Eur. J. Biochem. 145, 187-193 (1984) 0 FEBS 1984

The mechanism by which glucose increases fructose 2,6-bisphosphate concentration in Saccharomyces cerevisiae A cyclic-AMP-dependent activation of phosphofructokinase 2

Jean FRANCOIS, Emile VAN SCHAFTINGEN, and Henri-Gery HERS Laboratoire de Chimie Physiologique, Universite Catholique de Louvain, and International Institute of Cellular and Molecular Pathology, Brussels

(Received July 16, 1984) - EJB 84 D79b

When glucose was added to a suspension of Saccharomyces cerevisiae in stationary phase, it caused a transient increase in the concentration of cyclic AMP and a more persistent increase in the concentration of hexose 6- phosphate and of fructose 2,6-bisphosphate. These effects of glucose on cyclic AMP and fructose 2,6-bisphosphate but not that on hexose 6-phosphate were greatly decreased in the presence of 0.15 mM acridine orange or when a temperature-sensitive mutant deficient in adenylate cyclase was used at the restrictive temperature. Incubation of the cells in the presence of dinitrophenol and in the absence of glucose increased the concentration of both cyclic AMP and fructose 2,6-bisphosphate, but with a minimal change in that of hexose 6-phosphate. Glucose induced also in less than 3 min a severalfold increase in the activity of 6-phosphofructo-2-kinase and this effect was counteracted by the presence of acridine orange. When a cell-free extract of yeast in the stationary phase was incubated with ATP-Mg and cyclic AMP, there was a 10-fold activation of 6-phosphofructo-2-kinase. Finally, the latter enzyme was purified 150-fold and its activity could then be increased about 10-fold upon incubation with ATP-Mg and the catalytic subunit of cyclic-AMP-dependent protein kinase. This activation resulted from a 4.3-fold increase in V and a 2-fold decrease in K,. Both forms of the enzyme were inhibited by sn-glycerol 3- phosphate. From these results it is concluded that the effect of glucose in increasing the concentration of fructose 2,6-bisphosphate in S. cerevisiae is mediated by the successive activation of adenylate cyclase and of cyclic-AMP- dependent protein kinase and by the phosphorylation of 6-phosphofructo-2-kinase by the latter enzyme. In deep contrast with what is known of the liver enzyme, yeast 6-phosphofructo-2-kinase is activated by phosphorylation instead of being inactivated.

Fru(2,6)P2 is a recently discovered positive effector of PFK 1 and inhibitor of FBPase 1 of various origins (reviewed in [I]). Its concentration in the liver is greatly increased in the presence of high concentrations of glucose and decreased by glucagon. This effect of glucose is believed to be mediated by an increase in the concentration of Fru6P which is a substrate of PFK 2, the enzyme that forms Fru(2,6)P2, and a potent inhibitor of FBPase 2, which converts it back to Fru6P and Pi. The effect of glucagon is understood by the fact that both PFK 2 and FBPase 2 are substrates for cyclic-AMP- dependent protein kinase and that their phosphorylation causes the inactivation of the first enzyme and the activation of the second. Furthermore, there is a good indication that in the liver, PFK 2 and FBPase 2 are part of a single bifunctional protein (reviewed in [2] and [3]). The mechanism by which the hepatic concentration of Fru(2,6)P2 is controlled explains the well-known effect of glucagon in stimulating gluconeogenesis and inhibiting glycolysis in the liver and the reverse effect of glucose.

Fru(2,6)P2 is also present in Saccharomyces cerevisiae and its concentration in these cells is greatly increased upon addi-

Abbreviations. Fru(2,6)P2, fructose 2,6-bisphosphate; Fru6P, fructose 6-phosphate; PFK 2, 6-phosphofructo-2-kinase; PFK 1, 6- phosphofructo-I-kinase; FBPase 1, fructose 1,6-bisphosphatase; FBPase 2, fructose 2,6-bisphosphatase.

