- Email: [email protected]
Regulatory role of glucose-6 phosphatase in the repletion of liver glycogen during refeeding in fasted rats
Biochimica et Biophysica Acta 1452 (1999) 172^178 www.elsevier.com/locate/bba
Regulatory role of glucose-6 phosphatase in the repletion of liver glycogen during refeeding in fasted rats Carol Minassian, Sandrine Montano, Gilles Mithieux * INSERM U. 449, Faculte¨ de Me¨decine Laennec, rue Guillaume Paradin, 69372 Lyon, Ce¨dex 08, France Received 14 April 1999; received in revised form 19 August 1999; accepted 27 September 1999
Abstract We analyzed the biochemical mechanisms involved in the liver glycogen repletion upon refeeding for 360 min in 48 and 96 h-fasted rats. In 48 h-fasted rats, the glycogen synthesis involved a rapid and further sustained induction of glucokinase (GK) (increased twice from 90 min) and a rapid but transient activation of glycogen synthase a (GSa) (maximal increase by 150% at 90 min). It did not involve the inhibition of glycogen phosphorylase a (GPa). In 96 h-fasted rats, the glycogen repletion did not involve the induction of GK for the first 180 min of refeeding. It involved a slow activation of GSa (maximal 150% increase at 180 min) and a rapid inhibition of GPa (significant from 90 min, maximal 50% inhibition by 180 min). In both groups of rats, there was a progressive inhibition of the glucose-6 phosphatase (Glc6Pase) activity (maximal suppression by 30% in both groups at 360 min). These results highlighted a key role for the inhibition of Glc6Pase activity in the liver glycogen repletion upon refeeding. ß 1999 Elsevier Science B.V. All rights reserved. Keywords: Liver; Glycogen; Glucose-6 phosphatase; Glucokinase; Glycogen synthase; Glycogen phosphorylase
1. Introduction Liver glucose-6 phosphatase (Glc6Pase) is a key enzyme in glucose homeostasis, since it catalyzes the last biochemical reaction of glycogenolysis and gluconeogenesis [1,2]. Because the Km of the enzyme in vitro (in the millimolar range) far exceeds the concentration of glucose-6 phosphate (Glc6P) in the liver (50^200 nmol/g wet weight), it has long been believed that only the variations in Glc6P concentrations control the substrate £ux through Glc6Pase [3]. HowAbbreviations: Glc6Pase, glucose-6 phosphatase; GK, glucokinase; HK, hexokinase; GSa, glycogen synthase a; GPa, glycogen phosphorylase a; Glc6P, glucose-6 phosphate * Corresponding author. Fax: +33 (4) 78 77 87 62; E-mail: [email protected]
ever, recent evidence has accumulated that this enzyme, independently of its substrate, may play a crucial regulatory role in the liver glucose release in various situations, via mechanisms involving either variations in gene expression [4^6] or modulations of its enzymatic activity [7^10]. In this regard, the ¢rst indication that the inhibition of Glc6Pase activity might play a role in the suppression of hepatic glucose production and in the liver glycogen repletion after refeeding has been suggested by more than 10 years ago [11]. Our laboratory has then brought up the ¢rst direct evidence that an inhibition of Glc6Pase activity occurs during the post-prandial period in fasted rats [12]. More recently, we have emphasized the putative role that insulin might play in this phenomenon, via the activation of the phosphatidylinositol-3 kinase pathway [13,14]. During fast-
0167-4889 / 99 / $ ^ see front matter ß 1999 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 4 8 8 9 ( 9 9 ) 0 0 1 3 0 - 5
BBAMCR 14554 27-10-99
C. Minassian et al. / Biochimica et Biophysica Acta 1452 (1999) 172^178
ing, after a prior 48 h phase of glycogen depletion, we previously reported that glycogen stores rebound in the liver from the third day and that this could be due to a di¡erential control of the basal Glc6Pase £ux [9]. This might also suggest that the liver glycogen metabolism is di¡erent in both situations. To get further insight into the role of the inhibition of Glc6Pase in the liver glycogen storage, we have studied here the biochemical mechanisms involved during a dynamic phase of repletion of glycogen stores, i.e. during the refeeding in 48 and 96 h-fasted rats. 2. Materials and methods 2.1. Refeeding experiments Male Sprague-Dawley rats (IFFA CREDO, L'Arbresle, France) weighing 180^200 g were housed for 3 days with free access to water and non-puri¢ed chow diet [12]. Food was then withheld for a period of 48 or 96 h. Unfed rats were refed with the same non-puri¢ed chow diet for 90, 180 or 360 min. Rats were given free access to water throughout the experiment. They were anaesthetized with a single injection of pentobarbital (70 mg/kg body weight). The abdomen was rapidly incised to expose the liver which was rapidly freeze-clamped at liquid nitrogen temperature. The rest of the liver was removed and weighed. The frozen liver was stored at 380³C and used for the determination of metabolite concentrations and of enzymatic activities. Blood was collected for the assay of plasma glucose, insulin and glucagon. All animals used in this study were handled according to the rules de¢ned by our local ethic committee for animal experimentation. Anaesthetized rats were rapidly killed by heart excision after liver and blood sampling. 2.2. Assay of enzymes and metabolites Small pieces of freeze-clamped liver (to a total mass of 100^200 mg) were taken randomly, reduced in powder at liquid N2 temperature and homogenized rapidly by sonication in 0.25 M sucrose, 10 mM HEPES, pH 7.3. Speci¢c Glc6Pase activity was immediately determined as previously described [12,15].
