Eur. J. Biochem. 134, 269-273 (1983) 8 FEBS 1983

On the Mechanism of Inhibition of Neutral Liver Fructose 1,6-Bisphosphatase by Fructose 2,6-Bisphosphate

Jean FRANCOIS, Emile VAN SCHAFTINGEN, and Henri-Gery HERS

Laboratoire de Chimie Physiologique, Universite de Louvain and International Institute of Cellular and Molecular Pathology, Brussels

(Received March 17IApril 22, 1983) - EJB 83 0259

The inhibitory effect of fructose 2,6-biphosphate on fructose 1,6-bisphosphatase was reinvestigated in order to solve the apparent contradiction between competition with the substrate and the synergism with AMP, a strictly non- competitive inhibitor. The effect of fructose 2,6-bisphosphate was compared to that of other ligands of the enzyme, which, like the substrate and methyl (a + @fructofuranoside 1,6-bisphosphate bind to the active site or which, like AMP, bind to an allosteric site. An increase in temperature or pH, or the presence of sulfosalicylate, lithium or higher concentrations of magnesium as well as partial proteolysis by subtilisin increased [I],, for fructose 2,6-bisphosphate and AMP without affecting K,,,. With the exception of the pH change, all these conditions were also without effect on the affinity of the enzyme for the competitive inhibitor, methyl (a + P)fructofuranoside 1,6-bisphosphate. These observations can be explained by assuming that fructose 2,6-bisphosphate has no affinity for the active site of fructose 1,6-bisphosphatase but binds to an allosteric site which is different from the AMP site. Fructose 2,6-bisphosphate is therefore classified as an allosteric competitive inhibitor and a model is proposed which explains its synergism with AMP as well as the various cooperative effects.

Fructose 2,6-bisphosphate, initially discovered as a stimu- lator of liver phosphofructokinase [I, 21 (reviewed in [3]) is also an inhibitor of fructose 1,6-bisphosphatase from various sources [4-61. The mechanism of the inhibition of the liver enzyme remains, however, controversial. An allosteric type of in- teraction was suggested by the fact that fructose 2,6-bisphos- phate changes the saturation curve of the enzyme for its substrate from hyperbolic to sigmoidal [4]. Consequently, the inhibition by fructose 2,6-bisphosphate is much more impor- tant at low than at high substrate concentrations and V,,, is not significantly affected. This competitive aspect of the inhibition, together with the obvious structural similarity between the inhibitor and the substrate, has led some in- vestigators [5] to believe that the inhibition was due to the fixation of fructose 2,6-bisphosphate to the active site as in a truly competitive inhibition. This conclusion has been criticized on a technical basis [3]. It is also not in accordance with the fact [4, 51 that the effect of fructose 2,6-bisphosphate is synergistic with that of AMP, an allosteric and non-competitive inhibitor of the enzyme [7]. Nevertheless, several groups of workers [8- 101 have more recently reinvestigated the problem and favoured the competitive mechanism.

In the present work, we compare the effect of fructose 2,6- bisphosphate with that of other ligands of the enzyme, which, like the substrate or methyl fructofuranoside 1,6-bisphosphate [I 11, bind to the active site, or which, like AMP [7], bind to an allosteric site.

MATERIALS AND METHODS

Chemicals and Enzymes

Phenylmethylsulfonyl fluoride and auxiliary enzymes were from Boehringer (Mannheim, FRG) and subtilisin (Carls-

Enzymes. Fructose l,6-bisphosphatase (EC 3.1.3.1 1); glucosephos- phate isomerase (EC 5.3.1.9); glucose-6-phosphate dehydrogenase (EC 1.1.1.49); subtilisin (EC 3.4.21.14).