Enzymes. Fructose 1,6-bisphosphatase (EC 3.1.3.1 1); glucose-6- phosphate dehydrogenase (EC 1 .I .1.49); 6-phosphofructo-I-kinase (EC 2.7.1.11); trehalase (EC 3.2.1.28).

tion of glucose to a culture in the stationary phase [4]. Since the concentration of hexose 6-phosphate is even more rapidly increased under the same conditions, it was initially believed that the rise in Fru(2,6)P2 caused by the addition of glucose to yeast was, as in the liver, secondary to that of Fru6P [4]. However, other known short-term effects (occurring within minutes) of glucose in S. cerevisiae are to increase the con- centration of cyclic AMP [5], to partially inactivate FBPase 1 [6] and to activate trehalase [5] and the two latter changes could be attributed to a cyclic-AMP-dependent phosphoryla- tion of these enzymes [5, 7-10]. A slower effect of glucose is to cause, within 1-2 h, the proteolytic degradation of FBPase 1 [Ill. The idea has also been developed that the major role of cyclic AMP in lower eucaryotic cells could be to stimulate glycolysis rather than to inhibit it [12], as it does in the liver [I]. The present work was, therefore, undertaken in order to reinvestigate the respective roles of cyclic AMP and of Fru6Pin the control of the concentration of Fru(2,6)P2 in yeast. Part of this work was presented in a symposium [13].

MATERIALS AND METHODS

Materials

The C subunit of cyclic-AMP-dependent protein kinase type I1 from beef heart was kindly provided by Dr F. Hofmann (Heidelberg, FRG). Fru(2,6)P2 [ 141 and fructose 2,6-[2-32P]bisphosphate [15] were prepared as described. "'1- labelled cyclic AMP radioimmunoassay kit was from New

188

England Nuclear (Boston, MA, USA), acridine orange from Gurr (Pharbil, Rotterdam, The Netherlands), auramine 0 and safranine from Janssen Chemica (Beerse, Belgium), yeast extract and bactopeptone from Difco (Detroit, MI, USA), glucose, protamine sulfate and dithiothreitol from Fluka (Buchs, Switzerland), DEAE-cellulose from Whatman (Maid-

assays were performed at 30°C. One unit is the amount of enzyme that catalyzes the conversion 1 pmol of substrate in 1 rnin under the conditions of the assay. The concentration of protein was determined by the method of Bradford [18] with bovine serum albumin as a standard.

stone, England), Sephadex G-25 (fine) and blue Sepharose CL6-B from Pharmacia (Uppsala, Sweden). Auxiliary enzymes and other biochemicals were from Boehringer (Mannheim, FRG).

Organisms and growth conditions

Saccharomyces cerevisiae X2180, a gift of Dr C. Gancedo (Madrid, Spain), was grown with aeration at 30°C in a medi- um containing 1% yeast extract, 2% bactopeptone and 2% glucose. S. cerevisiae BE 333, a temperature-sensitive mutant deficient in adenylate cyclase, kindly provided by Dr F. Hilger (Gembloux, Belgium) was grown in the same medium but the temperature was 26°C. Yeasts were grown to the stationary phase (22 - 25 mg/ml) and collected by centrifugation, washed once with water and resuspended at approximately 25 mg/ml in 50 mM Mes buffer, adjusted to pH 6.0 with KOH. Cells were incubated at 30"C, except when indicated.

Preparation of extracts and determination of enzyme activities

Cell samples (100-150 mg in 5 ml) were poured on 5 ml ice-cold water, mixed and centrifuged for 1 min. The pellets were then frozen in a mixture of solid C 0 2 and acetone and stored at -20°C until used (no more than two days). Disrup- tion of the cells was obtained by shaking vigorously the pellets in 0.6 ml of 50 mM potassium phosphate, pH 7.5, containing 500 mM KCl, 2 mM EDTA, 1 mM dithiothreitol, 1 mM phenylmethylsulfonyl fluoride and 20 mM NaF, with 1 g glass beads (0.5 mm diameter) for four 1-min periods at 1-min intervals. During the intervals, the tubes were kept on ice. The extracts were then centrifuged in the cold for 3 rnin at 1000 x g and the supernatants were centrifuged again for 15 min at 5000 x g. The final supernatants were used for enzyme assays. This procedure was used in the experiments described in Fig. 2 and 4.

