Eur. J. Biochem. 174, 561 - 567 (1988) 0 FEBS 1988 The control of glycogen metabolism in yeast 2. A kinetic study of the two forms of glycogen synthase and of glycogen phosphorylase and an investigation of their interconversion in a cell-free extract Jean FRANCOIS and Henri-Gery HERS Laboratoire de Chimie Physiologique, Universite Catholique de Louvain and International Institute of Cellular and Molecular Pathology, Brussels (Received December 28, 1987/March 10,1988) - EJB 87 1430 Two interconvertible forms of glycogen synthase and glycogen phosphorylase, one active (a) or the other less active (b), were predominantly present in a thermosensitive adenylate-cyclase-deficient mutant that had been preincubated at the restrictive temperature of 35"C, either in the presence or in the absence of glucose. Glycogen phosphorylase was at least 20-fold less active after incubation of the cells in the presence of glucose, but this residual activity had kinetic properties identical to those of the active form of enzyme, obtained after incubation in the absence of glucose; this suggests that the b form might be completely inactive and that the low activity measured after glucose treatment must be attributed to a residual amount of phosphorylase a. By contrast, the kinetic properties of the two forms of glycogen synthase were very different. When measured in the absence of glucose 6-phosphate, the two forms of enzyme had a similar affinity for UDP-Glc but differed essentially by their V,,,. Glucose 6-phosphate had no effect on synthase a, but increased both V,,,,, and K, of synthase b; these effects, however, were in great part counteracted by sulfate and by inorganic phosphate, the latter also having the property of increasing the K, of the a form, without affecting V,,,. It was estimated that at physiological concentrations of substrates and effectors, synthase a was about 20-fold more active than synthase b. When an extract of cells that had been preincubated in the absence of glucose was gel-filtered and then incubated at 30 "C, phosphorylase was progressively fully inactivated and synthase was partially activated; these reactions were severalfold faster and, in the case of glycogen synthase, more complete in the presence of 10 mM glucose 6-phosphate. When a gel-filtered extract of cells that had been preincubated in the presence of glucose was incubated at 30°C in the presence of ATP-Mg and EGTA, phosphorylase became activated and synthase was inactivated; the first of these two reactions was severalfold stimulated by micromolar concentrations of CaZ+, whereas both reactions were completely inhibited by 10 mM glucose 6-phosphate and only slightly and irregularly stimulated by cyclic AMP. There seems to be a general agreement on the fact that in yeast, as in animal tissues, the rates of glycogen synthesis and degradation are controlled by the activities of glycogen synthase and glycogen phosphorylase, respectively. The ac- tivity of each of these enzymes can be modulated by the concentration of various effectors. Glc6P is indeed an inhibi- tor of yeast phosphorylase [l, 21 and a stimulator of glycogen synthase [3 - 71, whereas ATP, ADP and other nucleotides have been reported to inhibit glycogen synthase at pH 5.9 [3, 41. The activity is also controlled by phosphorylation and dephosphorylation, which allow the interconversion between an active (a) and a less active (b) form of each enzyme [1, 7, 81. Up to now, however, the a and b forms of the two yeast enzymes have not been completely separated from each other, preventing therefore a characterization of their kinetic proper- ties and of the control of their interconversion. In the preceding paper [9], we show that large reversible variations in the activity of glycogen synthase and glycogen Correspondence to H.-G. Hers, Laboratoire de Chimie Physio- logique de 1'Universite Catholique de Louvain, UCL 7539, Avenue Hippocrate 75, B-1200 Bruxelles, Belgium Abbreviations. Glc6P, glucose 6-phosphate; Glcl P, glucose 1-phosphate; PFK 2, phosphofructo 2-kinase. Enzymes. Phosphofructo 2-kinase (EC 2.7.1.105); trehalase (EC 3.2.1.28); glycogen synthase (EC 2.4.1.1 1); glycogen phosphory- lase (EC 2.4.1.1). phosphorylase are induced by treatment of yeast with glucose and that these changes are stable upon gel filtration, indicating that they result from a covalent modification. These changes were more pronounced in a thermosensitive mutant deficient in adenylate cyclase incubated at the restrictive temperature; preparations of this mutant could be obtained in which phos- phorylase and glycogen synthase were either nearly com- pletely inactive or greatly activated. The purpose of the present work was to use these prep- arations to investigate the kinetic properties of the two forms of each enzyme and the mechanism of their interconversion. MATERIALS AND METHODS Unless otherwise indicated, the materials and the methods used in the present work were those described in the preceding paper [9]. All experiments, except those reported in Fig. 1 and in Table 1, were performed with the thermosensitive mutant deficient in adenylate cyclase (cdc35 mutant, strain Be333). These cells were incubated for 60min at 35°C and, unless otherwise stated, were collected by filtration before or 30 min after addition of 100 mM glucose. The cells (150-200 mg) were extracted in 1.5 ml 20 mM Hepes pH 7.1, containing 1.5% glycogen, 100 mM KCl, 2 mM EDTA, 1 mM dithio- threitol and 1 mM phenylmethylsulfonyl fluoride. The homogenates were filtered at 4°C through 20 vol. Sephadex