173
Glucokinase (GK) and hexokinase (HK) were assayed in 10 000Ug supernatants of liver homogenates by the colorimetric method of Bontemps et al. [16]. Glycogen synthase a (GSa) was determined in total liver homogenates by the incorporation of UDP[U-14 C] glucose in glycogen [17]. Glycogen phosphorylase a (GPa) was assayed in whole liver homogenates in the direction of glycogen synthesis as described by Hue et al. [18]. The results are expressed as the total enzyme activities in Wmol of substrate transformed per min per g of wet liver at 30³C. Glc6P was measured according to Lang and Michal [19]. Glycogen was determined as described by Keppler and Decker [20]. Protein was assayed by the Lowry method [21] with bovine serum albumin as a standard. 2.3. Determination of plasma hormones and metabolites Plasma immunoreactive insulin and glucagon were assayed by radioimmunoassay [22,23]. Plasma glucose was determined according to Bergmeyer et al. [24]. Statistical analysis was performed by ANOVA. Statistical di¡erences were tested using the t test of Student-Fisher for unpaired data. 3. Results The e¡ects of refeeding on the evolution of body and liver weights and on the concentration of protein in the liver are given in Table 1. There was a progressive increase in both the body and liver weights as refeeding was ongoing. These weight gains were very similar in both groups: +12% and +11% for body weight gains after 360 min in 48 and 96 hfasted rats, respectively, +27% in both groups for liver weight gains, without attaining the pre-fast values, however (Table 1). Refeeding was also characterized by a progressive rehydratation of the liver (due to glycogen repletion, see below), resulting in a progressive decrease in the liver protein concentration. The latter decrease was quite similar in both 48 and 96 h-fasted rats: minus 17% at the end of the refeeding period (Table 1). This warranted the validity of the comparison between 48 and 96 h-fasted rats relating to the evolution of other hepatic param-
BBAMCR 14554 27-10-99
174
C. Minassian et al. / Biochimica et Biophysica Acta 1452 (1999) 172^178
Table 1 E¡ect of fasting and refeeding on body weight, liver weight and liver protein concentration Refeeding time (min) 0 90 180 360
Body weight (g)
Liver weight (g)
Liver protein concentration (mg/g)
Fed
48 h fast
96 h fast
Fed
48 h fast
96 h fast
Fed
48 h fast 96 h fast
186 þ 3.0
151 þ 1.3 164 þ 1.6a 164 þ 2.1a 169 þ 2.1a,b
136 þ 3.6 145 þ 2.4 151 þ 2.4a 151 þ 2.4a,b
8.4 þ 0.3
5.25 þ 0.2 5.37 þ 0.3 5.63 þ 0.2 6.68 þ 0.2a,b
4.33 þ 0.1 4.72 þ 0.2 4.95 þ 0.2a 5.53 þ 0.2a,b
143 þ 7
177 þ 8 160 þ 6 156 þ 6 147 þ 5a
183 þ 4 171 þ 5 156 þ 8a 152 þ 9a
Results are the means þ S.E.M. (n = 6). Signi¢cantly di¡erent from the value at time 0, P 6 0.05. b Signi¢cantly di¡erent from fed value, P 6 0.05. a
eters. The liver protein concentrations after the refeeding period were not di¡erent from the pre-fast values determined in fed rats (Table 1). Both groups of unfed rats had comparable very low plasma glucose (about 6 mM) and insulin (about 90 pM) before refeeding. Plasma glucose was increased to about 11 mM and plasma insulin to about 350 pM in both 48 and 96 h-fasted rats from 90 min of refeeding. Plasma glucose then tended to decrease at the end of the refeeding period (to about 9 mM) whereas plasma insulin remained constant throughout this time. In contrast, plasma glucagon (about 300 ng/l in both groups of fasted rats) was not altered upon refeeding in 96 h-fasted rats, whereas a weak (by about 25%) but signi¢cant (P 6 0.05) decrease was observed at the end of refeeding in the 48 h-fasted group. As a consequence, the insulin/ glucagon ratio was signi¢cantly increased about three times by refeeding and was not further signi¢cantly altered as refeeding was ongoing (data not shown). The liver glycogen content of 48 h-fasted rats was very low before refeeding. In contrast, the liver of 96 h-fasted rats contained a three times higher amount of glycogen: 4.6 þ 0.2 vs. 1.4 þ 0.5 mg/g net weight, means þ S.E.M., n = 6 (Fig. 1). This was in good agreement with our previous results [9]. Upon refeeding, the repletion of glycogen stores was almost linear in both groups up to 180 min. It was however less rapid in 96 h-fasted rats than in 48 h-fasted rats since the liver glycogen content in the former group was 18% lower than that in the latter at 180 min (Fig. 1). Glycogen repletion was less rapid in the second phase of refeeding (180^360 min). There
were no longer di¡erences in the glycogen stores in both groups at the end of the refeeding period. The glycogen contents attained at this time were quite similar to those observed in fed rats (Fig. 1). The basal liver Glc6P concentration was signi¢cantly higher in 96 h-fasted rats than in 48 h-fasted rats (Fig. 1). This was in agreement with previous data [9]. The Glc6P contents tended to be weakly increased in both groups in the course of refeeding, but a signi¢cant 70% higher Glc6P content in comparison with basal Glc6P was only observed in the 48 h-fasted group at the latest time of refeeding (Fig. 1). The ¢nal Glc6P contents tended to be lower (not statistically signi¢cant) than those in rats prior to fasting. In keeping with our previous results [12], the liver Glc6Pase speci¢c activity (about 12 Wmol/min/g before refeeding) was progressively inhibited during the refeeding period (to about 8 Wmol/min/g). The inhibition was however 17% less in the 96 h-fasted group than in the 48 h-fasted group at 180 min (Fig. 1). On the other hand, there was no more di¡erence at the end of the refeeding experiment. Noteworthy, the Glc6Pase activity at the end of the refeeding period was in both groups very close to Glc6Pase activity determined in fed rats. The GK activity was rapidly induced twice at the ¢rst time of refeeding in 48 h-fasted rats and was further induced at later times (Fig. 1). This was in agreement with previous published data [25]. In contrast, GK was not induced for 180 min and was only induced twice at the end of the refeeding period in the 96 h-fasted group (Fig. 1). The ¢nal GK activity in both groups was approximately twice less than the
BBAMCR 14554 27-10-99
C. Minassian et al. / Biochimica et Biophysica Acta 1452 (1999) 172^178
175
Fig. 1. E¡ect of fasting and refeeding on the liver metabolites and enzymes involved in glycogen metabolism. All values are expressed by g of wet liver. The results are expressed as the means þ S.E.M. of six animals at each time. The symbol * means that the value is signi¢cantly di¡erent from the corresponding fasting value (P 6 0.05). Signi¢cant di¡erences between the 48 and 96 h values are indicated in the di¡erent panels where appropriate.
