Actu Physiol Scund 1992, 145, 345-352
Effect of low glycogen on glycogen synthase in human muscle during and after exercise Z. Y A N , M. K. SPENCER and A. K A T Z Department of Kinesiology, University of Illinois at Urbana-Champaign, Urbana, Illinois 61801, USA
YAN,Z., SPENCER, M. K. & KATZ,A. 1992. Effect of low glycogen on glycogen synthase in human muscle during and after exercise. Actu Physiol Scund 145, 345-352. Received 30 October 1992, accepted 25 February 1992. ISSN 0001-6722. Department of Kinesiology, University of Illinois, Urbana, Illinois, USA. Subjects cycled at a work load calculated to elicit 75 yo of maximal oxygen uptake on two occasions: the first to fatigue (34.5k5.3 min; meankSE), and the second at the same workload and for the same duration as the first. Biopsies were obtained from the quadriceps femoris muscle before and immediately after exercise, and 5 min postexercise. Before the first experiment, muscle glycogen was lowered by a combination of exercise and diet, and before the second, experiment muscle glycogen was elevated. In the low glycogen condition (LG), muscle glycogen decreased from 169 f15 mmol glucosyl units kg-l dry wt at to rest to 1 3 k 6 after exercise. In the high glycogen condition (HG) glycogen decreased from 706 f52 at rest to 405 68 after exercise. Glycogen synthase fractional activity (GSF) was always higher during the L G treatment. During exercise in the HG condition, those subjects who cycled for < 35 min (n = 3) had GSF values in muscle which were lower than at rest, whereas those subjects who cycled for > 35 min ( n = 4) had values which were similar to or higher than at rest. Thus the change in GSF in muscle during HG was positively related to the exercise duration ( r = 0.94; y = 254-17x+0.3x2; P < 0.001) and negatively related to the glycogen content at the end of exercise ( r = - 0.82; y = 516-2x + 0 . 0 0 1 ~ ~ ; P < 0.05). During LG exercise GSF remained constant. GSF increased markedly after 5 min post-exercise in both HG and L G conditions. CAMP dependent protein kinase activity increased similarly during both L G and HG exercise and reverted to the preexercise values 5 min post-exercise. It is concluded that muscle contraction decreases GSF, but low glycogen levels can attenuate or abolish the decrease in GSF. The rapid increase of GSF during recovery from exercise does not require glycogen depletion during the exercise. Key words: cyclic AMP dependent protein kinase, glycogen synthase phosphatase, fatigue, recovery from exercise
Glycogen synthase (GS), a regularity enzyme for glycogenesis, is controlled by covalent phosphorylation/dephosphorylation (Friedman & Larner 1963). Phosphorylation (which results in a decrease in G S F ) is catalysed by specific kinases and dephosphorylation (which results in an increase in GSF) is catalysed by specific Correspondence : Abram Katz, Department of Clinical Physiology, Karolinska Institute, Karolinska Hospital, Box 60500, S-104 01 Stockholm, Sweden.
phosphatases (Cohen 1986, Larner & VillarPalasi 1971). Short-term intense exercise results in a decrease in GSF in human skeletal muscle (Chasiotis et al. 1983, Kida et al. 1989), and this is due to an activation of GS kinase(s) (Yan et al. 1991). However, prolonged submaximal exercise, which results in a marked degradation of muscle glycogen, results in an increase in GSF (Maehlum et al. 1977). I t is likely that the increase in GSF in the latter condition is due, at least in part, to an increased suitability of