The control of glycogen metabolism in yeast 2. A kinetic study of the two forms of glycogen synthase and of glycogen phosphorylase and an investigation of their interconversion in a cell-free extract
Jean FRANCOIS and Henri-Gery HERS Laboratoire de Chimie Physiologique, Universite Catholique de Louvain and International Institute of Cellular and Molecular Pathology, Brussels
(Received December 28, 1987/March 10,1988) - EJB 87 1430
Two interconvertible forms of glycogen synthase and glycogen phosphorylase, one active (a) or the other less active (b), were predominantly present in a thermosensitive adenylate-cyclase-deficient mutant that had been preincubated at the restrictive temperature of 35"C, either in the presence or in the absence of glucose. Glycogen phosphorylase was at least 20-fold less active after incubation of the cells in the presence of glucose, but this residual activity had kinetic properties identical to those of the active form of enzyme, obtained after incubation in the absence of glucose; this suggests that the b form might be completely inactive and that the low activity measured after glucose treatment must be attributed to a residual amount of phosphorylase a. By contrast, the kinetic properties of the two forms of glycogen synthase were very different. When measured in the absence of glucose 6-phosphate, the two forms of enzyme had a similar affinity for UDP-Glc but differed essentially by their V,,,. Glucose 6-phosphate had no effect on synthase a, but increased both V,,,,, and K, of synthase b; these effects, however, were in great part counteracted by sulfate and by inorganic phosphate, the latter also having the property of increasing the K , of the a form, without affecting V,,,. It was estimated that at physiological concentrations of substrates and effectors, synthase a was about 20-fold more active than synthase b.
When an extract of cells that had been preincubated in the absence of glucose was gel-filtered and then incubated at 30 "C, phosphorylase was progressively fully inactivated and synthase was partially activated; these reactions were severalfold faster and, in the case of glycogen synthase, more complete in the presence of 10 mM glucose 6-phosphate. When a gel-filtered extract of cells that had been preincubated in the presence of glucose was incubated at 30°C in the presence of ATP-Mg and EGTA, phosphorylase became activated and synthase was inactivated; the first of these two reactions was severalfold stimulated by micromolar concentrations of CaZ+, whereas both reactions were completely inhibited by 10 mM glucose 6-phosphate and only slightly and irregularly stimulated by cyclic AMP.
There seems to be a general agreement on the fact that in yeast, as in animal tissues, the rates of glycogen synthesis and degradation are controlled by the activities of glycogen synthase and glycogen phosphorylase, respectively. The ac- tivity of each of these enzymes can be modulated by the concentration of various effectors. Glc6P is indeed an inhibi- tor of yeast phosphorylase [l , 21 and a stimulator of glycogen synthase [3 - 71, whereas ATP, ADP and other nucleotides have been reported to inhibit glycogen synthase at pH 5.9 [3, 41. The activity is also controlled by phosphorylation and dephosphorylation, which allow the interconversion between an active (a) and a less active (b) form of each enzyme [1, 7, 81. Up to now, however, the a and b forms of the two yeast enzymes have not been completely separated from each other, preventing therefore a characterization of their kinetic proper- ties and of the control of their interconversion.
In the preceding paper [9], we show that large reversible variations in the activity of glycogen synthase and glycogen
Correspondence to H.-G. Hers, Laboratoire de Chimie Physio- logique de 1'Universite Catholique de Louvain, UCL 7539, Avenue Hippocrate 75, B-1200 Bruxelles, Belgium
phosphorylase are induced by treatment of yeast with glucose and that these changes are stable upon gel filtration, indicating that they result from a covalent modification. These changes were more pronounced in a thermosensitive mutant deficient in adenylate cyclase incubated at the restrictive temperature; preparations of this mutant could be obtained in which phos- phorylase and glycogen synthase were either nearly com- pletely inactive or greatly activated.
The purpose of the present work was to use these prep- arations to investigate the kinetic properties of the two forms of each enzyme and the mechanism of their interconversion.
