Rapid Reciprocal Changes in Rat Hepatic Glycolytic Enzyme and Fructose Diphosphatase Activities following Insulin and Glucagon Injection

SUMMARY Insulin and glucagon injected into the portal vein produced rapid reciprocal changes in the activities of certain hepatic glycolytic enzymes and fructose diphosphatase in the rat. a rapid increase in hepatic phosphofructokinase and pyruvate kinase activities and a decrease in fructose diphosphatase activity. but Fructose diphosphate aldolase activity was unaltered by either hormone. min following maximal by 10 rnin, and then gradually declined over the remaining 30 min of testing.

did not alter the response of the enzymes to insulin and glucagon, indicating that de novo protein synthesis was not responsible for the change in enzyme activities. Glucagon injected 5 min following insulin reversed the insulin effect; insulin given 5 min after glucagon also partially reversed the glucagon effect on cyclic adenosine 3':5'-monophosphate and the enzyme activities.
Intravenous insulin produced rapid changes in the activities of renal cortical, skeletal muscle, and epididymal fat glycolytic enzymes and fructose diphosphatase in the rat. Insulin (0.15 unit per kg) produced rapid increases in pyruvate kinase and phosphofructokinase and rapid decrease in fructose diphosphatase activities in all tissues. Fructose diphosphate aldolase activity was unchanged following insulin infusion.
Intravenous glucagon (0.15 mg) produced rapid changes, reciprocal to those seen with insulin, in fructose diphosphatase activity in all tissues. Glucagon significantly decreased epididymal fat phosphofructokinase activities but did not alter the activity of this enzyme in the renal cortex and skeletal muscle. Glucagon significantly decreased renal cortical pyruvate kinase but had no effect in the epididymal fat and skeletal muscle. Fructose diphosphate aldolase activity was unchanged in all tissues following glucagon infusion.
These data suggest that the glucagon responses seen in the four tissues studied are mediated by cyclic adenosine 3':5'monophosphate. toneally into rats weighing 200 to 220 g, 60 to 90 min before administration of glucagon or insulin. Actinomycin D, 0.66 rg per g of body weight, was injected intraperitoneally into rats weighing 200 to 250 g 2 hours before administration of glucagon or insulin. Control animals received injections of 0.9% NaCl solution but were otherwise treated identically with those pretreated with each drug.
Statistical analysis of the data was done using Student's t test (18). In all experiments, the means f standard error for three to five rats are given.

RESULTS
Insulin injection into the portal vein caused a rapid, significant (p < 0.01) increase in rat hepatic phosphofructokinase (at 5, 10, and 20 min) and pyruvate kinase (at 5 and 10 min) activities, a significant (p < 0.01) decrease in fructose diphosphatase activity (at 5, 10, 20, and 40 min), but no change in fructose diphosphate aldolase activity or cyclic AMP concentrations (Fig. 1). The insulin effect was time-dependent, being detectable within 5 min, and was maximal 10 min following the injection of 0.06 unit (0.3 unit per kg). The enzyme activities then returned toward control values. Forty minutes after insulin injection, pyruvate kinase and phosphofructokinase activities were back to control levels, but fructose diphosphatase activity was still significantly (p < 0.01) different from control values. The enzyme activities and cyclic AMP (not shown) in control animals given an injection of the NaCl vehicle were not significantly altered during the 40-min test period (Fig. 1).
The insulin effect was dependent on the amount of insulin injected (Table I). No effect was detected with 0.005 unit per kg; 0.015 unit per kg and greater produced a significant (p < 0.01) change in phosphofructokinase (0.015, 0.15, and 1.5 units per kg), pyruvate kinase (0.15 and 1.5 units per kg), and fructose diphosphatase (0.15 and 1.5 units per kg) activities. Enzyme activities in the control animals given injections of the NaCl vehicle were unaltered.
Puromycin and actinomycin D pret.reatment did not alter the significant (p < 0.01) enzyme responses to 0.15 unit per kg of insulin (Table II). When glucagon (150 pg) was injected 5 min after the insulin injection (0.15 unit per kg), the significant (p < 0.01) insulin effect on the activities of phosphofructokinase, pyruvate k&se, and fructose diphosphatase was reversed over the next 5 to 10 min (Table III).
