INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

IN SKELETAL MUSCLE, GLYCOGEN concentration is known to influence the activity of glycogen metabolizing enzymes, which act to maintain glycogen levels within a defined range. Increasing glycogen stores are known to decrease the rate of glycogen synthesis but increase the rate of catabolism by glycogen phosphorylase (GP) (24). Despite these findings, the precise mechanism by which glycogen regulates its own metabolism by a feedback mechanism is incompletely understood. Presently, debate exists on whether initial glycogen concentration plays a role in regulating glycogenolysis in exercising skeletal muscle because an equivocal number of studies have shown glycogen availability to either 1) mediate the rate and magnitude of glycogen catabolism (18, 20, 21, 32, 39) or 2) have no effect on glycogenolysis (4, 9, 30, 35-38). Part of this confusion may be because glycogen granules are not uniformly distributed within the muscle (15) and because glycogen granules are of varied composition. Glycogen is known to exist in a range of molecular sizes distinguished on the basis of solubility in acid. These two forms, termed proglycogen (PG) and macroglycogen (MG), have similar protein compositions but vary in their carbohydrate contents. PG represents the smaller continuum of glycogen granules (<400 kDa), whereas MG represents the upper range (~10⁷ Da). In resting muscle biopsies with normal glycogen concentrations [300-350 mmol glucosyl unit/kg dry wt (dw)], PG represents ~75% of the total glycogen concentration and MG accounts for the remaining 25%.

To date, few studies have examined the physiological role of these two forms of glycogen during exercise. Adamo et al. (2) have examined PG and MG resynthesis after an exhaustive exercise bout and have shown that PG and MG differ in both the timing and magnitude of their resynthesis. In the early phase of recovery, PG was resynthesized to a greater extent than MG, and only when total glycogen concentrations reached levels of ~300-350 mmol glucosyl units/kg dw did significant resynthesis of MG occur. This demonstrated that PG is the precursor to MG and that these two pools of glycogen appeared to differ in the rate and magnitude of their resynthesis. These findings suggest that glycogen synthase (GS) differentially acts on PG and MG depending on the level of glycogen within skeletal muscle and that the two pools were metabolically distinct.

To expand on these findings, the present investigation was undertaken to examine the influence of PG and MG concentration on GP as well as to examine catabolism in these two pools of glycogen. In skeletal muscle, the regulation of GP is complex and controlled by a number of factors, including covalent modification, allosteric modification, and substrate availability. GP exists in two interconvertible forms, a more active a form (GPa), and a less active b form (GPb). In vitro, skeletal muscle GP is present in excess of what is necessary for a high rate of glycogen catabolism with a Michaelis-Menton constant for glycogen of 1-2 mmol/l (27). This implies that GP is always saturated with its substrate glycogen and does not limit glycogenolysis even during high-intensity exercise. However, studies examining glycogen catabolism during both high-intensity and endurance exercise have suggested that preexercise glycogen concentration can influence the rate and extent of glycogen catabolism (7, 16, 20, 21, 23, 32, 39). Factors influencing the relationships between preexercise glycogen concentration, GP, and glycogenolysis are complex and poorly understood.

Given that previous studies examining PG and MG have shown PG to be more metabolically active compared with MG, our hypotheses were that PG would account for the majority of glycogen degraded and that PG rather than MG concentration would determine the rate and extent of glycogen catabolism. Because it is a smaller, more abundant molecule, we also hypothesized that PG would determine the extent of GP transformation and thus flux through the enzyme.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Subject characteristics. Seven male subjects volunteered for the study, age 25.4 ± 1.1 yr, height 182.6 ± 3.3 cm, weight 89.7 ± 2.4 kg, one-leg exercise maximal O₂ uptake (O_2 max) 31.9 ± 1.4 ml · kg¹ · min¹, two-leg exercise O_2 max 49.7 ± 1.5 ml · kg¹ · min¹ (means ± SE). The study received approval from the Human Ethics Committees of the University of Guelph and McMaster University. Subjects were informed of potential risks involved with the procedure, and consent was obtained. Participants were recreationally active, exercising two to three times per week. At least 1 wk after performing an incremental O_2 max test, subjects returned to the laboratory and underwent a muscle glycogen-loading and glycogen-depletion protocol to attain one leg with high glycogen levels (HL) and the other with low glycogen (LL).

