Fasting inhibits insulin-mediated glycolysis and anaplerosis in human skeletal muscle

CHARLES E. CASTILLO, ABRAM KATZ, MARK K. SPENCER, ZHEN YAN, AND BULANGU L. NYOMBA Clinical Diabetes and Nutrition Section, National Institute of Diabetes, Digestive, and Kidney Diseases, National Institutes of Health, Phoenix, Arizona 85106; and Department of Kinesiology, University of Illinois, Urbana, Illinois 61801

CASTILLO,~HARLES E., ABRAM KATZ,MARK K. SPENCER, ZHENYAN,ANDBULANGU L. NYOMBA. Fastinginhibitsinsulin- mediated glycolysis and anaplerosis in human skeletal muscle. Am. J. Physiol. 261 (Endocrinol. Metab. 24): E598-E605, 1991.-Euglycemic (~5.5 mM) hyperinsulinemic (60 mu. mS2 l

min-l) clamps were performed for 2 h after a 10-h fast and after a prolonged (72-h) fast. Biopsies were obtained from the quadriceps femoris muscle before and after each clamp. The rate of whole body glucose disposal was -50% lower during the clamp after the 72-h fast (P 5 0.001). The increase in carbo- hydrate (CHO) oxidation (which is proportional to glycolysis) during the clamp after the 10-h fast (to 13.8 t 1.5 pmolkg fat free mass-l l rein-l) was completely abolished during the clamp after the 72-h fast (1.7 t 0.6; P 5 0.001). During the clamp after the 10-h fast, postphosphofructokinase (PFK) interme- diates and malate in muscle increased, whereas glutamate de- creased (P 5 0.05-0.001 vs. basal) and citrate did not change. During the clamp after the 72-h fast, there were no significant changes in post-PFK intermediates or glutamate (P > 0.05 vs. basal), but there was a decrease in citrate (P 5 0.01 vs. basal). Euglycemic hyperinsulinemia increased glycogen synthase frac- tional activity in muscle under both conditions but to a greater extent after the 72-h fast (P 5 0.01). It is concluded that insulin (after 10-h fast) increases glycolytic flux and the content of malate in muscle, which is probably due to increased anaple- rosis. The decrease in citrate during hyperinsulinemia after the 72-h fast indicates that insulin activated the tricarboxylic acid (TCA) cycle but that the replenishment of TCA cycle inter- mediates normally observed (10-h fast) was blocked, possibly due to inadequate provision of three-carbon intermediates for anaplerosis. These data demonstrate that prolonged fasting can increase and decrease insulin-mediated activation of processes associated with CHO metabolism in human skeletal muscle.

lactate; malate; citrate; glycogen synthase; carbohydrate oxi- dation

IT IS WELL DOCUMENTED that in intact humans short- term starvation results in a decrease in whole body basal and insulin-mediated rates of glucose disposal (5). After starvation, basal and insulin-mediated glucose uptake by intact human skeletal muscle is reduced (28, 29, 37). Studies on rodents, using either the perfused rat hind- limb or isolated skeletal muscle in vitro, have demon- strated that glucose utilization in the absence or presence of insulin is not impaired after starvation (2, 11). In fact, at submaximal concentrations of insulin, glucose utili- zation is increased both in isolated mouse and rat soleus

muscle (6, 24), as well as in the perfused rat hindlimb (6). These latter studies showed that starvation resulted in increases in insulin-mediated glycolysis and glycogen- esis, demonstrating that both pathways for glucose uti- lization had become hypersensitive to insulin (6, 24). Consistent with these results were the findings that, after starvation, insulin binding (at submaximal concentra- tions of insulin) to rodent skeletal muscle was increased (6, 24), as was the insulin-mediated activation of 2- deoxy-D-glucose transport (24). Similarly, starvation re- sults in an increased expression of at least two distinct glucose transporters (adipose muscle and erythrocyte brain) in rat skeletal muscle (7). Insofar as starvation results in insulin resistance in intact humans, the appli- cability of the results from rodent preparations to the understanding of insulin-mediated carbohydrate (CHO) metabolism in humans is questionable.

The purpose of this study was to determine the effects of short-term starvation on basal and insulin-mediated CHO metabolism in humans. Specifically, we were inter- ested in determining whether the different pathways of glucose utilization (glycogenesis and CHO oxidation) were affected in a similar manner by starvation. Meas- urements of a number of metabolites and enzyme activi- ties in muscle were used to identify potential regulatory sites in the starvation-mediated effects.

METHODS

Subjects. Eight men, whose mean (range) age, height, weight, percent body fat, and plasma glucose concentra- tion 2 h after ingesting a solution containing 75 g of glucose were, respectively, 29 yr (22-40), 174 cm (169- 180), 71.7 kg (63.1-80.3), 15.7% (8.2-20.6), and 5.8 mmol/ 1 (4.7-6.9), participated in the study after giving volun- tary consent. The subjects’ health was assessed by phys- ical examination and routine hematologic, electrocardi- ograph, and urine tests. The experimental protocol was approved by the ethics committee of the National Insti- tutes of Health.

Experimental design. All subjects were studied on two occasions during their stay on the metabolic ward (-7 days). The first study was preceded by at least 3 days on a weight-maintenance diet (45% CHO, 40% fat, 15% protein). After an overnight fast (-10 h), a euglycemic hyperinsulinemic clamp was performed. Briefly, at 0600

E598 0193~1849/91 $1.50 Copyright 0 1991 the American Physiological Society

FASTING AND CARBOHYDRATE METABOLISM IN MUSCLE E599

h and after the subject had voided, a catheter was placed in an antecubital vein for infusion of insulin, glucose, and [3-3H]glucose. Another catheter was placed retro- grade in a dorsal hand vein of the contralateral hand for blood sampling. To arterialize the blood, the hand was kept in a warming box at 70°C. A primed (30 &i), continuous infusion of [3-3H]glucose (0.3 &i/min) was started at -120 min and was maintained for 240 min (i.e., until 120 min). At 0 min, a primed (958 mu/m’) continuous infusion of insulin (60 mU l m-’ l min-‘) was started and maintained for 120 min. A variable 20% glucose infusion was performed between 0 and 120 min to achieve an arterialized plasma glucose concentration of -5.5 mmol/l. Samples for plasma glucose were ob- tained every 5 min throughout the insulin infusion.