345
346
Z. Ynn et al.
lateral aspect of the quadriceps femoris muscle of both thighs ( - 3500 of the distance from the superior margin of the patella to the anterior iliac spine). After a blood sample and muscle biopsy (Bergstrom 1962) were obtained, the subjects cycled (70 rev min-.') for 2 min at a power output calculated to correspond to 5O0, i'02n,ax (125 F 8 W); thereafter, the workload was immediately increased to one calculated to correspond to 75°0V0,n,al (212f 15 W). Thesubjects cycled at this workload to fatigue (34.5 t- 5.3 min). A second biopsv was taken at fatigue (contralateral leg), and a third was taken at 5 min after termination of exercise (proximal to the post-exercise biopsy). Additional blood samples were taken at the end of exercise and 5 min post-exercise. This experiment constituted the low glycogen condition (LG). For the following 3 days subjects performed no exercise and consumed a diet rich in carbohydrate to increase the muscle glycogen content to high levels (HG condition) (Spencer & Katz 1991). The experimental protocol was then repeated with individual exercise durations and power outputs being identical to the LG condition. One subject who participated in the L G experiment did not participate in the HG experiment. T h e time between termination of exercise and freezing of the post-exercise biopsies averaged 2 0 F 2 and 17 1 s during the LG and HG trials (P > 0.05), and the time between termination of exercise and the freezing of the biopsy obtained during recovery averaged 5.30&0.02 and 5.22F0.18 min during the L G and IIG trials (P> 0.05), respectively. .4nalytical. All biopsies were quick-frozen in the needle by immersion into liquid freon, maintained at its melting point ( - 150 "C) with liquid N,. T h e biopsies were stored in liquid N, until preparation for analysis at which time they were lyophilized, dissected free of solid non-muscle constituents (connective tissue and blood), powdered and thoroughly mixed. M A T E R I 41,s AND 34ETHODS One aliquot of powder was digested in 1 M K O H (at Subjri-ts. Eight healthy men whose mean (range) age, 60 "C); the digest was neutralized with HCI, and height, weight and maximal oxygen uptake (VO,max) gll-cogen was hydrolyzed enzymatically (Harris el a / . were, respectively, 21 !-ears (20-22), 178 cm 1974). T h e released glucose was analysed enzymaticallv (changes in NADPH) with a fluorometric (172-18.54 73.0 kg (63.9-82.7) and 4.21 1 min-' (3.53--.5.07), participated in the study. The subjects technique (Lowry 8i Passonneau 19i2). .r\ second aliquot was homogenized (60 pl mg dry aere informed of the possible risks involJ-ed before giving voluntary consent. The experimental protocol wt) in a buffer containing 10 mM Tris, 10 mM EDTA, 100 mM K F and lo,, charcoal, p H 7.6 and then was approved by the Unit-ersity of Illinois Institutional centrifuged at 12800 g for 10 min at 4 "C (Walkenbach Review Board. Experimental Design. Vo,,,, was determined on an et al. 1980). Part of the supernatant was used for electrically-braked cycle ergometer (Bosch ERG-551, analysis of cyclic AMP-dependent protein kinase Germany) in the upright position as previously (cAMPPK) using Kemptide as a substrate in the described (Spencer & Katz 1991). The day before the presence or absence of 100 ,UM CAMP (see Kida et a / . first experiment the muscle glycogen content was 1991). T h e fractional activity is 0/100 ,//M CAMP. reduced by a combination of exercise and diet (Spencer Another part of the supernatant was diluted and used 8i Katz 1091). Subjects reported to the laboratory the for analysis of glycogen synthase (GS) in the presence next da! after an overnight fast and, after assuming of 0.17 (GS,,J and 7.2 mM glucose 6-P (GSk1,J (see the supine position, a plastic catheter was placed in an Kida et al. 1989). T h e fractional activity is antecubital vein and incisions were made over the GSIOR/GShlKh (GSF).
substrate (GS) for GS phosphatase ( G S P ) ; i.e., the low glycogen results i n a n increased availability of phosphorylated sites o n GS (Cohen 1986), and that this overrides the contractionmediated decrease in G S F . I t is not known, however, whether prior depletion of the muscle glycogen stores alters the changes in GSF during short-term exercise. I t might b e expected that the decrease in GSF during short-term exercise with a normal glycogen content would b e attenuate or abolished in the gl\-cogen depleted state, ow-ing to a high substrate suitability of GS for GSP. T h e r e is only a limited information on the role of glicogen in the rapid increase of GSF during recovery from exercise. Danforth (1965) showed that following a 20 s tetanus, the increase in GSF in mouse skeletal muscle deficient in phosphorylase b kinase a n d containing excessive glycogen both before a n d after the tetanus, is markedly attenuated compared with the increase in GSF in control mouse muscle. On the other hand. preliminary data i n humans indicate an increase in GSF in muscle during recovery after 30 s of moderate exercise performed with a low glycogen content (Hultman et al. 1971). T h e increase in GSF in the low glycogen condition was comparable to the increase observed in a normal glycogen condition (Hultman et al. 1971). T h e purpose of this study was to determine the effect of a low muscle glycogen content on GS activity in human muscle during and after exercise.