Glycogen-loading protocol. To induce glycogen supercompensation in both legs, subjects performed a two-legged exercise protocol on a cycle ergometer (Lode) designed to deplete glycogen stores in both slow- and fast-twitch fibers. This consisted of 1.5 h of cycling at 70% O_2 max followed by a 10-min rest period and then five 3-min of exercise bouts at 90% O_2 max with 1-min rest periods between bouts. After 5 min of rest, subjects then completed five 30-s sprints at maximal capacity or until exhaustion was achieved in each bout. Rest periods of 30 s were taken between bouts. After the exercise protocol, subjects consumed a carbohydrate supplement (Gatorlode) and a high-carbohydrate snack before leaving the laboratory (100 g of carbohydrate). A high-carbohydrate diet was then maintained for 2 days, followed by selective glycogen depletion of one leg.

Selective glycogen depletion. After 2 days of high-carbohydrate ingestion, subjects depleted glycogen in one leg only on a modified cycle ergometer (Monark). The one-leg depletion protocol consisted of 1.5 h at 70% one-leg O_2 max, five 1-min periods at 100% one-leg O_2 max, and five 30-s all-out sprints. Four subjects exercised their dominant leg, whereas three exercised their nondominant leg. After the exercise protocol, subjects maintained a low-carbohydrate diet for 1.5 days and refrained from exercise until the experiment was conducted. Dietary records for both the high- and low-carbohydrate diets showed that intakes of protein, carbohydrates, and fats were 9.2 ± 0.6, 78.7 ± 5.4, 12.1 ± 1.9%, respectively, after the glycogen-loading protocol and 35.8 ± 2.9%, 10.2 ± 5.5%, 54.0 ± 2.7%, respectively, in the glycogen-depletion protocol.

Experimental protocol. Subjects arrived at the laboratory after consuming a low-carbohydrate breakfast at least 3 h before testing. Muscle biopsies were obtained from the vastus lateralis of each leg by using the percutaneous needle biopsy technique. Under local anesthesia, two incisions were made over the vastus lateralis before commencement of exercise. The subjects then exercised at 130% O_2 max on a cycle ergometer for 90 s, and biopsies were obtained from each leg. After a 10-min rest period, this procedure was repeated and biopsies were taken after the second identical exercise bout. Previous studies have demonstrated the rate of glycogen resynthesis in the first 10 min after a bout of high-intensity exercise to be minimal (0.765 mmol glucosyl units · kg dw¹ · min¹) (3); therefore, glycogen resynthesis between exercise bouts was not a consideration in this study.

PG, MG, and total glycogen analyses. Muscle glycogen was analyzed as PG and MG as previously described (1). Briefly, PG and MG fractions were separated by the addition of 1.5 M perchloric acid to a 2- to 3-mg sample of freeze-dried muscle. The PG portion was insoluble in the acid, whereas MG was soluble, allowing their separation. Once separated into the PG and MG fractions, glucosyl units were determined by enzymatic measurement (2, 6). In the determination of MG, all forms of glucose and fructose are measured, which include the hexose monophosphates (HMP) (glucose 1-phosphate, glucose 6-phosphate, fructose 6-phosphate, glucose). Because the HMP are known to be elevated during intense exercise, they were determined as previously described (5) and subtracted from the MG concentration for all postexercise samples. As a result, all postexercise MG values are MG  HMP for that sample.

Glycogen phosphorylase analysis. Muscle GP measurements were performed on the exercise samples as previously described (11, 40). GP was not measured in resting samples because previous studies have shown an artificially high transformation of GPa as a result of elevated calcium release from the sarcoplasmic reticulum during the cutting of the muscle tissue during the biopsy (31). To obtain resting measurements of GP, the biopsies would have to sit at room temperature for 30 s, which would require an additional biopsy. Previous measurements of GPa have shown it to be ~10% during rest (12, 28).