This was accompanied by a significant (p < 0.01) increase in cyclic AMP concentration following the glucagon (Table III). Glucagon produced a significant (p < 0.01) decrease in phosphofructokinase and pyruvate kinase activities and a significant (p < 0.01) increase in fructose diphosphatase activity within 2 to 10 min after injection into the portal vein (Fig. 2). These significant (p < 0.01) enzyme changes were maximal after 5 to 10 min and persisted for at least 20 min following a single injection of 300 pg. The glucagon effect on the enzymes was preceded by a significant (p < 0.01) increase in cyclic AMP concentration, which occurred within 30 s after injection of the hormone (   detected. Injection of the glucagon vehicle had no effect on any of the enzyme activities (Fig. 2). The significant (p < 0.01) glucagon effect was dependent on the amount of hormone injected from about 15 pg to 300 pg (Table IV). No significant effects were seen with doses of 0.5 and 1.5 pg. The significant (p < 0.01) glucagon (150 pg) effect on these enzymes was unaltered by pretreatment of the rats with actinomycin D or puromycin (Table V). The significant (p < 0.01) glucagon (150 rg) effect on the enzyme activities was reversed when insulin (0.15 unit per kg) was injected 5 min after the glucagon (Table  VI). This was accompanied by a significant (p < 0.01) decrease in cyclic AMP concentration 5 min following the insulin (Table  VI). Table VII shows the effects of intravenous insulin (0.15 unit per kg) on rat epididymal fat, renal cortex, and skeletal muscle glycolytic enzyme and fructose diphosphatase activities. Insulin produced a rapid and significant (p < 0.01) increase in phosphofructokinase and pyruvate kinase activities in each of the three tissues, whereas fructose diphosphate aldolase activity was unchanged. In each of the three tissues, insulin produced a rapid and significant (p < 0.01) decrease in fructose diphosphatase activity.
The effect of intravenous glucagon (0.15 mg) on rat epididymal fat, renal cortex, and skeletal muscle glycolytic enzymes and fructose diphosphatase activities is shown in Table VIII.
The responses of phosphofructokinase (p < 0.01) and fructose diphosphatase (p < 0.01) activities in epididymal fat were similar to those in the liver, whereas epididymal fat pyruvate kinase and fructose diphosphate aldolase activities were unaffected by glucagon. In the renal cortex, glucagon significantly (p < 0.01) decreased pyruvate kinase activity and increased fructose diphosphatase activity but had no significant effects upon phosphofructokinase or fructose diphosphate aldolase activity. The only enzyme significantly altered by glucagon in skeletal muscle was fructose diphosphatase, which was significantly increased (p < 0.01) at 4 min.   Cyclic AMP (0.05 mmole per kg) given intravenously produced the same significant (p < 0.01) effect as glucagon on the hepatic enzyme activities (Table IX). Phosphofructokinase and pyruvate kinase activities decreased significantly (p < 0.01) within 5 min, were lowest at 10 min, and returned to normal by 15 min. Fructose diphosphatase activity increased significantly (p < 0.01) within 5 min and remained elevated for at least 15 min. No change was seen in fructose diphosphate aldolase activity.
The in tiuo effects of intravenous cyclic AMP on epididymal fat, renal cortex, and skeletal muscle enzymes are shown in Table IX. Cyclic AMP injection produced changes identical with those produced by glucagon in enzyme activities in the rena cortex and skeletal muscle. In the renal cortex, cyclic AMP significantly (p < 0.01) increased fructose diphosphatase activity, significantly (p < 0.01) decreased pyruvate kinase activity, but had no significant effect upon the other two enzyme activities. In skeletal muscle, the only enzyme significantly (p < 0.01) altered by cyclic AMP was fructose diphosphatase, which increased about 2-fold at 10 min postinjection. Cyclic AMP injections significantly (p < 0.01) decreased epididymal fat pyruvate kinase and phosphofructokinase activities, increased fructose diphosphatase activity, and had no significant effect upon the fructose diphosphate aldolase activity.
