Oscillation in fructose 2,6-bisphosphate levels and in the phosphorylation states of fructose 6-phosphate,2-kinase:fructose-2,6-bisphosphatase in ischemic rat liver.

In order to determine the role of fructose (Fru) 2,6-P2 in stimulation of phosphofructokinase in ischemic liver, tissue contents of Fru-2,6-P2, hexose-Ps, adenine nucleotides, and Fru-6-P,2-kinase:Fru-2,6-bisphosphatase were investigated during the first few minutes of ischemia. The Fru-2,6-P2 concentration in the liver changed in an oscillatory manner. Within 7 s after the initiation of ischemia, Fru-2,6-P2 increased from 6 to 21 nmol/g liver and decreased to 5 nmol/g liver within 30 s. Subsequently, it reached the maximum value at 50, 80, and 100 s and decreased to the basal concentration at 60, 90, and 120 s. Oscillatory patterns were also observed with Glc-6-P and Fru-6-P, but the ATP/ADP ratio decreased monotonically. Determination of Fru-6-P,2-kinase activity and the phosphorylation states of Fru-6-P,2-kinase:Fru-2,6-bisphosphatase demonstrated that at 7 and 50 s, where Fru-2,6-P2 was the highest, the enzyme was activated and mostly in a dephosphorylated form. On the other hand, at 0, 30, and 300 s, the enzyme was predominantly in the phosphorylated form. The concentration of cAMP in the liver also changed in an oscillatory manner between 0.5 to 1.3 nmol/g with varying frequency of 10 to 40 s. These results indicated that: (a) Fru-2,6-P2 was important in rapid activation of phosphofructokinase in the first few seconds and up to 2-3 min, and (b) the oscillation of Fru-2,6-P2 concentration was the result of activation and inhibition of Fru-6-P,2-kinase:Fru-2,6-bisphosphatase, which was caused by changes in the phosphorylation state of the enzyme.

208 activation of the phosphatase activities (8)(9)(10). The net result is a decrease in Fru-2,6-P2 concentration and inhibition of phosphofructokinase and glycolysis, which explains the action of a hormone such as glucagon on the regulation of glucose metabolism in liver. However, whether Fru-2,6-P2 plays any role in the regulation of carbohydrate metabolism in liver under other stimuli or interventions is still unclear. Hue (11) earlier investigated a possible role of Fru-2,6-P2 in activating phosphofructokinase and glycolysis in liver under anaerobiosis. He found that the concentration of Fru-2,6-P2 decreased in hepatocytes when measured 5 min after initiation of anoxia even though glycolysis was stimulated. Furthermore, Fru-B-P,B-kinase activity was not affected by anoxia. Thus, he concluded that Fru-2,6-P2 could not be responsible for stimulation of glycolysis in liver under anaerobiosis.
More recently, we investigated a possible role of Fru-2,6-P2 in brain under anaerobic condition (12). A brain-blowing device (13) allowed the measuring of the changes in glycolysis within the first few seconds and up to a few minutes of ischemia. Results showed that Fru-2,6-P2 does not change significantly in spite of the fact that phosphofructokinase and glycolysis are activated severalfold within the first 2-5 s (12). However, we found that a new activator of phosphofructokinase, Rib 1,5-P2 is formed within 2 s after the initiation of ischemia and reaches the maximum in 5 s (12). We concluded that Rib 1,5-P2 (also a potent activator of phosphofructokinase) rather than Fru-2,6-P2, is important during the first few seconds of ischemia in brain.
In contrast to brain, the preliminary studies have shown that Rib 1,5-P2 was not detectable in ischemic liver, ruling out its role in this tissue. Since the stimulation of glycolysis by ischemia in most tissues occurs within a few seconds, we investigated changes in Fru-2,6-P2 and the activity of Fru-6-P,f-kinase and phosphorylation states of the enzyme in liver from 5 s to 2 min after initiation of ischemia.
Rats-Male Sprague-Dawley rats weighing 200-250 g were ob-tained from Sasco Co. (Omaha, NE). Rats were fed ad libitum with the standard National Institutes of Health diet. Preparation of Ischemic Liver--Rats were anesthetized by injection of nembutal (7.5 mg/100 g body weight). The abdominal cavity was opened and arteria hepatica and venae porta were clamped stopping the blood flow. At a given time interval, a piece of left lobe of liver weighing approximately 1-3 g was cut out and immediately freezeclamped by tongs which were precooled in liquid nitrogen. The time between cutting and freezing was within 1 s. For zero time control samples, a piece of liver was quickly cut out and freeze-clamped without clamping of blood vessels. Frozen liver samples were ground in liquid nitrogen with a mortar and a pestle and were used immediately or stored at -70 "C. In all experiments 6-12 liver samples were used for each point unless otherwise stated.
