Regulation of rat liver fructose 2,6-bisphosphatase.

An enzyme activity that catalyzes the hydrolysis of phosphate from the C-2 position of fructose 2,6-bisphosphate has been detected in rat liver cytoplasm. The S0.5 for fructose 2,6-bisphosphate was about 15 microM and the enzyme was inhibited by fructose 6-phosphate (Ki 40 microM) and activated by Pi (KA 1 mM). Fructose 2,6-bisphosphatase activity was purified to homogeneity by specific elution from phosphocellulose with fructose by specific elution from phosphocellulose with fructose 6-phosphate and had an apparent molecular weight of about 100,000, 6-phosphofructo 2-kinase activity copurified with fructose 2,6-bisphosphatase activity at each step of the purification scheme. Incubation of the purified protein with [gamma-32P]ATP and the catalytic subunit of the cAMP-dependent protein kinase resulted in the incorporation of 1 mol of 32P/mol of enzyme subunit (Mr = 50,000). Concomitant with this phosphorylation was an activation of the fructose 2,6-bisphosphatase and an inhibition of the 6-phosphofructo 2-kinase activity. Glucagon addition to isolated hepatocytes also resulted in an inhibition of 6-phosphofructo 2-kinase and activation of fructose 2,6-bisphosphatase measured in cell extracts, suggesting that the hormone regulates the level of fructose 2,6-bisphosphate by affecting both synthesis and degradation of the compound. These findings suggest that this enzyme has both phosphohydrolase and phosphotransferase activities i.e. that it is bifunctional, and that both activities can be regulated by cAMP-dependent phosphorylation.

Fructose 2,6-bisphosphate plays an important role in the regulation of hepatic carbohydrate metabolism (1)(2)(3)(4)(5), Glucagon addition to isolated hepatocytes results in a decrease in the level of the compound (6)(7)(8)(9)(10)(11)(12) and to inactivation of the enzyme responsible for its synthesis, 6-phosphofructo 2-kinase (9-15). These results suggest that the effect of glucagon to lower fructose 2,6-P2' levels is due at least in part to a glucagon-induced phosphorylation and inactivation of 6-phosphofructo 2-kinase (10,ll). However, maximal concentrations of glucagon lower the level of fructose 2,6-P2 by greater than 90% within minutes (11) suggesting that the rate of degradation of the compound may also be affected. The enzyme responsible for the degradation of fructose 2,6-P2 has not yet been identified. The purpose of this study was to purify and characterize the enzyme which degrades fructose 2,6-P2 and to investigate its regulation by glucagon.
* This work was supported by National Institutes of Health Grant AM 18270. The costs of publication of this article were defrayed in part by the payment of page charges, This article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact. I The abbreviations used are: fructose 2,6-P2, fructose 2,6-bisphosphate; SDS, sodium dodecyl sulfate; TES, N-(tris[hydroxymethyl] methyl-2-amino}ethanesulfonic acid; PMSF, phenylmethanesulfonyl fluoride.

EXPERIMENTAL PROCEDURES
Preparation and Incubation of Isolated Hepatocytes-Isolated hepatocytes were prepared from fed rats (male, Sprague-Dawley, 175-225 g) as previously described (16). After a 10-min incubation the cells (approximately 0.3-0.5 g of liver/flask) were homogenized for 90 s (30 s, three times) with an ultraturrax homogenizer in 10 ml of cold homogenizing buffer that contained 50 mM N-{tris[hydroxymethyl] methyl-2-amino) ethanesulfonic acid, pH 7.5,50 mM KF, 2 mM EDTA, 1 mM dithiothreitol, 0.2 mM phenylmethanesulfonyl fluoride and 0.5 p g / d of leupeptin, and the homogenate was centrifuged at 30,000 X g for 30 min. Solid (NH4)2S04 was added to the supernatant fraction to achieve 30% saturation and the precipitate was discarded. The supernatant fraction was then made 70% saturated with (NH4)2S04 and the precipitate redissolved in 1 ml of buffer A which contained 20 mM TES, pH 7.5, 50 mM KCI, 0.1 mM EDTA, 1 mM dithiothreitol, 0.5 pg/ml of leupeptin and 0.2 mM PMSF and the (NH4)2S04 was removed by dialysis against the same buffer. The (NH&S04-treated hepatocyte extracts were then assayed for 6-phosphofructo 2-kinase activity, pyruvate kinase, and fructose 2,6-P2 degrading activity. Purification of Rat Liver Fructose 2,6-Bisphosphatase-The phosphatase was purified from rat liver by a modification of the method of El-Maghrabi et al. (11). This method was also used to purify 6-phosphofructo 2-kinase. The conditions of homogenization, polyethylene glycol fractionation and chromatography on DEAE sephadex A 50 were as previously described (11). The pooled enzyme fractions from the DEAE-Sephadex step were then subjected to (NH&S04 fractionation and the 30 to 60 percent (NH&SO, pellet dissolved in buffer A/50 m~ KC1 and subjected to gel fitration on Sephacryl S-200 superfine (11). The pooled enzyme fractions (616 mg of protein and 700 milliunits of 6-phosphofructo 2-kinase) were then equilibrated with 20 mM TES, pH 7.5,lO mM KCl, 1 mM dithiothreitol, 0.1 mM EDTA, 5% glycerol, 0.5 pg/ml of leupeptin, and 0.2 m~ PMSF (buffer B) by ultrafdtration and dialysis and applied to a phosphocellulose column (2.5 X 8 em). The column was washed with buffer until the absorbance at 280 nm fell below 0.02. 