GK activity determined prior to fasting. HK was not signi¢cantly altered by fasting and at any time throughout the refeeding, irrespective of the prior fasting time. GSa was rapidly activated by 150% from the ¢rst refeeding time in 48 h-fasted rats. However, this activation e¡ect was transient since GSa then progressively fell to a level close to its starting basal fasting level (Fig. 1). In contrast, the activation of GS was slow in 96 h-fasted rats since a signi¢cant increase in the GSa activity was only reached after 180 min of
refeeding. The GSa activity then plateaued up to the end of refeeding (Fig. 1). The values at the end of refeeding in both groups remained signi¢cantly different from that in fed rats. The GPa activity was not altered upon refeeding in 48 h-fasted rats. In marked contrast, the activity of GPa was rapidly decreased by 30% from 90 min in the 96 h-fasted group and by 50% after 180 min. The GPa-inhibited state was then maintained up to 360 min (Fig. 1). The ¢nal GPa activities were both different from the activity determined in fed rats.
BBAMCR 14554 27-10-99
176
C. Minassian et al. / Biochimica et Biophysica Acta 1452 (1999) 172^178
Fig. 2. Linear relationship between the glycogen content and the Glc6Pase activity during refeeding in both 48 and 96 hfasted rats.
4. Discussion In the present work, we have examined the key biochemical parameters involved in the repletion of liver glycogen in 48 and 96 h-fasted rats. In apparent discrepancy with the hypothesis underlying this work, i.e. that the intrahepatic mechanisms might be di¡erent in both types of fasted rats, there was only a subtle di¡erence in the time course of repletion which was slightly delayed in 96 h-fasted as compared to 48 h-fasted rats. However, in agreement with the hypothesis, the biochemical mechanisms involved in the repletion exhibited marked di¡erences. In 48 h-fasted rats, glycogen synthesis involved a rapid and further sustained induction of GK and a progressive and lasting inhibition of Glc6Pase activity. In contrast, it involved a rapid, but only transient, activation of GS, since the GSa activity was not di¡erent from the basal GSa activity after 6 h of refeeding. At this time, the increased concentration of Glc6P might play the key role in the glycogen repletion (`push' mechanism). On the other hand, the rapid activation of GSa at 90 min is in agreement with a `pull' mechanism operating at this time, as previously suggested to occur during the early phase
of glycogen repletion after glucose gavage in 20 hfasted rats [11]. The GPa activity was not decreased throughout the refeeding period. These results thus strongly suggest that the activation of GS is not necessarily dependent on the prior inhibition of GP. This is at variance with the previously suggested mechanism of liver glycogen synthesis in rodents [26], but in keeping with recent results in human [27]. The rapid induction of GK also suggests that a signi¢cant part of glycogen deposited in the liver during refeeding in 48 h-fasted rats may arise from a direct pathway, i.e. from glucose to glycogen, in addition to that arising from gluconeogenic origin, namely the indirect pathway, e.g. from lactate to glycogen [28,29]. In 96 h-fasted rats, as in their 48 h counterparts, the repletion of glycogen stores involved the inhibition of Glc6Pase. In contrast, it did not involve the induction of GK for the ¢rst 180 min of refeeding. A signi¢cant increase in GK activity was observed after 360 min only. Noteworthy, the time course of activation of GS was signi¢cantly lenghthened with regards to that of the 48 h-fasted group. The most striking di¡erence was related to the e¡ect of refeeding on the GPa activity. At variance with the absence of an e¡ect in the 48 h group, the GPa was rapidly and substantially inhibited upon refeeding. It is interesting to note that the absence of induction of GK for 3 h strongly suggests that the bulk of glycogen stored during the early period of refeeding in 96 hfasted rats likely arises from the indirect pathway. It has been suggested that intrahepatic translocation mechanisms of GK and GS may occur in response to the metabolic changes accompanying the fasted to fed transition and participate to the regulation of the glycogen repletion in this situation. On the one hand, GK has been reported to translocate from the nucleus to the cytosol in response to glucose [30,31]. These changes are only observable in intact cells and not in liver homogenates in which GK is only found in a soluble form. Since GK is absent in the livers from both 48 and 96 h-fasted rats [9] and since we assayed GK in 10 000Ug supernatants from liver homogenates, the rapid increase in the liver GK activity in 48 h-fasted rats after refeeding may be attributable to a rapid induction in the GK gene expression, which does not occur in 96 h-fasted rats. On the other hand, GS has been reported to
BBAMCR 14554 27-10-99
C. Minassian et al. / Biochimica et Biophysica Acta 1452 (1999) 172^178
translocate from the cytosol to the cell cortex in response to glucose in isolated hepatocytes [32] or to aggregate into clusters which, however, do not locate with glycogen particles, during the fasted to refed transition in vivo [33]. In contrast, the prior phases of glycogen synthesis have far been demonstrated to take place in close contact with the endoplasmic reticulum in the liver in vivo [34]. In our opinion, the extent to which the one or the other of these translocation mechanisms may participate to the regulation of glycogen repletion in vivo is yet to be established. So, we considered and studied the activation of total GSa in whole liver homogenates. It seems therefore obvious that there exists some marked divergences between the biochemical mechanisms of glycogen repletion in both types of fasted rats. The di¡erences are particularly related to the time course of induction of GK and of activation of GS and to the occurrence or not of the inhibition of GP. On the other hand, there is a parallelism in the inhibitions of Glc6Pase activity as the refeeding was ongoing. This inhibition is slightly less rapid in 96 h-fasted rats, since Glc6Pase activity remains by 17% higher than that of their 48 h-fasted counterparts at 180 min. Interestingly, the liver glycogen content at this time of refeeding is also by 18% less in the 96 h-fasted group than in the 48 h-fasted group. Taken as a whole, these results strongly suggest that the inhibition of Glc6Pase activity plays an important role in the repletion of liver glycogen during refeeding in the fasted states. Noteworthy, studying the mechanisms of glycogen repletion by various approaches in rats fasted for a shorter time than those studied here (e.g. 20 h), McGarry and co-workers previously inferred that an inhibition of Glc6Pase activity should occur during the repletion of liver glycogen [11,28,29]. The evidence obtained was indirect. The data provided herein bring up direct evidence that such an inhibition actually occurs and plays a key role in the glycogen repletion in rats fasted for various times and exhibiting markedly different capacities in terms of glycogen metabolism. The most astonishing result has appeared to be that this role may be as crucial (maybe more) as that of the variations in GS, GP and GK activities. In agreement with the latter proposal, considering both groups together, there exists a tight correlation between the glycogen content and the Glc6Pase ac-
177
tivity at a given time: R2 = 0.98, P 6 1034 (Fig. 2). The mean amount of glycogen deposited between two time points is also strongly correlated with the mean Glc6Pase activity observed during the same time interval (not shown). On the other hand, the glycogen content is less tightly correlated with the Glc6P concentration and the GK activity (R2 = 0.85, P 6 0.01 and R2 = 0.82, P 6 0.05, respectively), whereas there does not exist any correlation between the glycogen content and either the GSa or the GPa activities (not shown). The results presented herein thus further document in vivo the crucial regulatory role of Glc6Pase in liver glucose metabolism during the post-prandial period.
