Fructose bisphosphatase from Escherichia coli. Purification and characterization

ARCHIVES

OF BIOCHEMISTRY

Vol. 225, No. 2, September,

AND BIOPHYSICS

pp. 944-949, 1983

Fructose

Bisphosphatase from Escherichia Purification and Characterization’

JORGE BABUL’ Department

of Microbiology and Molecular

AND

co/i,

VICTORIA GUIXE’

Genetics, Harvard Received March

Medical

School, Boston, Massachusetts 02115

1, 1983

Escherichia coli fructose-l,&bisphosphatase has been purified for the first time, using a clone containing an approximately 50-fold increased amount of the enzyme. The procedure includes chromatography in phosphocellulose followed by substrate elution and gel filtration. The enzyme has a subunit molecular weight of approximately 40,000 and in nondenaturing conditions is present in several aggregated forms in which the tetramer seems to predominate at low enzyme concentrations. Fructose bisphosphatase activity is specific for fructose 1,6-bisphosphate (K, of approximately 5 PM), shows inhibition by substrate above 0.05 mM, requires Me for catalysis, and has a maximum of activity around pH 7.5. The enzyme is susceptible to strong inhibition by AMP (50% inhibition around 15 PM). Phosphoenolpyruvate is a moderate inhibitor but was able to block the inhibition by AMP and may play an important role in the regulation of fructose bisphosphatase activity in wivo. Fructose 2,6-bisphosphate did not affect the rate of reaction.

Fructose bisphosphatase (EC 3.1.3.11, D-fructose-1,6-bisphosphate l-phosphohydrolase) catalyzes an essential reaction of the gluconeogenic pathway and plays an important role in its regulation (1). The enzyme from several sources has been purified and characterized, but it is mainly the enzyme from mammalian tissues that has been extensively studied and whose structure and mechanism of action are relatively well known (2). The physiological role of the enzyme in Escherichia coli was shown through the use of mutants (3, 4), but the last report about the enzyme from i This work was supported by National Institutes of Health Grant 5ROl GM 21098-09 (to Dan G. Fraenkel). 2 To whom all correspondence should be addressed. Present address: Departamento de Quimica, Facultad de Ciencias Bdsicas y Farmaceuticas, Universidad de Chile, Casilla 653, Santiago, Chile. ‘Present address: Departamento de Biologia Facultad de Ciencias Basicas y Farmaceuticas, Universidad de Chile, Casilla 653, Santiago, Chile. 0003-9861/83 $3.00 Copyright All rights

0 1983 by Academic Press, Inc. of reproduction in any form reserved.

this bacterium, dealing with its partial purification and characterization, was published in 1966 (5). The possible involvement of the fructose bisphosphatase reaction in the impairment in gluconeogenic growth of an Eschtichiu coli mutant with an altered form of phosphofructokinase-2 (6,7), together with the availability of a strain carrying the fdp+ gene on a multicopy plasmid and containing approximately a 56-fold normal amount of the activity (J. Sedivy, F. Daldal, and D. G. Fraenkel, to be reported), prompted us to purify the enzyme and to study its properties. Part of this work has been presented (8). EXPERIMENTAL

PROCEDURES

Enzyme purification The strain used was a recA derivative of the original fructose bisphosphatase mutant strain (strain Qll, E. coli K 12, Hfrfl, fdpl (4)), carrying plasmid pZ12, pZ12 is a plasmid pBR322 with the wild type gene fdp+ cloned into it (J. Sedivy, F. Daldal, and D. G. Fraenkel, to be reported). The 944

Eschm-ichia coli FRUCTOSE TABLE

I

PURIFICATION OF FRUCTOSE BISPHOSPHATASE

Fraction Crude extract P-Cellulose Sephacryl S-300b Fractions 104-113 Fractions 91-103

Total activity (units)

Specific activity (units/mg)