Metabolite Measurements-Fru-2,6-P2 was extracted with 0.1 M glycine/NaOH buffer, pH 10.0 (1:3 weight/volume) and assayed by the method of Thomas and Uyeda (18). All other metabolites were assayed spectrophotometrically on neutralized perchloric acid (3.6%) extracts of livers using enzymatic methods described previously (19). Cyclic AMP level was determined by the CAMP '251-scintillation proximity assay as provided by the manufacturer except trichloroacetic acid extraction was substituted for perchloric acid extraction.
Assay Method for Fru-6-P,S-kinase-The reaction mixture contained in a final volume of 0.1 ml: 100 mM Tris-HC1, pH 7.5, 0.1 mM EDTA, 0.1 mM EGTA, 5 mM ATP, 5 mM phosphate, 10 mM MgC12, and indicated amounts of Fru-6-P. The mixture was incubated at 30 "C, and at timed intervals 10-pl aliquots were transferred into 90 pl of 0.02 M Tris-HC1, pH 9.0, and the solution was heated for 1 min at 80 "C to stop the reaction. Suitable aliquots of the heated reaction mixture were then assayed for Fru-2,6-P2 as described (18). One unit of activity is defined as the amount of enzyme that catalyzes the formation of 1 pmol of Fru-2,6-Pz/min under these conditions. Assay Method for Fru-2,6-P-bisphosphatase-The reaction mixture contained in a final volume of 0.1 ml: 100 mM Tris-HC1, pH 7.5, 0.1 mM EDTA, 0.1 mM EGTA, 5 mM MgC12, 50 p~ NADH, 0.4 unit of desalted glucose-6-P-dehydrogenase, 1 unit of phosphoglucose isomerase, and 20 p~ Fr~-2,6-[2~'P]P~ (8.2104 cpm/nmol). The reaction was initiated with the addition of the enzyme, and the reaction mixture was incubated at 30 'C. At suitable time intervals, aliquots were removed and transferred into 100 pl of ice-cold 0.1 N NaOH. This mixture was diluted by 1 ml of 0.02 M NH,OH and applied on a Dowex-l-Cl-X8 column (0.52 cm) which was equilibrated with 0.02 N NH4Cl. The column was washed with 20 ml of 0.02 N NH4OH and 0.5 M NaCl in NH4Cl.
[32P]Phosphate was then eluted with 5 ml of the same solution. A portion (1 ml) of the eluate was mixed with 15 ml of scintillation liquid (SCINT-A XG, Packard) and counted in a scintillation counter. One unit of activity is defined as the amount of enzyme that catalyzes the formation of 1 pmol of phosphate/min under these conditions. Assay for Glycogen Phosphorylase a-Glycogen phosphorylase a was assayed using a modification of the method described by Stalmans and Hers (20). After a given duration of ischemia, a piece of liver was cut out and immediately homogenized in ice-cold buffer containing 50 mM glycerol-P, pH 7.5, 150 mM KF, 5 mM EDTA, 1 mM dithiothreitol and centrifuged at 15,000 X g for 20 min. The supernatant solution was used for phosphorylase a assay.
Preparation of Tissue Extract and Partial Purification of Fru-6-P,2-kinase:Fru-2,6-bisphosphutase-Fresh ischemic liver was homogenized in ice-cold 50 mM Tris-phosphate, pH 8.0, 0.1 M KF, 5 mM EDTA, 5 mM EGTA, 1 mM dithiothreitol, 1 mM benzamidine, and 0.2 mM phenylmethylsulfonyl fluoride (buffer A) using a Polytron homogenizer. The homogenate was centrifuged at 33,000 X g for 30 min, and the supernatant solution (4 ml) was applied on a Blue Sepharose column (0.75 X cm) which was equilibrated with buffer A. The column was washed with 40 ml of buffer A, and the enzyme was eluted with buffer A containing 2 M NaCl (5 ml). Fractions (1 ml) were collected and assayed for Fru-B-P,2-kinase activity. Usually fraction 3 contained most of the enzyme, and the recovery of the enzyme was approximately 84 f 5% of that in the crude extract.
Polyacrylamide Gel Electrophoresis-Polyacrylamide slab gel electrophoresis in sodium dodecyl sulfate was carried out in 10% acrylamide containing 0.1% sodium dodecyl sulfate according to the procedure of Laemmli (21). The gels were stained with Coomassie Brilliant Blue and destained. For autoradiography of the gels Kodak X-Omat AR film was used.