berg, protease type VIII) from Sigma (St. Louis, USA). Sulfosalicylic acid was from Merck (Darmstadt, FRG) and was converted to its potassium salt with KOH. Fructose 1,6- [l-32P]bisphosphate was synthesized from [ Y - ~ ~ P I A T P [12] and purified as in [13]. D-Fructose 2,6-bisphosphate was prepared as in [14]. Methyl (a + fl)-D-fructofuranoside 1,6-bisphosphate was a generous gift from Dr S. J. Benkovic of the Pennsylvania State University. It was assayed spectrophotometrically by the reduction of NADP in the presence of fructose 1,6-bisphos- phatase, glucose phosphate isomerase, and glucosed-phos- phate dehydrogenase as the sum of fructose 1,6-bisphosphate and fructose 6-phosphate formed by hydrolysis in 0.1 M HC1 at 80°C for 4 h. Its inhibitory effect on rabbit liver fructose 1,6-bisphosphatase has been reported to be competitive with a Ki value of 7.1 & 3.0 pM at pH 9.2 [ll]. We found the same value with the rat liver neutral fructose 1,6-bisphosphatase measured at pH 7.4. The source of other chemicals and enzymes was as in [4].

Purification and Assay of' Fructose 1,6-Bisphosphatase

Fructose 1,6-bisphosphatase was purified from rat liver according to Tejwani et al. [15] except that the elution medium of the second CM-cellulose step contained 10 mM malonate, pH 6.0, 1 mM EDTA, 5 mM 2-mercaptoethanol, 1 mM fruc- tose 1,6-bisphosphate and 0.2 mM AMP. The active fractions were pooled and made 40 in glycerol; the preparation could be stored at - 20 "C for more than two months without loss of activity. The yield was 40 - 60 %. Before use, the preparation was filtered on Sephadex G-25 equilibrated with a solution containing 50 mM Tris/Cl, pH 7.4,O.l mM EDTA and 50 mM KCl and eluted with the same solution. The activity at pH 7.4 was 5-fold that at pH 9.5; this indicates that no proteolysis occurred during the purification. Proteolytic degradation of the purified fructose 1,6-bisphosphatase was performed with sub- tilisin in the conditions described by Traniello et al. [16]; the reaction was stopped by the addition of 2 mM phenylmethyl-

270

sulfonyl fluoride when the activity at pH 9.5 was 4.5-fold that at pH 7.4.

Fructose 1,6-bisphosphatase activity was assayed in the presence of 50 mM Tris/Cl, 100 mM KCl, 2 mM MgSO,, 0.1 mM EDTA and, unless otherwise indicated, in the presence of 25 pM fructose 1,6-bisphosphate at pH 7.4 and 30 "C. The reaction was started by addition of the substrate. In most experiments, the rate of reaction was measured spectrophoto- metrically in 1 ml containing also 0.2 mM NADP, 5 pg of glucose-6-phosphate dehydrogenase and 10 pg of phospho- glucose isomerase. Under these conditions, 1 mg of the purified enzyme preparation catalyzed the hydrolysis of 15 pmol substrate/min. In the experiments performed with methyl ( M + p)fructofuranoside 1,6-bisphosphate, the reaction was run in a total volume of 0.15ml and fructose 1,6-[l-"PI- bisphosphate was used as substrate. [32P]Pi was then measured as described by McClard [13]. & and [I],., values werc ob- tained graphically from Lineweaver-Burk and Hill plots respectively.

RESULTS

Effects of Temperature and p H

It is known that the inhibitory effect of AMP [7] and fructose 2,6-bisphosphate [4] is greatly increased with decreas- ing temperature. We show in Fig. 1 that the K, for fructose 1,6-bisphosphate and the affinity for the methyl fructofura- noside 1,6-bisphosphate were not significantly modified by varying the temperature between 18 " and 44 "C, whereas the affinities for AMP and fructose 2,6-bisphosphate were greatly affected.

Changes in pH also affect the activity of neutral fructose 1,6-bisphosphatase and its inhibition by both AMP [7] and fructose 2,6-bisphosphate [4, 51. The 5-fold greater activity of our preparation at pH 7.4 than at pH 9.5 was due to a change in V,,, since K,,, was not affected. By contrast, the affinity for AMP and that for fructose 2,6-bisphosphate were greatly decreased with increasing pH. The interpretation of this observation is, however, not unambiguous, since the affinity for methyl (a + P)fructofuranoside-1,6-bisphosphate was also decreased at alkaline pH. These pH effects are illustrated in Fig. 2.