6-Phosphofructo 2-kinase and fructose 2,6-bisphosphatase activities co-eluted with 5 mM fructose 6-phosphate in buffer B. The protein appeared homogeneous by the criterion of SDS-disc gel electrophoresis and the specific activities of the purified enzyme after specific elution from phosphocellulose were 50 milliunits of 6-phosphofructo 2-kinase/mg of protein and 20 milliunits of fructose 2,6-bisphosphatase/mg of protein (measured at submaximal fructose 2,6-bisphosphate concentrations). The enzyme pool (3 mg) was dialyzed extensively against buffer A containing 20% glycerol in order to remove fructose 6-phosphate and stored at -20 "C.
The reaction was stopped by adding NaOH to a final concentration of 0.25 N and heating for 30 min at 90 "C. The mixture was then neutralized, diluted 10-fold with 20 mM triethylamine HCO, (TEA-HCO,), pH 8.2, and applied to a DEAE-Sephadex column (0.7 x 5 cm). The column was then washed with 2 column volumes of 200 mM triethylamine HCOa, pH 8.2, and the eluate collected directly into scintillation vials and counted for 32P radioactivity. Fructose 2,6bisphosphatase activity was also assayed by measuring the production of fructose 6-phosphate spectrophotometrically as previously described for fructose 1,6-bisphosphatase (4). One unit of enzyme activity is the amount of fructose 2,6-bisphosphatase that catalyzes the hydrolysis of 1 pmol of fructose 2,6-P~/min.  (2).
Assay of Hepatocyte L-type Pyruvate Kinase Activity-Pyruvate kinase was assayed by previously described methods (16). Protein was assayed by the method of Lowry (18) with bovine serum albumin as a standard.

Identification and Isolation of Fructose 56-Bisphospha-
tase-Initial attempts to detect an activity which degraded fructose 2,6-P2 in crude rat liver extracts were unsuccessful because of the presence of large amounts of endogenous fructose 2,6-P2. An extract free of all low molecular weight effectors was prepared by (NH4)2S04 (0-70%) fractionation of a 100,000 X g supernatant fraction. When aliquots of this MINUTES FIG. 1. Time course of fructose 2,6-P, degradation catalyzed by an (NEt)SO&reated rat liver cytosol extract. The 100,OOO X g supernatant fraction of a rat liver homogenized in 3 volumes of buffer A in a Dounce homogenizer was brought to 70% saturation with (NH4)zS04. The precipitate was dissolved in the same buffer and dialyzed to remove the (NH4),S04. Two hundred microliter aliquots of the extract (0) or of buffer (0) were incubated with 10 p~ fructose 2,6-P~ and 5 m~ MgC12 in buffer A at 30 "C for the t i e s indicated. The reaction was stopped by the addition of NaOH to a concentration of 0.25 N and heating at 90 "C for 30 min and the amount of fructose 2,6-P2 remaining assayed in the neutralized extracts by the 6-phosphofructo 1-kinase activation assay as described under "Experimental Procedures." (NH&S04 treated extract were incubated with 10 p~ fructose 2,6-P2 and 5 mM MgClz it was possible to detect the disappearance of fructose 2,6-P2 with time by using the 6-phosphofructo 1-kinase activation assay (Fig. 1). No fructose 2,6-Pz was degraded in the absence of extract. In preliminary attempts to purify this degrading activity, it was found to copurify with 6-phosphofructo 2-kinase activity. We then decided to use the purification scheme, previously described by El-Maghrabi et al. (11) for the kinase, to purify the degrading activity. Both activities copurified at each step of this purification scheme. The purification of the enzyme to apparent homogeneity was made possible by our finding that the enzyme bound to phosphocellulose at pH 7.5 and could be specifically eluted with 5 mM fructose 6-phosphate. Both activities coeluted from the phosphocellulose column ( Fig.  2A) and they also eluted at the same position from a Sephadex G-100 superfine column (Fig. 2B) with an apparent molecular weight of about 100,000. Only a single peptide band was seen after SDS-disc gel electrophoresis (Fig. 3A).
