Studies on the Regulation of Rat Liver Pyruvate Kinase and Fructose-l,6=Bisphosphatase

Fructose-1.6-bis~hosuhatase from rat liver. A comuarison of the kinetics of the pnuhosuhorvlated enzvme and the e nzvme uhosphorvlated bv cvc lic AMP-deuendent protein kinase


INTRODUCTION
In the living cell there is a constant need for rapid adjustment of various anabolic and catabolic reactions in response to extracellular stimuli such as hormones.
Control of metabolic pathways can occur at several different levels using mechanisms such as substrate and cofactor availability, product removal, feedback regulation, allosteric effects, covalent modifications of the enzyme such as phosphorylation or proteolytic modification, and induction or repression of the total amount of the enzyme (56).

REGULATORY PHOSPHORYLATION
Hormones transmit information to the interior of the cell by activating transmembrane signaling systems. About half of the hormones found in mammals, e. g. glucagon and epinephrine, act by increasing the concentration of 3 , 5 cyclic-AMP (CAMP) (27,70).

I .
CAMP regulates protein kinase activity thus altering the phosphorylation states and biological properties of many intracellular enzymes (for a review of the current knowledge see 10).
The first enzyme of which the activity was reported to be regulated by phosphorylation was glycogen phosphorylase (73). The rapid increase in the number of enzymes reported to serve as substrates in phosphorylation-dephosphorylation reactions has made it necessary to formulate certain criteria for establishing that the phosphorylationdephosphorylation reactions have biological significance (44, 56). here with special reference to CAMP-dependent protein kinase: 1. The existence of a substrate for CAMP-dependent protein kinase which participates in a metabolic pathway influenced by CAMP. The rate of phosphorylation in vitro must be compatible with the speed with which the process occurs in vivo.
2. A reversible change in enzyme activity in vitra caused by phosphorylationdephosphorylation catalyzed by CAMP-dependent protein kinase and phosphoprotein uvruvate 7 . .
d in vivo farms.
Since the activity of rat liver pyruvate kinase of the L-type is regulated by reversible phosphorylation in vivQ the proportions between phosphorylated and unphosphorylated pyruvate kinase could be expected to differ in the liver of animals in different nutritional states.
The phosphorylated and unphosphorylated forms of an enzyme generally differ very little in isoelectric point, PI, (11) and it is hard to separate the two forms by ion-exchange chromatography. In order to overcome these difficulties the chromatofocusing system was used (14). Cell sap from livers of rats which had been fed a L U i b m , starved for 48 h, or which had been fed on a diet containing 40 % fructose, was subjected to chromatography on DEAE-cellulose. This procedure removed about 95 % of the cellular proteins. The remaining proteins, including pyruvate kinase of the L-type, were separated by chromatofocusing, a procedure which yielded three fractions exhibiting pyruvate kinase activity and having apparent PI'S of 5.0, 5.2 and 5.3, respectively.
These forms were identified as phosphorylated, unphosphorylated and proteolytically degraded pyruvate kinase, respectively, on the grounds of data on kinetics and phosphorylation. The identifications of the three forms of pyruvate kinase were further corroborated by separately subjecting unphosphorylated, phosphorylated and proteolytically digested pyruvate kinases to chromatofocusing under the same conditions. Since it was impossible to tell which protease had generated the third form of pyruvate kinase, an enzyme that had been degraded by a calcium activated protease was used as a model of digested pyruvate kinase. The calcium activated protease is known to remove a small peptide containing the phosphorylation site from the L-type of pyruvate kinase (12). The phosphorylated, unphosphorylated and digested pyruvate kinases eluted at pH 5.0, 5.2 and 5.3, respectively, and the degradation product showed an apparent Km which was higher than that of both phosphorylated and unphosphorylated pyruvate kinase. It was therefore concluded that the third pyruvate kinase fraction found in rat liver cell sap represented a form which had been proteolytically modified, and also that the proteolytic attack did not occur during purification, since the relative amount of this form did not increase during the course of preparation.
Finally the amounts of the three different forms were determined in the livers from rats that had been subjected to different diets. The enzyme form believed to be proteolytically modified was most abundant in livers from starved animals, where it accounted for about 15 % of the total activity. The corresponding values for the group of rats fed iidJhhm was 10 %, while it was 5 % for those held on a fructose-rich diet. The phosphorylated form of pyruvate kinase was predominant in all diet groups, amounting to between one half and two thirds of the total activity.
Fru-1,6-P2:ase forms part of a control point in the glycolytic and gluconeogenetic pathways. A close regulation of the enzyme is important since the simultaneous operation of both Fru-l,6-P2:ase and phosphofructokinase within a single system would lead to hydrolysis of ATP and thus to a short-circuit in metabolism (37).