1

1

Statistical analysis. Results are presented as means ± SE. A two-way repeated measures analysis of variance (ANOVA) was performed to compare absolute levels of G_t, PG, MG, phos a, phos a + b, and glycogenolytic rates within the LL and HL. To establish pairwise differences within the two-way repeated-measures ANOVA, a Tukey's post hoc test was used. To isolate differences in the net rates of glycogenolysis between PG and MG within a specific exercise bout, an unpaired t-test was employed. To test differences in total glycogenolysis between HL and LL for both exercise bouts together (1 and 2), a paired t-test was used. Linear regression analysis was performed to compare preexercise total glycogen concentration, PG, MG, and net glycogenolysis. Significance levels of P  0.05 were employed, and data are reported as means ±SE.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Muscle glycogen concentrations and HMP. The glycogen loading and glycogen depletion protocol produced significant differences in resting G_t, PG, and MG between HL and LL (P  0.05) (Fig. 1). G_t was 32% lower in the LL compared with the HL, whereas PG and MG were decreased by 33 and 29%, respectively (P  0.05). The ratio of MG to PG was 30:70 in both the legs before the onset of exercise and declined to 25:75 and 23:77 in the HL and LL, respectively, after the second exercise bout. After the first exercise bout, G_t and PG remained significantly different between the LL and HL (P  0.05) (Fig. 1), and, in absolute terms, 131.8 ± 47 mmol glucosyl units/kg dw of G_t were degraded in the HL, compared with 82.0 ± 26 mmol glucosyl units/kg dw in the LL. PG catabolism accounted for the majority of the degradation in the first exercise bout (67 and 60% of G_t in the HL and LL, respectively). After the second exercise bout, differences in G_t, PG, and MG were no longer found between the HL and LL. The initial preexercise resting PG and MG concentrations in the LL were very similar to those concentrations in the HL after the first exercise bout [223.7 ± 15 and 81.2 ± 20 in the HL (bout 1) and 208.1 ± 31 and 89.5 ± 43 mmol glucosyl units/kg dw in the LL (rest) for PG and MG, respectively]. There was no difference (P  0.05) between either PG or MG at these time points, which allowed comparison of the rates of glycogenolysis (bout 1 for LL vs. bout 2 for HL) when the starting concentrations were similar. HMP were quantified and subtracted from the MG measurement for all samples. Results of resting and exercise HMP are listed in Table 1.

View larger version (18K): [in this window] [in a new window]   Fig. 1.   A: total glycogen values are shown as the sum of proglycogen (PG) and macroglycogen (MG) in the high-glycogen leg (HL). B: total glycogen (Gt) values (PG + MG) in the low-glycogen leg (LL). Ratios above the histograms represent the percentage of MG to PG. Statistics are shown for PG and MG but not for Gt. Significant differences (P  0.05): *HL vs. LL for a given time point; arest vs. bout 1; bbout 1 vs. bout 2; crest vs. bout 2. dw, Dry weight.

Net glycogenolytic rates. Overall net glycogenolysis (G_t) for both exercise bouts in the HL and LL were 145 and 72 mmol glucosyl units · kg dw¹ · min¹, respectively (P  0.05), indicating that twice as much glycogen was degraded in the HL compared with the LL despite both legs performing the same exercise and having identical metabolic demands (Fig. 2). This suggests that initial glycogen concentration is a regulator of glycogenolysis and that more glycogen is degraded when initial concentrations are high. Analysis of net glycogenolytic rates during the first exercise bout indicated no difference between the HL and LL for G_t, PG, and MG (P  0.05) (Fig. 2). During the second exercise bout, the rate of PG glycogenolysis was 4.3 times greater in the HL vs. LL (P  0.05). Significantly more PG than MG was degraded in the HL during this bout compared with in the LL, which degraded both forms of glycogen to the same extent. Analysis of G_t in bout 2 indicated that the HL degraded significantly more G_t compared with the LL (P  0.05).

View larger version (15K): [in this window] [in a new window]   Fig. 2.   Rates of glycogenolysis in PG, MG, and Gt for the HL and LL. Values are means ± SE (in mmol glucosyl units · kg dw1 · min1). aSignificant difference (P  0.05) between HL and LL at a given exercise bout; bsignificant difference (P  0.05) between bout 1 and bout 2 within a leg; csignificant difference (P  0.05) between the rate of glycogenolysis between PG and MG within a given exercise bout.