Glucose-6-phosphate dehydrogenase and fructose 1,6- bisphosphatase were assayed as described by Gancedo and Gancedo [ 161. Trehalase was assayed in a mixture containing 50 mM Hepes (pH 7.1), 25 mM trehalose and 2.5 mM CaCl,. The reaction was stopped by heating the mixture for 3 rnin in boiling water. The glucose liberated was determined by the glucose oxidase method [17]. PFK 2 was assayed by the pro- duction of Fru(2,6)P2 in a mixture containing 50 mM Tris/ acetate, pH 7.8, 10 mM magnesium acetate, 2 mM potassium phosphate, 7 mM ATP, 8 mM Glc6P and 2 mM Fru6P. After various times, samples (0.125 ml) were mixed with 1 vol. of 0.1 M NaOH, heated at 80°C for 10 rnin and used for the determination of Fru(2,6)P2. Because in the crude extract the substrates of PFK 2 were also rapidly utilized by PFK 1 and other enzymes, the assay was only linear during 15 rnin even when performed with less than 0.25 mg protein/ml assay mixture. Under these conditions, the amount of Fru(2,6)Pz formed did not exceed 50 pmol/ml with the inactive form of PFK2 and 250pmol with the active form. FBPase 2 was assayed by the production of [32P]Pi from [2-32P]Fru(2,6)Pz [15] in a mixture containing 50 mM Hepes pH 7.1, 50 mM KC1, 2 mM MgCl, and 1 pM [2-32P]Fru(2,6)Pz. All enzymic

Measurement of Fru(2,6)P2, metabolites and cyclic AMP

A cell sample ( 5 ml containing about 100 mg yeast) was collected on a cellulose nitrate filter (47 mm diameter, pore size 0.45 pm; Gelman Sciences, Ann Arbor, MI, USA). The cake was scraped off and frozen in liquid nitrogen within less than 10 s after collection of the sample. The frozen pellet was weighed, mixed with 1 ml of 0.1 M NaOH, and heated for 15 min at 80°C. The mixture was clarified by centrifugation and Fru(2,6)P2 was determined as previously described [19,20].

Another sample (10 ml), collected as described above, was mixed with 2 ml of 15% perchloric acid and frozen and thawed four times [21]. The extract was centrifuged and the supernatant was neutralized and used for the determination of hexose-6-P [22] and of cyclic AMP after acetylation [23]. The concentration of metabolites are expressed as mol/g wet weight and can be converted concentrations in the cell sap by multiplying by 1.67 [24].

Purification of PFK 2

PFK 2 was purified by the following procedure performed in the cold. S. cerevisiae X2180 were harvested in the stationary phase of growth approximately 10 h after exhaus- tion of glucose. 10 g of cells were homogenized for two periods of 30 s in a cell homogenizer MSK (B. Braun, Melsungen, FRG) with 50 g glass beads (0.5 mm diameter) in 10 ml of the same buffer as described before. The beads were removed, washed with 20 ml of buffer. The washing and the extract were then pooled and centrifuged at 25000 x g for 20 min. The pellet was discarded (this type of extract was used in the experiment shown in Fig. 5). To the supernatant, a freshly prepared 1 % solution of protamine sulfate in 20 mM Tris/Cl, pH 7.4, 20 mM KCI, 0.5 mM dithiothreitol and 0.2 mM phenylmethylsulfonyl fluoride (buffer A) was added dropwise to a final concentration of 0.2%. After 10 rnin of stirring, the mixture was centrifuged at 25000 x g for 20 rnin and the precipitate discarded. The supernatant was passed through 25 vol. of Sephadex G-25 fine equilibrated in buffer A. The filtrate was chromatographed on a DEAE-cellulose column (1.4 x 16 cm) which had been equilibrated with the same buf- fer. Proteins were eluted from the column with a linear gradient of KC1 (0 - 0.5 M in a total volume of 100 ml). PFK 2 was eluted in one peak at bout 100mM KCl. The active fractions were pooled and applied on a column of blue Sepharose (0.9 x 6 cm) equilibrated with buffer A. The column was washed successively with 10 ml of buffer A, 10 ml of buffer A containing 100 mM KCl, 10 ml of buffer A con- taining 250 mM KC1 and finally buffer A containing 1 M KC1. Fractions of 1 ml were collected. PFK 2 was only eluted at the highest salt concentrations. The active fractions were pooled and stored at -80°C. The result of this purification is summarized in Table 1.