MATERIALS AND METHODS
Unless otherwise indicated, the materials and the methods used in the present work were those described in the preceding paper [9]. All experiments, except those reported in Fig. 1 and in Table 1, were performed with the thermosensitive mutant deficient in adenylate cyclase (cdc35 mutant, strain Be333). These cells were incubated for 60min at 35°C and, unless otherwise stated, were collected by filtration before or 30 min after addition of 100 mM glucose. The cells (150-200 mg) were extracted in 1.5 ml 20 mM Hepes pH 7.1, containing 1.5% glycogen, 100 mM KCl, 2 mM EDTA, 1 mM dithio- threitol and 1 mM phenylmethylsulfonyl fluoride. The homogenates were filtered at 4°C through 20 vol. Sephadex
5 62
control 0 +glucose fZ cdc35 mutant Wild type (X2180)
+DNP I
b 7.5
5.0
2.5
L5
30
15 I 0 b C d a
- I
a d o b c
Fig. 1. Activity of glycogen synthase and of glycogenphosphorylase in yeast incubated with or without glucose and collected by dqferentprocedures. Cells were incubated either at 30°C (wild type) or at 35°C (cdc35 mutant) in the absence (open columns) or the presence of 100 mM glucose (dashed columns) or 2 mM dinitrophenol (DNP, black columns), added 10 min before their isolation. The cell suspension (10 ml) was then treated as follows: (a) filtration under vacuum, scraping off the cake and freezing it in liquid nitrogen; (b) same as (a) except that one washing step with 10 ml cold water was performed before scraping; (c) centrifugation of 10 ml of the suspension during 2 min at 1000 x g and freezing the pellet in a mixture of solid CO2 and acetone; (d) same as (c) except that one washing step with 10 ml cold water was performed before freezing the pellet
(3-25 (fine) equilibrated in the extraction buffer but containing no glycogen. The Sephadex filtrate was used as a source of enzymes in all experiments, except that reported in Fig. 1 . The concentration of free calcium was calculated according to the equation of Bartfai [lo] using a value of 7.9 x lo6 M-' for the K of the Ca-EGTA complex at pH 7.1. Calmidazolium was purchased from Boehringer (Mannheim, FRG) and trifluoro- perazine from Sigma Chemicals (St Louis, MO, USA).
RESULTS
A cyclic-AMP-dependent interconversion of enzymes occurring during the isolation of cells
It was reported in the preceding paper [9] that the addition of glucose to a suspension of yeast caused a large activation of glycogen synthase and inactivation of glycogen phosphorylase and that these effects were much more pronounced in the cdc35 mutant incubated at the restrictive temperature. We show in Fig. 1 that these changes in enzymic activities, when measured in the wild type although not in the mutant, are dependent upon the mode of isolation of the cells. Indeed, these effects were seen when the cells were isolated by our rapid filtration procedure, but not when they were separated by centrifugation and washed once with water, as done by other authors [11 - 131. This difference is explained by the reversion of the enzyme interconversion during washing, as was illustrated in Fig. 2 of the preceding paper [9]. Since these effects of the isolation procedure were not seen with the cdc35 mutant maintained at 35"C, it appears likely that they were caused by a cyclic-AMP-dependent activation of phosphory- lase and inactivation of glycogen synthase occurring during centrifugation of the cells.
Treatment of the wild-type cells with 2,4-dinitrophenol in the absence of glucose (Fig. 1) or incubation of an extract of untreated cells for 15 min in the presence of 5 mM ATP-Mg, 10 pM cyclic AMP and 0.5 mM CaC12 (data not shown)
caused a three fold activation of phosphorylase, whereas only a slight activation, if any, was observed under the same con- ditions with the adenylate-cyclase-deficient mutant incubated at the restrictive temperature. Since this phosphorylase ac- tivity is the highest that could be observed, it probably rep- resents that of the fully activated enzyme.
Kinetic properties of the a and b forms of phosphorylase
Assuming that the residual phosphorylase activity mea- sured in an extract of glucose-treated cells corresponds to phosphorylase b, we found that the two forms of enzyme had a similar K, for Glcl P (close to 1 mM), but differed essentially by their V,,, (Fig. 2) . This difference was 14-fold in the exper- iment shown in Fig. 2. The activity of phosphorylase b was not affected by AMP and the two enzymic preparations dis- played a maximal activity at pH 5.5 and approximately 40% of this activity at pH 7.0 (not shown).
Kinetic properties of the a and b forms of glycogen synthase
In contrast to previous reports [6, 71, the saturation curve of the two forms of glycogen synthase for UDP-Glc was clearly hyperbolic (Fig. 3) and, as indicated in Table 1, there was only a slight difference in affinity in favor of synthase a. The major difference between the two forms was in their V,,,, at least when the activity was measured in the absence of Glc6P (Table 1). In the presence of this phosphate ester, the V,,, of the b form was greatly increased, reaching, at 20 mM Glc6P, a value identical to that of synthase a. Surprisingly, there was also a severalfold increase in K , for UDP-Glc (Table 1). By contrast, Glc6P had no effect on the activity of synthase a.