In order to determine whether the enzyme hydrolyzed phosphate from the C-2 position of fructose Z,6-P2, fructose 2,6-[2-32P]-P2 was incubated with the purified enzyme and the products of the reaction were separated by DEAE-Sephadex a n /

I1
Effect ofcAMP-dependent phosphorylation on the activity of purified rat Liver fructose 2,6-bisphosphatase and 6-phosphofructo 2-kinase Twenty microliters of the enzyme preparation used in Table I were incubated with ATP (0.2 mM for the bisphosphatase experiment or 2 mM for the kinase experiment) and 20 units of the catalytic subunit of CAMP-dependent protein kinase (+ catalytic subunit) or an equivalent volume of protein kinase buffer ("catalytic subunit) for 15 min at 30 "C. The enzymes were then assayed for their respective activities at the indicated substrate concentrations as described under "Experimental Procedures." Values represent averages of at least three experiments. chromatography (2). There was a time-dependent appearance of radioactivity in the Pi region and a corresponding decrease in fructose 2,6-[2-32P]P2. An equivalent amount of fructose 6phosphate could also be measured by the spectrophotometric assay (data not shown). These results suggest that the enzyme catalyzes the following reaction: fructose 2,6-P2 4 fructose 6- Properties of the Purified Fructose 2,6-Bisphosphatase- Table I shows that enzyme activity was dependent on the substrate Concentration with an apparent K,,, of 10 to 20 PM.
The enzyme was not affected by fructose 1,6-bisphosphate or AMP at concentrations up to 1 mM (data not shown) but was activated by phosphate. The enzyme was inhibited by fructose &phosphate, especially at low concentrations of substrate. The concentration of fructose 6-phosphate that gave half maximal inhibition was 40 pM with 10 p~ fructose 2,6-P2 as substrate. The reaction was completely inhibited by the addition of EDTA in excess of magnesium ion, indicating that fructose 2,6-bisphosphatase is dependent on divalent cations (data not shown). Phosphorylation of Fructose 2,6-Bisphosphatase/G-Phosphofructo 2-Kinase-When purified preparations of the enzyme were incubated with the catalytic subunit of CAMPdependent protein kinase and [y3'P]ATP, "P radioactivity was incorporated into the single peptide band ( M , = 50,000) as determined by SDS-disc gel electrophoresis (Fig. 3). About 2 mol of =P were incorporated/mol of dimeric enzyme or 1 mol of 32P/mol of enzyme subunit (inset Fig. 3B). Phosphorylation resulted in a %fold activation of fructose 2,6-bisphosphatase activity, and also caused an inhibition of 6-phosphofructo 2-kinase activity ( Table 11). The effect of phosphorylation on the enzyme activities was seen best when the assays were performed at low substrate concentrations and were reduced in magnitude at high concentrations of substrate.
Effect of Glucagon o n Fructose 2,6-P2 Levels, 6-Phosphofructo 2-Kinase, and Fructose2,6-Bisphosphatase in Isolated Rat Hepatocytes-Since both 6-phosphofructo 2-kinase and fructose 2,6-bisphosphatase activities are affected by phosphorylation by the CAMP-dependent protein kinase, it was of interest to determine whether they were affected by the addition of glucagon to intact hepatocytes. Fig. 4 shows the effect of increasing glucagon concentrations on the above two activities as well as on the level of fructose 2,6-P2, and on the GLUCAGON ( n M ) Hepatocytes from fed rats were incubated with increasing concentrations of glucagon. Hepatocytes were homogenized and treated with (NH4)&04 as described under "Experimental Procedures." 6-Phosphofructo 2-kinase was measured by following fructose 2,6-P2 producactivity of pyruvate kinase, an enzyme known to be regulated by cyclic AMP-dependent phosphorylation. The concentration of glucagon that caused a 50% reduction in fructose 2,6-P2 levels was 2 X lo-" M while that necessary for half-maximal inhibition of 6-phosphofructo 2-kinase activity was 7 X 10"' M. These concentrations of hormone are much lower than that necessary to produce half-maximal inhibition of pyruvate kinase (3 X M) (Fig. 4 ) or half-maximal elevation of cAMP levels (3 X 10"' M) (21). Fig. 4 also shows that glucagon enhanced the activity of fructose 2,6-bisphosphatase by about &fold and that the glucagon concentration giving half-maximal stimulation of this activity was the same as that necessary for half-maximal reduction of fructose 2,6-P2 levels. Since fructose 2,6-bisphosphatase activity was measured in (NH4)&04-treated extracts, it is likely that the glucagon effect involves covalent modification of the enzyme.

