European J. Biochem. 7 (1969) 588-593

Modification of Regulatory Properties of Phosphofructokinase by Acetylation A. CHAPMAN,T. SANNER, and A. PIHL Norsk Hydro’s Institutt for Kreftforskning, Montebello, Oslo (Received September 23, 1968)

The effect of acetylation and of sulfhydryl blocking agents on phosphofructokinase from rabbit muscle has been studied. Treatment of phosphofructokinase with N-acetylimidazole led t o rapid loss of its allosteric function, as measured by the ability of AMP to stimulate the ATP-inhibited enzyme. The catalytic activity was far less sensitive to acetylation. The loss of allosteric function on acetylation was due to the fact that the modified enzyme was insensitive to ATP inhibition. The presence of ATP during acetylation prevented the loss of allosteric activity. The results indicate that tyrosine residues may play an important role in the inhibitory binding sites of ATP. Incubation of phosphofructokinase with a 50-fold excess of para-chloromercuribenzoate led t o complete inactivation of the enzyme, an effect which could be reversed by the addition of excess thiol. Sulfhydryl blocking had no effect on the ability of AMP to stimulate the ATPinhibited enzyme. The results confirm that sulfhydryl groups are closely associated with the catalytic site of phosphofructokinase, and demonstrate that they are not involved in the inhibition of the enzyme by ATP and its reversal by AMP.

Phosphofructokinase is considered t o be a key enzyme in the regulation of glycolysis. Phosphofructokinase has been p u r s e d from a number of sources [l-51. The enzyme from rabbit muscle has been studied extensively, both with respect to physical and enzymatic properties [l-3,6-131. It is inhibited by high concentrations of one of its substrates, ATP, a n inhibition which is reversed by AMP [6]. Probably, the sensitivity of phosphofructokinase t o the ratio ATP/AMP is the decisive factor in the regulatory function of the enzyme [14,15]. A considerable amount of work has been carried out on the kinetic effect of various allosteric modifiers on phosphofructokinase. However, little information is available on the nature of the chemical groups which are active in the different sites of the enzyme. It has been shown that sulfhydryl groups are involved in, or closely associated with the catalytic sites of rabbit muscle phosphofructokinase [ll]. However, it is not known whether sulfhydryl groups play any role in the allosteric function of the enzyme, and studies of the importance of other amino acid residues for the different functions of phosphofructokinase do not seem to have been reported. The purpose of the present paper is to report on the effect of acetylation on the properties of rabbit Unusual Abbreviations. 5,5‘-Dithio-bis(2-nitrobenzoic acid), DTNB ; para-chloromercuribenzoate, pCMB. Enzyme. Phosphofructokinase or ATP: n-fructose 6-phosphate 1-phosphotransferase(EC 2.7.1.11).

muscle phosphofructokinase and to present further data on the role of sulfhydryl groups in the enzyme. MATERIALS AND XETHODS Materials Phosphofructokinase from rabbit muscle was purchased from C. F. Boehringer & Soehne (Mannheim, Germany). The enzyme had a specific activity of 40-50 pmole/min per mg of protein [I]. Aldolase, triosephosphate isomerase, and p-glycerophosphate dehydrogenase were obtained from Boehringer. The barium salt of fructose 6-phosphate (Schwartz BioResearch, Orangeburg, New York) was converted to the sodium salt. ATP, AMP, cyclic adenosine 3‘:5’monophosphate, NADH, p-chloromercuribenzoate (pCMB),5,5’-dithio-bis (2-nitrobenzoicacid) (DTNB), Cleland’s reagent and N-acetylimidazole were obtained from Sigma Chemical Co. (St. Louis, Missouri).