Fig. 4 illustrates the inhibitory effect of Pi on the two forms of enzyme, when measured at 0.25 mM UDP-Glc and in the presence of various concentrations of Glc6P. As shown in Table 2, this inhibition was related to a decreased affinity
563
of the two forms of enzyme for UDP-Glc, combined with an increased Hill coefficient. In the case of synthase 6, there was also a severalfold decrease in V,,, both in the presence and in the absence (not shown) of Glc6P. Sulfate had no action on synthase a, but, at concentrations above 5 mM, completely cancelled the positive effect of 2 mM Glc6P on synthase b and also decreased by twofold the low activity displayed by this enzyme in the absence of Glc6P (not shown).
Table 3 shows the inhibitory effect of ATP-Mg on the activity of glycogen synthase. In agreement with previous work [3, 41, the inhibition by ATP-Mg was much more pro- nounced at pH 6.0 than at pH 7.1 and was partially released by Glc6P. However, we did not find any significant difference in the sensitivity of the two forms of enzyme to this inhibition.
Stimulation of phosphorylase phosphatase and of synthase phosphatase by glucose 6-phosphate
It is shown in the preceding paper [9] that the effect of glucose in activating glycogen synthase and inactivating glyco- gen phosphorylase is most likely mediated by an increase in the concentration of Glc6P. In agreement with this hypoth- esis, we show in Fig. 5 that the inactivation of glycogen phos- phorylase and the activation of glycogen synthase occurring
/ O 7 C"
I $ 5 0 } E /
L
lo}: After glucose treatment 5 I,,-*-. Y I-.
" 0 1 2 5 10" 50 [Glucose I-phosphate1 (mM)
After glucose treatment
Q
0 1 2 3 L 5 IUDP-Glcl {mM)
Fig. 3. Saturation curve of the two forms of glycogen synthase for UDP-Glc in the presence of various concentrations of Glc6P. General procedure as in Fig. 2. The assay mixture did not contain Na2S04
L
- 3
1 1 I I 1 t-L- 0 2 L 6 8 1 2 5 0
1Glucose 6-phosphate1 (rnM1
Fig. 2. Saturation curve ofthe two forms of'glycogenphosphorylase,for GIcIP. An extract of the thermosensitive mutant, prepared before or after 30-min incubation of the cells in the presence of 100 mM glucose, was uscd as a source of enzyme
Fig. 4. efject of' Glc6P on the activity of the two forms of glycogen synthase measured at 0.25 mM UDP-Glc in the presence of different concentrations of Pi. General procedure as in Fig. 2. The assay mixture did not contain Na,S04
Table 1. Kinetic properties ufglycogen synthase
Source of enzyme [Glc6P] b form a form (before glucose treatment)
Km UDP-Glc vma, Km U DP-Glc vm,,
(after glucose treatment)
mM mM mU/mg protein mM mU/rng protein cdc35 incubated at 35°C 0 0.8 3.0 0.30 33
0.5 1.40 5 0.30 33 2 1.90 1 1 0.30 33
20 4 35 0.30 33 Wild type incubated at 30°C 0 0.55 4.1 0.25 25
20 1 .o 22 0.20 25
564
x c .- 2 6 0 - c
Table 2. Efect of' Pi on kinetic parameters of the a and b forms of glycogen synthase
[Glc6P] [Phosphate] b form a form (before glucose treatment) (after glucose treatment)
h 5'0.5 UDP-Glc Vmax h So.5 UDP-Glc Vmax
- C al .- c
mM
0 2 2 2 2 2
c C x ln
+Glc6P 10mM
1 'a
mM mM mU/mg protein mM 0 1.0 0.77 2.95 0 1.05 1.90 1 1 2 1.10 2.10 8.50 5 1.15 2.35 5.50
Table 3. Efiect of ATP-Mg on glycogen synthase activity The activity was measured in the absence or presence of 2 mM Glc6P by the standard assay, except that 50 mM Mes or 50 mM Hepes were used as buffers at pH 6.0 and 7.1 respectively and that Na2S04 was omitted
Fig. 5. E f i c t of' Glc6P on the inactivation of glycogen phosphorylase and the activation of glycogen synthase in a yeast extract. An extract, prepared from cells that had been incubated in the absence of glucose, was incubated in the presence of 20 mM Hepes, pH 7.1,3 mM MgCI2 with or without, 10 mM Glc6P. The presence of Na2S04 in the assay mixture prevented the stimulation of glycogen synthase by Glc6P (0.3 mM after dilution)
in a Sephadex-filtered extract of cells incubated in the absence of glucose, were severalfold faster in the presence of 10 mM Glc6P than in its absence. A half-maximal effect was obtained at approximately 5 mM Glc6P for the inactivation of phos- phorylase and 2 mM Glc6P for the activation of glycogen synthase (Fig. 6). These interconversions were completely in- hibited by 50 mM KF; they were only barely detectable in a