1

1

View larger version (17K): [in this window] [in a new window]   Fig. 3.   Initial preexercise glycogen concentration in relation to net Gt breakdown in the HL and LL. Each data point represents the data from an individual muscle sample from either the HL or the LL. Solid line, linear regression analysis for all data in plot: y = 0.15 x  8.97, r = 0.93, n = 12. Dotted lines, linear regressions for HL: y = 0.05x  92.20, r = 0.94, n = 6 and LL: y = 0.30x  16.54, r = 0.97, n = 6.

Glycogen phosphorylase. Results of the GP measurements are depicted in Table 2. There was no difference in total V_max between legs or between exercise bouts. V_max a was higher after the first exercise bout in the HL compared with the LL (P  0.05). A significant decrease in V_max a was also apparent between bout 1 and bout 2 in the HL (P  0.05). These differences translated into an elevated GP transformation (phos a %) between the HL and LL for bout 1 and between bout 1 and bout 2 in the HL (P  0.05) (Fig. 4). Interestingly, this increase in phos a % between the HL and LL occurred, whereas there were no significant differences in glycogen degradation between HL and LL after the first exercise bout.

View larger version (11K): [in this window] [in a new window]   Fig. 4.   Phosphorylase transformation during 90 s of cycling at 130% maximal O2 uptake. Transformation was calculated from (Vmax a/total Vmax) × 100, where Vmax is maximal velocity. Values are means ± SE. aSignificant difference (P < 0.05) between HL and LL in a given exercise bout; bsignificant difference between bout 1 and bout 2 for a given leg (P < 0.05).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

In the present study, the different levels of glycogen in the legs before exercise allowed us to investigate the effects of preexercise glycogen concentration on the rate of glycogenolysis in skeletal muscle. Although the relationship between glycogen content and glycogenolysis has been studied extensively in both human and rodent models, no clear consensus exists. The purpose of this study was to examine the catabolism of two pools of glycogen, PG and MG, over a range of glycogen concentrations. To date, minimal information exists on glycogen metabolism in these pools, which appear to be differentially regulated (1). Major findings of the present investigation include the following. 1) Initial glycogen concentration is a regulator of glycogenolysis in the PG and MG pools. 2) PG is preferentially degraded over MG at the onset of high-intensity exercise when glycogen concentrations are in the high to normal range. When glycogen concentrations are lowered and repeated exercise is performed, PG and MG contribute equally to G_t glycogenolysis. 3) GP transformation is elevated in the presence of high glycogen concentrations, but it does not necessarily predict glycogenolysis in either pool.

During exercise bout 1, there was no significant difference between legs for PG and MG breakdown despite an increased GP transformation (phos a %). The ratio of PG to MG catabolism was 2:1 and 1.5:1 in the HL and LL, respectively, indicating that PG is preferentially degraded over MG during the first exercise bout. In the second exercise bout, the PG-to-MG catabolic ratio was 2.2:1 in the HL and 1.1:1 in the LL, showing that, as glycogen concentration decreased in the LL, both forms of glycogen were degraded. A recent study by Derave et al. (13) also showed differences in PG and MG catabolism over a range of glycogen concentrations in electrically stimulated rat skeletal muscle. They showed PG to be the major contributor of glycogenolysis when glycogen levels were in the low to normal range but that MG was more readily broken down when muscles were glycogen supercompensated. This finding differs from the present study, in which PG rather than MG contributed more to glycogenolysis when glycogen levels were elevated.