Effects of Sulfosalicylate

The effect of salicylate to release the inhibition of fructose 1,6-bisphosphatase by AMP was first reported by Marcus [17]. Salicylate had no effect on the activity of the enzyme and was assumed to bind at an hydrophobic adenosine-binding pocket, presumably identical to the AMP site.

As shown in Fig. 3 - 5, we could confirm the observation of Marcus, using sulfosalicylate (which had an effect similar to that of salicylate but at five-times lower concentrations), and extended our investigation to the interaction with fructose 2,6- bisphosphate and the methyl fructofuranoside 1,6-bisphos- phate. I t is apparent from Fig. 3A and B that sulfosalicylate released the inhibition both by AMP and by fructose 2,6- bisphosphate, although much more completely in the first case than in the second. Indeed, whereas the inhibition by even high doses of AMP could be released by an excess of sulfosalicylate, this was not true for the inhibition by fructose 2,6-bisphos- phate, indicating a different mechanism of action. In the presence of 5 mM sulfosalicylate, the residual inhibitory power of fructose 2,6-bisphosphate kept the characteristic properties

? 15

0- 15 25 35 15

T e m p e r a t u r e ( " C i

Fig. 1. Effect o f temperature on the K,, of fructose 1,h-bisphospIiarase and [I]o,5 of its inhibitors

30 I

0 7.0 7.5 8.0 8.5 9.0 9.5

PH

Fig. 2. EjJect ofpHon the K, of:fiuctose I ,b-bisphosphatase and the illo of its inhibitors

observed in the absence of sulfosalicylate. This includes the shift of the substrate saturation curve from hyperbolic to sigmoidal (Fig. 4), the synergism with AMP (Fig. 5 ) and a 10-fold increase in affinity for fructose 2,6-bisphosphate when lowering the temperature from 44" to 18°C (not shown). The inhibition of fructose 1,6-bisphosphatase by the methyl fructo- furanoside 1,6-bisphosphate was not significantly affected by sulfosalicylate (Fig. 3 B).

Effect hf Lithium and of' Magnesium

Lithium is a known inhibitor of fructose 1,6-bisphos- phatase [I81 which has also the property to decrease the affinity of the enzyme for AMP [19]. We found that it decreases V,,, but does not affect K,,, nor the affinity for methyl fructofura- noside 1,6-bisphosphate. It released the inhibition by fructose 2,6-bisphosphate, essentially by decreasing the affinity of the inhibitor for the enzyme (Table 1).

Magnesium is an essential cofactor of fructose 1,6-bisphos- phatase to which it binds cooperatively without interference

211

15

10

- - 'c 5

0 c

Q

m ; o C ._ E -

15 1

1 I

10

5

0

n o U 0

J I I 0 10 2(

lSulfosalicylotel (mM1

Fig. 3. Effect of sulfosalicylate on the inhibition of fructme I,6-his- phosphatase by ( A ) AMP and ( B ) fructose 2,6-bisphosphate and methyl fructofuranoside 1,6-bisphosphaie

control A c- +5mM sulfosalicvlate B I

I +10pM Fru(2,6)% I ~~

0 10 20 30 LO 50 0 100 200 300 LOO 500 [AMPI (pM)

Fig. 5. Effect of fructose 2,6-bisphosphute on the inhibition of fructose 1,6-bisphosphatase by AMP in the absence ( A ) and in the presence ( B ) of 5 mM sulfosalicylate. In the absence of AMP the activities (expressed in pmol . min-' . mg protein-') were 16 in the absence fructose 2,6-bis- phosphate (0) with or without sulfosalicylate, 8.8 with 3 pM fructose 2,6-bisphosphate alone and 9.3 in the presence of 10 pM fructose 2,6- bisphosphate and 5 mM sulfosalicylate

Table 1. Effect of lithium on the K, offructose 1,6-bisphosphatase and on [l]o.5 of its inhibitors