" 0 k z O'O 4 Ib 1; ;o -Is'
[Glucose 6-phosphate] (mM)
Fig. 6. Dose-dependency of the glucose-6-phosphate effect of stimul- ating the activation ofglycogen synthase and the inactivation of glyco- genphosphoryEase in a yeast extract. The experimental conditions were as in Fig. 5 except that a different yeast extract was used and the duration of incubation was 15 min
non-gel-filtered extract, possibly because of the presence of ATP in that preparation. By analogy with the mammalian enzymes, it is most likely that these two reactions were catalyzed by specific protein phosphatases. They could be reversed by the subsequent addition of ATP-Mg, cyclic AMP and calcium (not shown). We have also checked that glucose, caffeine, nicotinamide or AMP, which are known to affect the rate of phosphorylase inactivation in a liver extract [14, 151, were without action on this reaction measured in a yeast extract.
565
100 -
LO CI +5 mM ATP-Mg. 10pM CAMP
" 0 10 20 30 Time of incubation at 3OoC (min)
Fig. 7. Ejyect of ATP-Mg, cyclic AMP and calcium on the reactivation of phosphorylase in a yeast extract. An extract from cells that had been preincubated in the presence of glucose was used as a source of enzyme. EGTA was substituted for EDTA in the gel filtration buffer. The preparation was incubated at 30°C in the presence of 20 mM Hepes pH 7.1, 5 mM EGTA and 3 mM MgClz and, as indicated, 5 mM ATP-Mg, 10 pM cyclic AMP or 5.1 mM CaClz
The ATP-Mg-dependent activation of phosphorylase and inactivation of glycogen synthase
Because calcium and cyclic AMP are known to be involved in the activation of mammalian phosphorylase [16], we have investigated the action of these two effectors on the activation of the yeast enzyme. It is shown on Fig. 7 that when a gel- filtered extract from glucose-treated cells was incubated at 30 "C with 5 mM EGTA and 3 mM MgC12, there was a 15- fold increase in the activity of phosphorylase upon addition of ATP-Mg and that this increase reached 70-fold if 5.1 mM CaClz (giving a calculated concentration of free Ca2+ of approximately 0.1 mM) was also present. This effect of the CaZ+ ion was already maximal at a calculated concentration of 0.1 pM (not shown) and, accordingly, could only be demon- strated in the presence of EGTA. It was 75% and 50% inhibited by 0.2 mM trifluoroperazine and 0.01 mg/ml calmidazolium respectively, two potent calmodulin inhibitors [I71 (results not shown). It is also shown in Fig. 7 that, if added in the absence of calcium, cyclic AMP caused a slightly (20%) greater activation of phosphorylase. This cyclic AMP effect, however, was somewhat irregular, since it was only observed in eight experiments out of ten, performed in the presence or the absence of EGTA and was never observed when calcium was added to the incubation medium. The activation of phos- phorylase by ATP-Mg and calcium was almost completely prevented by 10 mM Glc6P and a half-maximal inhibition was observed at 1 mM Glc6P (not shown).
By contrast, the ATP-Mg-dependent inactivation of glyco- gen synthase a, performed in the presence or the absence of 5 mM EGTA, was affected neither by calcium nor by cyclic AMP, except for one experiment (in ten) in which a slight effect of the nucleotide was observed (results not shown). We have verified that, under these conditions and as previously reported [IS], the activation of PFK 2 and of trehalase was entirely dependent on the presence of cyclic AMP (not shown). As shown in Fig. 8, 10 mM Glc6P completely prevented the a to b transformation of glycogen synthase while sulfate, at the same concentration, could only reduce the rate of this transformation by 30% (not shown). When the hexose phos-
" 0 10 20 30 Time after addition of ATP-Mg (mid
Fig. 8. Eflect of Glc6P on the ATP-Mg-dependent inactivation of glyco- gen synthase. General procedure as in Fig. 7, except that 10 mM Glc6P was added (0 ) or not (0) at the times indicated (arrows)
phate was added after the inactivation had already progressed, it prevented further inactivation and caused a slight reacti- vation of glycogen synthase. A half-maximal inhibitory effect was obtained with 2 mM Glc6P (not shown).