Activation of PFK 2

Purified PFK 2 (0.28 mg/ml) was incubated in 10 mM potassium phosphate (pH 7.5) with 10 pg/ml of catalytic sub- unit of cyclic-AMP-dependent protein kinase in the presence

189

Table 1. Purification of PFK 2 from Saccharomyces cerevisiae

Step Volume Protein Total Specific Puri- Yield activity activity fication

ml mg mU mU/mg protein -fold % Extract 26 495 5.33 0.01 1 1 100 Protamine sulfate supernatant 25 250 5.52 0.022 2 104 DEAE-cellulose chromatography - 10 52 4.55 0.087 7.9 85 Blue Sepharose chromatography 2.8 1.5 2.4 1.60 145 45

Fig. 1. Effect of glucose, added with or without acridine orange, on the Concentration of cyclic A M P ( A ) , hexose-6-P and Fru(2.6)P2 ( B ) in S. cerevisiae. The cell suspension was incubated at 30°C for 15 min. Glucose was then added at a final concentration of 0.1 M, with or without 0.15 mM acridine orange. The zero time samples were withdrawn immediately before the addition of glucose

of 1 mM ATP-Mg. After 15 min, the reaction mixture was diluted 15-fold with buffer A containing 1 mg/ml bovine serum albumin. Samples of this mixture were used for PFK 2 assays.

RESULTS

The concentration of phosphate esters in yeast cells incubated under various conditions

As shown in Fig. 1 and in confirmation of previous work, the addition of glucose to a suspension of Saccharomyces cerevisiae in stationary phase caused a rapid and sustained increase in the concentration of hexose-6-P and Fru(2,6)P2 [4] and a more transient increase in the concentration of cyclic AMP [5] . The same experiments were also performed in the presence of 0.15 mM acridine orange because this dye had been found to prevent the activation of trehalase under similar conditions (personal communication of Dr J. Thevelein, Leuven, Belgium); the effect of glucose in increasing cyclic AMP and Fru(2,6)P2 concentrations was then greatly decreased whereas its effect on hexose-6-P was unaltered. These results suggested that the effect of glucose in increasing the concentration of Fru(2,6)P2 in yeast is mediated by cyclic AMP rather than by Fru6P. The glucose effect was therefore studied in a temperature-sensitive mutant unable to form cyclic AMP at 35°C but well able to at 26°C. Fig. 2 shows that in this mutant, Fru(2,6)P2 was formed at 26°C but not at 35 "C, despite the fact that hexose-6-P increased to a similar level at both temperatures. The same figure also shows that, as expected, trehalase could be activated at 26°C but not at 35°C.

Incubation of S. cerevisiae in the presence of dinitrophenol is another way to increase its content in cyclic AMP [25]. This is confirmed by Fig. 3 which, in agreement with a previous

Fig. 2. Effect of glucose on the concentration of Fru(2,6)P2 ( A ) and hexose-6-P (Bj and on the activity of trehaiase ( C ) in a temperature- sensitive mutant of s. cerevisiae, lacking adenylate cyclase activity at 35°C. The cells were grown at 26"C, resuspended in Mes buffer and incubated at the indicated temperature. Glucose was added after 1 h at a final concentration of 0.1 M

observation made by Dr C. Gancedo (personal communica- tion), shows a moderate but definite increase in Fru(2,6)P2 concentration and a minimal change in the concentration of hexose-6-P, associated with a very large increase in cyclic AMP.

190

Fig. 3. Efect of 2,4-dinitrophenol on the concentration of cyclic AMP (A), hexose-6-P and Fru(2,6)P2 ( B ) in S. cerevisiae. General procedure as in Fig. 1 except that dinitrophenol was added instead of glucose, as a 0.1 M ethanolic solution to reach a final concentration of 2 mM

Fig. 4. Effect of ( A ) glucose and ( B ) dinitrophenol on the activity of PFK 2 and FBPase 1 in S. cerevisiae. Same procedure as in Fig. 1 and 3

The activation of PFK 2 by glucose and other means in S. cerevisiae

The addition of gIucose (Fig. 4A) or of dinitrophenol (Fig. 4B) to the incubation medium caused a progressive in- crease in the activity of PFK 2 and a more rapid decrease in the activity of FBPase 1 but no change in the activity of glucose-6-phosphate dehydrogenase (not shown) which was taken as a reference enzyme. The larger activation of PFK 2 by glucose than by dinitrophenol (Fig. 4) was also observed in two other similar experiments. The effects of glucose, although not those of dinitrophenol, were in great part pre- vented by the addition of 0.15 mM acridine orange to the cell suspension (see Fig. 4 in [13]). Auramine 0 and safranine had effects similar to those of acridine orange (not shown). The activation of PFK 2 by glucose could not be reversed by an extensive gel filtration of the extract (not shown).