Occurrence
Fru-1,6-P2:ase has been isolated and characterized from many systems e.g. mammal liver, kidney, intestine, muscle, mammary glands, brown adipose tissue, brain and placenta (for a complete review see 78). The highest activity of Fru-1,6-P2:ase is seen in gluconeogenetic tissues such as liver and kidney cortex.
Mammalian Fru-1,6-P2:ase is a tetrameric enzyme having identical subunits and having a molecular weight of about 140 000, The complete amino acid sequence has been determined for Fru-1,6-P2:ase from pig kidney cortex (48) and sheep liver (29). These sequences show a high degree of homology with 90 % being identical (36).
Native Fru-1,6-P2:ase has a neutral pH-optimum (e. g. 79). It was shown that digestion of native Fru-l,6-P2:ase with papain (61) and subtilisin (80) yielded enzyme forms with alkaline pH-optima by removal of the N-terminal part of the polypeptide chains.
Zn2+ has a dual role concerning the activity of Fru-1,6-P2:ase; it inhibits the activity at low concentrations, while at higher concentrations it can replace Mg2+ or Mn2+ as the activating cation (58).
Intact Fru-1,6-P2:ase is inhibited by AMP at millimolar levels (55, 75). Subtilisin digestion of the enzyme abolishes the inhibitory effect of AMP (61). It has also been reported that Fru-1,6-P2ase is inhibited by ADP and ATP but higher concentrations are needed than of AMP for a similar effect (30,51,75).
Fru-2,6-P2 in the micromolar range is a potent inhibitor of Fru-1,6-P2:ase activity. Its effect is synergistic with that of AMP (for a review see 38). Hormonal studies have shown that the concentration of Fru-2,6-P2 in rat liver can be lowered by both glucagon and epinephrine (31,62).
By and large, present knowledge of the molecular interactions between Fru-1,6-P2:ase and the catalytic metal ion, AMP, Fru-2,6-P2 and Fru-1,6-P2 may be summed up as follows: Binding of the metal ion and AMP is of competitive nature: binding of Fru-2,6-P2 and binding of Fru-1,6-P2 are also mutually exclusive; and binding of Fru-2,6-P2 increases the affinity of the allosteric site for AMP (68).
The activity of Fru-1,6-P2:ase is also affected by a number of activating substances such as the chelating agents EDTA, histidine and citrate (59) and nonchelating agents like fatty acids (1) and phospholipids (7).
-r at li ver F r u -m It was reported in 1977 (64) that purified rat liver Fru-1,6-P2:ase is a substrate for CAMP-PK in the presence of Mg2+ and with (32P)ATP as phosphate donor, and it was seen that phosphorylation occurs both in vitra and in viva. About 4 mol of phosphate were incorporated per mol of enzyme tetramer and bound to serine (64). It was observed that trypsin digestion removed peptides from the enzyme containing all the incorporated phosphate (60). The amino acid sequence around the phosphorylated serine was determined (60) as Ser-Arg-Pro-Ser(P)-Leu-Pro-Leu-Pro. The same sequence except with Tyr in position 3 was found (40) and a synthetic peptide with the same sequence proved to be a substrate for CAMP-PK although with a comparatively high K, .
The phosphorylatable amino acid sequence was shown to be located close to the C-terminal end of the enzyme subunits (39) and its exact location was determined as Set-341 (65).
Reports of a second phosphorylated amino acid in the C-terminal sequence of Fru-1,6-P2:ase appeared when it was found that another serine residue beyond Arg-348 was phosphorylated (65). This residue was later identified as Ser-356 (8) within the sequence

Ser-Arg-Ala-Arg-Glu-Ser-Pro-Val-His-Ser(P)-Ile.
The influence of phosphorylation on the activity of Fru-1,6-P2:ase has been the subject of some dispute. A few authors have not detected any change in activity ( e. g. 65) while others have detected an increase in Vmax (64), or a decrease in apparent Km (20,50).