The activation of phosphorylase by ATP-Mg and calcium and the ATP-Mg-dependent inactivation of glycogen synthase were not observed or were very weak in extracts of cells that had been incubated in the absence of glucose or in the presence of dinitrophenol, indicating that, under these conditions, these enzymes were fully phosphorylated.
DISCUSSION
Our method of cell isolation, based on a 15-s filtration, has allowed us to establish very large differences in the activity of glycogen phosphorylase and of glycogen synthase accord- ing to the conditions under which the cells were incubated. Because these differences persisted upon gel filtration of the extracts, they can be attributed to a covalent modification of the enzymes. From the work of others [l, 5, 7, 81 and also by analogy with the property of the two enzymes in mammalian tissues, we can assume that they reflect an interconversion between an active (a ) and a less active (b) form of each enzyme by phosphorylation and dephosphorylation, phosphorylation causing the activation of phosphorylase and the inactivation of glycogen synthase. These differences in enzymic activities were much more pronounced in the cdc35 mutant incubated at the nonpermissive temperature; this preparation was there- fore used as a source of active and inactive forms of the two enzymes. A nearly complete activation of phosphorylase and inactivation of glycogen synthase were also obtained by incubating the wild-type cells in the presence of dinitrophenol.
The two forms of yeast glycogen phosphorylase
Because the activity of the phosphorylase present in the cdc35 mutant incubated at 35°C in the absence of glucose could not be further increased upon incubation of the cells in the presence of dinitrophenol or of the extract in the presence of ATP-Mg, cyclic AMP and calcium, we can assume that this enzyme was maximally active, i.e. was in its a, phosphorylated, form. The small amount of phosphorylase present in the same
566
cells after their incubation in the presence of glucose had the same affinity for GlclP and the same pH dependency as the a form. Since some interconversion could have occurred during the 15-s filtration and during the extraction procedure, we cannot discard the possibility that this residual activity could be due to contaminating phosphorylase a and that the b form of yeast phosphorylase is actually completely inactive. This hypothesis is supported by the fact that phosphorylase a was nearly completely inactivated during incubation of the extracts in the presence of Glc6P (Fig. 5) and that a 50- 100-fold activation could be obtained upon incubation of an inactive preparation in the presence of ATP-Mg and calcium (Fig. 7). It therefore appears likely that yeast phosphorylase b has only a very low, if any, activity. Since this activity, contrary to that of muscle phosphorylase b, cannot be in- creased by AMP, there is no known procedure to measure the b form in a cell-free extract. As explained in the preceding paper [9] and as illustrated in Fig. 1, the total amount of phosphorylase in a yeast suspension can be known by activating the enzyme in vivo upon incubation of the cells in the presence of 2 mM dinitrophenol and the same activation can be obtained in a cell-free extract upon addition of ATP- Mg and calcium (see Fig. 7). The amount of phosphorylase b in the cell can be estimated by substracting that of phosphoryl- ase a from the total amount of enzyme.
The possibility that the low activity present in phosphory- lase b preparations could be due to contamination by some active a form was already considered by Fosset et al. [l], who could not reach a definite conclusion because of the potential formation of partially phosphorylated hybrids. Nearly inac- tive forms of phosphorylase were also isolated from baker’s yeast by Becker et al. [8] who, furthermore, reported that the activity of the enzyme correlated with its phosphate content.
The two,forms of yeast glycogen synthase
By contrast with the situation observed for phosphorylase, the kinetic properties of the glycogen synthase present in the cdc35 cells incubated with or without glucose were very different from each other. The a form, obtained in the presence of glucose, was completely insensitive to Glc6P and to sulfate, indicating that it was not contaminated by an appreciable amount of synthase b, the activity of which is markedly af- fected by these two anions. Conversely, the greatest part of the activity of cells incubated without glucose could be inhibit- ed by 10 mM sulfate, and could therefore not be attributed to a contaminating amount of synthase a. This, however, may not be true for the small part of the activity which was not inhibited by sulfate.
Synthase a and synthase b have a hyperbolic saturation curve for UDP-Glc and a similar K, for this substrate, but present an approximately 10-fold difference in Vm,, in favor of the a form. Glc6P increases V,,, but also K, of synthase band the first of these effects is deeply counteracted by physio- logical concentrations of Pi. One can therefore estimate that, under the ionic conditions prevailing in the cell, the specific activity of synthase b does not exceed 5% of that of synthase a and is only slightly affected by variations in the concen- tration of Glc6P.