-CONTROL -GLUCAGON When a breakdown product of cyclic AMP, 5'-AMP (0.05 mmole per kg), was injected intravenously, there was no significant change in the activities of phosphofructokinase, pyruvate k&se, or fructose diphosphate aldolase (Table X). Although the mean value of pyruvate kinase was decreased at 5 min and increased at 15 min, these changes were not statistically significant. However, fructose diphosphatase activity was significantly (p < 0.05) decreased after 5 min but was the same as the control after 10 and 15 min (Table X).
Glucagon (1.0 mg per ml) and insulin (1 .O mg per ml) added to the 104,000 x g supernatant had no effect on the activities of any of the enzymes. In all animals, glucagon (15 to 300 pg) produced a rise in plasma glucose levels and insulin (0.005 to 1.5 units per kg) caused a drop in plasma glucose levels.

DISCUSSION
The antagonistic effects of insulin and glucagon on glycogen metabolism and glycolysis and gluconeogenesis in liver are well known, but the mechanisms by which these hormones produce their effects are not completely understood. In recent years these hormones have been found to exert their opposing effects on glycogen metabolism by controlling the activity of phos-phorylase and glucogen synthetase through phosphorylationdephosphorylation mechanisms, It seems clear that the glucagon effect is mediated by cyclic AMP and that this nucleotide regulates the activity of glycogen synthetase and phosphorylase by activating a protein kinase (1). A series of phosphorylations then occurs, finally resulting in phosphorylation of these two regulatory enzymes. Phosphorylase activity increases and glycogen synthetase activity decreases with phosphorylation.
Insulin, on the other hand, increases glycogen synthetase activity and decreases phosphorylase activity, although the mechanism is less well understood. Insulin rapidly increases the activity of a phosphatase which dephosphorylates glycogen synthetase (19). This effect does not seem to be mediated by a decrease in the cyclic AMP level. Although insulin does decrease cyclic AMP levels in the liver under certain conditions, it does not usually cause a drop below basal levels. It has, therefore, been suggested that insulin action might be mediated by a chemical messenger other than cyclic AMP and that this mediator could affect the cyclic AMP system (20, 21).
The mechanisms by which insulin increases glycolysis and decreases gluconeogenesis (22-26) and by which glucagon produces the opposite effects on these pathways (24, 25) have not  been defined. The long terms (hours) effects of insulin on glycolysis and gluconeogenesis seem to be due, at least in part, to its effects on synthesis of key regulatory enzymes such as pyruvate carboxylase, phosphofructokinase, pyruvate kinase, and fructose diphosphatase (4-6). Glucagon also has been shown to alter the synthesis of certain key regulatory enzymes in these pathways (7).
A rapid effect of these hormones on glycolysis and gluconeogenesis in the liver also has been demonstrated, but the mechanism of these effects is not understood.
The data which we present indicate that these two hormones rapidly alter the activities of key regulatory enzymes in these two opposing pathways in a reciprocal manner and could explain the rapid changes in gluconeogenesis and glycolysis produced by these two hormones.
Since the rate-limiting step in the glycolytic pathway in the vow.
------ liver seems to be at the phosphofructokinase step (23, 27), the rapid change in the activity of this key enzyme could explain the effects of glucagon and insulin on glycolysis.
It is of interest that the activity of other key glycolytic enzymes, including glucokinase (28) and pyruvate kinase, are also rapidly altered in a reciprocal manner by glucagon and insulin.
The significance of these multiple sites of control by the two hormones is unclear but would be consistent with the possibility that under different conditions the rate of the limiting enzyme in glycolysis might be different.
The rate-limiting step in the gluconeogenic pathway seems to be between the conversion of pyruvate to phosphoenolpyruvate (29)(30)(31).