Enzyme Assay The enzyme activity was measured a t room temperature by coupling the reaction:

+

fructose 6-phosphate ATP --f fructose 1,6-diphosphate ADP

+

with aldolase, triose phosphate isomerase, and a-glycerophosphate dehydrogenase, and recording NADH

A. CHAPMAN, T. SINNER, and A.

V01.7, So.4, 1969

oxidation a t 340mp. The standard assay mixture, slightly modified from Ling et al. [l], contained, in a total volume of 3 ml: 33 mM Tris-C1, p H 8; 2 mM fructose 6-phosphate; 2 mM ATP; 4 mM MgSO,; 10mM Cleland's reagent; 5mM KC1; 0.16mM NADH; and an excess of auxiliary enzymes. The reaction was initiated by the addition of phosphofructokinase. When indicated in the text, the enzyme was preincubated with an excess of Cleland's reagent for 5 min a t room temperature prior to the assay, and the reaction was started by adding the rest of the assay mixture. The absorbance a t 340 mp was read a t 1 min intervals in a Zeiss PMQ I1 spectrophotometer. A linear decrease in the absorbance was observed after an initial lag period. The slope of the linear part of the curve was taken as a measure of the enzyme activity. When the ATP concentration was varied, the concentration of MgSO, was varied accordingly to keep a constant Mg++/ATP ratio of 2 to 1. The effect of AMP on the activity of phosphofmctokinase was determined by assaying the enzyme activity a t p H 6.7 (Tris-C1) in the presence of inhibiting concentrations of ATP (3 mM) and low concentrations of fructose 6-phosphate (0.2 mM). The concentration of the other components of the assay mixture was the same as above. The p H of the final assay mixture was adjusted t o 6.7. The allosteric activity is expressed as the per cent increase in the activity observed in the presence of 1 mM AMP.

PIHL

589

noAMP

001 10 20 30 40 [Fru-6-P] (mM)

001

01 10 [Fru-6-P] (rnM)

10

Fig. I. Effect of A M P on the activity of ATP-inhibited phosphofructokinase. The enzyme activity was measured at p H 6.7 in the presence of 3 mM ATP, and 6 mM MgSO,. Where indicated, 1 mM AMP was added. The Concentration of the other components of the assay mixture was a s described in the Methods section. (A) Reaction rate versu.s fructose 6-phosphate concentration. (B) Hill plots of the data. The slopes (n)were 2.1 and 1.1, in the absence and presence of AMP, respectively. The corresponding half maximum substrate concentrations, Km,were 0.6 mM and 0.2 mM

RESULTS

I n the present paper the effect of acetylation on the catalytic activity has been studied as well as the effect on one of its allosteric functions, the ability of AMP to reverse the ATP inhibition of the enzyme. The catalytic activity has been measured under optimal conditions a t p H 8 . At this p H the reaction follows first order kinetics with respect to both fructlose-6-phosphateand ATP [1,3]. The ability of AMP to reverse the ATP inhibition was studied a t p H 6.7, as the enzyme is sensitive to the AMP/ATP ratio only a t pH around or slightly below neutrality. The data in Fig. 1A show that under these conditions the fructose 6-phosphate saturation curve in the presence of inhibiting concentrations of ATP is sigmoidal, in agreement with the finding of previous authors [6]. When AMP was added, the saturation curve changed into a hyperbolic response curve. I n separate experiments it was found that activation by AMP follows first order kinetics and that maximal stimulation was obtained a t an AMP concentration of 1 mM. I n the experiments reported below the effect of AMP was tested a t a fructose 6-phosphate concentration of 0.2mM. Under these conditions a 300-400°/0 increase in the activity was found. When the data in Fig. 1A were plotted according t o Hill (Fig. I B) the slope of the line, n, had values of 2.1 and 1.1, in the

O F 0

20

.: 40

-

60

Incubation time (min)

I

80

20

40

60

80

Incubation time (rnin)

Fig.2. Effect of acetylation o n Catalytic and allosteric activity of phosphofructokinase. The enzyme (200 pg/ml), in 0.2 M borate buffer, p H 7 . 5 was incubated at 0" with N-acetylimidazole (0.5 mglml). (A) Remaining catalytic and allosteric activity, as a function of incubation time. The catalytic activity was measured under optimal conditions a t p H 8.0, as described in the Methods section. The allosteric activity (the relative increase in activity in the presence of 1 mM AMP) is expressed in per cent of t h a t of the native enzyme. This assay was carried out a t pH6.7, and the reaction mixture contained 3 mM ATP, 6 mM Mg++, and 0.2 Mm fructose 6-phosphate. (B) Activity in presence and absence of AMP, measured at pH6.7, as described above

absence and presence of AMP, respectively. I n contrast, Atkinson et al. [16] calculated, on the basis of data of Passonneau and Lowry [6], that AMP does not alter the value of n for the rabbit muscle enzyme. The effects of acetylation on the catalytic and allosteric activity of phosphofructokinase are shown in Fig.2. When the enzyme was incubated with N-acetylimidazole, under relatively mild conditions a t O", the optimal catalytic activity decreased only slowly with time (Fig.2A). However, the allosteric