Samples for determination of plasma [ 3-3H] glucose specific activity were drawn every 10 min from -40 to 0 min and from 80 to 120 min. Rates of glucose infusion and appearance and CHO and fat oxidation (indirect calorimetry) were estimated at the same time. Glucose disposal (i.e., disappearance of glucose from plasma) was estimated from the specific activity of [3-3H]glucose in plasma and its rate of infusion using Steele’s non-steady- state equations (34), assuming a glucose distribution volume of 100 ml/kg. The total rate of glucose appear- ance was never higher than the rate of glucose infusion in any subject after the 10-h fast; therefore the rate of glucose infusion is used to reflect glucose disposal. After the 72-h fast (see below), however, the rate of glucose appearance was slightly higher than the glucose infusion rate in four subjects (suggesting incomplete inhibition of endogenous glucose production). In these four subjects the rate of glucose appearance is used to reflect glucose disposal, whereas in the other four subjects the rate of glucose infusion is used to reflect glucose disposal. Fur- ther details on the clamp procedure, determination of body composition (underwater weighing), methods for analysis of [ 3-3H] glucose specific activity, indirect calo- rimetry measurements, and calculations are provided elsewhere (25).

Biopsies from the lateral aspect of the quadriceps femoris muscle were obtained with the needle biopsy technique. Briefly, after local anesthesia (lidocaine 10 mg/ml), incisions at biopsy sites on both thighs were made, and biopsies were obtained before (0 min) and at the end of the clamp. After the clamp the subjects were allowed to eat until nightfall. The next morning they ate breakfast; thereafter, the subjects started a 72-h fast. They were allowed to drink all fluids that did not contain any calories. After 72 h, at the same time of day (i.e., 0600 h), the same clamp procedure, as described above, was repeated. Thus the clamps were not randomized. This is because we did not know how long it would take to completely recover from the 72-h fast. Therefore the 10-h fast experiments were always performed first. It is not likely that lack of randomization significantly af- fected the results, because, had the subjects not fasted for 72 h after the initial clamp (i.e., if they would have consumed the controlled diet described above), virtually all of the measured values measured in the basal state 72 h after the initial clamp would have been the same as

those obtained before the clamp after the 10-h fast (31, 33)

dnalytical methods. All biopsies were frozen in liquid N2 and stored at -70°C until analysis. The samples were freeze-dried, dissected free from nonmuscle constituents (blood, fat, connective tissue), and powdered. The powder was thoroughly mixed and, when sufficient material was available, was divided into six aliquots. One aliquot was extracted with 0.5 M perchloric acid and was neutralized with KHC03. The neutralized extract was used for fluo- rometric enzymatic analyses of metabolites (3, 26). The second aliquot was digested in 50 mM NaOH (SOOC) for 20 min. The extract was neutralized with acetic acid and was assayed spectrophotometrically for fructose 2,6-di- phosphate (F-2,6-&) using pyrophosphate-dependent phosphofructokinase (PFK) from potato tubers (35). The third aliquot was digested in 1 M KOH (6O”C), neutral- ized with HCl, and used for enzymatic analysis of total carnitine (spectrophotometrically) (3). In another part of this neutralized extract, glycogen was hydrolyzed en- zymatically and was assayed enzymatically for glucose (13). Glycogen was expressed as millimoles glucosyl units per kilogram dry weight. The fourth aliquot was used for measurement of glycogen synthase (GS) activity using filter paper (see Ref. 22). Briefly, powder was homoge- nized with a buffer containing 30% (vol/vol) glycerol, 10 mM EDTA, and 50 mM KF, pH 7, at 4°C. The homog- enate was centrifuged at 10,000 g for 20 min at 4°C. The supernatant was diluted with a buffer containing 50 mM tris(hydroxymethyl)aminomethane (Tris), 20 mM EDTA, and 130 mM KF, pH 7.8, at 4°C and was then used for assay of GS. Enzyme (25 ,ul) was added to 50 ~1 of a reaction buffer that contained a low glucose 6- phosphate (G-6-P) concentration (0.17 mM; GSI,,) and a high G-6-P concentration (7.2 mM; GShigh). UDP- glucose concentration in the buffer was 0.20 mM. En- zyme activity was estimated from the incorporation of UDP-[14C]glucose into glycogen at 30°C. The fractional activity (GSF) is GSlow/GShigh.’

The fifth aliquot was used for analysis of GS phospha- tase (GSP), which was assayed as previously described (see Ref. 22). Commercial rabbit GS-D (i.e., dependent on G-6-P; Sigma G2259) was purified as described else- where (22) and was -95% G-6-P dependent when as- sayed in the presence of 7.0 mM UDP-glucose with and without 7.2 mM G-6-P. Muscle powder was homogenized with a ground-glass homogenizer in a buffer (133 pl/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 10,000 g for 20 min at 4°C. Purified GS-D (25 mu, 8 ~1) were preincubated with 0.3% glyco- gen, 10 mM Tris, 1 mM EDTA, and 5 mM dithiothreitol, pH 7.8 (total volume 75 ~1) for 5 min at 30°C. The GSP reaction was then started by transferring 100 ~1 of extract supernatant into the preincubated reaction mixture. The reaction was stopped right after the start of reaction and after 15 min incubation by diluting 50 ,ul of the incuba- tion mixture with 2,000 ,ul of 130 mM KF, 50 mM Tris, and 20 mM EDTA, pH 7.8, at 4°C. Twenty-five micro- liters were then used to determine GSP activity using the GS assay (i.e., the change in GSI,,). Values are expressed as micromoles of glucose from UDP- [ 14C]glu-

E600 FASTING AND CARBOHYDRATE METABOLISM IN MUSCLE

case incorporated into glycogen per kilogram dry weight per minute squared.