Experiments with a cell-free system

When an extract of S. cerevisiae (1 1 mg proteinlml) was incubated in. the Dresence of ATP-MP and cvclic AMP. the

Fig. 5 . Cyclic-AMP-dependent activation of PFK2 in a cell-free extract of S. cerevisiae. The incubation mixture contained 11 mg protein of yeast extract/ml, 2 mM ATP, 5 mM MgS04 and, as indicated, 10 pM cyclic AMP. The same concentration of ATP and cyclic AMP were added again at 20 min, as indicated. Samples (20 pl) were mixed with 20 vol. cold solution containing 10 mM Pi, 1 mM dithiotheitol, 5 mM EDTA and 20 mM NaF, pH 7.5. 0.1 ml of this mixture was used for the assay of PFK 2 performed in a total volume of 0.5 ml

activity of PFK 2 was increased more than 10-fold within 1-2 min (Fig. 5) and then decreased again with an half-life of about 10 min. The enzyme could be reactivated if cyclic AMP was added again and still better if cyclic AMP and ATP-Mg were both reintroduced in the mixture. No activation occurred in the absence of cyclic AMP and a half-maximal effect was reached with a concentration of cyclic AMP close to 1 pM (not shown). Other similar experiments also confirmed that, as previously observed by others [5 ] , trehalase was also greatly activated in a cell-free extract incubated in the presence of cyclic AMP and ATP. This activation was however not followed by inactivation within the following 30 min (see Fig. 6 in 1131).

Experiments with partially purified PFK 2 The incubation of the purified PFK 2 with subunit C of

cyclic-AMP-dependent protein kinase in the presence of ATP- Mg changed several of its kinetic properties. As shown on Fig. 6, there was a 4.3-fold increase in V and a 2-fold decrease in K,,, for Fru6P whereas K , for ATP was not affected. Fig. 7

" shows that both forms were inhibited by sn-glycerol-3-P lbut

191

Fig. 6. Kinetic properties of PFK 2. The enzyme was purified as reported in Table 1 and used either as such (0) or after incubation in the presence of ATP-Mg and subunit C of cyclic-AMP-dependent protein kinase (0). The activity was measured at 5 mM ATP in (A) or 5 mM Fru6P in (B). Calculated V for the non-activated and the activated enzymes was 3 and 13 mU/mg protein and K,,, for Fru6P was 1.3 and 0.6 mM. K, for ATP was 0.5 mM for both enzymes

Fig. 7. Inhibition of PFK 2 by sn-glycerol-3-P. The activity was mea- sured in the presence of 5 mM ATP and 0.5 mM Fru6P

that the activated enzyme was less sensitive to this inhibition than the non-activated one. In both cases, sn-glycerol-3-P caused a decrease in Vand a slight increase in K, (not shown). Citrate, AMP or P-enolpyruvate (2mM) had no effect on either form of PFK 2 (not shown). Maximal activity was observed at pH 7.8 with both the non-activated and the activated PFK 2 and approximately 50% of this value was observed at pH 6.0. In agreement with observations made by M. Laloux, E. Van Schaftingen and H. G. Hers (unpublished results) on liver PFK 2, the yeast enzyme was completely inactive in the near absence of Pi.

FBPase 2 The activity of FBPase 2 in the yeast extract was barely

detectable (about 1 pU/mg protein). The purified non-activat- ed preparation of PFK 2 still contained 30 yU/mg protein of the same enzyme.

