleted leg. Indeed, previous studies in both rats (Kou et al. 1999) and humans (Greiwe et al. 2000) have demonstrated that acute exercise increases GLUT4 protein expression in skeletal muscle. The data from the present study contrast these previous studies. Despite a small increase in GLUT4 mRNA as a result of exercise the previous day, the GLUT4 protein was remarkably similar (Fig. 1). The reason for the discrepancy between our results and those of Greiwe et al. (2000) is not readily apparent. In the previous study subjects performed 60 min of two-legged cycling, whereas in the current study subjects performed one-legged exercise of a similar duration and intensity. Although further research is required to clarify this anomaly, we are confident that in our study the prior exercise had no influence on GLUT4 protein expression and that the augmented glucose uptake in the depleted leg cannot be attributed to an increase in crude membrane transporter protein expression. Despite the fact that crude membrane GLUT4 protein expression was not increased as a result of prior exercise, it is possible that GLUT4 translocation from the intracellular compartment to the plasma membrane and/or the intracellular signalling cascade responsible for trafficking GLUT4 was influenced by the prior exercise. In the current study we were unable to make such measurements, but based on previous results this seems unlikely. It has been previously demonstrated that acute exercise increases GLUT4 translocation from the intracellular pool to the plasma membrane in human skeletal muscle (Kristiansen et al. 1996), but the time course for transport from the plasma membrane back to the intracellular stores in recovery from acute exercise has not been measured in human skeletal muscle. Nonetheless, Goodyear et al. (1991) have made such measurements in rodent skeletal muscle and have demonstrated that the translocation of GLUT4 to the plasma membrane as a result of acute exercise is transient, with GLUT4 protein content in the plasma membrane returning to basal levels within 3 h. It is also possible that the activity of intracellular signalling molecules associated with the trafficking of GLUT4 to the plasma membrane could be up-regulated such that, at the onset of subsequent exercise, GLUT4 translocation is increased. While the precise signalling pathway for contraction-mediated glucose uptake has not been fully elucidated, it has been suggested that 5′-AMP-activated protein kinase (AMPK) may have a regulatory role in contraction-mediated glucose uptake (Hayashi et al. 1998). In this regard, Wojtaszewski et al. (2000) have recently demonstrated that the exercise-induced activation of the specific isoform α2-AMPK in human skeletal muscle is transient with levels returning to baseline within 3 h of the cessation of exercise. Taken together, our GLUT4 data and those of Goodyear et al. (1991) and Wojtaszewski et al. (2000) argue against the prior exercise being the regulator of the augmented glucose uptake observed during exercise in the depleted leg in Series 1. Rather, we propose that the lower glycogen per se in the depleted leg gave rise to the activation of intracellular processes ultimately resulting in an increase in GLUT4 protein in the plasma membrane and an increase in glucose uptake. Indeed, while the effect of glycogen content on GLUT4 translocation has not been measured in human muscle, Derave et al. (1999) have demonstrated, in rodent skeletal muscle, that both contraction-stimulated glucose uptake and cell surface GLUT4 protein expression are dependent upon glycogen content.
It is also possible that the difference in glucose uptake when comparing EL with CL in Series 1 was due to the effect of prior exercise in EL on insulin sensitivity. It has been well demonstrated that prior exercise can enhance insulin sensitivity (for review see Richter et al. 2001a), an effect that may persist for up to 48 h (Mikines et al. 1988). In addition, it must be noted that subjects consumed a low carbohydrate diet in the intervening period between the prior exercise and the experiment. Since diet can influence glucose disposal (Richter et al. 2001a), it is possible that glucose uptake was reduced somewhat in both EL and CL due to the low carbohydrate intake. Hence, while our data suggest that pre-exercise muscle glycogen influences glucose uptake during exercise, we cannot categorically rule out the effect of prior exercise on our measures.
Even though glycogen content may have affected glucose uptake during exercise in Series 1, neither direct (Fig. 3) nor tracer-determined (Fig. 4) glucose disposal was higher when comparing LCHO with HCHO in Series 2. In fact, glucose Rd was lower in LCHO compared with CHO late in exercise (Fig. 4). It is important to note that the magnitudes of difference in glycogen content when comparing Series 1 with Series 2 were remarkably similar (Table 1). Since both plasma glucose and glucose Ra was lower throughout exercise in LCHO it is possible that the lower delivery of glucose led to a lower glucose disposal. However, if glucose delivery were the sole reason for abolishing differences in glucose disposal when comparing LCHO with HCHO, then the MCR would have been higher in LCHO, because MCR measures the amount of plasma that is cleared for a given amount of glucose. Rather, MCR was very similar when comparing these two trials, indicating that other factors might have contributed to the regulation of glucose disposal in Series 2. Indeed, arterial FFA and adrenaline were higher, and insulin lower in LCHO compared with HCHO throughout exercise (Fig. 5). Circumstances of altered insulin (Lavoie et al. 1997) and FFA (Hargreaves et al. 1991) concentration can markedly affect glucose uptake during exercise. In addition, an increase in adrenaline concentration has recently been demonstrated to decrease glucose uptake during exercise (Watt et al. 2001). Hence, it appears from our data that changes in circulating hormones and substrates have the capacity to override any possible effect of glycogen content on glucose uptake during exercise. It must be noted, however, that the present data contrast those of a recent study. Richter et al. (2001b) demonstrated that leg glucose uptake was twice as high during bicycling when subjects performed exercise the preceding day followed by a low carbohydrate diet compared with a high carbohydrate diet. It must be noted, however, that in this recent study (Richter et al. 2001b) subjects performed bicycle exercise at 70 % maximal oxygen consumption for a much shorter duration (1 h), while the magnitude of difference in pre-exercise glycogen content (185 vs. 800 mmol (kg dry wt)−1 was much higher than in the present study. These methodological differences are likely to account for the different results when comparing the data of Richter et al. (2001b) with the current data.
In summary, our results suggest that pre-exercise glycogen content may affect glucose uptake during subsequent exercise. However, this is only the case when delivery of substrates and hormones remains constant. When delivery of substrates and hormones is altered, the possible effect of glycogen on glucose uptake is negated.