Glycogen and glycogen synthase in muscle When available, a third aliquot was used for analysis of GSP (see Kida et al. 1989). Commercial rabbit GS-D (i.e., dependent on glucose 6-P) (G2259, Sigma Chemical Co, St Louis, MO, USA) was purified as described elsewhere (Kida et al. 1989), and was 95 glucose 6-P-dependent when assayed in the presence of 7.0 mM UDPG with and without 7.2 mM glucose-6-P'. Muscle powder was homogenized with a ground-glass homogenizer in a buffer (133 yl mg-' dry wt) containing 50 mM Tris, 10 mM EDTA and 50 mM 2-mercaptoethanol, pH 7.8, at 4 "C. The homogenate was centrifuged at 10000 g for 20 rnin at 4 "C. Twenty-five milliunits of purified GS-D (8 pl) were pre-incubated with 0.3% glycogen, 10 mM Tris, 1 mM EDTA and 5 mM dithiothreitol, pH 7.8 (total volume = 75 PI) for 5 rnin at 30 "C. The GSP reaction was then started by transferring 100yl of extract supernatant into the pre-incubated reaction mixture. The reaction was stopped right after the start of reaction and after 15 rnin incubation by diluting 50 yl of the incubation mixture with 2000 yl of 130 mM KF, 50 mM Tris and 20 mM EDTA, pH 7.8, at 4 "C. Twenty-five microlitres were then used to determine GSP activity using the GS assay (i.e., the change in GS,,_). Values are expressed as p n o l of glucose from UDP-( "C]G incorporated into glycogen kg-' dry wt min. All enzymes were assayed at 30 "C and were linear with time and extract volume (data not shown). Part of the GS supernatant, after extraction with PCA and neutralization with KHCO,, was also used for analysis of phosphocreatine (PCr) and creatine (Cr) (Lowry & Passoneau 1972). The sum of PCr and Cr (TCr) averaged 116.0$4.3, 105.9k4.2 and 111.5 k 6 . 7 mmol kg-' dry wt at rest, end of exercise and 5 min post-exercise in the LG condition ( P > 0.05), and 107.7k4.2, 90.2k4.3 ( P < 0.01 vs. 107.7& 4.2) and 108.5f4.7 at rest, end of exercise and 5 min post-exercise in the H G condition, respectively. The reason for the decrease in TCr after HG exercise is not clear, but to account for this decrease all enzyme activities and glycogen contents were divided by the individual TCr values and then multiplied by the mean TCr content for the whole material (106.4 mmol kg-' dry wt). The latter manipulation allows for expression of data in a physiologically meaningful manner without altering the relative differences in values within a given subject. Blood was drawn anaerobically and immediately injected into ice-cold tubes containing EDTA. The samples were left on ice for 10 min, centrifuged at 4 "C, and the plasma was aspirated, stored at - 80 "C, and subsequently used for analysis of free fatty acids, glucose, lactate and insulin as previously described (Spencer et al. 1992). Statistics. Significant differences between means were determined with a one way repeated measures analysis of variance (ANOVA) or a paired t-test where appropriate. When the ANOVA yielded a significant
347
F ratio, the location of significance was identified with the Newman-Keuls test. Significance was set at P < 0.05. Values are reported as means+SE unless otherwise indicated.
RESULTS Cardiorespiratory Cardiorespiratory results and estimates of whole body substrate utilization rates at rest and during exercise are presented elsewhere (Spencer et al. 1992). Briefly, during exercise, heart rate and whole body Voz were significantly lower, while the respiratory exchange ratio (Vco2/VoI) and the rate of whole body carbohydrate oxidation were significantly higher during HG vs. LG, respectively. T h e higher Vo, during LG exercise is probably due to a higher fat oxidation (per litre of 0, consumed the caloric yield of fat is less than that of carbohydrate).
Muscle
As a result of the experimental manipulations the muscle glycogen content at rest was 4-fold higher during HG vs. LG (Table 1). Glycogen was essentially completely depleted after LG exercise and remained depleted during recovery. Although approximately twice as much glycogen was degraded during HG exercise, the values at the end of exercise and during recovery were high and similar to those in untrained human muscle under basal conditions (Harris et al. 1974). T h e activity of GShighwas similar between treatments at rest. Although the activity of GShighdid not change significantly during HG exercise, there was a significant decrease during LG exercise, and this decrease persisted during the 5 rnin recovery period. Interestingly, in two subjects there was no decrease in GShighduring LG exercise. These two subjects had the highest muscle glycogen contents at the end of LG exercise (27 and 49 mmol glucosyl units kg-' dry wt). Within 3 days of ingesting a high carbohydrate diet the activity had increased back to the initial value ( H G rest). As expected on the basis of earlier work (Danforth 1965), GSF was markedly higher at rest in the LG condition. There were no significant changes in GSF during exercise under both conditions, but there were significant increases during the 5 min recovery under both
348
%. l i i n et al.
Table 1. Effect\ of exercise on gl!cogen 2nd GS acti\it! in muscle
I IG I IG \'dues are means +SE: for .iX (glycogen) or i-8 (GS and GSl,,~,/GS,~,K,,) individuals. Glycogen is giwn in mrnol gl!cos!l tinits kg-'dr! u t , and GS,,,,,,in mmol Ig-'dr? t+t min-' at 30 "C. GS, glycogen synthase; GS,,,,,, GS measured at IOU glucose 6-P concentration; GS,,,pl,,GS measured at high glucose 6-1' concentration; TAG, low gl!-cogen; IfG high gl!cogen. * P < 0.0.5; **P < 0.01 : *** P < 0.001 vs. respective rest value. tP < (1.05; .tt P < 0.01 ; j-tt P < 0.001 VS. LG.
(a) Low glycogen
400
0
m
200
-3
......... ,Q.
9
$ L
8
........a... ..........0.
8
c
.............. ...................................
9 400
:b) High glycogen
v
?...a ....................
0
L (I)
200
0
o i
0
....