590

Acetylation of Phosphofructokinase

activity decreased rapidly. The loss in allosteric function on acetylation is apparently due to the fact that the activity in the absence of AMP increased upon acetylation and approached that measured in the presence of AMP (Fig.2B). Thus, after 20min incubation, AMP had virtually no stimulating effect on the enzyme. The activity measured in the presence of AMP disappeared a t a rate similar to that of the catalytic activity measured under optimal conditions (Fig.2A). The increased activity of the acetylated phosphofructokinase in the absence of AMP could conceivably be due to an increased affinity for the substrate, fructose 6-phosphate. The substrate saturation curves in Fig.3A confirm that this is the case. The corresponding Hill plots (Fig.3B) show that acetylation

12-

A

'

European J. Biochein.

'

'

' -

16

20

2 '

04

08 12 [ATPI (mM)

00001 0001 001 [ATPI (mM)

01

Fig.4. Effect of ATP in native and acetylated phosphofructokinase. The enzyme was acetylated as described in Fig.3. The activity was measured at p H 6.7 in the absence of AMP. The fructose 6-phosphate concentration was 0.2 mM. The Mg++/ATP ratio was kept equal to 2. (A) Reaction rate versus ATP concentration for native and acetylated phosphofor the fructokinase. (B) Hill plots of data in (A). V,,, acetylated enzyme was estimated from the curve in Fig.4A. Vmax for the native enzyme was determined by extrapolation to l/[S] = 0 of the linear part of the curve of a LineweaverBurk plot

I n"

'

"

"

"

1

01

0 5 10 15 20 40 [Fru-6-P] (mM)

[Fru-6-P] (mM)

Fig.3. Effect of acetylation om the fructose 6-phosphate saturation curve. The enzyme (200 pg/ml) in 0.2 M borate buffer, p H 7.5, was acetylated by incubation with N-acetylimidazole (1 mg/ml) for 30 min a t 0". The enzyme activity was measured at p H 6.7 in the absence of AMP, under the conditions given in Fig. 1 A. (A) Reaction rate versus fructose 6-phoshate concentration in native and acetylated phosphofructokinase. (B) Hill plots of results in (A). The slopes ( n ) were 2.1 and 1.1 in native and acetylated enzyme, respectively. The corresponding half maximum substrate concenwere 0.6 mM and 0.1 mM trations (Km) ~

40 Incubation time (min)

X)

not only increased the affinity for the substrate, but also decreased the kinetic order of the reaction. The response of the acetylated enzyme to increasing fructose 6-phosphate concentrations (Fig.3A) resembled closely that of the native enzyme, assayed in the presence of AMP (Fig.1A). I n fact, both the half maximum substrate values and the n values were similar in the two cases. Since the effect of AMP is to reverse the inhibition by ATP [6], the effect of acetylation observed in Fig.2B and 3A could be explained if acetylation reduced the ATP inhibition of the enzyme. This possibdity was studied by measuring the effect of increasing ATP concentration on the activity of the native and acetylated enzyme (Fig.4). The results demonstrate that the inhibition by high ATP concentrations was abolished in the acetylated enzyme Fig.4A). When the data (the ascending part of the

~~

60

Fig.5. Effect of substrates and of A M P on inactivation of phosphofructokinase by acetylation. The enzyme (200 pg/ml) in 0.2 M borate buffer, p H 7.5, was incubated with N-acetylirnidazole (3mg/ml) at 25". The added compounds were present during acetylation in concentrations of 2 mM. When nucleotides were added, the Mg++ concentration was 4 m&l. The enzyme activity was measured a t p H 8.0, after preincubation for 5 m i n a t room temperature with Cleland's reagent

curves) were plotted according to Hill (Fig.4B)it was found that acetylation decreased the affinity of ATP for the enzyme, and that it reduced the n-value, suggesting a decreased interaction between the ATP binding sites. I n the preceding experiments in which the acetylation was carried out under relatively mild conditions, the catalytic activity was affected only to a moderate extent. It was found, however, that the catalytic ac-