The last aliquot was used for analysis of adenosine 3’,5’-cyclic monophosphate (CAMP)-dependent protein kinase (CAMP-PK), which was assayed using kemptide as substrate (see Ref. 24). Muscle powder was homoge- nized in a buffer (60 pl/mg dry wt) containing 10 mM Tris, 10 mM EDTA, 100 mM KF, and 1% charcoal, pH 7.6, using a ground-glass homogenizer. The homogenate was centrifuged at 12,800 g for 10 min at 4°C. Extract supernatant (15 ~1) were incubated with 85 ~1 of reaction mixture, giving a final concentration of 200 PM kemptide (Sigma, K1127), 0.3 mM [32P]ATP (sp act -20 mCi/ mmol; Amersham), 50 mM Z-(N-morpholino)ethane- sulfonic acid, and 8 mM MgC12, pH 6.5, in the absence or presence of 0.2 and 100 PM CAMP at 30°C for 5 min. Preliminary data showed that CAMP-PK was saturated at 100 PM CAMP (data not shown). To stop the reaction, 60 ~1 of the incubation mixture were applied to a disk of Whatman-P81 paper. The disks were washed with 150 mM phosphoric acid, rinsed with 95% ethanol, and dried. The radioactivity was determined in a Packard scintil- lation counter. CAMP-PK activity was expressed as mi- cromoles 32P incorporated into kemptide per kilogram dry weight per minute. All enzyme activities were linear with time and extract volume (data not shown). The interassay coefficient of variation for muscle metabolites and enzymes is generally <lo% (see Ref. 13 and Katz, unpublished observations).

All metabolite contents (except for pyruvate, lactate, and glucose, which are also present in the extracellular compartment) were adjusted to the mean total creatine content (sum of phosphocreatine and creatine) for the whole material [ 125.7 t 1.8 (SE) mmol/kg dry wt]. This adjustment was performed to correct for variability in blood, connective tissue, or other solid nonmuscle con- stituents between biopsies.

Plasma glucose was determined with the glucose oxi- dase method using a Beckman glucose analyzer (Beck- man Instruments, Fullerton, CA), and plasma glucose was determined by radioimmunoassay using a Concept 4 automated analyzer (Micromedic Systems, Norsham, PA). Plasma free fatty acids were determined enzymati- tally (see Ref. 19).

Statistics. Significant differences (P IS 0.05) were de- termined with the paired t test. Values are presented as means t SE unless otherwise indicated.

RESULTS

The 72-h fast resulted in a significant decrease in body weight (from 71.7 t 1.9 to 68.5 t 1.9 kg; P 5 0.001). No significant difference in urinary nitrogen output was detected as a result of fasting [control, 7.2 t 0.6 mg*kg fat-free mass (FFM)-lomin-l; fasting, 8.3 t 0.7; P > 0.051. Percent body fat was only determined before the first clamp, and it was not deemed valid to use this value to estimate FFM after the 72-h fast (this would have resulted in an overestimate in the loss of lean body mass). To determine the loss of FFM during the 3-day fast, the urinary nitrogen excretion was corrected for insensible nitrogen loss, and total protein loss was estimated using

a value of 6.25 g protein/g nitrogen. FFM after the 3-day fast was then calculated by subtracting protein loss from the initial FFM. FFM was estimated to decrease from 60.4 t 1.9 kg before the 3-day fast to 59.3 t 1.8 kg after the fast (P 5 0.001).

The 72-h fast resulted in a significant decrease in basal endogenous glucose production and insulin-mediated glucose disposal (Table 1). During the clamp after the 10-h fast, endogenous glucose production was completely suppressed, and there was no significant difference be- tween treatments during the clamp. These data support the previous conclusion that fasting-mediated insulin resistance is relegated to peripheral tissues (5). The 72- h fast resulted in a marked decrease in basal CHO oxidation and completely blocked the increase in CHO oxidation during the clamp. The 72-h fast did not affect the basal rate of nonoxidative glucose disposal, and al- though nonoxidative glucose disposal during the clamp was lower after the 72-h fast in six of the eight subjects, the relative effect of the 72-h fast was not nearly as dramatic as that on CHO oxidation. Similarly, the 72-h fast resulted in increased rates of fat oxidation in the basal state and completely abolished the insulin-me- diated decrease in fat oxidation.

The effects of short-term starvation on the concentra- tions of metabolites and insulin in plasma in the basal state (Table 2) are as previously described under similar

TABLE 1. Effects of 72-h fast and insulin on whole body metabolic rates

Basal Clamp

Endogenous glucose production, pmol . kg FFM-l. min-l

10 h 72 h

Glucose disposal, pmol l kg FFM-1 l min-l

10 h 72 h

CHO 0X, pmol . kg FFM-1 a min-’

10 h 72 h

CHO nonox, pmol . kg FFM-l. min-l

10 h 72 h

Fat 0X, ymol . kg FFM-1 l min-’ 10 h 72 h

RER (VcoJVo2) 10 h 72 h

vog, ml l kg FFM-l. min-’ 10 h 72 h

EE, Cal. kg FFM-l. min-’ 10 h 72 h

16.3t0.4 O.l+O.lf 11.6t0.4" 1.7_t0.7f

16.3t0.4 11.6t0.4"

8.4t0.7 2.320.4"

13.8t1.5f 1.7t0.6"

7.9t0.7 9.320.3

25.3t3.2" 18.1t1.4f~a

2.9t0.4 4.5kO.3"

1.7kO.4" 4.720.3"

0.835t0.009 0.756t0.005"

0.899t0.017" 0.748t0.006"

3.18t0.10 3.35+0.0gb

3.33t0.09 3.36kO.10

15.2t0.5 15.7t0.4

16.1+0.5d 15.7kO.4

Values are means t SE for 8 individuals. FFM, fat-free mass. Fat,,, fat oxidation, assuming palmitate is being oxidized (mol wt = 258). CHO,,,, carbohydrate oxidation; CHO,,,,, nonoxidative carbohydrate disposal; ko2, CO2 production; %702, volume of 02 consumed; RER, respiratory exchange ratio; EE, energy expenditure. a P I 0.05, b P s 0.01, 'P IS 0.001, vs. 10 h. d P 5 0.05, e P 5 0.01, f P 5 0.001, vs. respective basal value.