586 358 306 225 81

0.85 30

Yield” (%I 100 61 53

52 38

“Yield was not corrected for the presence of gluconate-&phosphate dehydrogenase in the crude extract. The P-cellulose fraction was devoided of this activity. “13% of the total units of the previous fraction were applied to the Sephacryl column. The indicated fractions correspond to Fig. 1A. strain was grown aerobically in medium 63 (9) supplemented with 1% Bacto-tryptone (Difco) and 0.4% Bacto yeast (Difco), and ampicillin, 125 pg/ml (to maintain selective pressure for retention of the plasmid), and harvested at Am of 3.5 by centrifugation for 10 min at 13,200~. Eight grams of cells were suspended in 32 ml of 10 rnre sodium malonate (pH 6.0) containing 5 mM MgClz, 0.1 mM EDTA, 1 mM dithiothreitol, and 5% ethanol (buffer A). The cells were then disrupted in a French pressure cell press (American Instrument) at 20,000 psi at 4°C and centrifuged for 60 min at 143,OOOg. The supernatant fluid was dialyzed for 2 h versus 500 ml of buffer A with one change and then adjusted to pH 6.0 with 10 rnM sodium malonate, pH 5.0, and applied to an g-ml (bed volume) column of P-cellulose” (Whatman Pll) equilibrated with buffer A. The column was washed with the equilibrating buffer until the absorbance of the eluate at 280 nm was approximately 0.01 and the fructose bisphosphatase activity was then eluted with a ZOO-ml linear gradient of fructose-1,6-q from 0 to 0.6 mM in buffer A (the peak of activity was at approximately 0.15 mM fructose-1,6-P*). Fractions containing two or more units of enzyme were pooled (Table I, P-cellulose fraction) and concentrated to 3 ml using an Amicon ultrafiltration cell with a PM 10 membrane. A portion of this fraction corresponding to 13% of the enzyme units was then applied to a Sephacryl S300 (Pharmacia) column (1.4 X 81.5 cm) equilibrated with buffer A supplemented with 50 mM malonate and 0.2 mM fructose-1,6-P*, and the enzyme was eluted

’ Abbreviations used: P-cellulose, phosphocellulose; fructose-1,6-P,, fructose 1,6-bisphosphate; SDS, sodium dodecyl sulfate.

945

BISPHOSPHATASE

with the same buffer. The fractions indicated in legend to Fig. 1 were pooled (Table I, Sephacryl S-300 fraction) and concentrated as above. The enzyme preparation was then made 50% in glycerol and stored at -15°C. Enzyme assays. Fructose bisphosphatase activity was measured spectrophotometrically by coupling the fructose 6-phosphate formation to the reduction of NADP in the presence of phosphoglucose isomerase and glucose-6-phosphate dehydrogenase (Boehringer Mannheim) as described by Fraenkel et aL (5). The assay mixture contained (final concentrations in 1 ml): 50 mM Tris-HCl, pH 8.0, 10 mM MgClz, 0.2 m M NADP, 0.1 mM EDTA, 1.0 mM dithiothreitol, and 2 pg each of phosphoglucose isomerase and glucose-6phosphate dehydrogenase. The reaction was initiated by addition of the enzyme. Activity was also measured by phosphate released using the method of Chen et aL (10) with the assay mixture described above except for the absence of the coupling system: 10% ascorbic acid and 0.42% ammonium molibdate in 1 N HzSO, were mixed 1:6 and then 2.3 ml of this mixture were added to 1 ml of the

A

40

t

:I& 80

90

100

I23 $44

4 4

110

120

130

FRACTION NUMBER

FIG. 1. Gel filtration of the P-cellulose fraction on Sephacryl S-300 at two enzyme concentrations. The Sephacryl S-300 column (1.4 X 81.5 cm) was equilibrated with 50 mM sodium malonate, 5 mM MgClz, 0.1 mM EDTA, 1 mM dithiothreitol, 5% ethanol, and 0.2 mM fructose-1,6-P,, pH 6.0. (A) 32 U of enzyme in 0.5 ml of sample. (0) Enzyme activity expressed as change in absorbance at 340 nm/min/ml enzyme; (-), protein concentration. Fractions 91 to 103 and 104 to 113 were pooled separately. The arrows indicate the elution volumes of four standard proteins of the following molecular weights: (1) catalase (230,000); (2) aldolase (158,000); (3) lactate dehydrogenase (130,000); and (4) albumin (67,000). The Blue Dextran elution peak was fraction 74, and the glucose peak was fraction 173. (B) 9 U of enzyme in 0.5 ml of sample.