0
0
30 Exercise duration (rnin)
0
60
Fig. 1. Relationship between the change in muscle GSF during esercise and esercise duration. (a) During low glyxgen condition (0). (h) During high glycogen condition 1.). r = 0.94; P < 0.001: 1' = 2.54.2117.53u t0.331'. - - represents d u e at rest (i.e., no change in GSF). ~~
conditions. Examination of the indii idual data revealed a good relationship between the change in GSF during HG exercise and the exercise duration (Fig. 1). I t is noteworthy that those subjects who cycled for the shortest durations ( < 3.5 min, n = 3) had marked decreases in GSF (the post-exercise values ranged from 16-73 of the resting value), while those who cycled for > 3.5 min (n = 4) showed no decrease or an increase in GSF. O n the other hand, G S F remained fairly constant during LG exercise.
0
400 800 Glycogen (rnrnol glucosyl units kg-' dry wt)
Fig. 2. Relationship between the change in muscle GSF during exercise and glycogen at the end of esercise in high glycogen condition. r = - 0 . 8 2 ; P < 0.05; y = ~ 1 ~ . 9 ( ~ ~ - ~ . 6 0 4--~ ~represents 0.0~1~~. ~ a l u ca t rest (i.c., no change in GSF). There was also an inverse relationship between the change in GSF during HG exercise and the glycogen content in the muscle at the end of exercise (Fig. 2 ) . Exercise resulted in an activation of c A M P P K (0/100 ,//xi CAMP) without altering the activity ~ ) both conat saturating CAMP ( 1 0 0 , ~ under ditions (Table 2 ) . During recovery c A M P P K fractional aetivitj- reverted toward the preexercise values. There was not sufficient material for meaningful analyses of GSP at rest (only one paired observation which is not reported). However, GSP was higher during HG exercise and recovery.
Glycogen and glycogen synthase in muscle
349
Table 2. Effects of exercise on cAMPPK and GSP activity in muscle Rest cAMPPK (0/100 p ~ ) LG 0.15 f0.01 HG 0.13 f0.02 cAMPPK (100 p ~ ) LG 135f8 IIG 152f 13 GSPI I,G I IG ~
Values are means k SE for 6-8
(a =
Exercise
5-min post-exercise
0.25 f 0.03" 0.22 f 0.03"*"
0.18 f0.03 0.16 f0.02
128 f 12 152f11
139f 17 146k 18
430f41 706 k 5 4 t t t
476 f36 616k70t
5 for GSP, 5 min post-exercise, LG) individuals and are given in pmol
kg-' dry wt min-' 30 "C for cAMPPK and in pmol kg-' dry wt (min-') at 30 "C for GSP. cAMPPK was measured in the absence and presence of 100 ,UM CAMP. cAMPPK, cyclic AMP dependent protein kinase;
GSP, glycogen synthase phosphatase; LG, low glycogen; HG, high glycogen. 1Insufficient material was available for analysis of GSP in the resting state under both conditions. * P < 0.05; ""I P < 0.001 vs. respective rest value. t P = 0.05; t t t P < 0.001 vs. LG. Table 3. Effects of exercise on plasma metabolites and insulin
Glucose LG HG Insulin LG HG Lactate LG HG FFA LG HG
Rest
Exercise
5-min post-exercise
4.59 rt 0.16 5.55 + O . l l t t t
3.64 f 0.28" 5.22fO.lOj-t
4.13 f0.32 5.80 f 0.22tt
6f 1 12 5 3 t
4 51 16f5t
3*1* 6f1t
0.78 & 0.05 0.96 f0.071.
5.39 f0.66*** 8.05 & O.jO***+tt
4.93 f0.59*** 7.565 O.SS***ttt
0.87 fO.ll 0.35 k 0.05t-t
0.77 f0.08 0.25 0.04Qyf-t
1.51 f0.07" 0.57 fO.l2*"*ttt
*
Values are means k SE for 7-8 individuals and are given in mmol 1.' or pU ml-' (insulin). FFA, free fatty acids. " P < 0.05 ; "*" P < 0.001 vs. respective rest value. t P < 0.01 ; tt P < 0.05 ; ttt P < 0.001 vs. LG.
Plasma At rest, plasma glucose, insulin and lactate were higher during HG, while FFA was higher during LG (Table 3 ) . Glucose decreased during LG exercise but not during HG exercise, and increased during recovery in HG ( P < 0.01 vs. end of exercise value). Insulin decreased during exercise under both conditions. Insulin increased in six of seven subjects during HG recovery, but remained depressed during the LG recovery. Lactate increased to a greater extent during HG exercise. FFA were not markedly altered during LG exercise but decreased during HG exercise, and then increased during recovery under both
-
conditions. T h e increase in FFA during recovery twice that observed in the LG condition was in the HG condition. This may in part relate to the higher plasma insulin concentration in the HG condition at this time, which should blunt lipolysis.
DISCUSSION
GSF I t was first shown by Friedman & Larner (1963) that GS is regulated by covalent phosphorylation/dephosphorylation. T h e GS
3.30
%.