LiCl K , for [I],., for

Fru(l,6)P2 Fru(2,6) P, MeFru(l,h)P,

mM WM

0 4.65 3.60 31.5 0.05 5.10 6.10 30 0.10 5.30 11.50 31.5

Table 2. Effect of magnesium on the K, of fructose 1,6-bisphosphatuse and on of its inhibitors

Fru(l,6)P, Fru(2,6)P, AMP MeFru(l,6)P2)

mM pM

0.5 5 .0 2.15 20 34.5 2.0 5.0 3.15 35 21 5.0 5.6 6.15 50 30.5

0 25 so'ioo [Fructose 1.6-bisphosphatel IpM)

Fig. 4. Inhibition offmciose 1,6-bisphosphatase by fructose 2,6-bisphosphate at various concentrations of subsirate in the presence of 5 mM sulfosalicylute (closed symbols). The open symbols refer to the activity measured in the absence of sulfosalicylate

with the binding of fructose 1,6-bisphosphate whereas the apparent affinity for AMP decreases with increasing Mg2+ concentrations [20]. We observe a similar decrease in affinity for fructose 2,6-bisphosphate but no change in affihity for methyl fructofuranoside 1,6-bisphosphate (Table 2).

Effect of Proteolysis

Proteolysis of fructose 1,6-bisphosphatase by papain [7] or subtilisin [I61 is known to decrease its inhibition by AMP. A similar effect of subtilisin to decrease the inhibition by fructose 2,6-bisphosphate has also been briefly reported, although not

documented, by Pontremoli et al. [lo]. We show in Fig. 6 that when proteolysis by subtilisin had increased [Ao, for AMP 10-fold [Ilo,5 for fructose 2,6-bisphosphate was increased only 2.4-fold. The residual inhibition by fructose 2,6-bisphosphate kept its characteristic properties, including inducing cooper- ativity for the substrate (Fig. 7) and being synergistic with AMP (Fig.6). The K,,, (Fig.7), as well as the Ki for the methyl fructofuranoside 1,6-bisphosphate (not shown), were not modified.

Lack of Synergism between the Inhibitory Action of A M P and qf Methyl Fructofuranoside 1,h-Bisphosphate

Benkovic et al. [21] reported that the inhibitions by methyl fructofuranoside 1,6-bisphosphate and AMP are simply ad- ditive, indicating an absence of synergism. In accordance with this, we found no difference in the relative inhibition by methyl fructofuranoside 1,6-bisphosphate when the assay was per-

212

" 0 5 10 15 20 25 0 100 200 300 LOO 500

lFru(2.6)P21 IpM) IAMPl IpM)

Fig.6. Eflect of proteolysis by subtilisin on the inhibition of fructose 1,6-bisphosphatase by fructose 2,6-bisphosphate ( A ) and by AMP (B) and by an association of the two inhibitors ( A ) . In the absence of inhibitors, the activities were 13.6 and 3.4 pmol . min-' . mg protein-' for the control and proteolysed enzymes respectively

10 ~r

n 0.1 0.2 0.3 0.L

1//Fruil.6)f21 (pM)

Fig. I . Double-reciprocal plot of the rate of proteolysed fructose 1,6- bisphosphatase in the presence of increasing concentrations of fructose 2,6-bisphosphate

formed in the presence or the absence of 65 pM AMP (not shown).

DISCUSSION

The proposal initially made by Pilkis et al. [5] and sub- sequently supported by other groups of investigators [8 - 101 that fructose 2,6-bisphosphate inhibits neutral liver fructose 1,6-bisphosphatase by competing with the substrate for the active site appears untenable for the following reasons. Firstly, as stressed previously [3, 41, the effect of a truly competitive inhibitor cannot be synergistic with that of AMP, which acts allosterically and in a non-competitive manner, and indeed, we found no synergism between AMP and methyl fructofura- noside 1,6-bisphosphate. Secondly, various factors, like changes in temperature (Fig. 1) or the presence of sulfosali- cylate (Fig. 3 and 4), lithium (Table 1) or magnesium (Table 2), which have no or little effect on X, or on the competitive inhibition by methyl fructofuranoside 1,6-bisphosphate, great- ly modify the affinity for fructose 2,6-bisphosphate and AMP. Thirdly, fructose 2,6-bisphosphate and AMP, but not the substrate or its methyl glycoside [21] protect the AMP allosteric