FASTING AND CARBOHYDRATE METABOLISM IN MUSCLE E601

TABLE 2. Effects of 72-h fast and insulin TABLE 4. Effects of 72-h fast and insulin on on plasma metabolites glycogenolytic and glycolytic intermediates in muscle

Basal Clamp

Glucose, mmol/l 10 h 5.3kO.l 5.4t0.1 72 h 3.9t0.2” 5.5t0.1f

FFA, mmol/l 10 h 0.25t0.02 0.05+0.01f 72 h 0.87t0.06" 0. 10+0.02a~f

Lactate, mmol/l 10 h 0.71-r-0.08 1.02-c-0.06” 72 h 0.70t0.04 0.76+0.07b

Insulin, &J/ml 10 h 5tl 76t4f 72 h 2tlb 72k4f

Values are means & SE for 8 individuals. FFA, free fatty acids. a P 5 0.05, bP 5 0.01, "P 5 0.001, vs. 10 h. e P 5 0.01, f P 5 0.001, vs. respective basal value.

Basal Clamp

Glycogen 10 h 72 h

Glucose 10 h 72 h

UDP-Glucose 10 h 72 h

Glucose l-phosphate 10 h 72 h

Glucose 6-phosphate 10 h 72 h

Fructose 6-phosphate 10 h 72 h

Fructose 1,6-diphosphate 10 h 72 h

DHAP 10 h 72 h

Glucose 1,6-diphosphate 10 h 72 h

Fructose 2,6-diphosphate 10 h 72 h

379tl6 328k15b

390tl7 322tl2"

2.520.2 2.3t0.2

2.420.4 3.0+0.3d

0.86t0.12 l.llt0.16

1.03-eo.10 0.96t0.11

1.5t0.3 1.7-co.3

2.2t0.2 1.8-eO.3

TABLE 3. Effects of 72-h fast and insulin on high-energy phosphates in muscle

0.63t0.08 0.64t0.09

0.74t0.08 0.62t0.09

Basal Clamp

PCr 10 h 80.2t2.1 78.6tl.5 72 h 78.2tl.4 79.1t2.0

Cr 10 h 45.51t2.1 47.1tl.4 72 h 47.5kl.4 46.6t2.0

PCr/Cr 10 h 1.80t0.11 1.69t0.08 72 h 1.6620.08 1.7320.12

ATP 10 h 24.2t0.6 23.9t0.7 72 h 24.320.5 24.1t0.8

Values are means t SE for 8 individuals and are given in mmol/kg dry wt. PCr, phosphocreatine; Cr, creatine.

0.27t0.05 0.33kO.06

0.28t0.05 0.27t0.05

0.17t0.03 0.21t0.03

0.2220.03 0.18t0.03

83_t4 96k9

94&6 86t9

11.4kl.3 15.4+l.6b

11.8t0.9 12.4tl.O

Values are means t SE for 6-8 individuals and are given in mmol/ kg dry wt, except for glucose 1,6-diphosphate and fructose 2,6-diphos- phate, which are in pmol/kg dry wt. DHAP, dihydroxyacetone phos- phate. b P 5 0.01, ’ P s 0.001, vs. 10 h. d P i 0.05 vs. respective basal value.

conditions (5). Starvation completely inhibited the in- sulin-mediated increase in plasma lactate.

TABLE 5. Effects of 72-h fast and insulin on carnitines and free amino acids in muscle

Neither starvation nor insulin affected the contents of high-energy phosphates in muscle (Table 3). Starvation resulted in an -15% decrease in the basal content of muscle glycogen (Table 4) and a significant increase in F-2,6& but did not affect any of the other sugar phos- phates. There were no significant changes in any of the sugar phosphates or glycogen during either clamp. It could be calculated, however, that the increase in muscle glycogen during the clamp after the 10-h fast would amount to only -5% of the basal value. Starvation resulted in a small but significant decrease in free car- nitine and a threefold increase in acetylcarnitine, as well as decreases in glutamate and aspartate in the basal state (Table 5). After the 10-h fast, hyperinsulinemia resulted in a significant decrease in glutamate. There was a trend for acetylcarnitine to decrease during the clamp after starvation, but the decrease was not significant.

Basal Clamp .

Total carnitine 10 h 19.0tl.l 18.521.2 72 h 18.6kl.5 18.8kl.4

Free carnitine 10 h 18.3tl.2 17.6kl.4 72 h 16.521.2” 16.6tl.3

Acetylcarnitine 10 h 0.21t0.05 0.22t0.04 72 h 0.69t0.11” 0.53t0.10”

Glutamate 10 h 11.5tl.O 8.7t0.gf 72 h 7.9kO.9” 7.2k0.8

Aspartate 10 h 1.7-r-0.1 1.5kO.l 72 h 0.9tO.l” 0.7kO.l"

Values are means t SE for 8 individuals and are given in mmol/kg dry wt. a P 5 0.05, ’ P 5 0.001, vs. 10 h. f P 5 0.001 vs. respective basal value.