References [1] J.D. Foster, B.A. Pederson, R.C. Nordlie, Proc. Soc. Exp. Biol. Med. 215 (1997) 314^332. [2] G. Mithieux, Eur. J. Endocrinol. 136 (1997) 137^145. [3] L. Hue, H.G. Hers, Ann. Rev. Biochem. 52 (1983) 617^653. [4] B.A. Haber, S. Chin, E. Chuang, W. Buikhuisen, A. Naji, R. Taub, J. Clin. Invest. 95 (1995) 832^841. [5] C. Minassian, C. Zitoun, G. Mithieux, Mol. Cell. Biochem. 155 (1996) 37^41. [6] G. Mithieux, H. Vidal, C. Zitoun, N. Bruni, N. Daniele, C. Minassian, Diabetes 42 (1996) 1614^1620. [7] A.M. Bode, J.D. Foster, R.C. Nordlie, J. Biol. Chem. 267 (1992) 2860^2863. [8] L.B. Gardner, Z. Liu, E.J. Barrett, Diabetes 42 (1993) 1614^ 1620. [9] C. Minassian, A. Ajzannay, J.P. Riou, G. Mithieux, J. Biol. Chem. 269 (1994) 16585^16588. [10] C. Minassian, S. Tarpin, G. Mithieux, Biochem. Pharmacol. 55 (1998) 1213^1219. [11] C.B. Newgard, D.W. Foster, J.D. McGarry, Diabetes 33 (1984) 192^195. [12] C. Minassian, N. Daniele, J.C. Bordet, C. Zitoun, G. Mithieux, J. Nutr. 125 (1995) 2727^2732. [13] G. Mithieux, N. Daniele, B. Payrastre, C. Zitoun, J. Biol. Chem. 273 (1998) 17^19. [14] N. Daniele, F. Rajas, B. Payrastre, G. Mauco, C. Zitoun, G. Mithieux, J. Biol. Chem. 274 (1999) 3597^3601. [15] F. Chatelain, J.P. Pegorier, C. Minassian, N. Bruni, S. Tarpin, J. Girard, G. Mithieux, Diabetes 47 (1998) 882^889. [16] F. Bontemps, L. Hue, H.G. Hers, Biochem. J. 174 (1978) 603^611. [17] J.A. Thomas, K.K. Schlender, J. Larner, Anal. Biochem. 25 (1968) 486^489. [18] L. Hue, F. Bontemps, H.G. Hers, Biochem. J. 152 (1975) 105^114. [19] G. Lang, G. Michal, in: H.U. Bergmeyer (Ed.), Methods of
BBAMCR 14554 27-10-99
178
[20]
[21] [22] [23]
[24]
[25] [26]
C. Minassian et al. / Biochimica et Biophysica Acta 1452 (1999) 172^178 Enzymatic Analysis, vol. 3, VCH Publishers, New York, 1974, pp. 1238^1242. Keppler, D., Decker, C. (1974) in: H.U. Bergmeyer (Ed.), Methods of Enzymatic Analysis, vol. 3, VCH Publishers, New York, pp. 1127^1131. O.H. Lowry, N.J. Rosebrough, A.R. Farr, R.J. Randall, J. Biol. Chem. 193 (1951) 265^275. C.M. Hales, C.A. Randle, Biochem. J. 88 (1963) 137^138. J. Harris, G.R. Faloona, R.H. Unger, in: B.M. Jo¡e, H.R. Behrman (Eds.), Methods of Hormone Immunoassay, Academic Press, San Diego, CA, 1979, pp. 643^671. H.U. Bergmeyer, E. Bernt, F. Schmidt, H. Stork, in: H.U. Bergmeyer (Ed.), Methods of Enzymatic Analysis, vol. 3, VCH Publishers, New York, 1974, pp. 1196^1201. F. Eilers, H. Bartels, K. Jungerman, Histochemistry 99 (1993) 133^140. W. Stalmans, H. DeWulf, L. Hue, H.G. Hers, Eur. J. Biochem. 41 (1974) 117^134.
[27] K.F. Petersen, D. Laurent, D.L. Rothman, G.W. Cline, G.I. Schulman, J. Clin. Invest. 101 (1998) 1203^1209. [28] C.B. Newgard, S.V. Moore, D.W. Foster, J.D. McGarry, J. Biol. Chem. 259 (1984) 6958^6963. [29] M. Kuwajima, C.B. Newgard, D.W. Foster, J.D. McGarry, J. Clin. Invest. 74 (1984) 1108^1111. [30] Y. Toyoda, I. Miwa, S. Satake, M. Anai, Y. Oka, Biochem. Biophys. Res. Commun. 215 (1995) 467^473. [31] K.S. Brown, S.S. Kalinowsky, J.R. Megill, S.K. Durham, K.A. Mookhtiar, Diabetes 46 (1977) 179^186. [32] J.M. Fernandez-Novell, D. Bellido, J.J. Guinovart, Biochem. J. 321 (1997) 227^231. [33] J.M. Fernandez-Novell, A. Roca, D. Bellido, S. Vilaro, J.J. Guinovart, Eur. J. Biochem. 238 (1996) 570^575. [34] R.R. Cardell Jr., Int. Rev. Cytol. 48 (1977) 221^229.