ATP is another effector, the role of which in the control of glycogen metabolism in yeast has been greatly emphasized by Cabib and coworkers [3 - 71. We could confirm that ATP is a potent inhibitor of both synthase a and synthase b at pH 6. We found, however, that this effect was negligible under
physiological conditions, such as pH 7 and an ATP concen- tration of 2 mM.
The control of protein phosphatases by glucose 6-phosphate
The spontaneous inactivation of phosphorylase a in a Sephadex filtrate incubated at 30°C and the acceleration of this conversion by Glc6P, a known inhibitor of yeast phos- phorylase [ l , 21, are reminiscent of the similar conversion occurring in a liver Sephadex filtrate and of its stimulation by glucose, an inhibitor of mammalian phosphorylase [14, 151. Furthermore, the K, of 5 mM that we have calculated for the Glc6P effect is similar to the Ki (10 mM) reported by Fosset et al. [l] for the inhibition of baker’s yeast phosphorylase. It appears therefore likely that Glc6P acts, like glucose in the mammalian system, by binding to phosphorylase a, the sub- strate of the reaction, rather than to phosphorylase phospha- tase itself. In agreement with this hypothesis is the fact that a Glc6P-binding site has been detected in the primary structure of yeast phosphorylase [19]. In the case of the mammalian enzyme, the effect of glucose is to expose the phosphorylated site of phosphorylase to the action of phosphorylase phospha- tase [20]. The effect of Glc6P to inhibit phosphorylase could also play a role in the arrest of glycogenolysis after glucose addition to a yeast suspension.
The effect of Glc6P to favor the reactivation of synthase b in a cell-free extract is presumably due, as in the case of phosphorylase, to the binding of Glc6P to glycogen synthase rather than to the protein phosphatase itself. A similar effect of Glc6P in stimulating synthase phosphatase has been report- ed previously for the enzyme from various sources [21-231.
The control of protein kinases by calcium, cyclic AMP and glucose 6-phosphate
The fact that the reactivation of phosphorylase b in the presence of ATP-Mg is greatly stimulated by calcium is remi- niscent of the well-known property of muscle and liver non- activated phosphorylase kinases to be in great part calcium- dependent [24]. The sensitivity of this effect to trifluoro- perazine and calmidazolium suggests that yeast phosphoryl- ase kinase could, like the mammalian enzymes, contain calmodulin as one of its subunits. Calmodulin is known to be present in yeast [25, 261 and a Ca2 +/calmodulin-dependent protein kinase has recently been purified from Saccharomyces cerevisiae [27] but its activity on phosphorylase b has not been determined. Nothing is known at the present time of factors that could affect the concentration of Ca2+ in yeast.
The minimal effect of cyclic AMP on both the activation of phosphorylase and the inactivation of glycogen synthase in a cell-free extract is, at a first glance, difficult to reconcile with biochemical (see preceding paper [9]) and genetic data [28, 291 indicating an important role for this nucleotide in the interconversion of these enzymes in vivo. The lack of cyclic AMP effect cannot be attributed to a contamination of ATP by traces of the cyclic nucleotide, since it has been checked that, under similar conditions, the activation of trehalase and of PFK 2 was entirely cyclic-AMP-dependent. A possible ex- planation for the discrepancy, at least as far as the activation of phosphorylase is concerned, could be that cyclic AMP itself induces a mobilization of intracellular calcium which in turn activates phosphorylase through its action on phosphorylase kinase. The action of dinitrophenol in activating phosphory- lase when added to the cdc35 mutant incubated at the restric- tive temperature [9] could also be explained by an effect of
567
calcium on phosphorylase kinase, since it is known that in hepatocytes this uncoupler promotes the intracellular mobi- lization of calcium from mitochondria [30].
A kinase able to phosphorylate phosphorylase specifically has been isolated from yeast by Pohlig et al. [31]. This enzyme was not sensitive to cyclic AMP nor to calcium and, consider- ing its low molecular mass (30 kDa), could correspond to the catalytic subunit of a more complex structure.
The effect of Glc6P in preventing both the activation of phosphorylase and the inactivation of glycogen synthase by their respective protein kinase would reinforce its property of stimulating the antagonistic protein phosphatases. The rela- tively greater sensitivity of phosphorylase kinase to Glc6P as compared to that of phosphorylase phosphatase would be in agreement with a greater affinity of phosphorylase b than phosphorylase a for the phosphate ester [l]. As already dis- cussed in the preceding paper [9], a decrease in Glc6P concen- tration secondary to an activation of PFK 2 , a synthesis of Fru(2,6)P2 and a stimulation of glycolysis may represent an indirect mechanism by which cyclic AMP could cause the activation of phosphorylase and the inactivation of glycogen synthase.