and AMP on t h m i v i t v of uho-
Ugphosuhormed fruc -from rat liver.
The reaction was interrupted by removal of (32P)ATP on a Sephadex G-50 column before the phosphorylation reaction was estimated to be completed. The partially phosphorylated Fru-1,6-P2:ase was subjected to chromatofocusing which yielded two fractions with enzyme activity: one with an apparent PI of about 5.0 and the other of about 4.5. The former fraction corresponded to unphosphorylated and the latter to phosphorylated Fru-l.6-PZ:ase (15).
The effects of Fru-2,6-P2 and AMP on the two enzyme forms were examined and it was found that unphosphorylated Fru-l,6-P2:ase was more sensitive than the phosphorylated enzyme to both inhibitors. AMP acted by decreasing the VmaX of the enzyme; the presence of 25 pM AMP decreased the Vmax to 70 % of the uninhibited value for the phosphorylated and to 40 % for the unphosphorylated form of Fru-l,6-P2:ase. The effect of Fru-2,6-P2 was to lower Vmax as well as to increase the apparent K, for Fru-1,6-P2. The amount of Fru-2,6-P2 needed for a 50 % decrease in Vmax was 11 pM for the phosphorylated and 4 pM for the unphosphorylated enzyme. It was also shown that the combined effect of AMP and Fru-2,6-P2 on the VmaX of Fru-1,6-P2:ase was more than additive.
The finding that the phosphorylated and unphosphorylated forms of Fru-l,6-P2:ase differ in their sensitvity to the two important inhibitors Fru-2,6-P2 and AMP, and the observation that those inhibitors act synergistically, might provide a way to amplify the effect of phosphorylation. Administration of glucagon to hepatocytes induces phosphorylation of both Fru-1,6-P2:ase (9) and the bifunctional enzyme 6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase (21,67). The effect of phosphorylation of the latter enzyme is an inhibition of the kinase and a stimulation of the phosphatase activity. So the administration of glucagon gives a decrease in the concentration of Fru-2,6-P2 as well as in the amount of unphosphorylated and more easily inhibited Fru-1,6-P2:ase.
These observations give rise to speculations that the physiologically most important effect of phosphorylation of Fru-l,6-P2:ase is to decrease the sensitivity to AMP and Fru-2,6-P2 rather than to affect the apparent K, for Fru-1,6-P2. These improvements resulted in an increase in the incorporation of phosphate and an increased pH-ratio, the activity of the enzyme at pH 7.5 divided by that at pH 9.2, (16).
It was investigated whether the presence of the inhibitor Fru-2,6-P2 affected the rate of phosphorylation of Fru-1,6-P2:ase, as in the yeast enzyme (32), but no such effect could be seen.
The effects of various factors on the activity of phosphorylated and unphosphorylated Fru-1,6-P2:ase were determined. It was confirmed that unphosphorylated Fru-1,6-P2:ase was more susceptible to inhibiton by AMP and Fru-2,6-P2, while both forms were inhibited equally but less efficiently by ADP and ATP. implying that the two former inhibitors might be more important for the regulation of Fru-1,6-P2:ase.
Both phosphorylated and unphosphorylated Fru-1,6-P2:ase was fully active at 1 mM M g 2 + and no inhibition was seen at concentration of up to 5 mM. The two forms also had neutral pH-optima although the profile for phosphorylated Fru-1,6-P2:ase was more level.
Citrate, 3 mM, or histidine, 1 mM, did not affect the activity of either of the forms.
Under conditions resembling the metabolic state during gluconeogenesis the phosphorylated form of Fru-1,6-P2:ase had twice the activity of the unphosphorylated.
Effects of li n on the acti vi ty and-ofhorvlation insu ' -- Hepatocytes were kept in primary culture for 24 h. After that time they were incubated with or without radioactive (32P)orthophosphate for 30 min to label endogenous ATP, and fresh medium containing the hormones to be tested was added. After a suitable period of incubation the medium was sucked off and the cells were frozen in liquid nitrogen (17).
The activity of Fru-l,6-P2:ase was analyzed in the thawed and centrifuged cell extract.
Changes in Km were monitored as changes in the activity at 12.5 p M Fru-1,6-P2 divided by that at 83 p M (corresponding to an approximately saturating substrate concentration).
The degree of phosphorylation of Fru-1,6-P2:ase was determined after removal of the other labeled cell components by chromatography on a column with anti-Fru-1,6-P2:ase coupled to CNBr-activated Sepharose.
It was seen that glucagon caused a decrease in apparent Km as judged from the change of the activity ratio from 0.57 to 0.66. Maximal effect was obtained with a glucagon concentration of 2 x 10-l' M , the effect was complete within 10 min and lingered on for at least 2 hours.
Epinephrine increased the activity ratio of Fru-1,6-P2:ase even more, to a ratio of 0.76, and more quickly. The concentration of hormone needed for maximal effect was 2 x M; activation was complete within 2-5 min and lasted for less than 15 min. Treatment of the (32P)labeled hepatocytes with glucagon increased the incorporation of phosphate from 2.5 mol per mol of tetrameric enzyme (found in Fru-1,6-P2:ase from cells 2 which had not been exposed to hormone) to 4.2 mol per mol. Epinephrine stimulated phosphorylation to 3.5 mol of phosphate per mol of enzyme. The effect of the two hormones in this respect was additive. In contrast, insulin brought about a decrease in the degree of phosphorylation to 2.0 mol per mol enzyme tetramer.