946

BABUL

AND

reaction mixture, incubated 20 min at 45°C and the absorbance at 820 nm was measured. Fructose-1,6-Pz was measured spectrophotometritally in the presence of NADH, aldolase, triosephosphate isomerase, and a-glycerophosphate dehydrogenase (11). Phosphoenolpyruvate was assayed with the use of pyruvate kinase and lactate dehydrogenase (11) and AMP was measured in the same system containing also myokinase and ATP. Electroph,uresis Polyacrylamide gel electrophoresis in nondenaturing conditions was performed in 10% cylindrical gels as described by Davis (12), at 3 mA per gel and 5°C with the following additions to the running buffer: 1 mM dithiothreitol, 1 mM MgClz, and 0.1 mM EDTA. Proteins were stained with Coomassie brilliant blue G-250 (Serva) (13) and enzyme activity was detected using the coupled assay described above with 0.2 mM fructose-1,6-Pz plus the addition of 0.03 mg/ml of phenazine methosulfate and 0.3 mg/ml of nitroblue tetrazolium (14). Electrophoresis of the denatured protein in the presence of SDS was according to Laemmli (15) in 10 X 14 X 0.15-cm slab gels with 5% acrylamide stacking gel and 8-14% separating gel. The enzymes in 5% 2-mercaptoethanol were incubated for 2 min in boiling water and after the addition of glycerol and bromphenol blue were applied to the gel and electrophoresis performed for 16 h at 20 mA. The proteins were stained as described by Fairbanks et al. (16) using 3.46% sulfosalicylic acid and 11.5% trichloroacetic acid as fixing solution. Molecular weight okterminations. The molecular weight of the native protein was estimated by correlating the distribution coefficients and molecular weights of well known standard proteins after filtration in a Sephacryl S-300 column according to Andrews (1’7). The column and equilibrating buffer used are described in the purification procedure. The standard proteins are listed in Fig. 1. The subunit molecular weight was determined by polyacrylamide slab gel electrophoresis in the presence of SDS using a mixture of low-molecular-weight standard proteins (Pharmacia) (Fig. 2B, lane 3). Protein ileterwbinatim Protein concentration was determined with Coomassie brilliant blue G-250 (18) with bovine serum albumin (Sigma) as standard.

RESULTS

Enzyme puri$catim The purification procedure used for fructose bisphosphatase is summarized in Table I. Gel filtration of the enzyme preparation after the P-cellulose step reveals the presence of enzyme activity in fractions corresponding to a wide range of molecular weight. This phenomenon was observed at two enzyme concentrations (Fig. 1) and a similar pattern

GUIXE A

FIG. 2. Polyacrylamide gel electrophoresis of fructose bisphosphatase in the absence and in the presence of dodecyl sulfate. (A) Nondenaturing conditions, 10% cylindrical gels. (1) 42 pg P-cellulose fraction stained for activity (see Experimental Procedures). Although in the photographic reproduction the stained bands are difficult to see, in the original gel several bands were clearly observed. (2) ‘7 pg P-cellulose fraction. (3) 4 pg enzyme after the gel filtration step (pooled fractions 91-103, Fig. 1A). (4) 6 pg pooled fractions 104-113, Fig. 1A. (B) In the presence of SDS, slab gels with 8-14% acrylamide in the separating gel. (1) 2 pg P-cellulose fraction; (2) 2 Fg pooled fractions 104113 Fig. 1A; (3) standard proteins of the following subunit molecular weights (from top to bottom): phosphorylase b, 94,000; albumin, 67,000; ovalbumin, 43,000; carbonic anhydrase, 30,000; trypsin inhibitor, 20,100.