BBAMCR 14554 27-10-99
Regulatory role of glucose-6 phosphatase in the repletion of liver glycogen during refeeding in fasted rats Carol Minassian, Sandrine Montano, Gilles Mithieux * INSERM U. 449, Faculte¨ de Me¨decine Laennec, rue Guillaume Paradin, 69372 Lyon, Ce¨dex 08, France Received 14 April 1999; received in revised form 19 August 1999; accepted 27 September 1999
Abstract We analyzed the biochemical mechanisms involved in the liver glycogen repletion upon refeeding for 360 min in 48 and 96 h-fasted rats. In 48 h-fasted rats, the glycogen synthesis involved a rapid and further sustained induction of glucokinase (GK) (increased twice from 90 min) and a rapid but transient activation of glycogen synthase a (GSa) (maximal increase by 150% at 90 min). It did not involve the inhibition of glycogen phosphorylase a (GPa). In 96 h-fasted rats, the glycogen repletion did not involve the induction of GK for the first 180 min of refeeding. It involved a slow activation of GSa (maximal 150% increase at 180 min) and a rapid inhibition of GPa (significant from 90 min, maximal 50% inhibition by 180 min). In both groups of rats, there was a progressive inhibition of the glucose-6 phosphatase (Glc6Pase) activity (maximal suppression by 30% in both groups at 360 min). These results highlighted a key role for the inhibition of Glc6Pase activity in the liver glycogen repletion upon refeeding. ß 1999 Elsevier Science B.V. All rights reserved. Keywords: Liver; Glycogen; Glucose-6 phosphatase; Glucokinase; Glycogen synthase; Glycogen phosphorylase
1. Introduction Liver glucose-6 phosphatase (Glc6Pase) is a key enzyme in glucose homeostasis, since it catalyzes the last biochemical reaction of glycogenolysis and gluconeogenesis [1,2]. Because the Km of the enzyme in vitro (in the millimolar range) far exceeds the concentration of glucose-6 phosphate (Glc6P) in the liver (50^200 nmol/g wet weight), it has long been believed that only the variations in Glc6P concentrations control the substrate £ux through Glc6Pase [3]. HowAbbreviations: Glc6Pase, glucose-6 phosphatase; GK, glucokinase; HK, hexokinase; GSa, glycogen synthase a; GPa, glycogen phosphorylase a; Glc6P, glucose-6 phosphate * Corresponding author. Fax: +33 (4) 78 77 87 62; E-mail: [email protected]
ever, recent evidence has accumulated that this enzyme, independently of its substrate, may play a crucial regulatory role in the liver glucose release in various situations, via mechanisms involving either variations in gene expression [4^6] or modulations of its enzymatic activity [7^10]. In this regard, the ¢rst indication that the inhibition of Glc6Pase activity might play a role in the suppression of hepatic glucose production and in the liver glycogen repletion after refeeding has been suggested by more than 10 years ago [11]. Our laboratory has then brought up the ¢rst direct evidence that an inhibition of Glc6Pase activity occurs during the post-prandial period in fasted rats [12]. More recently, we have emphasized the putative role that insulin might play in this phenomenon, via the activation of the phosphatidylinositol-3 kinase pathway [13,14]. During fast-
0167-4889 / 99 / $ ^ see front matter ß 1999 Elsevier Science B.V. All rights reserved. PII: S 0 1 6 7 - 4 8 8 9 ( 9 9 ) 0 0 1 3 0 - 5
BBAMCR 14554 27-10-99
C. Minassian et al. / Biochimica et Biophysica Acta 1452 (1999) 172^178
ing, after a prior 48 h phase of glycogen depletion, we previously reported that glycogen stores rebound in the liver from the third day and that this could be due to a di¡erential control of the basal Glc6Pase £ux [9]. This might also suggest that the liver glycogen metabolism is di¡erent in both situations. To get further insight into the role of the inhibition of Glc6Pase in the liver glycogen storage, we have studied here the biochemical mechanisms involved during a dynamic phase of repletion of glycogen stores, i.e. during the refeeding in 48 and 96 h-fasted rats. 2. Materials and methods 2.1. Refeeding experiments Male Sprague-Dawley rats (IFFA CREDO, L'Arbresle, France) weighing 180^200 g were housed for 3 days with free access to water and non-puri¢ed chow diet [12]. Food was then withheld for a period of 48 or 96 h. Unfed rats were refed with the same non-puri¢ed chow diet for 90, 180 or 360 min. Rats were given free access to water throughout the experiment. They were anaesthetized with a single injection of pentobarbital (70 mg/kg body weight). The abdomen was rapidly incised to expose the liver which was rapidly freeze-clamped at liquid nitrogen temperature. The rest of the liver was removed and weighed. The frozen liver was stored at 380³C and used for the determination of metabolite concentrations and of enzymatic activities. Blood was collected for the assay of plasma glucose, insulin and glucagon. All animals used in this study were handled according to the rules de¢ned by our local ethic committee for animal experimentation. Anaesthetized rats were rapidly killed by heart excision after liver and blood sampling. 2.2. Assay of enzymes and metabolites Small pieces of freeze-clamped liver (to a total mass of 100^200 mg) were taken randomly, reduced in powder at liquid N2 temperature and homogenized rapidly by sonication in 0.25 M sucrose, 10 mM HEPES, pH 7.3. Speci¢c Glc6Pase activity was immediately determined as previously described [12,15].