Other Methods-Protein concentration was determined with the Bradford method (22) using crystalline bovine serum albumin as a standard.

RESULTS
Oscillation of Fru-2,6-Pz-The concentration of Fru-2,6-P2 in the liver underwent rapid pulsatory changes after the initiation of ischemia (Fig. 1). After a slight initial (3 s) decrease, it reached the maximum value of 21 nmol/g within 7 s and decreased to 5 nmol/g in 30 s. After the initial cycle the Fru-2,6-P2 level increased to the maximum concentration of 13, 17, 16 nmol/g at 50, 80, and 100 s, respectively, and decreased to the basal values at 60,90, and 120 s. Beyond 120 s the oscillatory change ceased and the Fru-2,6-P2 concentration remained constant at 3 nmol/g up to 5 min. In contrast, the adenine nucleotides concentration changed rather monotonically ( Fig. 2 A ) . ATP decreased rapidly to 1.2 pmol/g from 2.6 pmol/g within the first 30 s and continued to decrease more slowly, while AMP increased rapidly for the first 30 s, followed with a slow increase. The concentrations of Glc-6-P and Fru-6-P changed in an oscillatory manner, reaching the maximum values at 7, 30, and 70 s, and the minimum at 10, 40,80, and 120 s. It should be noted that the mean values of Glc-6-P and Fru-6-P in the peaks and valleys of oscillation are significantly less than the overall curves. Unlike Fru-2,6-Pz, however, the concentrations of these hexose phosphates did not decrease to the basal level at the minima (Fig. 2B). Both Fru-1,6-P~ and lactate continued to accumulate. However, it is difficult to decide whether Fru-l,6-P2 underwent oscillation because of inaccuracy of the assay due to extremely low level in uiuo (Fig. 2C). Comparison of the oscillations of Fru-2,6-Pz and hexose-Ps revealed that their oscillatory pe- riods were similar (approximately 20-30 s), but the periods were out of phase except at 7 s of ischemia. Both hexose-Ps and Fru-2,6-P2 increased severalfold within 7 s, but Fru-1,6-P2 did not increase until 10 s. These changes in hexose-Ps suggest the activation and the inhibition, i.e. oscillation of phosphofructokinase activity.
Oscillation of Fru-6-P,&kinase Activity-In order to determine if the oscillation of Fru-2,6-P2 is as a result of activation and inhibition of Fru-6-P,2-kinase by phosphorylation and dephosphorylation of the bifunctional enzyme rather than by changes in allosteric effectors, the activity of Fru-6-P,Z-kinase at two different concentrations of Fru-6-P was determined in the ischemic livers. The kinase was assayed at 0.05 mM (v) and 1.0 mM ( Vmax) of Fru-6-P, where v/VmaX = 1 for a fully dephosphorylated and v/Vm., = 0.4 for a fully phosphorylated form of the bifunctional enzyme (23). For this assay the enzyme in the crude extract of a fresh ischemic liver was partially purified by Blue Sepharose chromatography and assayed. As shown in Table I, the activity ratios of the enzyme at 0 and 30 s of ischemia (at which times the Fru-2, 6-P2 concentration was minimal) were 0.5 & 0.05 and 0.4 & 0.11, respectively. The ratios increased to 0.98 & 0.03 at 7 s and 50 s, at which time Fru-2,6-P2 increased to the maximum level (Fig. 1). These results strongly suggested that the Fru-6-P,2kinase:Fru-2,6-Pase was mostly in dephospho form at 7 and 50 s, while it was in phospho form at 0 and 30 s.
Phosphorylation and Dephosphorylation of a Bifunctional Enzyme-To demonstrate directly whether the phosphorylation and dephosphorylation of the enzyme had occurred during the ischemia-induced oscillation, the partially purified enzymes from livers at various time after ischemia were phosphorylated by CAMP-dependent protein kinase in vitro in the presence of [32P]ATP, precipitated with specific antibodies, and analyzed by polyacrylamide gel electrophoresis.