site against acetylation [6] and proteolysis [lo, 221. Fourthly, limited proteolysis decreases inhibition by both AMP [7, 161 and fructose 2,6-bisphosphate [lo] (Fig. 6) whereas it has little effect on K, (Fig.7) or on the affinity for the competitive inhibitor. It is, therefore, remarkable that in numerous con- ditions fructose 2,6-bisphosphate behaved much more like AMP than like the substrate and the competitive inhibitor methyl (a + p)fructofuranoside 1,6-bisphosphate.

Binding of fructose 2,6-bisphosphate to the active site is, by contrast, suggested by the protection that, similarly to fructose 1,6-bisphosphate, the inhibitor exerts on this site under various experimental conditions [6, 8, lo]. This protection could, however, result from a conformational change of the protein caused by the binding of fructose 2,6-bisphosphate at an allosteric site. Furthermore, Ganson and Fromm [9], in an investigation of the reverse reaction, concluded that there was an inhibition by fructose 2,6-bisphosphate which was com- petitive with both fructose 6-phosphate and Pi, but, surpris- ingly, the inhibition constant was severalfold greater for the first substrate than for the second. The authors, however, admitted that their data were kinetically indistinguishable from a situation in which fructose 2,6-bisphosphate binds exclusively to an allosteric site, causing simultaneously release of product or substrates at the active site.

The effect of sulfosalicylate reveals the complexity of the inhibitory mechanism. As already apparent from the work of Marcus [17], the release of the AMP inhibition can be essentially understood as a competitive binding of sulfosali- cylate to the AMP site. The effect of the inhibition by fructose 2,6-bisphosphate is clearly different since the only partial release of inhibition could best be explained by a limited conformational alteration of the site that binds the inhibitor. On the other hand, this incomplete release also suggests that the residual inhibitory effect could be caused by a binding of fructose 2,6-bisphosphate to the catalytic site. The experiments shown in Fig. 4 and 5 as well as the temperature effect, allow this simple hypothesis to be discarded since the residual activity measured in the presence of sulfosalicylate presented the typical characteristics of the inhibition observed in the absence of sulfosalicylate.

As a conclusion, it appears that the interaction of fructose 2,6-bisphosphate with fructose 1,6-bisphosphatase has two major characteristics which must be taken into account in any model that could be proposed. Firstly, fructose 2,6-bisphos- phate does not bind to the active site since, in most conditions tested, its behaviour was quite different from that of the substrate and of a truly competitive inhibitor. Secondly, fructose 2,6-bisphosphate binds to an allosteric site different from the AMP site since the characteristic synergism observed between the two inhibitors requires their simultaneous binding and since their inhibitory effect is quite different in nature. Binding of fructose 2,6-bisphosphate to its specific site excludes to various degrees the occupancy of the active site by the substrate and the two products, Pi and fructose 6-phosphate (see above), and reciprocally. These interactions are respon- sible for the competitive aspect of the inhibition observed when the reaction is measured both in the forward and in the reverse direction and allow fructose 2,6-bisphosphate to be classified as an 'allosteric competitive inhibitor' [23].

The cooperativity between the two inhibitors and also for the substrate when the reaction is measured in the presence of fructose 2,6-bisphosphate can be explained in the terms of the model proposed by Nimmo and Tipton [20]. In this model, two states of the enzyme, called A and M, preferentially bind AMP or magnesium, respectively, but have the same affinity for the

273

substrate. Binding to AMP excludes in great part binding to magnesium and, because of that, decreases the activity. If one assumes that fructose 2,6-bisphosphate binds preferentially to A, the apparent F& of that form would then be increased in the presence of this inhibitor as compared to that of the other form; in other words, fructose 2,6-bisphosphate induces a difference in the affinity of the two forms for the substrate. The model has the advantage of giving a simple explanation for the synergism between AMP, a strictly non-competitive inhibitor, and fruc- tose 2,ti-bisphosphate which decreases the affinity of the enzyme for the substrate. The shift in the equilibrium between the two forms, induced by fructose 2,6-bisphosphate, explains that cooperativity for AMP, as expressed by the Hill coef- ficient, is decreased under these conditions [6]. The model also explains the antagonism between fructose 2,6-bisphosphate and magnesium, and, as said above, the various cooperative effects .