was observed upon filtration of the crude enzyme preparation. Polyacrylamide gel electrophoresis in the absence and in the presence of SDS of the enzyme after each purification step is shown in Fig. 2. Electrophoresis in nondenaturing conditions either after the P-cellulose or gel filtration steps showed the presence of several molecular species (Fig. 2A), whereas in the presence of SDS (Fig. 2B) mainly one component was observed in each case. Activity staining after electrophoresis (Fig. 2A, lane 1) also showed the presence of several bands.

Eschm-ichia coli FRUCTOSE

947

BISPHOSPHATASE

~~~:Ij---qq 0

05

10

15

20

FRUCTOSE-1,6-P2,mM

0

loo

3co

FRUCTOSE-1,6-q,

500

mM4

FIG. 3. Fructose bisphosphatase activity as a function of fructose-1,6-P, concentration. Enzyme activity was determined with the coupled enzyme assay and the reaction was initiated by addition of purified fructose bisphosphatase (pooled fractions 104-113, Fig. 1A). The saturation curve is presented in (A) and the reciprocal data are presented in (B).

The size of the dissociated protein was determined in the presence of SDS using the conditions described by Laemmli. From Fig. 2B a subunit molecular weight of around 40,000 could be estimated. Only an approximate molecular weight could be estimated for the native enzyme since an asymmetrical peak was obtained after gel filtration and the elution volume of the main fraction varied with the enzyme concentration (Fig. 1). The predominant species at low enzyme concentration appears to be tetrameric. TABLE SUBSTRATE

II

SPECIFICITY OF FRUCTOSE BISPHOSPHATASE

Concentration Compound Fructose-1,6-Pz Fructose-1,6-P* Fructose-l-P Glucose-1,6-Pz Fructose-6-P Glucose-6-P Gluconate-6-P Mannose-6-P Ribose-5-P Sedoheptulose-1,7-P*

mM

% Vb

0.05 1.0 1.0 0.5 1.0 1.0 1.0 1.0 1.0 0.5

100 72 19 14 4 0.0 0.0 0.0 0.0 0.0

(LEnzyme activity was determined by the Pi release purified fructose bisphosphatase (pooled fractions 104-113, Fig. 1A). b Activity is expressed as percentage of the velocity obtained with 0.05 mM fructose-1,6-P,.

assayusing

Catalytic measurements. All measurements were performed with the main fraction of Fig. 1A. The activity of purified fructose bisphosphatase as a function of the fructose-1,6-Pz concentration is shown in Fig. 3. Maximum activity was observed at a concentration of substrate of around 0.05 mM. The same behavior was observed whether the activity was determined by the coupled enzyme or the phosphate release assay. From Fig. 3A a K, of 5 PM was determined. Table II shows the ability of fructose bisphosphatase to hydrolyze various compounds. Like virtually all fructose bisphosphatases, the E. coli enzyme requires a divalent cation for catalytic activity. Several ions were tested and, as reported earlier for the partially purified enzyme (5), maximal rates were obtained above 5 mM Mgz+, and no activity was detected at similar concentrations of Mn’+, Ba2+, Cd2+, and Zn2+. The enzyme showed its maximum of activity around pH ‘7.5 at 40 PM fructose-1,6-P2. AMP is a strong inhibitor of the enzyme (Fig. 4). In the presence of AMP the rate of the reaction decreased with time, apparently due to inactivation of the enzyme by AMP. Thus, in order to measure the fructose bisphosphatase activity the rates were determined from the increase in absorbance occurring in the first minute of reaction. The reaction was initiated by the addition of the enzyme since, as reported before (5), much lower rates are observed

948

BABUL

AND