DISCUSSION

The determination of PFK 2 activity in yeast extracts

Because of the high activity of PFK 1 and possibly of other enzymes that consume ATP and Fru6P, the rate of

PFK 2 reaction in a yeast extract remains linear for only a short period of time, during which maximally 250 pmol Fru(2,6)Pz can be formed per ml (see Materials and Methods). In the present work, PFK 2 could be measured with great precision thanks to the exquisitely sensitive analytical proce- dure used for the determination of Fru(2,6)Pz. Clifton and Fraenkel [26] could avoid this problem by removing PFK 1 by immunoprecipitation with specific antibodies. The activity that they reported is in good agreement with ours, if one takes into account that these authors performed their assay at lower Fru6P concentrations and at lower temperature. Other in- vestigators [27,28] did not take this problem into account and for this reason markedly underestimated PFK 2 activity in the crude extract [27] or were even unable to measure it [28]. It is probably because they largely underestimated the activity of their starting material that Yamashoji and Hess [28] could report a 4000-fold purification of a PFK 2 preparation, the specific activity of which was similar to that reported in Table 1.

It must also be recalled that the fact that Fru(2,6)Pz is formed in a crude extract in the presence of ATP and Fru6P is not by itself a demonstration of PFK 2 activity, since this conversion could occur through various indirect mechanisms. The fact that this activity has now been purified 150-fold with a yield of 45% strongly argues in favour of the same mecha- nism operating in the crude preparation.

Factors that determine the concentration of Fru(2,6)P2 in Saccharomyces cerevisiae

When Fru(2,6)Pz was first found in yeasts cells [4], its formation was closely related to glucose availability and, by analogy with what was known for the liver, was interpreted as a consequence of the large increase in Fru6 P concentration. More recently, Clifton and Fraenkel[26] also reported a corre- lation between the levels of Fru(2,6)P2 and Fru6P. From the data of Fig. 1 and 3, it is now apparent that the concentration of these two phosphoric esters can be dissociated under various experimental conditions and that cyclic AMP rather than Fru6P is clearly the determining factor. Indeed, when the formation of cyclic AMP was inhibited by the presence of acridine orange in the incubation medium or by a repressive temperature in a temperature-sensitive mutant, little Fru(2,6)P2 was formed, despite a large increase in hexose-6- P. Acridine orange was in this respect a most valuable tool,

192

which allowed the cyclic-AMP-mediated effects of glucose to be inhibited. Its inability to counteract the dinitrophenol effect may be due to insufficient dosage and requires more investiga- tion.

Fru(2,6)P2 was also formed, although in smaller amount, in the absence of glucose and in the presence of dinitrophenol; under these conditions, some glucose was presumably formed from trehalose, since trehalase was activated [29].

Our observation that, despite a striking parallel (or even sequential [4]) increase in the concentration of hexose-6-P and Fru(2,6)Pz, these two changes are not causally related, might also apply to other circumstances in which such a parallelism was observed. This is the case for fungal spores incubated in the presence of glucose or 6-deoxyglucose after the breakage of their dormancy [30] and also for seeds after imbibition [13]. It must also be recalled that in slices of Jerusalem artichoke tuber, there was a dramatic increase of Fru(2,6)P2 without a change in Fru6P concentration [20]. All this indicates that Fru(2,6)Pz is not simply a signal of glucose availability but may have still unsuspected roles in biology.

Fig. 8. Control of glycolysis and gluconeogenesis by Fru(2,6)P2 and cyclic AMP in S. cerevisiae

Yeast PFK 2, an interconvertible enzyme

It was already apparent from the work of Clifton and Fraenkel [26], that the activity of PFK 2 in S. cerevisiae can vary as much as 10-fold according to the incubation con- ditions and that the presence of glucose was an important factor which increased the enzymic activity. From experiments performed with intact cells (Fig. 4), cell-free extract (Fig. 5) and purified enzyme (Fig. 6), it is now clearly apparent that yeast PFK 2, like its liver counterpart, is a substrate for cyclic- AMP-dependent protein kinase and that its phosphorylation can greatly affect its activity. However, in deep contrast with the liver enzyme, yeast PFK 2 is activated by phosphorylation, allowing cyclic AMP to be a potent stimulator of glycolysis. This phosphorylation is most likely reversed by a phosphatase, since, in the cell-free extract, the activation is only transient but can be repeated upon a new addition of cyclic AMP. This protein phosphatase is relatively specific for PFK 2, since it does not cause a simultaneous inactivation of trehalase or reactivation of FBPase 1. It is remarkable, however, that the reverse situation was observed in intact cells, in which the activation of PFK 2 was more persistent than that of trehalase. The fact that the activation of PFK 2 was larger in the presence of glucose than in the presence of dinitrophenol (Fig. 4), although more cyclic AMP is formed in the second condition than in the first, further indicates that factors other than cyclic AMP can play a role in the activation mechanism.