173
Glucokinase (GK) and hexokinase (HK) were assayed in 10 000Ug supernatants of liver homogenates by the colorimetric method of Bontemps et al. [16]. Glycogen synthase a (GSa) was determined in total liver homogenates by the incorporation of UDP[U-14 C] glucose in glycogen [17]. Glycogen phosphorylase a (GPa) was assayed in whole liver homogenates in the direction of glycogen synthesis as described by Hue et al. [18]. The results are expressed as the total enzyme activities in Wmol of substrate transformed per min per g of wet liver at 30³C. Glc6P was measured according to Lang and Michal [19]. Glycogen was determined as described by Keppler and Decker [20]. Protein was assayed by the Lowry method [21] with bovine serum albumin as a standard. 2.3. Determination of plasma hormones and metabolites Plasma immunoreactive insulin and glucagon were assayed by radioimmunoassay [22,23]. Plasma glucose was determined according to Bergmeyer et al. [24]. Statistical analysis was performed by ANOVA. Statistical di¡erences were tested using the t test of Student-Fisher for unpaired data. 3. Results The e¡ects of refeeding on the evolution of body and liver weights and on the concentration of protein in the liver are given in Table 1. There was a progressive increase in both the body and liver weights as refeeding was ongoing. These weight gains were very similar in both groups: +12% and +11% for body weight gains after 360 min in 48 and 96 hfasted rats, respectively, +27% in both groups for liver weight gains, without attaining the pre-fast values, however (Table 1). Refeeding was also characterized by a progressive rehydratation of the liver (due to glycogen repletion, see below), resulting in a progressive decrease in the liver protein concentration. The latter decrease was quite similar in both 48 and 96 h-fasted rats: minus 17% at the end of the refeeding period (Table 1). This warranted the validity of the comparison between 48 and 96 h-fasted rats relating to the evolution of other hepatic param-
BBAMCR 14554 27-10-99
174
C. Minassian et al. / Biochimica et Biophysica Acta 1452 (1999) 172^178
Table 1 E¡ect of fasting and refeeding on body weight, liver weight and liver protein concentration Refeeding time (min) 0 90 180 360
Body weight (g)
Liver weight (g)
Liver protein concentration (mg/g)
Fed
48 h fast
96 h fast
Fed
48 h fast
96 h fast
Fed
48 h fast 96 h fast
186 þ 3.0
151 þ 1.3 164 þ 1.6a 164 þ 2.1a 169 þ 2.1a,b
136 þ 3.6 145 þ 2.4 151 þ 2.4a 151 þ 2.4a,b
8.4 þ 0.3
5.25 þ 0.2 5.37 þ 0.3 5.63 þ 0.2 6.68 þ 0.2a,b
4.33 þ 0.1 4.72 þ 0.2 4.95 þ 0.2a 5.53 þ 0.2a,b
143 þ 7
177 þ 8 160 þ 6 156 þ 6 147 þ 5a
183 þ 4 171 þ 5 156 þ 8a 152 þ 9a
Results are the means þ S.E.M. (n = 6). Signi¢cantly di¡erent from the value at time 0, P 6 0.05. b Signi¢cantly di¡erent from fed value, P 6 0.05. a
eters. The liver protein concentrations after the refeeding period were not di¡erent from the pre-fast values determined in fed rats (Table 1). Both groups of unfed rats had comparable very low plasma glucose (about 6 mM) and insulin (about 90 pM) before refeeding. Plasma glucose was increased to about 11 mM and plasma insulin to about 350 pM in both 48 and 96 h-fasted rats from 90 min of refeeding. Plasma glucose then tended to decrease at the end of the refeeding period (to about 9 mM) whereas plasma insulin remained constant throughout this time. In contrast, plasma glucagon (about 300 ng/l in both groups of fasted rats) was not altered upon refeeding in 96 h-fasted rats, whereas a weak (by about 25%) but signi¢cant (P 6 0.05) decrease was observed at the end of refeeding in the 48 h-fasted group. As a consequence, the insulin/ glucagon ratio was signi¢cantly increased about three times by refeeding and was not further signi¢cantly altered as refeeding was ongoing (data not shown). The liver glycogen content of 48 h-fasted rats was very low before refeeding. In contrast, the liver of 96 h-fasted rats contained a three times higher amount of glycogen: 4.6 þ 0.2 vs. 1.4 þ 0.5 mg/g net weight, means þ S.E.M., n = 6 (Fig. 1). This was in good agreement with our previous results [9]. Upon refeeding, the repletion of glycogen stores was almost linear in both groups up to 180 min. It was however less rapid in 96 h-fasted rats than in 48 h-fasted rats since the liver glycogen content in the former group was 18% lower than that in the latter at 180 min (Fig. 1). Glycogen repletion was less rapid in the second phase of refeeding (180^360 min). There
were no longer di¡erences in the glycogen stores in both groups at the end of the refeeding period. The glycogen contents attained at this time were quite similar to those observed in fed rats (Fig. 1). The basal liver Glc6P concentration was signi¢cantly higher in 96 h-fasted rats than in 48 h-fasted rats (Fig. 1). This was in agreement with previous data [9]. The Glc6P contents tended to be weakly increased in both groups in the course of refeeding, but a signi¢cant 70% higher Glc6P content in comparison with basal Glc6P was only observed in the 48 h-fasted group at the latest time of refeeding (Fig. 1). The ¢nal Glc6P contents tended to be lower (not statistically signi¢cant) than those in rats prior to fasting. In keeping with our previous results [12], the liver Glc6Pase speci¢c activity (about 12 Wmol/min/g before refeeding) was progressively inhibited during the refeeding period (to about 8 Wmol/min/g). The inhibition was however 17% less in the 96 h-fasted group than in the 48 h-fasted group at 180 min (Fig. 1). On the other hand, there was no more di¡erence at the end of the refeeding experiment. Noteworthy, the Glc6Pase activity at the end of the refeeding period was in both groups very close to Glc6Pase activity determined in fed rats. The GK activity was rapidly induced twice at the ¢rst time of refeeding in 48 h-fasted rats and was further induced at later times (Fig. 1). This was in agreement with previous published data [25]. In contrast, GK was not induced for 180 min and was only induced twice at the end of the refeeding period in the 96 h-fasted group (Fig. 1). The ¢nal GK activity in both groups was approximately twice less than the
BBAMCR 14554 27-10-99
C. Minassian et al. / Biochimica et Biophysica Acta 1452 (1999) 172^178
175
Fig. 1. E¡ect of fasting and refeeding on the liver metabolites and enzymes involved in glycogen metabolism. All values are expressed by g of wet liver. The results are expressed as the means þ S.E.M. of six animals at each time. The symbol * means that the value is signi¢cantly di¡erent from the corresponding fasting value (P 6 0.05). Signi¢cant di¡erences between the 48 and 96 h values are indicated in the di¡erent panels where appropriate.