if it is initiated with the substrate. Inactivation occurs when the enzyme is diluted in the absence of its substrate. Phosphoenolpyruvate at high concentrations caused partial inhibition of fructose bisphosphatase (Fig. 5), but it was able to block the inhibition by AMP (Fig. 4). No effect was observed when fructose bisphosphatase was assayed in the presence of 1 PM fructose 2,6-bisphosphate at concentrations of substrate between 0.02 and 2 mM. DISCUSSION

Enzyme pur&.xxtion. The use of substrate elution of fructose bisphosphatase from P-cellulose columns is a common step in the purification of the enzyme from many sources. In the purification of the E. co& enzyme in this work, after the P-cellulose purification step the enzyme preparation shows several protein components as detected by gel electrophoresis in nondenaturing conditions, while after electrophoresis in the presence of SDS shows mainly one component. These findings indicate that the enzyme probably forms aggregates. The same phenomenon was observed in gel filtration experiments where the enzyme eluted in fractions corresponding to a range of molecular weights between 145,000 and 450,000. The aggregation has not been described for the well known mammalian fructose

II Oo

I IO

I 20

I 30

I 40

II 50

AMP, pM FIG. 4. Fructose bisphosphatase activity as a function of AMP concentration in the presence and in the absence of phosphoenolpyruvate. For enzyme fractions see Fig. 3; all assays were performed in the presence of 0.1 mM fructose-1,6-P2. 0, In the presence, and 0, in the absence of 1 mM phosphoenolpyruvate.

GUIXfi

8 01 0

2

4

6

8

10

PHOSPHOENOLPYRUVATE,mM FIG. 5. Fructose bisphosphatase activity as a function of phosphoenolpyruvate concentration. Details as in Fig. 3, all solutions contained 0.04 mM fructose1,6-Ps.

bisphosphatases, most of which are tetramers of molecular weights around 140,000 (1). The enzyme from B. lichenifmis (19) has a higher molecular weight, 500,000, and apparently also does not aggregate, while the enzyme from B. subtilis (20) showed aggregates in sucrose density gradient experiments, but not in the presence of phosphoenolpyruvate. It would be of interest to characterize the aggregated species of the E. coli enzyme, especially under the enzymatic reaction conditions and in the presence of metabolites in order to establish their physiological significance. Catalytic measurements. Purified E. coli fructose bisphosphatase is subjected to substrate inhibition at concentrations above 0.05 mM (about 50% at 1.5 IIIM). This behavior has also been observed for the bovine and rabbit liver enzymes and has been attributed to allosteric sites on the enzyme (21). AMP is a noncompetitive inhibitor of fructose bisphosphatase from a wide variety of sources (1,22) and this action is believed to be of importance in the control of carbohydrate metabolism, especially in E. coli considering the apparently constituitive nature of the enzyme (3). Stone and Fromm (23) have investigated the interaction between fructose-1,6-Pz, fructose 6-phosphate, Pi, AMP, and the enzyme from bovine liver by means of progress curve analysis and concluded that AMP controls the activity of the enzyme acting as an inhibitor and that also affects the amount of substrate available to fruc-

Escherichia

coli FRUCTOSE