Treatment of the hepatocytes with insulin resulted in an increase in apparent
It was also shown that the effect of glucagon and epinephrine was not synergistic since addition of the two hormones together affected the activity no more than epinephrine alone.
It was shown that phosphorylation of Ser-356 required the intact three-dimensional structure of the enzyme as evidenced by the observation that this site was not phosphorylatable after unfolding of the enzyme with 4 M urea prior to phosphorylation. (The urea was diluted in the phosphorylation mixture to a final concentration of 1.6 M, which had no effect on the activity of the protein kinase under the experimental conditions.) It was also impossible to phosphorylate a synthetic peptide with the same sequence at a significant rate. This might indicate that the basic amino acid(s) needed close to the phosphorylation site are located in another part of the amino acid sequence.
It was concluded that the Km-values for phosphorylation of the three different phosphorylation sites were very similar, since all three sites were phosphorylated at the same rates in the intact enzyme. Fru-1,6-P2:ases incorporating between 0.30 and 2.60 mol (32P)phosphate per mol subunit were all found to contain 1/3 of the radioactivity bound to each serine.
It was shown that peptides containing the three different phosphorylatable serine residues could be sequentially cleaved off by first incubating the enzyme with chymotrypsin, which removed a peptide containing Ser-356, then with 4 M urea followed by a higher concentration of chymotrypsin, which removed a peptide containing Ser-341, and finally with trypsin, which digested a tripeptide containing Ser-338 (Figure 1). This stepwise progressive digestion method provided a tool for investigating at which site or sites Fru-l,6-P2:ase isolated from hepatocytes is phosphorylated in response to different hormones. As seen earlier (15, 16) phosphorylated Fru-1.6-P2:ase had a lower Km for Fru-1,6-P2 and was less readily inhibited by AMP and Fru-2,6-P2. Removal of Ser-356 did not affect any of these. parameters. When Ser-341 was removed both phosphorylated and unphosphorylated Fru-1,6-P2:ase were affected equally by both inhibitors, but the phosphorylated enzyme still had a lower Km for Fru-1,6-P2.
In vitro DhosDhorvlation of fructose-1.6-bis~hosuhatase from rabbit and pie liver with qvclic AMP-deuendent Drotein kinase, Fru-1,6-P2:ase was purified from rat, rabbit, pig, mouse and human liver using the method described earlier (16). Fru-1,6-P2:ases from all species were tested as substrates for CAMP-PK with (32P)ATP as the phosphate donor (19).After incubation they were subjected to polyacrylamide gel electrophoresis and the radioactivity was located by autoradiography. It was found that, in addition to rat liver Fru-1,6-P2:ase, the enzymes from pig and rabbit liver were phosphorylatable, both with up to 1 mol of phosphate per mol enzyme subunit. The incorporation of (32P)phosphate found in Fru-1,6-P2:ase from mouse and human liver was less than 0.05 mol per mol enzyme subunit.
Phosphorylation of pig and rabbit liver Fru-l,6-P2ase decreased the apparent Km for Fru-1,6-P2 but, in contrast to the rat liver enzyme, had no effect on the degree of inhibition by AMP and Fru-2,6-P2.
The subunit molecular weight (Mr) for rat liver Fru-1,6-P:ase was 41 kd. For all other species investigated it was 37 kd. It has earlier been reported that trypsin digestion of phosphorylated Fru-1,6-P2:ase removes peptides containing all of the phosphorylated serine residues from the C-terminal of the polypeptide chain (60). leaving a core protein with Mr 37 kd. No decrease in molecular size could be seen after trypsin digestion of Fru-l,6-P2:ase from any of the sources except rat liver. However, digestion of ( 3 'P)-labeled pig and rabbit liver Fru-l,6-P2:ase removed the protein-bound radioactivity quantitatively.
, These observations suggest that Fru-1,6-P2:ase from pig and rabbit liver is phosphorylated on at least one amino acid residue very close to the C-terminal end of the enzyme subunits. So it would seem that regulation of the activity of Fru-1,6-P2:ase by reversible phosphorylation is not a feature unique to the rat liver enzyme.