GK activity determined prior to fasting. HK was not signi¢cantly altered by fasting and at any time throughout the refeeding, irrespective of the prior fasting time. GSa was rapidly activated by 150% from the ¢rst refeeding time in 48 h-fasted rats. However, this activation e¡ect was transient since GSa then progressively fell to a level close to its starting basal fasting level (Fig. 1). In contrast, the activation of GS was slow in 96 h-fasted rats since a signi¢cant increase in the GSa activity was only reached after 180 min of
refeeding. The GSa activity then plateaued up to the end of refeeding (Fig. 1). The values at the end of refeeding in both groups remained signi¢cantly different from that in fed rats. The GPa activity was not altered upon refeeding in 48 h-fasted rats. In marked contrast, the activity of GPa was rapidly decreased by 30% from 90 min in the 96 h-fasted group and by 50% after 180 min. The GPa-inhibited state was then maintained up to 360 min (Fig. 1). The ¢nal GPa activities were both different from the activity determined in fed rats.
BBAMCR 14554 27-10-99
176
C. Minassian et al. / Biochimica et Biophysica Acta 1452 (1999) 172^178
Fig. 2. Linear relationship between the glycogen content and the Glc6Pase activity during refeeding in both 48 and 96 hfasted rats.
4. Discussion In the present work, we have examined the key biochemical parameters involved in the repletion of liver glycogen in 48 and 96 h-fasted rats. In apparent discrepancy with the hypothesis underlying this work, i.e. that the intrahepatic mechanisms might be di¡erent in both types of fasted rats, there was only a subtle di¡erence in the time course of repletion which was slightly delayed in 96 h-fasted as compared to 48 h-fasted rats. However, in agreement with the hypothesis, the biochemical mechanisms involved in the repletion exhibited marked di¡erences. In 48 h-fasted rats, glycogen synthesis involved a rapid and further sustained induction of GK and a progressive and lasting inhibition of Glc6Pase activity. In contrast, it involved a rapid, but only transient, activation of GS, since the GSa activity was not di¡erent from the basal GSa activity after 6 h of refeeding. At this time, the increased concentration of Glc6P might play the key role in the glycogen repletion (`push' mechanism). On the other hand, the rapid activation of GSa at 90 min is in agreement with a `pull' mechanism operating at this time, as previously suggested to occur during the early phase
of glycogen repletion after glucose gavage in 20 hfasted rats [11]. The GPa activity was not decreased throughout the refeeding period. These results thus strongly suggest that the activation of GS is not necessarily dependent on the prior inhibition of GP. This is at variance with the previously suggested mechanism of liver glycogen synthesis in rodents [26], but in keeping with recent results in human [27]. The rapid induction of GK also suggests that a signi¢cant part of glycogen deposited in the liver during refeeding in 48 h-fasted rats may arise from a direct pathway, i.e. from glucose to glycogen, in addition to that arising from gluconeogenic origin, namely the indirect pathway, e.g. from lactate to glycogen [28,29]. In 96 h-fasted rats, as in their 48 h counterparts, the repletion of glycogen stores involved the inhibition of Glc6Pase. In contrast, it did not involve the induction of GK for the ¢rst 180 min of refeeding. A signi¢cant increase in GK activity was observed after 360 min only. Noteworthy, the time course of activation of GS was signi¢cantly lenghthened with regards to that of the 48 h-fasted group. The most striking di¡erence was related to the e¡ect of refeeding on the GPa activity. At variance with the absence of an e¡ect in the 48 h group, the GPa was rapidly and substantially inhibited upon refeeding. It is interesting to note that the absence of induction of GK for 3 h strongly suggests that the bulk of glycogen stored during the early period of refeeding in 96 hfasted rats likely arises from the indirect pathway. It has been suggested that intrahepatic translocation mechanisms of GK and GS may occur in response to the metabolic changes accompanying the fasted to fed transition and participate to the regulation of the glycogen repletion in this situation. On the one hand, GK has been reported to translocate from the nucleus to the cytosol in response to glucose [30,31]. These changes are only observable in intact cells and not in liver homogenates in which GK is only found in a soluble form. Since GK is absent in the livers from both 48 and 96 h-fasted rats [9] and since we assayed GK in 10 000Ug supernatants from liver homogenates, the rapid increase in the liver GK activity in 48 h-fasted rats after refeeding may be attributable to a rapid induction in the GK gene expression, which does not occur in 96 h-fasted rats. On the other hand, GS has been reported to
BBAMCR 14554 27-10-99
C. Minassian et al. / Biochimica et Biophysica Acta 1452 (1999) 172^178
translocate from the cytosol to the cell cortex in response to glucose in isolated hepatocytes [32] or to aggregate into clusters which, however, do not locate with glycogen particles, during the fasted to refed transition in vivo [33]. In contrast, the prior phases of glycogen synthesis have far been demonstrated to take place in close contact with the endoplasmic reticulum in the liver in vivo [34]. In our opinion, the extent to which the one or the other of these translocation mechanisms may participate to the regulation of glycogen repletion in vivo is yet to be established. So, we considered and studied the activation of total GSa in whole liver homogenates. It seems therefore obvious that there exists some marked divergences between the biochemical mechanisms of glycogen repletion in both types of fasted rats. The di¡erences are particularly related to the time course of induction of GK and of activation of GS and to the occurrence or not of the inhibition of GP. On the other hand, there is a parallelism in the inhibitions of Glc6Pase activity as the refeeding was ongoing. This inhibition is slightly less rapid in 96 h-fasted rats, since Glc6Pase activity remains by 17% higher than that of their 48 h-fasted counterparts at 180 min. Interestingly, the liver glycogen content at this time of refeeding is also by 18% less in the 96 h-fasted group than in the 48 h-fasted group. Taken as a whole, these results strongly suggest that the inhibition of Glc6Pase activity plays an important role in the repletion of liver glycogen during refeeding in the fasted states. Noteworthy, studying the mechanisms of glycogen repletion by various approaches in rats fasted for a shorter time than those studied here (e.g. 20 h), McGarry and co-workers previously inferred that an inhibition of Glc6Pase activity should occur during the repletion of liver glycogen [11,28,29]. The evidence obtained was indirect. The data provided herein bring up direct evidence that such an inhibition actually occurs and plays a key role in the glycogen repletion in rats fasted for various times and exhibiting markedly different capacities in terms of glycogen metabolism. The most astonishing result has appeared to be that this role may be as crucial (maybe more) as that of the variations in GS, GP and GK activities. In agreement with the latter proposal, considering both groups together, there exists a tight correlation between the glycogen content and the Glc6Pase ac-
177
tivity at a given time: R2 = 0.98, P 6 1034 (Fig. 2). The mean amount of glycogen deposited between two time points is also strongly correlated with the mean Glc6Pase activity observed during the same time interval (not shown). On the other hand, the glycogen content is less tightly correlated with the Glc6P concentration and the GK activity (R2 = 0.85, P 6 0.01 and R2 = 0.82, P 6 0.05, respectively), whereas there does not exist any correlation between the glycogen content and either the GSa or the GPa activities (not shown). The results presented herein thus further document in vivo the crucial regulatory role of Glc6Pase in liver glucose metabolism during the post-prandial period.