The activation of PFK 2 by phosphorylation results from both a change of V and a decrease in K, for Fru6P. The relatively high value of K, even after the phosphorylation of the enzyme, explains that in yeast incubated in the presence of dinitrophenol, the formation of Fru(2,6)P2 in the cells appears to be limited by the availability of its substrate, Fru6P (see Fig. 1 B).

The control of glycolysis in S. cerevisiae

Fig. 8 summarizes the mechanism by which glucose (and possibly other agents) stimulates glycolysis and inhibits gluconeogenesis in S. cerevisiae. Glucose enters the cells by a mechanism which has been described as facilitated diffusion [31] but could also be more complex (reviewed in [29]); it then becomes available for glycolysis, the rate of which is regulated

at the level of PFK 1, essentially by Fru(2,6)P2. The passage of glucose through the membrane appears to cause a transient activation of adenylate cyclase, which is believed to be second- ary to a depolarization of the membrane [25]. This activation is counteracted by acridine orange and other basic dyes and is also obtained in the presence of dinitrophenol or other ionophores [25]. Cyclic AMP is then formed in the cell and activates cyclic-AMP-dependent protein kinase. In yeast, as in mammalian cells, this enzyme is made of catalytic and regulatory subunits which dissociate in the presence of cyclic AMP [32]. The free catalytic subunit can then phosphorylate a series of substrate proteins. One of them is trehalase [5, 101 which is simultaneously activated and can then form free glucose from endogenous trehalose, explaining the formation of hexose phosphate in the absence of exogenous glucose (Fig. 3). Another substrate protein is PFK 2 which, in deep contrast with the liver enzyme, is activated by phosphoryla- tion and can then catalyze the formation of Fru(2,6)Pz from Fru6P. Fru(2,6)Pz stimulates PFK 1 and initiates glycolysis; it also inhibits FBPase 1 and arrests gluconeogenesis. A third substrate for cyclic-AMP-dependent protein kinase is FBPase 1, but remarkably, its phosphorylation also requires Fru(2,6)Pz [9, 331. The phosphorylation of FBPase 1 causes its partial (30-50%) inactivation and is also believed to be a prerequisite for further proteolysis and complete inactivation of that enzyme [8]. There are, therefore, three mechanisms by which FBPase 1 is rendered inactive in the presence of glucose: (a) inhibition by Fru(2,6)P2, (b) phosphorylation and partial inactivation dependent on cyclic AMP and Fru(2,6)Pz and (c) proteolysis.

In summary, it appears that, in yeast, glucose and cyclic AMP act synergistically to form Fru(2,6)P2, initiate glycolysis and arrest gluconeogenesis. This effect of cyclic AMP is, there- fore, in deep contrast with the general idea that cyclic AMP is a hunger signal that signifies the absence of glucose, a concept which seems to apply to procaryotes and to the liv& [l], but apparently not to lower eucaryotes [12].

The authors are grateful to Dr C. Gancedo and F. Hilger for the yeast strains and to Dr F. Hofmann for subunit C of cyclic-AMP- dependent protein kinase. This work was supported by the Fonds de la Recherche Scientifique medicale and by the United States Public Health Service (grant AM 9235). J. F. is Aspirant and E. V. S. Charge de Recherches of the Belgian Fonds National de la Recherche Scientifique.

193

REFERENCES 1. Hers, H. G. &Van Schaftingen, E. (1982) Biochem. J . 206,l-12. 2. Hers, H. G. &Van Schaftingen, E. (1984) Adv. Cyclic Nucleotide

Res. 17, 343 - 349. 3. Claus, T.-H., El-Maghrabi, M. R., Regen, D. M., Steward, H.

B., McGrane, M., Kountz, P. D., Nyfeler, F., Pilkis, J. & Pilkis, S. J. (1984) Curr. Top. Cell. Regul. 23, 57-86.