I'UtI
er
ill.
assay used in the present study has been shown to bc. sensitive to the phosphoryhtion state IJf the
cnzynie (Ciuinovart tl id. 1979). Thus an increase in
Effect of low glycogen on glycogen synthase in human muscle during and after exercise Z. Y A N , M. K. SPENCER and A. K A T Z Department of Kinesiology, University of Illinois at Urbana-Champaign, Urbana, Illinois 61801, USA
YAN,Z., SPENCER, M. K. & KATZ,A. 1992. Effect of low glycogen on glycogen synthase in human muscle during and after exercise. Actu Physiol Scund 145, 345-352. Received 30 October 1992, accepted 25 February 1992. ISSN 0001-6722. Department of Kinesiology, University of Illinois, Urbana, Illinois, USA. Subjects cycled at a work load calculated to elicit 75 yo of maximal oxygen uptake on two occasions: the first to fatigue (34.5k5.3 min; meankSE), and the second at the same workload and for the same duration as the first. Biopsies were obtained from the quadriceps femoris muscle before and immediately after exercise, and 5 min postexercise. Before the first experiment, muscle glycogen was lowered by a combination of exercise and diet, and before the second, experiment muscle glycogen was elevated. In the low glycogen condition (LG), muscle glycogen decreased from 169 f15 mmol glucosyl units kg-l dry wt at to rest to 1 3 k 6 after exercise. In the high glycogen condition (HG) glycogen decreased from 706 f52 at rest to 405 68 after exercise. Glycogen synthase fractional activity (GSF) was always higher during the L G treatment. During exercise in the HG condition, those subjects who cycled for < 35 min (n = 3) had GSF values in muscle which were lower than at rest, whereas those subjects who cycled for > 35 min ( n = 4) had values which were similar to or higher than at rest. Thus the change in GSF in muscle during HG was positively related to the exercise duration ( r = 0.94; y = 254-17x+0.3x2; P < 0.001) and negatively related to the glycogen content at the end of exercise ( r = - 0.82; y = 516-2x + 0 . 0 0 1 ~ ~ ; P < 0.05). During LG exercise GSF remained constant. GSF increased markedly after 5 min post-exercise in both HG and L G conditions. CAMP dependent protein kinase activity increased similarly during both L G and HG exercise and reverted to the preexercise values 5 min post-exercise. It is concluded that muscle contraction decreases GSF, but low glycogen levels can attenuate or abolish the decrease in GSF. The rapid increase of GSF during recovery from exercise does not require glycogen depletion during the exercise. Key words: cyclic AMP dependent protein kinase, glycogen synthase phosphatase, fatigue, recovery from exercise
Glycogen synthase (GS), a regularity enzyme for glycogenesis, is controlled by covalent phosphorylation/dephosphorylation (Friedman & Larner 1963). Phosphorylation (which results in a decrease in G S F ) is catalysed by specific kinases and dephosphorylation (which results in an increase in GSF) is catalysed by specific Correspondence : Abram Katz, Department of Clinical Physiology, Karolinska Institute, Karolinska Hospital, Box 60500, S-104 01 Stockholm, Sweden.
phosphatases (Cohen 1986, Larner & VillarPalasi 1971). Short-term intense exercise results in a decrease in GSF in human skeletal muscle (Chasiotis et al. 1983, Kida et al. 1989), and this is due to an activation of GS kinase(s) (Yan et al. 1991). However, prolonged submaximal exercise, which results in a marked degradation of muscle glycogen, results in an increase in GSF (Maehlum et al. 1977). I t is likely that the increase in GSF in the latter condition is due, at least in part, to an increased suitability of