Starvation did not affect any of the measured three- carbon glycolytic intermediates (including alanine) in muscle in the basal state (Fig. 1). After the 10-h fast, hyperinsulinemia resulted in significant increases in py- ruvate, lactate, and alanine. On the other hand, starva- tion abolished the insulin-mediated increases in the three-carbon intermediates. Starvation did not affect the tricarboxylic acid (TCA) cycle intermediates in muscle

in the basal state (Fig. 2). After the 10-h fast, insulin infusion did not affect the muscle citrate content, which is consistent with previous observations (18, 33), but did result in a significant increase in malate. However, after starvation, insulin infusion resulted in a significant de- crease in citrate. There was also a trend for malate to decrease, but the decrease did not reach statistical sig- nificance.

E602

O-4

10

8

FASTING AND CARBOHYDRATE METABOLISM IN MUSCLE

0.50 Citrate Pyruvate -*1 -NS-

0.25

-NS- Lactate

f---*- --S- Malate l-

0.25

-NS- Alanine -*1

I

I

0.00

Basal Clamp

Control

Clamp Basal

Fasted

FIG. 2. Effects of 72-h fast and insulin on citrate and malate in muscle. Values are means t SE for 8 individuals. Control, after 10-h fast; fasted, after 72-h fast; NS, not significant. None of basal values are significantly different between treatments. * P d 0.05; ** P 5 0.01.

Basal Clamp

Control

Basal Clamp

Fasted

FIG. 1. Effects of 72-h fast and insulin on glycolytic intermediates in muscle. Values are means t SE for 6-8 individuals. Control, after 10-h fast; fasted, after 72-h fast; NS, not significant. None of basal values are significantly different between treatments. * P 5 0.05; ** P 5 0.01.

TABLE 6. Effects of 72-h fast and insulin on glycogen synthase activity in muscle

Basal Clamp

Starvation resulted in a significant decrease in GSF in the basal state, as well as a slight but significant decrease in GShigh (Table 6). I nsulin activated GS under both conditions. However, the insulin-mediated increase in @SF after starvation was significantly greater than that after the 10-h fast (A0.25 t 0.03 vs. A0.17 t 0.03; P 5 0.01).

The activity of CAMP-PK (measured at 100 PM CAMP) was statistically lower in the basal state after the 72-h fast (n = 4; Table 7). However, the difference was small and is not considered to be of physiological signif- icance. There was no indication of an altered activation of the enzyme under any condition (O/100 and 0.2/100 CAMP). Complete data on GSP were available for only three subjects; therefore significance testing was not performed. The activity of GSP was lower after the 72- h vs. the 10-h fast in the basal state for all three subjects.

10 h 72 h

Gf%igh 10 h 72 h

GSlow/GShigh 10 h 72 h

1.4tO.l 2.2+0.3d 0.870.2' 2.1t0.2f

5.5t0.2 5.0t0.3 4.8+0.3b 5.1k0.2

0.25t0.02 0.42t0.04f 0.16+0.02b 0.42+0.04f

Values are means t SE for 8 individuals and are given in mmol kg dry wt-l l min-l at 30°C. GS, glycogen synthase. GSl,, was measured at a glucose 6-phosphate concentration of 0.17 mM and GShigh at 7.2 mM. b P 5 0.01, ’ P 5 0.001, vs. 10 h. d P 5 0.05, f P 5 0.001, vs. respective basal value.

is not affected. Starvation did not affect the basal rate of nonoxidative glucose disposal. However, it is not clear what this component represents (glycogenesis, glycolysis, or the pentose shunt pathway), or whether there were any changes in the flux through these different pathways within or between different organs.

The decrease in muscle glucose uptake in the basal state after starvation (28, 29) is probably due to a de- crease in substrate availability, as well as to a decrease in plasma insulin, and possibly to changes in other hormones (e.g., increase in growth hormone) (37). Based on experiments in rats, there is no evidence to suggest

DISCUSSION

Glucose disposal. This study confirms that starvation results in a decrease in basal hepatic glucose production (5). We now also demonstrate that fasting results in a marked decrease in whole body CHO oxidation and an increase in fat oxidation, such that caloric expenditure

FASTING AND CARBOHYDRATE METABOLISM IN MUSCLE E603

TABLE 7. Effects of 72-h fast and insulin on CAMP protein kinase and GSP activities in muscle

Basal Clamp

CAMP-PK (O/100 PM) 10 h 0.21t0.02 0.23zkO.01 72 h 0.22t0.02 0.21t0.01

CAMP-PK (0.2/100 pM) 10 h 0.62t0.04 0.64t0.01 72 h 0.67t0.03 0.71t0.04

CAMP-PK (100 /JM) 10 h 97.1t12.3 102.2t12.6 72 h 95.4t13.8" 96.2t14.0

GSP 10 h 4641k21 347t19 72 h 332-c-46 278t32

Values are means & SE for 3-4 individuals and are given in ,umol. kg dry wt-l. min-’ [CAMP protein kinase (CAMP-PK), measured at 0, 0.2, and 100 pM CAMP] or pmol . kg dry wt-‘. min-’ [glycogen synthase phosphatase (GSP)] at 30°C. a P 5 0.05 vs. 10 h.

that fasting affects the capacity of the glucose transport system, as judged by the increased expression of glucose transporters (7) and rates of 2-deoxy-D-glucose uptake (6,24). The increase in FFA concentration and oxidation may also be contributing to the decrease in muscle glu- cose uptake (30). If this is correct, the decrease in glucose uptake still cannot be explained by the classic glucose- fatty acid cycle (30), because neither muscle citrate nor G-6-P were increased in the basal state after the 72-h fast.

This study confirms that starvation in humans results in a marked decrease in insulin-mediated glucose dis- posal. During euglycemic hyperinsulinemia -90% of the infused glucose is taken up by skeletal muscle (9). It follows therefore that the decrease in insulin-mediated glucose disposal after starvation is due primarily to a lower glucose uptake by skeletal muscle.