As expected, the HL degraded more glycogen than did the LL. When both exercise bouts were considered together, the HL degraded 106 mmol glucosyl units/kg dw more total glycogen than the LL, which is surprising considering that both legs performed the same exercise (P  0.05). Between exercise bouts 1 and 2, the rate of glycogenolysis decreased 38% in the HL and 70% in the LL, indicating that glycogenolysis was blunted to a greater extent when starting concentrations were low (P  0.05). Attenuated rates of glycogenolysis between exercise bouts were seen in both PG and MG pools with decreases of 35 and 44% in the HL and 73 and 66% in the LL for PG and MG, respectively. This preferential degradation of PG suggested that these two forms of glycogen may be differentially regulated in skeletal muscle. To explore this further, the rate-limiting enzyme in glycogen catabolism, GP, was examined. Results showed that V_max a and total V_max were not reliable predictors of PG or MG glycogenolysis and that no clear relationship existed between glycogenolysis in either the PG or MG pools and GP for either leg or exercise bout (results not shown). This is not surprising considering numerous factors, including preexercise glycogen is known to regulate the enzyme (19). It should also be recognized that PG is not degraded to a greater extent because of its higher concentration in skeletal muscle. If PG and MG were degraded equally, there would be ~70% and 30% catabolism in each pool, which was not consistently observed in the present study. For example, in the last exercise bout in the LL, the concentration of PG to MG was 77:23, yet the catabolism of PG to MG was 1.1:1.

Although this study shows that catabolism of these two pools of glycogen are distinctly regulated, it cannot describe the exact nature of PG and MG. Presently, debate exists as to whether PG is a separate glycogen entity, as proposed by Lomako and colleagues (25, 26), who have shown that PG is a distinct 400-kDa glycogen species, or a continuum of smaller glycogen particles, as shown by Roach and Skurat (34). The data obtained from this study are not dependent on the nature of PG itself and do not support either side of this controversy but rather examine the metabolism associated with the two types of glycogen. Another controversial issue is the methodology used to separate PG and MG. Although it is intuitively expected that the separation of PG and MG may depend on such factors as acid concentration, Jansson (22) has shown that the separation PG and MG is not influenced by the strength of acid used to separate PG and MG in the range of 0.5-3 M, the type of acid used (perchloric acid vs. trichloroacetic acid), the freeze-drying of muscle, or the weight of the muscle sample (0.2-2 mg) (22). This early work clearly demonstrates that the separation of PG and MG is not simply artifact and that these two pools of glycogen were separable on the basis of acid solubility, and the present study illustrates that the pools are metabolically distinct.

The concept that glycogen is a regulator of its own metabolism is not new but is a point of contention in the literature (4, 14, 16, 20, 30, 32, 36, 38, 39). In the present study, linear regression analysis demonstrated a significant (P  0.05) relationship between initial, preexercise glycogen concentration and net glycogenolysis for both types of glycogen in bout 1 in the HL and for MG but not PG in the LL. In contrast, Derave et al. (13) have shown that the initial concentration of MG, but not PG, was significantly correlated to glycogenolysis in electrically stimulated rat skeletal muscle in all muscle fiber types across a range of glycogen concentrations. This led the authors to speculate that GP may be differentially regulated between the two pools of glycogen. When GP was examined in the present study, results showed that phos a % was increased by 26% in the HL compared with the LL in the first exercise bout (P  0.05). This is in agreement with the studies of Richter and colleagues (20, 32, 39), who found glycogenolysis and GP transformation (phos a %) to be higher in rodent skeletal muscle with elevated glycogen concentrations during intermittent electrical stimulation. One possible explanation for the apparent coupling of GP transformation with initial glycogen concentration may be the location and association-dissociation of GP and related enzymes within skeletal muscle. For example, it is known that glycogen granules are complexed to glycogen-metabolizing enzymes in vivo, including GS, glycogenin, targeting proteins, phosphorylase kinase, phosphatases, and GP (33). This glycogen-protein complex is not static, and proteins will associate and disassociate depending on the metabolic state of the myocyte (29). Therefore, a possible explanation for the finding that V_max a and phos a % were increased in the first exercise bout in the HL may be that GP is predominantly bound to glycogen-protein complex when glycogen levels are elevated, resulting in rapid degradation of attached glycogen on Ca²⁺ release. In comparison, there may be a higher proportion of unbound (not attached to glycogen) GP in the cytosol when glycogen levels are lowered, which results in decreased GP transformation and a greater uncoupling of the Ca²⁺-mediated phosphorylase kinase signal. This would explain why less glycogen catabolism is observed when initial glycogen concentrations are lowered. Thus the proposed mechanism would conserve glycogen stores and promote the use of other substrates to support energy demands within the working muscle. To date, little is known about GP and its physical association-dissociation and its relation to the glycogen granule.