345
346
Z. Ynn et al.
lateral aspect of the quadriceps femoris muscle of both thighs ( - 3500 of the distance from the superior margin of the patella to the anterior iliac spine). After a blood sample and muscle biopsy (Bergstrom 1962) were obtained, the subjects cycled (70 rev min-.') for 2 min at a power output calculated to correspond to 5O0, i'02n,ax (125 F 8 W); thereafter, the workload was immediately increased to one calculated to correspond to 75°0V0,n,al (212f 15 W). Thesubjects cycled at this workload to fatigue (34.5 t- 5.3 min). A second biopsv was taken at fatigue (contralateral leg), and a third was taken at 5 min after termination of exercise (proximal to the post-exercise biopsy). Additional blood samples were taken at the end of exercise and 5 min post-exercise. This experiment constituted the low glycogen condition (LG). For the following 3 days subjects performed no exercise and consumed a diet rich in carbohydrate to increase the muscle glycogen content to high levels (HG condition) (Spencer & Katz 1991). The experimental protocol was then repeated with individual exercise durations and power outputs being identical to the LG condition. One subject who participated in the L G experiment did not participate in the HG experiment. T h e time between termination of exercise and freezing of the post-exercise biopsies averaged 2 0 F 2 and 17 1 s during the LG and HG trials (P > 0.05), and the time between termination of exercise and the freezing of the biopsy obtained during recovery averaged 5.30&0.02 and 5.22F0.18 min during the L G and IIG trials (P> 0.05), respectively. .4nalytical. All biopsies were quick-frozen in the needle by immersion into liquid freon, maintained at its melting point ( - 150 "C) with liquid N,. T h e biopsies were stored in liquid N, until preparation for analysis at which time they were lyophilized, dissected free of solid non-muscle constituents (connective tissue and blood), powdered and thoroughly mixed. M A T E R I 41,s AND 34ETHODS One aliquot of powder was digested in 1 M K O H (at Subjri-ts. Eight healthy men whose mean (range) age, 60 "C); the digest was neutralized with HCI, and height, weight and maximal oxygen uptake (VO,max) gll-cogen was hydrolyzed enzymatically (Harris el a / . were, respectively, 21 !-ears (20-22), 178 cm 1974). T h e released glucose was analysed enzymaticallv (changes in NADPH) with a fluorometric (172-18.54 73.0 kg (63.9-82.7) and 4.21 1 min-' (3.53--.5.07), participated in the study. The subjects technique (Lowry 8i Passonneau 19i2). .r\ second aliquot was homogenized (60 pl mg dry aere informed of the possible risks involJ-ed before giving voluntary consent. The experimental protocol wt) in a buffer containing 10 mM Tris, 10 mM EDTA, 100 mM K F and lo,, charcoal, p H 7.6 and then was approved by the Unit-ersity of Illinois Institutional centrifuged at 12800 g for 10 min at 4 "C (Walkenbach Review Board. Experimental Design. Vo,,,, was determined on an et al. 1980). Part of the supernatant was used for electrically-braked cycle ergometer (Bosch ERG-551, analysis of cyclic AMP-dependent protein kinase Germany) in the upright position as previously (cAMPPK) using Kemptide as a substrate in the described (Spencer & Katz 1991). The day before the presence or absence of 100 ,UM CAMP (see Kida et a / . first experiment the muscle glycogen content was 1991). T h e fractional activity is 0/100 ,//M CAMP. reduced by a combination of exercise and diet (Spencer Another part of the supernatant was diluted and used 8i Katz 1091). Subjects reported to the laboratory the for analysis of glycogen synthase (GS) in the presence next da! after an overnight fast and, after assuming of 0.17 (GS,,J and 7.2 mM glucose 6-P (GSk1,J (see the supine position, a plastic catheter was placed in an Kida et al. 1989). T h e fractional activity is antecubital vein and incisions were made over the GSIOR/GShlKh (GSF).
substrate (GS) for GS phosphatase ( G S P ) ; i.e., the low glycogen results i n a n increased availability of phosphorylated sites o n GS (Cohen 1986), and that this overrides the contractionmediated decrease in G S F . I t is not known, however, whether prior depletion of the muscle glycogen stores alters the changes in GSF during short-term exercise. I t might b e expected that the decrease in GSF during short-term exercise with a normal glycogen content would b e attenuate or abolished in the gl\-cogen depleted state, ow-ing to a high substrate suitability of GS for GSP. T h e r e is only a limited information on the role of glicogen in the rapid increase of GSF during recovery from exercise. Danforth (1965) showed that following a 20 s tetanus, the increase in GSF in mouse skeletal muscle deficient in phosphorylase b kinase a n d containing excessive glycogen both before a n d after the tetanus, is markedly attenuated compared with the increase in GSF in control mouse muscle. On the other hand. preliminary data i n humans indicate an increase in GSF in muscle during recovery after 30 s of moderate exercise performed with a low glycogen content (Hultman et al. 1971). T h e increase in GSF in the low glycogen condition was comparable to the increase observed in a normal glycogen condition (Hultman et al. 1971). T h e purpose of this study was to determine the effect of a low muscle glycogen content on GS activity in human muscle during and after exercise.