Glycolysis. The effect of starvation on muscle glycolysis can be evaluated as follows. If it is assumed that glycogen decreased linearly over 72 h of fasting, then the glyco- genolytic rate is estimated to be 0.01 mmol. kg dry wt-’ l

min. This value is in agreement with the idea that muscle glycogenolysis is negligible in the resting state (8). The basal rate of glucose uptake by muscle has been determined to be between 0.02 and 0.08 mmol. kg dry wt-l . min-’ (1, 17). If it is assumed that the amount of glucose going to glycogen is negligible (and this assump- tion becomes especially reasonable during the 3-day fast because the activity of GS decreased), then glucose up- take can be estimated to account for 67-89% of glycolytic flux in the basal state. Therefore, the finding that a 3- day fast decreases basal muscle glucose uptake to a value that is -35% of the control value (overnight fast) (28, 29), indicates that muscle glycolysis was lower after starvation. Similarly, the findings that the rates of lac- tate and pyruvate release from the forearm are not al- tered during starvation, whereas the rate of CHO oxi- dation is decreased (28, 29), are also consistent with a decreased glycolytic flux after starvation. The decrease in muscle glycolysis may be due to the decrease in plasma insulin, resulting in a decreased activation of PFK, a regulatory enzyme for glycolysis. In this context, the increase in F-2,6-P2 after starvation is puzzling, since

this should, assuming all other factors are constant, activate PFK (35).

After the 10-h fast, insulin infusion resulted in an increase in whole body CHO oxidation as well as in- creases in the muscle contents of post-PFK intermedi- ates, which indicate that muscle glycolysis was acceler- ated (i.e., PFK was activated). In a recent study, in which insulin was infused at a lower rate (40 mUa rnm2. min-‘), we could not detect significant increases in post-PFK intermediates in muscle (33). However, at higher insulin infusion rates (60 mu. rnm2. min-l), the increase in py- ruvate and lactate become readily detectable (present data). In contrast to the situation during prolonged in- sulin infusion (16), the activation of PFK during eugly- cemic hyperinsulinemia after the overnight fast in the present study could not be attributed to increases in the steady-state level of glucose 1,6-diphosphate, nor could the activation of PFK be associated with increases in any of the other measured activators of PFK [fructose mono- and diphosphates and free ADP and AMP (based on the phosphocreatine-to-creatine ratio; see Ref. 36)] or decreases in the inhibitors of PFK (citrate and ATP). Either the changes in these metabolites necessary to activate PFK are too small to be detected with our techniques or insulin activates PFK by some other mech- anism(s).

One of the dramatic effects of starvation was what appeared to be a complete inhibition of insulin-mediated glycolysis (no change in 3-carbon post-PFK intermedi- ates or CHO oxidation). Because the concentrations of fructose 6-phosphate and ATP were similar between treatments during the clamp, the low glycolytic rate after the 72-h fast is probably due to inadequate insulin- mediated activation of PFK.

CHO oxidation. Starvation resulted in a significant decrease in the basal rate of CHO oxidation. One possible explanation for this is that the increased reliance on fat oxidation resulted in an increase in acetyl-CoA, judged from the increase in acetylcarnitine (2, 11). This should result in an activation of pyruvate dehydrogenase kinase and therefore an inhibition of pyruvate dehydrogenase (14). In addition, a decrease in the availability of post- PFK three-carbon intermediates (i.e., pyruvate), due to a low glycolytic flux may also have contributed to the low rate of CHO oxidation.

After the 10-h fast, insulin infusion did not alter the content of citrate in muscle. However, there was a sig- nificant increase in malate. The increase in malate may be due to an increased activation of anaplerotic reactions (i.e., increased formation of TCA cycle intermediates), which are dependent on the availability of three-carbon intermediates such as pyruvate. The increases in pyru- vate and alanine, coupled with the decrease in glutamate (during the clamp after the 10-h fast) are consistent with an increased flux through the alanine aminotransferase reaction, resulting in the production of 2oxoglutarate. 2Oxoglutarate can then be used to increase malate and other TCA cycle intermediates (32,33). It is also possible that increased flux through the pyruvate carboxylase reaction could have contributed to the increase in malate and could have maintained levels of citrate. The activity of pyruvate carboxylase is probably quite low in human

E604 FASTING AND CARBOHYDRATE METABOLISM IN MUSCLE

skeletal muscle, but because the postclamp biopsy was taken 2 h after the basal biopsy, flux through pyruvate carboxylase, or other anaplerotic reactions, may have been sufficient to contribute to the increase in malate.

The finding that citrate decreased during the clamp after starvation indicates that insulin activated the TCA cycle. It has been previously documented that insulin activates the TCA cycle and respiration in mitochondria from diaphragm and heart muscle, as well as liver (4, 15, 27). [We did not measure a significant increase in whole body oxygen consumption. However, a major increase, even in the affected tissue, is not expected (4).]

The decrease in citrate during the clamp after starva- tion may be consequential to lack of insulin-mediated anaplerosis or decreased availability of acetyl-CoA within the mitochondria. In either case, this could be attributed to inadequate availability of three-carbon in- termediates. Consistent with this explanation are recent findings on two patients with PFK deficiency in skeletal muscle. During euglycemic hyperinsulinemia (400 mu. me2 .min-‘) there were decreases in acetylcarnitine and citrate in muscle. There were no increases in muscle pyruvate or lactate (20). We have recently described other conditions (exercise and epinephrine infusion) where apparently increased glycolytic flux is required for expansion of TCA cycle intermediates (32,33). Thus the present data suggest that starvation attenuates insulin- mediated stimulation of CHO oxidation and anaplerosis in muscle.