A. CHAPMAN, T. SANNER, and A. PIHL

Vol.7, N0.4,1969

tivity could be completely inhibited by using higher concentrations of N-acetylimidazole or by raising the incubation temperature. I n Fig.5 it is shown that under such conditions the presence during acetylation of the substrates, fructose 6-phosphate or ATP, or the allosteric modifier M P provided only slight protection against the loss of catalytic activity. The results indicate that the residues modified are not directly involved in the catalytic site, and that the extensive loss of catalytic activity observed after vigorous acetylation may be due to a general denaturation of the enzyme. Attempts were then made to see whether the effects of acetylation on the allosteric properties of the enzyme could be altered by carrying out the reaction in the presence of the above substances. It was found (Fig.6A) that the presence of ATP during

591

to inhibit the enzyme. As expected, it did not prevent the effect of acetylation (Fig.6A). On the other hand, GTP which, like ATP, inhibited the enzyme a t high concentrations likewise protected against acetylation. However, unexpectedly, UTP and CTP failed to prevent the effect of acetylation (Fig.6A) in spite of the fact that they both inhibited the enzyme. Since the two pyrimidine nucleotides did not protect, even though they are probably bound to the inhibitory site, it appears that binding of a nucleotide is not sufficient for protection of this site against acetylation. It was also of interest to study whether the presence of nucleotides during acetylation would influence the activity measured in the presence of AMP. The results in Fig.6B show that the presence during acetylation of AMP, TTP, UTP, CTP or fructose Allosteric activity

A

80

20

40

60

Incubation time (rnin)

80

20 40 60 Incubationtime (min)

ao

Fig.6. Effect of substrate and different nucleotides on loss of allosteric function upon acetylation. The enzyme (200 pg/ml) in 0.2 M borate buffer, pH 7.5, was incubated with N-acetylimidaeole (1 mg/ml) at 0". The concentrations of reagents during acetvlationwas 2 mM. When nucleotides were present, the ckcen"tration of Mgff was 4mM. The enzyhe was assayed at pH 6.7 in the presence of 3 mM ATP. (A) Enzyme assay in absence of AMP. (B) Enzyme assay in presence of 1 mM AMP

acetylation completely prevented the characteristic increase in the activity measured in the absence of AMP (Fig.2B). Presumably, binding of ATP t o the inhibitory site blocked the access of the acetylating agent. I n contrast, AMP or fructose 6-phosphate had no protective effect. Previous studies of phosphofructokinase isolated from various sources have shown [14,17] that certain nucleotides can serve as phosphate donors but that they, unlike ATP, are unable to inhibit the reaction when present in high concentrations. This might be due to lack of binding t o the inhibitory sites. I n our system, nucleotides which do not inhibit phosphofructokinase in high concentrations, would be expected to be unable to protect the enzyme against the effects of acetylation. It was found that TTP, although it served as a phosphate donor, was unable

0

30

[DTNB]/[Enzyme]

50

I

I\

10 30 50 [pCMB]l[Enzyme]

Fig. 7. Effect of sulfhydryl reagents on the catalytic and allosteric activities of phosplwfructolciwe. The enzyme (200 pgglml) in 20 mM phosphate buffer, pH 8.0, was incubated at 0' with increasing concentrations of DTNB (for 60 min) or with pCMB (for 20 min). The catalytic activity was measured at pH 8.0. The allosteric activity was expressed as in the legend to Fig.2

6-phosphate had no effect, while the presence of ATP, or GTP increased the rate of inactivation of the enzyme. It was pointed out above that when the catalytic activity was assayed under optimal conditions (Fig.5), ATP had a slight protective effect. One possible explanation of these apparently contradictory results might be that acetylation altered the pH-activity profile of the enzyme. Previously Lardy et al. [ll,121 have reported that phosphofructokinase consists of 4 identical subunits (protomers) each of which contains 16-18 thiol groups. Approximately 7001, of the activity was lost when the enzyme was incubated with excess oxidized glutathione or with monoiodoacetamide [ll], indicating that SH groups are closely associated with the catalytic site of the enzyme. I n order to obtain further information on the role of SH groups in the function of phosphofructokinase the inhibiting effect of DNTB and pCMB was studied.