References [1] J.D. Foster, B.A. Pederson, R.C. Nordlie, Proc. Soc. Exp. Biol. Med. 215 (1997) 314^332. [2] G. Mithieux, Eur. J. Endocrinol. 136 (1997) 137^145. [3] L. Hue, H.G. Hers, Ann. Rev. Biochem. 52 (1983) 617^653. [4] B.A. Haber, S. Chin, E. Chuang, W. Buikhuisen, A. Naji, R. Taub, J. Clin. Invest. 95 (1995) 832^841. [5] C. Minassian, C. Zitoun, G. Mithieux, Mol. Cell. Biochem. 155 (1996) 37^41. [6] G. Mithieux, H. Vidal, C. Zitoun, N. Bruni, N. Daniele, C. Minassian, Diabetes 42 (1996) 1614^1620. [7] A.M. Bode, J.D. Foster, R.C. Nordlie, J. Biol. Chem. 267 (1992) 2860^2863. [8] L.B. Gardner, Z. Liu, E.J. Barrett, Diabetes 42 (1993) 1614^ 1620. [9] C. Minassian, A. Ajzannay, J.P. Riou, G. Mithieux, J. Biol. Chem. 269 (1994) 16585^16588. [10] C. Minassian, S. Tarpin, G. Mithieux, Biochem. Pharmacol. 55 (1998) 1213^1219. [11] C.B. Newgard, D.W. Foster, J.D. McGarry, Diabetes 33 (1984) 192^195. [12] C. Minassian, N. Daniele, J.C. Bordet, C. Zitoun, G. Mithieux, J. Nutr. 125 (1995) 2727^2732. [13] G. Mithieux, N. Daniele, B. Payrastre, C. Zitoun, J. Biol. Chem. 273 (1998) 17^19. [14] N. Daniele, F. Rajas, B. Payrastre, G. Mauco, C. Zitoun, G. Mithieux, J. Biol. Chem. 274 (1999) 3597^3601. [15] F. Chatelain, J.P. Pegorier, C. Minassian, N. Bruni, S. Tarpin, J. Girard, G. Mithieux, Diabetes 47 (1998) 882^889. [16] F. Bontemps, L. Hue, H.G. Hers, Biochem. J. 174 (1978) 603^611. [17] J.A. Thomas, K.K. Schlender, J. Larner, Anal. Biochem. 25 (1968) 486^489. [18] L. Hue, F. Bontemps, H.G. Hers, Biochem. J. 152 (1975) 105^114. [19] G. Lang, G. Michal, in: H.U. Bergmeyer (Ed.), Methods of
BBAMCR 14554 27-10-99
178
[20]
[21] [22] [23]
[24]
[25] [26]
C. Minassian et al. / Biochimica et Biophysica Acta 1452 (1999) 172^178 Enzymatic Analysis, vol. 3, VCH Publishers, New York, 1974, pp. 1238^1242. Keppler, D., Decker, C. (1974) in: H.U. Bergmeyer (Ed.), Methods of Enzymatic Analysis, vol. 3, VCH Publishers, New York, pp. 1127^1131. O.H. Lowry, N.J. Rosebrough, A.R. Farr, R.J. Randall, J. Biol. Chem. 193 (1951) 265^275. C.M. Hales, C.A. Randle, Biochem. J. 88 (1963) 137^138. J. Harris, G.R. Faloona, R.H. Unger, in: B.M. Jo¡e, H.R. Behrman (Eds.), Methods of Hormone Immunoassay, Academic Press, San Diego, CA, 1979, pp. 643^671. H.U. Bergmeyer, E. Bernt, F. Schmidt, H. Stork, in: H.U. Bergmeyer (Ed.), Methods of Enzymatic Analysis, vol. 3, VCH Publishers, New York, 1974, pp. 1196^1201. F. Eilers, H. Bartels, K. Jungerman, Histochemistry 99 (1993) 133^140. W. Stalmans, H. DeWulf, L. Hue, H.G. Hers, Eur. J. Biochem. 41 (1974) 117^134.
[27] K.F. Petersen, D. Laurent, D.L. Rothman, G.W. Cline, G.I. Schulman, J. Clin. Invest. 101 (1998) 1203^1209. [28] C.B. Newgard, S.V. Moore, D.W. Foster, J.D. McGarry, J. Biol. Chem. 259 (1984) 6958^6963. [29] M. Kuwajima, C.B. Newgard, D.W. Foster, J.D. McGarry, J. Clin. Invest. 74 (1984) 1108^1111. [30] Y. Toyoda, I. Miwa, S. Satake, M. Anai, Y. Oka, Biochem. Biophys. Res. Commun. 215 (1995) 467^473. [31] K.S. Brown, S.S. Kalinowsky, J.R. Megill, S.K. Durham, K.A. Mookhtiar, Diabetes 46 (1977) 179^186. [32] J.M. Fernandez-Novell, D. Bellido, J.J. Guinovart, Biochem. J. 321 (1997) 227^231. [33] J.M. Fernandez-Novell, A. Roca, D. Bellido, S. Vilaro, J.J. Guinovart, Eur. J. Biochem. 238 (1996) 570^575. [34] R.R. Cardell Jr., Int. Rev. Cytol. 48 (1977) 221^229.
BBAMCR 14554 27-10-99