Limitations of this study include prior exercise of the LL but not the HL 1.5 days before the experimental protocol. Although the effects of 1.5 days previous exercise have not been extensively studied, it has the potential to alter glycogen metabolizing enzymes as well as patterns of motor recruitment. Grisdale et al. (17) examined the effects of 1 day previous exercise on subjects' ability to perform maximal voluntary contractions in glycogen-depleted, glycogen-repleted, and control conditions. They showed that 1 day previous exercise resulted in a decreased endurance capacity but that fiber recruitment patterns and lactate accumulation were unchanged in the three conditions. Similarly, Blomstrand and Saltin (8) performed an experiment in which subjects depleted glycogen in one leg 12 h before an experimental protocol in which they were required to perform 60 min of exercise on a cycle ergometer at 69% O_2 max. Results showed no difference in blood flow, oxygen extraction, or blood flow between the legs. Considering the results of these studies and that the rest period in the present study was longer in duration (1.5 days vs. 12-24 h), we believe the effects of the previous exercise on motor recruitment and force production to be negligible. Alternatively, a crossover design with multiple trials could have been used to conduct this study; however, this design has potential problems because differences between days in hormones, blood-borne substrates, and blood flow would have been introduced. Because the activity of GP is affected by such factors, the authors felt that the present study design was most appropriate. Another possible limitation of this study is the 10-min rest period between exercise bouts. Although it is possible that some glycogen resynthesis from lactate and blood glucose may have occurred during this time, previous studies have demonstrated the rate of glycogen resynthesis in the first 10 min after a bout of high-intensity exercise to be minimal (0.765 mmol glucosyl units · kg dw¹ · min¹) (3). This would predict that a total of 7.65 mmol glucosyl units may have been synthesized for the 10-min period that would mainly be in the PG form (1). Because this value is minimal, glycogen resynthesis between the exercise bouts was not a consideration in this study.

New information provided by this investigation includes the finding that PG and MG are degraded at different rates depending on initial glycogen concentration. In the presence of normal to high glycogen concentrations, PG was always preferentially degraded. There are a number of possible explanations for this observation. First, it is possible that PG and MG may vary in their subcellular locations within the muscle. Friden et al. (15) have shown glycogen granules to be distributed in five distinct subcellular locations and that, during exercise, glycogen granules located near Z disks and I band were preferentially degraded during exercise. This suggests that separate depots of glycogen granules are functionally independent and may have different physical interactions with GP. It may also be that PG molecules are localized to different subcellular locations compared with MG, which makes them more susceptible to degradation by GP. Another explanation for the preferential degradation of PG in this study may be an altered enzyme-substrate relation between PG, MG, and GP in these two pools. For example, at its maximal size, MG is very densely branched in its outer tiers, which may limit how efficiently GP and debranching enzyme can remove glucosyl units from this structure. In comparison, the outer branches of the smaller PG molecule may be more readily accessible to the enzyme. Therefore, having abundant, smaller PG granules with more accessible glycogen would be a reason that a high proportion of PG is maintained despite its inferior glucose storage capacity compared with the larger MG molecule. Brammer et al. (10) showed that large glycogen molecules contain regions that are resistant to degradation. However, at the present time, no data exist for GP activity in these two pools of glycogen.

In conclusion, little is known about the functional significance of PG and MG in human skeletal muscle. This study examined PG and MG under metabolically demanding circumstances to enhance our understanding of how these two pools of glycogen are regulated. We found that a higher preexercise concentration of muscle glycogen before intense exercise was associated with a higher rate of glycogenolysis and the main type of glycogen catabolized was PG. V_max a was higher in the presence of elevated glycogen concentrations, and results suggest that GP has an enhanced activity for catabolizing glycogen in the PG pool. In essence, the PG pool may represent a pool of glycogen that is readily available for immediate use, whereas the MG pool may represent a reserve of glycogen that is only used under extreme metabolic demands.