Acetylation of Phosphofrucbkinase

592

I n Fig. 7 A is shown that even small amounts of DTNB inhibited the enzyme effectively. As with oxidized glutathione and monoiodoacetamide [ll] it was not possible to inactivate the enzyme completely. Thus, a great excess of DTNB reduced the catalytic activity only by 60-700/0. pCMB was far less effective, per SH group blocked, in inhibiting the enzyme (Fig. 7 B). However, with this reagent the catalytic activity could be entirely inhibited. That this inhibition was due to a specific blocking of SH groups, and not to an irreversible denaturation of the enzyme, follows from the fact that the enzyme could be completely reactivated by addition of an excess of Cleland's reagent. It is of interest to note that with both reagents, blocking of SH groups, causing extensive loss of catalytic activity had no effect on the allosteric function. It thus appears that the sulfhydryl groups of phosphofructokinase are not involved in the inhibition of the enzyme by ATP and its reversal by AMP. DISCUSSION

The results presented in this paper show that the allosteric activity of phosphofructokinase, as defined by its sensitivity to AMP activation, can be preferentially inactivated by acetylation, while it is unaffected by sulfhydryl blocking which selectively reduces the catalytic activity of the enzyme. The acetylating agent used here, N-acetylimidazole, can react with amino- and sulfhydryl groups of proteins. However, it reacts far more readily with tyrosine hydroxyl groups [18,19] and it is likely that the loss of AMP activation after acetylation under relatively mild conditions was due primarily t o the modification of tyrosine groups. An acetylation of sulfhydryl groups cannot account for these results, since treatment of phosphofructokinase with specific SH reagents had no effect on the allosteric activity of the enzyme. Whether the loss of catalytic activity observed after more vigorous acetylation was due to modification of tyrosine groups, SH groups, or both, cannot be decided from the present data. The data presented indicate that the loss of the allosteric function of phosphofructokinase upon acetylation involves the loss of sensitivity of the enzyme to ATP inhibition. Recently we have found [20] that the X-ray induced inhibition of allosteric function is likewise due t o a decreased sensitivity of the enzyme to ATP. The finding that the presence of ATP prevented the effect of acetylation supports the view that tyrosine residues may be involved in the allosteric function of phosphofructokinase and may be closely associated with the inhibitory binding sites for ATP. Also, in the case of the related enzyme, fructose 1,6-diphosphatase, tyrosine groups seem t o be involved in the allosteric function [21,22].

European J. Biochem.

Binding studies with labeled compounds have indicated that phosphofructokinase will bind 1 molecule of fructose-6-phosphate, 3 molecules of ATP and 1 molecule of AMP per protomer [8].This suggests that each protomer may contain 1, or possibly2, inhibitory binding sites for ATP. The question has been raised whether AMP activates the enzyme by displacing ATP from its inhibitory site, or whether AMP and ATP bind a t different sites [8,15]. The view that a common site is involved is supported by the finding of Atkinson et al. [15] that the significant parameter for the phosphofructokinase activity is the reatio between ATP and AMP, rather than their absolute concentrations, and the fact that no instance seems to have been reported where loss of AMP activation occurs without concurrent loss of ATP inhibition. On the other hand, the fact that AMP was unable to prevent acetylation of the allosteric site is more easily explained by assuming that different sites are involved. Obviously, more data are needed to decide whether AMP and ATP bind to the same sites of phosphofructokinase. Since ATP was able to prevent the acetylation of the allosteric site, it was expected that other nucleotides inhibiting the enzyme would likewise prevent the loss of allosteric activity on acetylation. This was indeed found to be the case with GTP. However, CTP, and UTP failed to protect the allosteric function against inactivation, although these nucleotides did cause enzyme inhibition. Presumably, their failure to protect could be due to the fact that the pyrimidine nucleotides are smaller than the purine nucleotides, and may be unable to cover the relevant tyrosine groups. The results obtained in this paper suggest certain conclusions concerning the significance of the sulfhydryl groups for the catalytic function of the enzyme. It was found that the weakly reactive sulfhydryl reagent DTNB was particularly effective in inhibiting the enzyme. Thus, more than half the activity was lost after addition of 4 equivalents of DTNB. Since the enzyme consists of 4 protomers, this suggests the possibility that each protomer may contain one particularly reactive SH group, which is essential for the enzyme activity. The finding that in order to obtain the same degree of inhibition more extensive blocking was required with monoiodoacetamide and pCMB than with DTNB may be due to the fact that the former reagents are more reactive and probably interact also with SH groups which are not involved in the catalytic activity. This work was supported by the National Center for Radiological Health, U. S. Public Health Service and by the Nansen Foundation. T. Sanner is a Fellow of the Norwegian Cancer Society. REFERENCES 1. Ling, K.-H., Marcus, F., and Lardy, H. A., J. Bid. Chem. 240 (1965) 1893.