Activiw ratio of Fru-b-P,S-kinase in ischemic liver and p2PJ phosphate incorporation in Fru-6-P,2-kinase:Fru-2,6-Pase
For the determination of phosphate incorporation, Fru-B-P,Z-kinase:Fru-2,6-Pase in ischemic liver was partially purified, phosphorylated by CAMP protein kinase, and immunoprecipitated as in Fig. 2A. The [32P]phosphate content of the enzyme was determined by the filter assay method of Corbin and Reimann (24). The amount of the enzyme was estimated from Fru-&P,B-kinase activity using the specific activity of 55 milliunits/mg and M, of subunit = 55,000 (23). The amount of 32P incorporation was also determined by the filter assay for protein kinase according to the procedure of Corbin and Reimann (24). The Coomassie Blue-stained gel indicated that the protein band corresponding to the Fru-6-P,2-kinase:Fru-2,6-Pase in addition to heavy IgG bands was visible in all the samples (Fig. 3A). The autoradiogram of the same gel showed significant 32P incorporation into the enzymes after 7 s (lane 2 ) and 50 s (lane 4 ) of ischemia, indicating that these enzymes were mostly in dephospho forms in vivo. This was in contrast to the enzymes after 0,30, and 300 s of ischemia, in which little 32P was incorporated, indicating that the enzymes were in phospho forms in these livers. Interestingly, the degree of incorporation decreased with time, indicating that the amount of phosphorylated enzyme increased with time of ischemia, and after 300 s of ischemia the enzyme was fully phosphorylated. The determination of the amount of [32P]phosphate incorporation demonstrated that at zero time 0.33 mol of phosphate incorporated/mol of enzyme subunit which indicated that 33% of the enzyme was in dephospho form in vivo, but within 7 s of ischemia 70% of the enzyme was in dephospho form. Similarly, 27,50, and 0% of the enzymes were in dephospho forms at 30, 50, and 300 s, respectively (Table I).
Changes in CAMP in Ischemic Liver-Since the above results demonstrated that the phosphorylation of the bifunctional enzyme was involved when the Fru-2,6-P2 decreased, and thus far only CAMP-dependent protein kinase is known to phosphorylate the enzyme, it was important to determine whether cAMP also undergoes oscillation. The results (Fig.  4) showed that oscillatory changes in cAMP concentration occurred in the ischemic livers. For comparison the changes in Fru-2,6-P2 from Fig. 1 were also included in Fig. 4. The concentrations of cAMP reached the maximum level at 5,10, 50, and 80 s and the minimum level at 7, 20, and 60 s under these conditions. One would expect that the oscillation of cAMP and Fru-2,6-P2 is out of phase because of the opposite relationship between these compounds in liver. The expected opposite effect was observed only at 7 s at which time Fru-2,6-Pz reached the maximum as cAMP decreased to the minimum value. At other time points, however, this opposite relationship was not maintained. For example, at 50 and 80 s both cAMP and Fru-2,6-P2 reached the maxima and at 60 s both reached the minima.
Change in Glycogen Phosphorylase a in Ischemic Liuer-Glycogen phosphorylase a activity in fresh ischemic liver was also determined. Results showed that activity increased from 6.5 units/g liver to 8.5 units/g within 7 s, but subsequent changes were relatively small (data not shown).

DISCUSSION
These results showed for the first time that sudden ischemic conditions in rat liver induced oscillatory changes in Fru-2,6-Pz and glycolytic intermediat.es. The oscillatory cycle began within a few seconds after the initiation of ischemia and lasted up to 2 min. The oscillation period of Fru-2,6-Pz was 20-30 s, which is extremely short compared to othar known oscillations such as glycolysis in yeast (25)(26)(27), heart (28), and muscle (29,30) extracts, insulin secretion from islets (31)(32)(33), and Ca2+ release in signal transduction (for a review, see Ref. 34). The oscillation periods in these systems range from 5 min to hours.
These results further demonstrated that Fru-2,6-P2 appeared to play an important role in triggering the rapid activation of phosphofructokinase in liver following initiation of ischemic conditions. There was no significant amount of ribose 1,5-P2 in any of the ischemic liver samples unlike ischemic rat brain (12). The observed changes in hexose phosphates were consistent with the activation of phosphofructokinase by the changes in the Fru-2,6-P2 concentrations. Hue (11) concluded that Fru-2,6-P2 is not responsible for the activation of glycolysis in ischemic liver because he failed to detect any change in Fru-2,6-P2 in hepatocytes or in liver under ischemia. The likely explanation for the discrepancy is that he measured Fru-2,6-P2 only at 5 and 10 min after initiation of ischemia. As shown in Fig. 1, the oscillation of Fru-2,6-Pz and major changes in hexose-Ps and adenine nucleotides occurred within the first few seconds and persisted only for 2 min. As is the case in brain (12,35), the present results with the ischemic liver emphasize the importance of determining the metabolic changes within the first 30 s in order to understand the mechanism of activation of phosphofructokinase and glycolysis. The Fru-2,6-P2 oscillation ceased within 2 min, and beyond this time it remained at a basal level (Fig. 1). The question then is does the activation of phosphofructokinase beyond 2 min continue under these conditions? It is possible that a combination of increasing Frul,6-P2, which activates the enzyme in an autocatalytic manner (especially in the presence of Fru-2,6-Pz and increased AMP), and decreased ATP (an inhibitor), could account for the sustained activation of phosphofructokinase and glycolysis.