tose bisphosphatase by complexing with fructose-1,6-Pz. The low Ki of the enzyme for AMP suggests that under physiological conditions the enzyme would be inhibited to a great extent. Nevertheless, phosphoenolpyruvate, a mild inhibitor of the E. coli enzyme, was able to block the inhibition by AMP at levels known to exist in vivo under gluconeogenic conditions (24). A similar effect has been observed with the enzyme from B. lichenifmis (19), although in that case phosphoenolpyruvate also acted as an activator. It seems possible that phosphoenolpyruvate might have an important role in the regulation of gluconeogenesis in E. coli,for its amount is known to be higher than in growth on glucose (24), and it would not only in effect be activating fructose bisphosphatase, but also, since the major phosphofructokinase of E. coli is inhibited by phosphoenolpyruvate (25, ll), would be preventing a futile cycle of reformation of fructose-1,6-Pz from fructose 6-phosphate. Daldal and Fraenkel (7) investigated the existence of futile cycling under gluconeogenie conditions and Chambost and Fraenkel(26) in growth on glucose and the findings of both studies were that futile cycling was marginal. The availability of a simple purification procedure for fructose bisphosphatase from a high level enzyme strain together with the kinetic characterization of this essential gluconeogenic enzyme will now also make it possible to assess the involvement of the reaction catalyzed by fructose bisphosphatase in the impairment in gluconeogenic growth observed in E. coli strains with an altered form of phosphofructokinase-2 (6). ACKNOWLEDGMENTS

Weexpressour sincerethanks

to Dan G. Fraenkel and the members of his group for their help andhaspitality. REFERENCES 1. BENKOVIC, S. J., AND DEMAINE, M. M. (1981) in “Advances in Enzymology” (A. Meister, ed.), Vol. 53, pp. 45-82, Interscience, New York.

BISPHOSPHATASE

949

2. MARCUS, F. (1981) in “Regulation of Carbohydrate Formation and Utilization in Mammals” (C. M. Veneziale, ed.), pp. 269-290, University Park Press, Baltimore. 3. FRAENKEL, D. G., AND HORECKER, B. L. (1965) J. Bacterial 90, 837-842. 4. FRAENKEL, D. G., (196’7) J. Bacterid 93, 15821587. 5. FRAENKEL, D. G., PONTREMOLI, S., AND HORECKER, B. L. (1966) Arch. Biochem. Biophys. 114,4-12. 6. DALDAL, F., BABUL, J., GUIX~, V., AND FRAENKEL, D. G. (1982) Eur. 3: Biochem. 126. 373-379. 7. DALDAL, F., AND FRAENKEL, D. G. (1983) J. Batteriol. 153, 390-394. 8. GUIX~, V., AND BABUL, J. (1981) Arch. Bid Med Exp. 14, 268. 9. COHEN, G. N., AND RICKENBERG, H. V. (1956) Ann Inst. Pasteur (Paris) 91,693-720. 10. CHEN, P. S., TORIBARA, T. Y., AND WARNER, H. (1956) Anal. Chewz. 28,1756-1758. 11. BABUL, J. (1978) J. Bid Ch.em. 253, 4350-4355. 12. DAVIS, B. J. (1964) Ann N. Y. Acad Sci 121,404427. 13. REISNER, A. H., NEMES, P., AND BUCHOLTZ, C. (1975) And Chem. 64, 509-516. 14 GABRIEL, 0. (1971) in “Methods in Enzymology” (W. B. Jakoby, ed.), Vol. 22, pp. 578604, Academic Press, New York. 15. LAEMMLI, U. K. (1970) Nature (London) 227,680685. 16. FAIRBANKS, G., STECK, T. L., AND WALLACH, D. F. H. (1971) Biochemistry 10, 2606-2617. 17. ANDREWS, P. (1964) B&hem J. 91,222-233. 18. BRADFORD, M. M. (1976) Anal. B&hem. 72,248254. 19. OPHEIM, D. J., AND BERNLOHR, R. W. (1975) J. Biol. Chem. 254, 3024-3033. 20. FUJITA, Y., AND FREESE, E. (1979) J. Biol. C~WTL. 254, 5340-5349. 21. PONTREMOLI, S., MELLONI, E., DEFLORA, A., AND HORECKER, B. L. (1973) Arch. Biochem Biophys 156, 255-260. 22. PONTREMOLI, S., AND HORECKER, B. L. (1971) in “The Enzymes” (P. D. Boyer, ed.), Vol. IV, pp. 611-646, Academic Press, New York. 2.3. STONE, S. R., AND FROMM, H. J. (1980) Biochemistry 19, 620-625. 24. LOWRY, 0. H., CARTER, J., WARD, J. B., AND GLASSER, L. (1971) J. Biol Chem 246. 6511-6521. 25. BLANGY, D., But, H., AND MONOD, J. (1968) J. Md Bid 31, 13-35. 26. CHAMBOST, J-P., AND FRAENKEL, D. G. (1980) J. Biol. Chem 2.55, 2867-2869.