DISCUSSION
The present results provide evidence for the existence of a fructose 2,6-P2 degrading activity in rat liver. This activity can be clearly designated as a fructose 2,6-bisphosphatase since a highly purified preparation of the enzyme was shown to hydrolyze phosphate from the C-2 position of the compound. Others (9,10,22) have also reported the existence of a fructose 2,6-P2 degrading activity but did not show that it specifically removed phosphate from the C-2 position. These authors also claimed that glucagon had no effect on this activity. This activity was not due to fructose 1,6-bisphosphatase since 1) the purified fructose 2,6-bisphosphatase did not contain any measurable fructose I,6-bisphosphatase activity (data not shown) and 2) fructose 1,6-bisphosphatase has been shown not to hydrolyze fructose 2,6-P2 (1, 4). The fructose 2,6-bisphosphatase reaction is also dependent on divalent cations and is inhibited by micromolar concentrations of fructose 6phosphate. Various estimates place the level of fructose 6phosphate in liver cytosol at 100-200 p~ and that of fructose 2,6-P2 at 8-16 p~ (12,17). From the data in Table 11, one would expect that fructose 6-phosphate would be an effective inhibitor of fructose 2,6-bisphosphatase in vivo. tion and is expressed as the ratio of activity at 0.5 mM and 5 mM fructose 6-phosphate. Pyruvate kinase was measured with 0.4 mM and 4 mM phosphoenolpyruvate. Fructose 2,6-bisphosphatase was measured with 10 PM fructose 2,6-P~. Disappearance of fructose 2,6-P 2 was followed with the 6-phosphofructo I-kinase activation assay. Hepatic fructose 2,6-P2 levels were measured as described under "Experimental Procedures." Fig. 4 provides support for the hypothesis that glucagon not only regulates the enzyme responsible for synthesis of fructose 2,6-P2, 6-phosphofructo 2-kinase, but that the hormone also activates fructose 2,6-bisphosphatase. This may explain the great sensitivity of hepatic fructose 2,6-P2 levels to glucagon. Since these changes occur at concentrations of hormone which have no measurable effect on cAMP levels (1,21) or on the inactivation of pyruvate kinase (16), glucagon may act on these enzymes by a CAMP-independent mechanism or by a CAMP-dependent mechanism where only very small, perhaps localized changes in CAMP, are necessary to induce phosphorylation of these enzymes. The results shown in Table I1 provide additional support for a role for cAMP in the regulation of these enzyme activities. Incubation of purified 6-phosphofructo 2-kinase/fructose 2,6-bisphosphatase with Mg-ATP and catalytic subunit of the CAMP-dependent protein kinase resulted in activation of the hydrolytic activity and inhibition of the kinase. Regardless of the mechanism, the ability of glucagon to affect these activities in a reciprocal manner provides an extremely rapid, efficient, and sensitive means for regulating hepatic glycolytic and gluconeogenic flux.
The results shown in Fig. 2 suggest that either the two enzymes are different proteins but very similar in terms of charge and size or that the two activities are present in the same enzyme ( i e . the enzyme is bifunctional). Support for the latter alternative includes the following: 1) both activities coeluted from a phosphocellulose column by fructose 6-phosphate; 2) the resulting preparation contained a single peptide band following SDS-disc gel electrophoresis; 3) both activities coeluted on a Sephadex G-100 column with an apparent molecular weight of about 100,000 (14); and 4) incubation of the enzyme preparation with catalytic subunit and [ Y -~P J ATP resulted in the phosphorylation of the peptide (M, = 50,000), concomitant with a reciprocal effect on the enzyme activities. However, the specific activity of the enzyme is low compared to that of other phosphoryl transferring enzymes, suggesting that the enzyme may have a low turnover number. Further studies on the characterization of the two activities and their interrelationship are in progress.