In this study we have shown a stimulatory effect of glucagon on fructose diphosphatase activity and suggest that this enzyme may also play an important role   in the control of gluconeogenesis by this hormone. Other investigators have recently presented indirect evidence that glucagon can increase fructose diphosphatase activity (32,33). It is possible, therefore, that under certain conditions fructose diphosphatase may be rate-limiting in the gluconeogenic pathway. This is supported by the finding that low activity of this enzyme in human liver is associated with hypoglycemia (34-37). Preliminary findings from our laboratory suggest that two other key enzymes in the gluconeogenic pathway, pyruvate carboxylase and phosphoenolpyruvate carboxykmase, are not rapidly altered by glucagon and insulin in ~ioo.~ The glucagon effect in the liver on phosphofructokinase, pyru- vate kinase, and fructose diphosphatase was preceded by a significant increase in cyclic AMP concentration and suggests that the effect is mediated by this nucleotide. Additional support for this suggestion includes the fact that cyclic AMP mimics the effect of glucagon on gluconeogenesis in the isolated perfused rat liver (25,31). Epinephrine, the action of which is mediated by cyclic AMP, given in t&o, causes changes in the activities of these enzymes similar to those produced by glucagon (28). Finally, cyclic AMP injected into the rat portal vein produced changes identical with those seen with glucagon, whereas 5'-AMP had only a minimal, and opposite, effect on fructose diphosphatase. The lack of correlation between the effect of insulin on these enzymes and a change in cyclic AMP suggests that the insulin   effect is not mediated by this nucleotide. However, the possibility has not been excluded that the free (unbound) concentration of intracellular cyclic AMP was altered by insulin or that there was a change in cyclic AMP concentration in a specific intracellular compartment (38). The reversal of the glucagon effect when insulin was injected 5 min after glucagon was accompanied by a decrease in the elevated cyclic AMP level produced by glucagon. This lowering effect of insulin on elevated cyclic AMP levels produced by glucagon is similar to that reported by others (28,39). The rapid effect of these hormones was not altered by pretreatment with inhibitors of protein synthesis (at concentrations known to inhibit protein syntheses (4, 40-42)) 1 to 2 hours before the injection of the hormones. This finding and the fact that the hormone effects were maximal at 5 to 10 min indicate that de novo protein synthesis is probably not involved. The mechanism by which these hormones rapidly alter the activity of these enzymes is not known. However, it seems sufficiently plausible to consider tentatively that mechanisms similar to those involved in the control of glycogen synthetase and phosphorylase, i.e. phosphorylation-dephosphorylation, might exist. This would indicate that, through the same mechanisms, insulin and glucagon reciprocally regulate carbohydrate metabolism in several opposing pathways.
Although this POSsibility is most attractive, other possibilities should be considered; these include (a) allosteric effects of cyclic AMP on the enzymes and (b) alteration of the concentration of certain metabolic intermediates, such as fatty acids, which alter the activity of these enzymes in vitro (43). The latter possibility seems unlikely, since the intermediates would be diluted lo-to 2OO-fold in concentration in the 104,000 x g supernatant during the analysis of the enzyme activities. No single metabolic inter-    mediate is known at present which would account for all of the enzyme changes seen after the hormone injections. Particularly, it is difficult to envision one or more ligands generated by a hormone the action of which can be reversed by the antagonistic hormone.
In this connection, the work of Mansour (44), concerning the effect of epinephrine on rabbit skeletal muscle phosphofructokinase, should be considered. Mansour (44) found it essential to prepare his muscle extracts in the presence of caffeine in order to demonstrate the difference between the control and epinephrine-treated muscle enzyme. A concentration of 4.25 x lo-+ M cyclic AMP was also included in his assay mixture. The stimulatory effect of cyclic AMP and glucose 1,6-diphosphate on rabbit skeletal muscle phosphofructokiise required the presence of caffeine. It may well be that his epinephrine effect was related to the generation by epinephrine of one or more substances (ligands) which overcame the inhibitory effect of caffeine. In our studies, intravenous glucagon and cyclic AMP had no effect on rat skeletal muscle phosphofructokinase at a time when significant changes were occurring in hepatic phosphofructokinase, whereas insulin did stimulate phosphofructokinase in muscle and liver.
At least with regard to glucagon and cyclic AMP, no effect on skeletal muscle phosphofructokinase was found which would be consistent with the findings of Mansour (44) in the experiments in which caffeine was not used. Although we saw a small, significant change in skeletal muscle fructose diphosphatase activity following intravenous cyclic AMP, this may be of no physiological significance.
The studies with cyclic AMP suggest that muscle enzymes are not particularly responsive to cyclic AMP, because cyclic AMP is not generated by glucagon, or it does not enter muscle readily, or it is rapidly destroyed despite the injection of cyclic AMP.