Glycogen and glycogen synthase in muscle When available, a third aliquot was used for analysis of GSP (see Kida et al. 1989). Commercial rabbit GS-D (i.e., dependent on glucose 6-P) (G2259, Sigma Chemical Co, St Louis, MO, USA) was purified as described elsewhere (Kida et al. 1989), and was 95 glucose 6-P-dependent when assayed in the presence of 7.0 mM UDPG with and without 7.2 mM glucose-6-P'. Muscle powder was homogenized with a ground-glass homogenizer in a buffer (133 yl mg-' dry wt) containing 50 mM Tris, 10 mM EDTA and 50 mM 2-mercaptoethanol, pH 7.8, at 4 "C. The homogenate was centrifuged at 10000 g for 20 rnin at 4 "C. Twenty-five milliunits of purified GS-D (8 pl) were pre-incubated with 0.3% glycogen, 10 mM Tris, 1 mM EDTA and 5 mM dithiothreitol, pH 7.8 (total volume = 75 PI) for 5 rnin at 30 "C. The GSP reaction was then started by transferring 100yl of extract supernatant into the pre-incubated reaction mixture. The reaction was stopped right after the start of reaction and after 15 rnin incubation by diluting 50 yl of the incubation mixture with 2000 yl of 130 mM KF, 50 mM Tris and 20 mM EDTA, pH 7.8, at 4 "C. Twenty-five microlitres were then used to determine GSP activity using the GS assay (i.e., the change in GS,,_). Values are expressed as p n o l of glucose from UDP-( "C]G incorporated into glycogen kg-' dry wt min. All enzymes were assayed at 30 "C and were linear with time and extract volume (data not shown). Part of the GS supernatant, after extraction with PCA and neutralization with KHCO,, was also used for analysis of phosphocreatine (PCr) and creatine (Cr) (Lowry & Passoneau 1972). The sum of PCr and Cr (TCr) averaged 116.0$4.3, 105.9k4.2 and 111.5 k 6 . 7 mmol kg-' dry wt at rest, end of exercise and 5 min post-exercise in the LG condition ( P > 0.05), and 107.7k4.2, 90.2k4.3 ( P < 0.01 vs. 107.7& 4.2) and 108.5f4.7 at rest, end of exercise and 5 min post-exercise in the H G condition, respectively. The reason for the decrease in TCr after HG exercise is not clear, but to account for this decrease all enzyme activities and glycogen contents were divided by the individual TCr values and then multiplied by the mean TCr content for the whole material (106.4 mmol kg-' dry wt). The latter manipulation allows for expression of data in a physiologically meaningful manner without altering the relative differences in values within a given subject. Blood was drawn anaerobically and immediately injected into ice-cold tubes containing EDTA. The samples were left on ice for 10 min, centrifuged at 4 "C, and the plasma was aspirated, stored at - 80 "C, and subsequently used for analysis of free fatty acids, glucose, lactate and insulin as previously described (Spencer et al. 1992). Statistics. Significant differences between means were determined with a one way repeated measures analysis of variance (ANOVA) or a paired t-test where appropriate. When the ANOVA yielded a significant
347
F ratio, the location of significance was identified with the Newman-Keuls test. Significance was set at P < 0.05. Values are reported as means+SE unless otherwise indicated.
RESULTS Cardiorespiratory Cardiorespiratory results and estimates of whole body substrate utilization rates at rest and during exercise are presented elsewhere (Spencer et al. 1992). Briefly, during exercise, heart rate and whole body Voz were significantly lower, while the respiratory exchange ratio (Vco2/VoI) and the rate of whole body carbohydrate oxidation were significantly higher during HG vs. LG, respectively. T h e higher Vo, during LG exercise is probably due to a higher fat oxidation (per litre of 0, consumed the caloric yield of fat is less than that of carbohydrate).
Muscle
As a result of the experimental manipulations the muscle glycogen content at rest was 4-fold higher during HG vs. LG (Table 1). Glycogen was essentially completely depleted after LG exercise and remained depleted during recovery. Although approximately twice as much glycogen was degraded during HG exercise, the values at the end of exercise and during recovery were high and similar to those in untrained human muscle under basal conditions (Harris et al. 1974). T h e activity of GShighwas similar between treatments at rest. Although the activity of GShighdid not change significantly during HG exercise, there was a significant decrease during LG exercise, and this decrease persisted during the 5 rnin recovery period. Interestingly, in two subjects there was no decrease in GShighduring LG exercise. These two subjects had the highest muscle glycogen contents at the end of LG exercise (27 and 49 mmol glucosyl units kg-' dry wt). Within 3 days of ingesting a high carbohydrate diet the activity had increased back to the initial value ( H G rest). As expected on the basis of earlier work (Danforth 1965), GSF was markedly higher at rest in the LG condition. There were no significant changes in GSF during exercise under both conditions, but there were significant increases during the 5 min recovery under both
348
%. l i i n et al.
Table 1. Effect\ of exercise on gl!cogen 2nd GS acti\it! in muscle
I IG I IG \'dues are means +SE: for .iX (glycogen) or i-8 (GS and GSl,,~,/GS,~,K,,) individuals. Glycogen is giwn in mrnol gl!cos!l tinits kg-'dr! u t , and GS,,,,,,in mmol Ig-'dr? t+t min-' at 30 "C. GS, glycogen synthase; GS,,,,,, GS measured at IOU glucose 6-P concentration; GS,,,pl,,GS measured at high glucose 6-1' concentration; TAG, low gl!-cogen; IfG high gl!cogen. * P < 0.0.5; **P < 0.01 : *** P < 0.001 vs. respective rest value. tP < (1.05; .tt P < 0.01 ; j-tt P < 0.001 VS. LG.
(a) Low glycogen
400
0
m
200
-3
......... ,Q.
9
$ L
8
........a... ..........0.
8
c
.............. ...................................
9 400
:b) High glycogen
v
?...a ....................
0
L (I)
200
0
o i
0
....