A. CHAPMAN,T. SANNER,and A.

Vo1.7, No 4,1969

2. Uyeda, K., and Racker, E., J . Biol. Chem. 240 (1965) 4682. 3. Parmeggiani, A,, Luft. J. H., Love, D. S., and Krebs, E. G., J. Biol. Chem. 241 (1966) 4625. 4. Griffin, C. C., Houck, B. N., and Brand, L., Biochem. Biophys. Res. Commun. 27 (1967) 287. 5. Lindell. T. J.. and Stellwaeen. E.. J . Biol. Chem. 243 (1968) 907: 6. Passonneau. J. V.. and Lowrv. 0. H.. Biochem. Bionhus. Res. Commun. i (1962) 10." 7. Passonneau, J. V., and Lowry, 0. H., Biochem. Biophys. Res. Commun. 13 (1963) 372. 8. Kemp, R. G., and Krebs, E. G., Biochemistry, 6 (1967) 423. 9. Paetkau, V., and Lardy, H. A., J . Biol. Chem. 242 (1967) 2035. 10. Ui, BI., Biochim. Biophys. Acta, 159 (1968) 50. 11. Younathan, E. S., Paetkau, V., and Lardy, H. A., J . Biol. Chem. 243 (1968) 1603. 12. Paetkau, V. H., Younathan, E. S., and Lardy, H. A., J . Mol. Biol. 33 (1968) 721. 13. Pogell, B. M., Tanaka, A., and Siddons, R. C., J . Biol. Chem. 243 (1968) 1356. 14. Ramaiah, A., Hathaway, J. A., and Atkinson, D. E., J. Biol. Chem. 239 (1964) 3619. A

'

39 European J. Biochem., Vol. 7

"

PIEIL

593

15. Atkinson, D. E., and Walton, G. AT., J . Biol. Chem. 242 (1967) 3239. 16. Atkinson, D. E., Hathaway, J. A., and Smith, E. C., Bioehem. Biophys. Res. Commun. 18 (1965) 1. 17. Vifiuela, E., Salas, M. L., and Sols, A., Biochem. Biophys. Res. Commun. 12 (1963) 140. 18. Riordan, J. F., Wacker, W. E. C., and Vallee, B. L., Biochemistry, 4 (1965) 1758. 19. Pontremoli, S., Graxi, E., and Accorsi, A., Biochemistry, 5 (1966) 3072. 20. Chapman, A., Sanner, T., and Pihl, A., Biochim. Biophys. Acta, in press. 21. Pontremoli, S., Grazi, E., and Accorsi, A., Bioc?Lemistry, 5 (1966) 3568. 22. Little, C., Sanner, T., and Pihl, A., European J . Biochem., in press. A. Chapman's present address: Veterans Administration Center Los Angeles, California 90073, U.S.A.

T. Sanner and A. Pihl Xorsk Hydro's Institutt for Kreftforskning Radiumhospitalet, Montebello, Oslo 3, Norway