One may conclude from the studies of Mansour (44) that ligands generated by hormones may stimulate or inhibit enzyme activity.
Conceivably, then, physiologically generated substances (ligands) might be implicated in the hormone-induced enzyme changes.
In many in vitro studies in which ligands have been shown to have direct effects on enzymes, the concentrations used have been higher than those usually found physiologically intracellularly (e.g. in liver, cyclic AMP levels range from 10 to 20 pmoles per mg of protein).
Fructose diphosphatase and phosphofructokinase activities have been reported to be altered by cyclic AMP in vi&o. Fructose diphosphatase activity from swine kidney was decreased by 0.6 mM cyclic AMP (45), and phosphofructokinase activity from liver fluke and sheep heart was increased by cyclic AMP (44, 46). These effects are opposite to those we observed with glucagon and cyclic AMP in Go.
The reason for these differences is not clear but may be related to the type of tissue, different species, and different conditions used in the different studies. When we add 1 mM cyclic AMP to the 104,000 x g supernatant from rat liver we observe a decrease in fructose diphosphatase activity.4 Insulin and glucagon affect certain enzymes of extrahepatic tissues (epididymal fat, renal cortex, and skeletal muscle) very much like that in the liver.
In each of the four tissues, intravenous insulin produced similar effects. Insulin significantly increased the activities of phosphofructokinase and pyruvate kinase, significantly decreased the activity of fructose diphosphatase, and had no effect upon fructose diphosphate aldolase activity.
The responses of the four enzymes to glucagon were not as consistent in each of the four tissues as those seen with insulin. This is not surprising, however, since glucagon is not known to have physiological effects in each of these tissues. Two of the four enzymes (fructose diphosphate aldolase and fructose diphosphatase) did respond uniformly in each of the four tissues. The fructose diphosphate aldolase activity was unchanged in each of the four tissues following glucagon injection.
In the extrahepatic tissues, glucagon produced a rapid and significant rise in fructose diphosphatase activity, which was mimicked by cyclic AMP.
It is interesting that, even in the two tissues which are nongluconeogenic (fat and skeletal muscle), glucagon still exerted a significant effect upon fructose diphosphatase activity. However, the activity of the enzyme in these two tissues is so low that the significance of the glucagon effect on fructose diphosphatase activity is not apparent.
Ghcagon produced a significant decrease in pyruvate kinase activity in renal cortex but had no effect upon fat or skeletal muscle pyruvate kinase activity. In fat, glucagon significantly decreased the activity of phosphofructokinase.
Phosphofructokinase activity was unchanged in the renal cortex and skeletal muscle.
Insulin produced uniform changes in the enzymes studied in all of the tissues tested, whereas glucagon did not. The lack of responsiveness of certain tissues to glucagon might be a reflection of the isozyme distribution of pyruvate kinase and phosphofructokinase.
Based upon the immunoreactivity of the two isozyme forms, the L (liver), or adaptive, form of pyruvate kinase has been found only in rat liver and erythrocytes (47). However, despite the failure of neutralization with the type L antibody, the crude kidney pyruvate kinase showed two peaks on starch zone electrophoresis, one of which had the same mobility as the type L enzyme (47). Since pyruvate kinase in the two glucagonresponsive tissues (liver and kidney) probably contains both the L and M isozyme forms whereas the pyruvate kinase in the two non-glucagon-responsive tissues exists only in the M form (47), it would seem that glucagon exerts its effect upon the L form of the enzyme.
Presumably, liver and adipose tissue may possess inducible forms of phosphofructokinase (called the L, or liver, form), whereas the skeletal muscle and kidney do not.
In recent years it has been shown that the activity of key enzymes in other metabolic pathways is altered by phosphorylation. These include the hormone-sensitive lipase in adipose tissue (48), acetyl-CoA carboxylase (49), and pyruvate dehydrogenase (50). Certainly these results provide, sufficient precedent to make consideration of a phosphorylation-dephosphorylation mechanism most tempting. Nevertheless, at least in the case of skeletal muscle phosphofructokinase, attempts to demonstrate such a mechanism have not been successful (44).