0
0
30 Exercise duration (rnin)
0
60
Fig. 1. Relationship between the change in muscle GSF during esercise and esercise duration. (a) During low glyxgen condition (0). (h) During high glycogen condition 1.). r = 0.94; P < 0.001: 1' = 2.54.2117.53u t0.331'. - - represents d u e at rest (i.e., no change in GSF). ~~
conditions. Examination of the indii idual data revealed a good relationship between the change in GSF during HG exercise and the exercise duration (Fig. 1). I t is noteworthy that those subjects who cycled for the shortest durations ( < 3.5 min, n = 3) had marked decreases in GSF (the post-exercise values ranged from 16-73 of the resting value), while those who cycled for > 3.5 min (n = 4) showed no decrease or an increase in GSF. O n the other hand, G S F remained fairly constant during LG exercise.
0
400 800 Glycogen (rnrnol glucosyl units kg-' dry wt)
Fig. 2. Relationship between the change in muscle GSF during exercise and glycogen at the end of esercise in high glycogen condition. r = - 0 . 8 2 ; P < 0.05; y = ~ 1 ~ . 9 ( ~ ~ - ~ . 6 0 4--~ ~represents 0.0~1~~. ~ a l u ca t rest (i.c., no change in GSF). There was also an inverse relationship between the change in GSF during HG exercise and the glycogen content in the muscle at the end of exercise (Fig. 2 ) . Exercise resulted in an activation of c A M P P K (0/100 ,//xi CAMP) without altering the activity ~ ) both conat saturating CAMP ( 1 0 0 , ~ under ditions (Table 2 ) . During recovery c A M P P K fractional aetivitj- reverted toward the preexercise values. There was not sufficient material for meaningful analyses of GSP at rest (only one paired observation which is not reported). However, GSP was higher during HG exercise and recovery.
Glycogen and glycogen synthase in muscle
349
Table 2. Effects of exercise on cAMPPK and GSP activity in muscle Rest cAMPPK (0/100 p ~ ) LG 0.15 f0.01 HG 0.13 f0.02 cAMPPK (100 p ~ ) LG 135f8 IIG 152f 13 GSPI I,G I IG ~
Values are means k SE for 6-8
(a =
Exercise
5-min post-exercise
0.25 f 0.03" 0.22 f 0.03"*"
0.18 f0.03 0.16 f0.02
128 f 12 152f11
139f 17 146k 18
430f41 706 k 5 4 t t t
476 f36 616k70t
5 for GSP, 5 min post-exercise, LG) individuals and are given in pmol
kg-' dry wt min-' 30 "C for cAMPPK and in pmol kg-' dry wt (min-') at 30 "C for GSP. cAMPPK was measured in the absence and presence of 100 ,UM CAMP. cAMPPK, cyclic AMP dependent protein kinase;
GSP, glycogen synthase phosphatase; LG, low glycogen; HG, high glycogen. 1Insufficient material was available for analysis of GSP in the resting state under both conditions. * P < 0.05; ""I P < 0.001 vs. respective rest value. t P = 0.05; t t t P < 0.001 vs. LG. Table 3. Effects of exercise on plasma metabolites and insulin
Glucose LG HG Insulin LG HG Lactate LG HG FFA LG HG
Rest
Exercise
5-min post-exercise
4.59 rt 0.16 5.55 + O . l l t t t
3.64 f 0.28" 5.22fO.lOj-t
4.13 f0.32 5.80 f 0.22tt
6f 1 12 5 3 t
4 51 16f5t
3*1* 6f1t
0.78 & 0.05 0.96 f0.071.
5.39 f0.66*** 8.05 & O.jO***+tt
4.93 f0.59*** 7.565 O.SS***ttt
0.87 fO.ll 0.35 k 0.05t-t
0.77 f0.08 0.25 0.04Qyf-t
1.51 f0.07" 0.57 fO.l2*"*ttt
*
Values are means k SE for 7-8 individuals and are given in mmol 1.' or pU ml-' (insulin). FFA, free fatty acids. " P < 0.05 ; "*" P < 0.001 vs. respective rest value. t P < 0.01 ; tt P < 0.05 ; ttt P < 0.001 vs. LG.
Plasma At rest, plasma glucose, insulin and lactate were higher during HG, while FFA was higher during LG (Table 3 ) . Glucose decreased during LG exercise but not during HG exercise, and increased during recovery in HG ( P < 0.01 vs. end of exercise value). Insulin decreased during exercise under both conditions. Insulin increased in six of seven subjects during HG recovery, but remained depressed during the LG recovery. Lactate increased to a greater extent during HG exercise. FFA were not markedly altered during LG exercise but decreased during HG exercise, and then increased during recovery under both
-
conditions. T h e increase in FFA during recovery twice that observed in the LG condition was in the HG condition. This may in part relate to the higher plasma insulin concentration in the HG condition at this time, which should blunt lipolysis.
DISCUSSION
GSF I t was first shown by Friedman & Larner (1963) that GS is regulated by covalent phosphorylation/dephosphorylation. T h e GS
3.30
%.
I'UtI
er
ill.
assay used in the present study has been shown to bc. sensitive to the phosphoryhtion state IJf the
cnzynie (Ciuinovart tl id. 1979). Thus an increase in