This work was supported by the Fonds de la Recherche Scientifique Medicale and by the U.S. Public Health Service (Grant AM 9235). J.F. is Aspirant and E.V.S. Chargt de Recherches of the Belgian Fonds Nationalde la Recherche Scientifique.

REFERENCES 1. Van Schaftingen, E., Hue, L. & Hers, H. G. (1980) Biochem. J. 192,

2. Van Schaftingen, E., Hue, L. & Hers, H. G. (1980) Biochem. J . 192,

3. Hers, H . G. & Van Schaftingen, E. (1982) Biochem. J . 206, 1 - 12.

887- 895.

897-901.

J. Franqois, E. Van Schaftingen, and H.-G. Hers, Laboratoire de Chimie Physiologique de 1'Universiti Catholique de Louvain, UCL 7539, Avenue Hippocrate 75, B-1200 Bruxelles, Belgium

4. Van Schaftingen, E. & Hers, H. G. (1981) Proc. Natl Acad. Sci. USA,

5 . Pilkis, S. J., El-Maghrabi, M. R., Pilkis, J. & Claus, T. (1982) J . Biol.

6. Pilkis, S . J., El-Maghrabi, M. R., McGrane, M. M., Pilkis, J. & Claus,

7. Taketa, K. & Pogell, B. M. (1965) J . Biol. Chem. 250, 651 -662. 8. Gottschalk, M. E., Chatterjee, T., Edelstein, I. & Marcus, F. (1982) J .

9. Ganson, N. J. & Fromm, H. J. (1982) Biochem. Biophys. Res. Commun.

10. Pontremoli, S. , Melloni, E., Michetti, M., Salamino, F., Sparatore, B. & Horecker, B. L. (1982) Arch. Biochem. Biophys. 218, 609-613.

11. Benkovic, S. J., DeMaine, M. M. & Kleinschuster, J. J. (1970) Arch. Biochem. Bwphys. 139, 248- 251.

12. Walseth, T. F. &Johnson, R. A. (1979) Biochim. Biophys. Acta, 562,

13. McClard, R. W. (1979) Anal. Biochem. 96, 500- 503. 14. Van Schaftingen, E. &Hers, H. G. (1981) Eur. J . Biochem. 117,319-

15. Tejwani, G. A,, Pedrosa, F. D., Pontremoli, S. & Horecker, B. L.

16. Traniello, S., Melloni, E., Pontremoli, S., Sia, C. L. & Horecker, B. L.

17. Marcus, F. (1976) FEBS Lett. 70, 159- 162. 18. Black, W. J., Van Tol, A., Fernando, J. &Horecker, B. L. (1972) Arch.

19. Nakashima, K. & Tuboi, S. (1976) J . Biol. Chem. 251, 4315-4321. 20. Nimmo, H. G. & Tipton, K. F. (1975) Eur. J . Biochem. 58,575-585. 21. Benkovic, S. J., Kleinschuster, J. J., de Maine, M. M. & Siewers, I. J.

(1971) Biochemistry 10, 4881 -4887. 22. Pontremoli, S., Melloni, E., Balestrero, F., De Flora, A. & Horecker,

B. L. (1973) Arch. Biochem. Biophys. 156, 255-260. 23. HalsaU, H. B. (1980) Trends Biochem. Sci. 5, No. 11, IX-X.

78, 2861 -2863.

Chem. 256, 3619- 3622.

T . H. (1981) J . Biol. Chem. 256, 11489-11495.

Biol. Chem. 257, 8016-8020.

108, 233 - 239.

11-31.

323.

(1976) Arch. Biochem. Biophys. 177,253 - 264.

(1972) Arch. Biochem. Biophys. 149, 222-231.

Biochem. Biophys. 151, 576- 590.