Glycogen synthase. Starvation resulted in a small but significant decrease in the activity of GShigh. This has also been observed in the soleus muscle of mouse after 48 h of starvation (24) and may be related to an increase in muscle protein catabolism (12). Starvation also re- sulted in a significant decrease in GSF in the basal state. The mechanism for this is not clear. It may be due to changes in the activities of the phosphatases and kinases that control the phosphorylation state of GS and/or changes in the substrate suitability (i.e., affinity) of GS for its phosphotransferases. Indeed it has been shown that 24 h of fasting results in a decreased activity of GS phosphatase and probably a decrease in the substrate suitability of endogenous GS for GSP in rat heart (10). The present data, although limited, support the idea that a decrease in GSP specific activity can at least partly explain the decrease in GSF in the basal state after the 72-h fast. Moreover, the data also indicate that this decrease in GSF is not due to an increased activation state of CAMP-PK. It might be argued that the decrease in GSF is solely a consequence of the small decrease in plasma insulin. We recently showed that, when glucose infusion is superimposed on somatostatin infusion (in- creasing plasma glucose from -4 to 26 mM), plasma insulin rises from 1 to 7 pU/ml (i.e., the increase is similar to the decrease observed after starvation), but there is no significant change in GSF (19). These findings suggest that the decrease in GSF is not due simply to acute effects by small changes in insulin.

Interestingly, GSF and GSI,, increased to the same values in the two clamps and to larger extents during the clamp after the 72-h fast. These data indicate that the activation of muscle GS by insulin is not a consequence

of increases in glucose uptake and/or utilization (24). We could not associate the increase in GSF during the clamp with increases in GSP or decreases in CAMP-PK activity. It should be noted that the clamp biopsies were taken 2 h after onset of insulin infusion, and thus the presumably early changes in the kinase and phosphatase activities were not detected (21, 23).

In conclusion, these data demonstrate that starvation causes bidirectional changes in insulin-mediated activa- tion of enzymes and metabolic pathways associated with CHO metabolism in human skeletal muscle. Specifically, starvation attenuates insulin-mediated glucose disposal, glycolysis, CHO oxidation, and anaplerosis but enhances activation of GS.

We thank the nursing, dietary, and technical staff of the Clinical Diabetes and Nutrition Section, National Institute of Diabetes and Digestive and Kidney Diseases, for their contributions and Deb Shilts for secretarial assistance.

This research was supported by Arizona Department of Health Services Grant 827-OOOOOO-l-0-YD-8358.

Address for reprint requests: A. Katz, Dept. of Clinical Physiology, Karolinska Institute, Karolinska Hospital, Box 60500, S-104 01 Stock- holm, Sweden.

Received 22 March 1991; accepted in final form 16 July 1991.

REFERENCES

1.

2.

3.

4.

5.

6.

7.

8.

9.

10.

11.

12.

ANDRES, R., G. CADER, AND K. L. ZIERLER. The quantitatively minor role of carbohydrate in oxidative metabolism by skeletal muscle in intact man in the basal state. Measurements of oxygen and glucose uptake and carbon dioxide and lactate production in the forearm. J. CZin. Inuest. 34: 671-682, 1956. BERGER, M., S. A. HAGG, M. N. GOODMAN, AND N. B. RUDERMAN. Glucose metabolism in perfused skeletal muscle. Effects of star- vation, diabetes, fatty acids, acetoacetate, insulin and exercise on glucose uptake and disposition. Biochem. J. 158: 191-202,1976. BERGMEYER, H. U. (Editor). Methods of Enzymatic Analysis. New York: Academic, 1974. BESSMAN, S. P., C. MOHAN, AND I. ZAIDISE. Intracellular site of insulin action: mitochondrial Krebs cycle. Proc. N&Z. Acad. Sci. USA 83: 5067-5070,1986. BJORKMAN, O., AND L. S. ERIKSSON. Influence of a 60-hour fast on insulin-mediated splanchnic and peripheral glucose metabolism in humans. J. Clin. Invest. 76: 87-92, 1985. BRADY, L. J., M. N. GOODMAN, M. N. KALISH, AND N. B. RUD- ERMAN. Insulin binding and sensitivity in rat skeletal muscle: effect of starvation. Am. J. Physiol. 240 (Endocrinol. Metub. 3): El84- E190,1981. CHARRON, M. J., AND B. B. KAHN. Divergent molecular mecha- nisms for insulin-resistant glucose transport in muscle and adipose cells in vivo. J. BioZ. Chem. 265: 7994-8000, 1990. CHASIOTIS, D., AND E. HULTMAN. The effect of circulatory occlu- sion on the glycogen-phosphorylase-synthetase system in human skeletal muscle. J. Physiol. Lond. 345: 167-173, 1983. DEFRONZO, R. A., R. GUNNARSON, 0. BJ~RKMAN, M. OLSSON, AND J. WAHREN. Effects of insulin on peripheral and splanchnic glucose metabolism in noninsulin dependent (type II) diabetes mellitus. J. CZin. Inuest. 76: 149-155, 1985. GANNON, M. C., A. W. TAN, AND F. W. NUTTALL. Effect of starvation and insulin treatment on glycogen synthase D and synthase D phosphatase activity in rat heart. MOL. CeZZ. Biochem. 34: 31-34, 1981. GOODMAN, M. N., M. BERGER, AND N. B. RUDERMAN. Glucose metabolism in rat skeletal muscle at rest. Effect of starvation, diabetes, ketone bodies, and free fatty acids. Diabetes 23: 881-888, 1974. GOODMAN, M. N., M. A. MCELANEY, AND N. B. RUDERMAN. Adaptation to prolonged starvation in the rat: curtailment of skeletal muscle proteolysis. Am. J. Physiol. 241 (Endocrinol. Metub. 4): E321-E327,1981.

FASTING AND CARBOHYDRATE METABOLISM IN MUSCLE El605

13.

14.

15.

16.

17.

18.

19.

20.