This rapid activation and inhibition of the bifunctional enzyme caused mainly by the phosphorylation and dephosphorylation of the enzyme rather than changes in concentration of Fru-6-P or other metabolites. The following lines of evidence support this conclusion: ( a ) phosphorylation of the bifunctional enzyme, catalyzed by CAMP-dependent protein kinase, causes inhibition of Fru-6-P,B-kinase and activation of Fru-2,6-Pase activities and results in a decrease in Fru-2,6-P2 concentration. Dephosphorylation of the bifunctional enzyme by a protein phosphatase causes the opposite effect resulting in increased Fru-2,6-P2 levels. This study demonstrated that when the Fru-2,6-Pz level was low at 0, 30, and 300 s post-ischemia, the bifunctional enzyme was in the phosphorylated state, but when the level was high at 7 and 50 s, the enzyme was in a dephosphorylated state (Fig. 3). ( b ) The concentration of Fru-6-P, the substrate for the bifunctional enzyme, did not oscillate in conjunction with Fru-2,6-P2 levels. For example, Fru-2,6-P2 reached a maximum concentration of 21 ~L M at 7 s, while Fru-6-P increased from 70 to 100 nmol/g, and then continued to increase reaching 200 nmol/g at 70 s. ( c ) The Fru-6-P level was at least four to five times higher than the estimated K, value of 16 ~L M as determined with purified enzyme in vitro (24). ( d ) Rapid decrease in Fru-2,6-P2, which is catalyzed by Fru-2,6-Pase, was observed in the oscillation in spite of the presence of a high concentration of Fru-6-P, a potent inhibitor of Fru-2,G-bisphosphatase with Ki = 1.2 p~ (36). All these results argue against the possibility that Fru-6-P concentration regulates the Fru-6-P,2-kinase activity in liver under these conditions. Oscillation in the phosphorylation state of the bifunctional enzyme indicates that either the activities of both protein kinase A and protein phosphatase are coordinately regulated in a reciprocal manner or one of these enzymes undergoes oscillation while the other remains at a constant activity. To our knowledge such a coordinated regulation of protein kinase A and a protein phosphatase has not been reported. Closer examination of the oscillation of cAMP and Fru-2,6-P2 (Fig.  4) showed a reciprocal relationship between these two compounds at 7 s. At 7 s of ischemia cAMP dropped rapidly, presumably inhibiting protein kinase A, which resulted in activation of Fru-g-P,B-kinase and inhibition of Fru-2,6-bisphosphatase activities and increased levels of Fru-2,6-P2. Whether the protein phosphatase was activated or remained the same during this period remains to be determined. Beyond 7 s of ischemia, the reciprocal relationship between cAMP and Fru-2,6-P2 did not occur, but instead they changed in synchrony. The changes in the protein kinase and the protein phosphatase during this period also need to be elucidated.
The physiological significance of the oscillation of Fru-2,6-Pz in liver under ischemia is not clear at present. However, oscillatory behavior usually serves to trigger rapid and efficient stimulation of physiological and metabolic changes. It is possible that the Fru-2,6-Pz oscillation serves to trigger the activation of phosphofructokinase and thus glycolysis in liver during the initial phase of aerobic to anaerobic shift. The oscillation of glycolysis caused by a series of bursts of phosphofructokinase activity in crude muscle extract has been shown to respond to ATP/ADP, and thus the tissues can maintain high ATP/ADP (37). This is not the case in the ischemic liver since the ratio and the ATP concentration continued to decrease in the tissue. In the liver it appears that Fru-2,6-Pz is the major triggering factor for the rapid activation of phosphofructokinase. As Berridge (34) and Rapp (38) pointed out that the advantage of oscillation in signal transduction is higher signal to background ratio. Relatively infrequent large bursts of Fru-2,6-P2 as in the case of the oscillation may be more effective than steady smaller increases in activation of phosphofructokinase in ischemic liver. The important question remains to be answered, i.e. the mechanism of regulation of protein kinase A and the protein phosphatase, whose activities control the relative activity of Fru-6-P,2kinase and Fru-2,6-Pase and ultimately the level of Fru-2,6-PP.