4. Lederer, B., Vissers, S., Van Schaftingen, E. & Hers, H. G. (1981) Biochem. Biophys. Res. Commun. 103, 1281-1287.

5. Van der Plaat, J. B. & Van Solingen, P. (1974) Biochem. Biophys. Res. Commun. 56, 580- 587.

6. Lenz, A. G. & Holzer, H. (1980) FEBSLett. 109,271 -274. 7. Londesborough, J. (1982) FEBS Lett. 144,269-272. 8. Purvin, C., Leidig, F. & Holzer, H. (1982) Biochem. Biophys. Res.

9. Gancedo, J. M., Mazon, M. J. & Gancedo, C. (1983) J . Biol.

10. Uno, I., Matsumoto, K., Adachi, K. & Ishikawa, T. (1983) J .

11. Funayama, S., Gancedo, J. M. & Gancedo, C. (1980) Eur. J .

12. Pall, M. L. (1984) Mol. Cell. Biochem. 58, 187-191. 13. Hers, H. C., FranGois, J. &Van Schaftingen, E. (1984) Curr. Top.

14. Van Schaftingen, E. & Hers, H. G. (1981) Eur. J . Biochem. 117,

15. Van Schaftingen, E., Davies, D. R. & Hers, H. G. (1982) Eur. J .

16. Gancedo, J. M. & Gancedo, C. (1971) Arch. Microbiol. 76,133-

Commun. 107, 1482- 1489.

Chem. 258, 5998 - 5999.

Biol. Chem. 258, 10867-10872.

Biochem. 109,61-66.

Cell. Regul.. in press.

319 - 323.

Biochem. 124, 143 - 149.

138.

17. 18. 19.

20.

21.

22.

23.

24.

25. 26.

27.

28. 29. 30.

31.

32.

33.

Huggett, A. St. G. & Nixon, D. A. (1957) Biochem. J . 66, 12p. Bradford, M. (1976) Anal. Biochem. 72,248-254. Van Schaftingen, E., Lederer, B., Bartrons, R. & Hers, H. G.

(1982) Eur. J. Biochem. 129, 191 - 195. Van Schaftingen, E. & Hers, H. G. (1983) FEBS Lett. 164, 195-

200. Saez, M. J. & Lagunas, R. (1976) Mol. Cell. Biochem. 13, 73-

78. Hohorst, H. J. (1963) in Methods oJEnzymatic Analysis (Berg-

meyer, H. U. ed.) pp. 134-139, Verlag Chemie, Weinheim/ Bergstr.

Steiner, A. L., Parker, C. W. & Kipnis, D. M. (1972) J . Biol. Chem. 247,1106-1113.

Gancedo, J. M. & Gancedo, C. (1973) Biochimie (Paris) 55,

Trevillyan, J. M. & Pall, M. L. (1979) J. Bacteriol. 138,397 -403. Clifton, D. & Fraenkel, D. G. (1983) J . Biol. Chem. 258,9245-

Furuya, E., Kotaniguchi, H. & Hagihara, B. (1982) Biochem.

Yamashoji, S. & Hess, B. (1984) FEBS Lett. 172, 51 -54. Thevelein, J. M. (1984) Microbiol. Rev. 48,42- 59. Van Laere, A., Van Schaftingen, E. & Hers, H. G. (1983) Proc.

Heredia, C. F., Sols, A. & DelaFuente, G. (1968) Eur. J. Biochem.

Hixson, C. S . & Krebs, E. G. (1980) J. Biol. Chem. 255, 2137-

Pohlig, G., Wingender-Drissen, R., Noda, T. & Holzer, H. (1983)

205 - 21 1.

9249.

Biophys. Res. Commun. 105, 1519 - 1523.

Natl Acad. Sci. USA 80,6601 -6605.

5,321 - 329.

2145.

Biochem. Biophys. Res. Commun. 115,317-324.

J. FranGois, E. van Schaftingen, and H.-G. Hers, Laboratoire de Chimie Physiologique de YUniversite Catholique de Louvain, and International Institute of Cellular and Molecular Pathology, UCL 7539, Avenue Hippocrate 75, B-I 200 Bruxelles, Belgium

* The abbreviation as typed by WGB corresponds to the official abbreviation listed in Biosis.