HARRIS, R. C., E. HULTMAN, AND L.-O. NORDESJ~. Glycogen, glycolytic intermediates and high energy phosphates determined in biopsy samples of musculus quadriceps femoris of man at rest. Methods and variance of values. Stand. J. CZin. Lab. Invest. 33: 109-120,1974. HENNIG, G., G. L~FFLER, AND 0. H. WIELAND. Active and inactive forms of pyruvate dehydrogenase in skeletal muscle as related to the metabolic and functional state of the muscle cell. FEBS Lett. 59: 142-145, 1975. IDA, T., M. SATO, Y. YAMAOKA, H. TAKEDA, Y. KAMIYAMA, K. KIMURA, K. OZAWA, AND I. HONJO. Effect of insulin on mitochon- drial oxidative phosphorylation and energy charge of the perfused guinea pig liver. J. Lab. Clin. Med. 87: 925-933, 1976. KATZ, A., AND C. BOGARDUS. Relationship between carbohydrate oxidation and G-1,6-& in human skeletal muscle during euglycemic hyperinsulinemia. Am. J. Physiol. 260 (Regulatory Integrative Comp. Physiol. 29): R113-R119, 1991. KATZ, A., S. BROBERG, K. SAHLIN, AND J. WAHREN. Leg glucose uptake during maximal dynamic exercise in humans. Am. J. Phys- iol. 251 (Endocrinol. Metab. 14): E65-E70, 1986. KATZ, A., B. L. NYOMBA, AND C. BOGARDUS. Euglycemic hyper- insulinemia increases glucose 1,6-bisphosphate in human skeletal muscle. Int. J. Biochem. 21: 1079-1082, 1989. KATZ, A., I. RAZ, M. K. SPENCER, R. RISING, AND D. M. MOTT. Hyperglycemia induces accumulation of glucose in human skeletal muscle. Am. J. Physiol. 260 (Regulatory Integrative Comp. Physiol. 29): R698-R703,1991. KATZ, A., M. K. SPENCER, S. LILLIOJA, Z. YAN, D. M. MOTT, R. G. HALLER, AND S. F. LEWIS. Basal and insulin-mediated carbo- hydrate metabolism in human muscle deficient in phosphofructo- kinase 1. Am. J. Physiol. 261 (Endocrinol. Metab. 24): EOOO-EOOO, 1991.

21.

22.

23.

KIDA, Y., A. ESPOSITO-DEL PUENTE, C. BOGARDUS, AND D. M. MOTT. Insulin resistance is associated with reduced fasting and insulin-stimulated glycogen synthase phosphatase activity in hu- man skeletal muscle. J. CZin. Inuest. 85: 476-481, 1991. KIDA, Y., A. KATZ, A. D. LEE, AND D. M. MOTT. Contraction- mediated inactivation of glycogen synthase is accompanied by inactivation of glycogen synthase phosphatase in human skeletal muscle. Biochem. J. 259: 901-904, 1989. KIDA, Y., B. L. NYOMBA, C. BOGARDUS, AND D. M. MOTT. Defec- tive insulin response of cyclic adenosine monophosphate-depend- ent protein kinase in insulin-resistant humans. J. Clin. Invest. 87: 673-679,199l.

24. LE MARCHAND-BRUSTEL, Y., AND P. FREYCHET. Effect of fasting

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

and streptozotocin diabetes on insulin binding and action in the isolated mouse soleus muscle. J. CZin. Inuest. 64: 1505-1515, 1979. LILLIOJA, S., C. BOGARDUS, D. M. MOTT, A. L. KENNEDY, W. C. KNOWLER, AND B. V. HOWARD. Relationship between insulin- mediated glucose disposal and lipid metabolism in man. J. CZin. Invest. 75: 1106-1115, 1985. LOWRY, 0. H., AND J. V. PASSONEAU. A Flexible System of Enzy- matic Analysis. New York: Academic, 1972. MCKEE, E. E., AND B. L. GRIER. Insulin stimulates mitochondrial protein synthesis and respiration in isolated perfused rat heart. Am. J. Physiol. 259 (Endocrinol. Metab. 22): E413-E421, 1990. OWEN, 0. E., AND G. A. REICHARD, JR. Human forearm metabo- lism during progressive starvation. J. CZin. Invest. 50: 1536-1545, 1971. POZEFSKY, T., R. G. TANCREDI, R. T. MOXLEY, J. DUPRE, AND J. D. TOBIN. Metabolism of forearm tissues in man: studies with glucagon. Diabetes 25: 128-135, 1976. RANDLE, P. J., E. A. NEWSHOLME, AND P. B. GARLAND. Regula- tion of glucose uptake by muscle. 8. Effects of fatty acids, ketone bodies and pyruvate and of alloxan diabetes and starvation on the uptake and metabolic fate of glucose in rat heart and diaphragm muscles. Biochem. J. 93: 652-665, 1964. RAZ, I., A. KATZ, AND M. K. SPENCER. Epinephrine inhibits insulin-mediated glycogenesis but enhances glycolysis in human skeletal muscle. Am. J. Physiol. 260 (Endocrinol. Metab. 23): E430- E435,1991. SAHLIN, K., A. KATZ, AND S. BROBERG. Tricarboxylic acid cycle intermediates in human muscle during prolonged exercise. Am. J. Physiol. 259 (Cell Physiol. 28): C834C841, 1990. SPENCER, M. K., A. KATZ, AND I. RAZ. Epinephrine increases tricarboxylic acid cycle intermediates in human skeletal muscle. Am. J. Physiol. 260 (Endocrinol. Metab. 23): E436-E439, 1991. STEELE, R. Influences of glucose loading and of injected insulin on hepatic glucose output. Ann. NY Acad. Sci. 82: 420-430, 1959. VAN SCHAFTINGEN, E., B. LEDERER, R. BARTRONS, AND H.-G. HERS. A kinetic study of pyrophosphate: fructose-6-phosphate phosphotransferase from potato tubers. Application to a microas- say of fructose 2,6-bisphosphate. Biochem. J. 129: 191-195, 1982. VEECH, R. L., J. W. R. LAWSON, N. W. CORNELL, AND H. A. KREBS. Cytosolic phosphorylation potential. J. Biol. Chem. 254: 6538-6547,1979. ZIERLER, K. L., AND D. RABINOWITZ. Roles of insulin and growth hormone, based on studies of forearm metabolism in man. Medicine Baltimore 42: 385-402, 1963.