An onloff mechanism of control
The control of glycogen metabolism in yeast is, therefore, as in the liver [32], an on/off mechanism, occurring essentially by the interconversion of two forms of enzymes, one active and the other one inactive under the ionic conditions prevail- ing in the cell. The control is essentially exerted at the level of the interconverting protein kinases and phosphatases and not by the concentration of effectors of synthase and phosphoryl- ase. As discussed above, Glc6P, calcium and cyclic AMP appear to be the main regulators of the converter enzymes in yeast.
This work was supported by the Fond7 de la Recherche Scientrfique Medicale and by the US Public Health Service (grant DK 09235). J.F. is Charge de Recherche of Fonds National de la Recherche Scientifique.
REFERENCES 1. Fossct, M., Muir, L. W., Nielscn, L. D. & Fischer, E. H. (1971)
2. Sagardia, F., Gotay, I. &Rodriguez, M. (1971) Biochem. Biophys. Biochemistry 10,4105-4113.
Res. Commun. 42,829 - 835.
3. Rothman, L. B. & Cabib, E. (1966) Biochem. Biophys. Res.
4. Rothman, L. B. & Cabib, E. (1967) Biochemistry 6,2107-2112. 5. Rothman-Denes, L. B. & Cabib, E. (1970) Proc. Nut1 Acad. Sci.
6. Rothman-Denes, L. B. & Cabib, E. (1971) Biochemistry 10,
7. Huang, K. P. & Cabib, E. (1974) J . Biol. Chem. 249,3851 -3857. 8. Becker, J. U., Wingender-Drissen, R. & Schiltz, E. (1983) Arch.
9. FranGois, J., Villanueva, M. E. & Hers, H.-G. (1988) Eur. J .
Commun. 25,644- 650.
USA 66,967 - 974.
1236- 1242.
Biochem. Biophys. 255,667 - 678.
Biochem. 174, 551 - 559. 10. Bartfai, T. (1979) Adv. Cyclic Nucleotides Res. 10, 219-242. 11. Rothman, L. B. & Cabib, E. (1969) Biochemistry 8, 3332-3341. 12. Becker, J. U. (1977) Arch. Microbiol. 115, 181 -184. 13. Becker, J. U. (1982) J . Gen. Microbiol. 128, 447-454. 14. Stalmans, W., De Wulf, H., Lederer, B. & Hers, H.-G. (1970) Eur.
15. Stalmans, W., Laloux, M. & Hers, H.-G. (1974) Eur. 1. Biochem.
16. Krebs, E. G. & Beavo, J. A. (1979) Annu. Rev. Biochem. 48,923-
17. Vanbelle, H. (1984) Adv. Cyclic Nucleotides Res. 17, 557-567. 18. FranGois, J., Van Schaftingen, E. & Hers, H.-G. (1984) Eur. J .
19. Hwang, P. K. & Fletterick, R. J. (1986) Nature (Lond.) 324,80-
20. Fletterick, R. J. & Madsen, N. B. (1980) Annu. Rev. Biochem. 49,
21. Hizukuri, S. & Takeda, Y. (1970) Biochim. Biophys. Acta 212,
22. Kato, K. & Bishop, J. S. (1972) J. Biol. Chem. 247, 7420-7429. 23. Gilboe, D. P. & Nuttall, F. Q. (1982) Arch. Biochem. Biophys.
24. Pieckett-Gies, C. A. & Walsh, D. (1986) in The enzymes, 3rd edn (Boyer, P. D., ed.) vol. 7, pp. 395-459, Academic Press, New York.
25. Hubbard, M., Bradlcy, M., Sullivan, P., Shepherd, M. & Forrester, I. (1982) FEBS Lett. 137, 85-88 .
26. Davis, T. N., Urdea, M. S., Masiarz, F. R. & Thorner, J. (1987) Cell 47, 423 - 43 1.
27. Londesborough, J. & Nuutinen, M. (1987) FEBS Lett. 219,249- 253.
28. Tatchell, K., Robinson, L. L. & Breitenbach, M. (1985) Proc. Nut1 Acad. Sci. USA 82, 3785 - 3789.
29. Fraenkel, D. (1985) Proc. Nut1 Acad. Sci. USA 82,4740-4744. 30. Williamson, J. R., Cooper, R. H. & Hoek, J . B. (1981) Biochim.
31. Pohlig, G., Wingender-Drissen, R. & Becker, J. U. (1983) Bio-
32. Mersmann, B. J. & Segal, H. L. (1967) Proc. Natl Acad. Sci. USA