GENERAL DISCUSSION
Studies concerning the phosphorylation state of rat liver pyruvate kinase in vivQ used to give somewhat contradictory results. Pyruvate kinase in hepatocytes isolated from fed rats has been reported to be phosphorylated and inactive (4) while other studies showed that pyruvate kinase in whole livers from rats fed an enriched sucrose diet was inhibited and phosphorylated with 2-3 mol phosphate/mol tetrameric enzyme (22). The latter report is compatible with the results in (14) where it is seen that phosphorylated pyruvate kinase is the predominant form in livers from rats irrespective of diet.
These discrepancies were later explained by the observation that pyruvate kinase of the L-type is dephosphorylated during perfusion of the liver, as demonstrated by a change in the apparent Km and by increased phosphorylatability of the enzyme after perfusion (63).
Earlier studies have shown that phosphorylated pyruvate kinase is more prone to proteolytic degradation than the unphosphorylated enzyme in vitro (3). The detection of a form of pyruvate kinase in v i v a which has been degraded proteolytically but still has enzymatic activity may indicate that phosphorylation of pyruvate kinase not only regulates the activity but also initiates degradation of the enzyme.
It was recently shown that pyruvate kinase was phosphorylated in vitrQ with 1.7 mol phosphate/mol enzyme subunit by a Ca2+-calmodulin dependent protein kinase (69). This caused an increase in apparent Km which was greater than that brought about by phosphorylation with CAMP-PK. It was also shown that one site phosphorylated by C a2+-calmodulin activated protein kinase was identical to that phosphorylated by CAMP-PK. This report invites speculations about the intriguing nature of hormone interaction in the process of pyruvate kinase regulation, in particular the interaction between glucagon, which acts via CAMP-PK, and hormones such as epinephrine, which affect the Ca2+ flux to thus activate Ca2+-calmodulin activated protein kinase.
Phosphorylation of rat liver Fru-1,6-P2:ase, probably at more than one site, makes the enzyme less sensitive to inhibition by AMP and Fru-2,6-P2. This difference between phosphorylated and unphosphorylated Fru-1,6-P2:ase is enhanced since the two inhibitors act synergistically (e.g. 38), giving an overall activating effect on Fru-1,6-P2:ase by phosphorylation due both to a decreased apparent Km and the decrease in sensitivity to the inhibitors. The effect of glucagon on the activity of Fru-1,6-P2:ase in the cell is further amplified by the CAMP-PK-dependent phosphorylation of the enzyme responsible for Fru-2,6-P2 homeostasis, which is conducive to a decrease in concentration (21,62).
The identification of a third phosphorylation site for CAMP-PK in the C-terminal region of rat liver Fru-1,6-P2:ase immediately gives rise to speculations whether the enzyme is phosphorylated at different sites in response to different hormones (18). Speculations of this nature started when the second phosphorylation site was discovered (8, 65) but no conclusive evidence was given. Such an explanation might furnish an explanation of the data concerning Fru-1,6-P2 :ase form hepatocytes (17) where it is seen that both epinephrine and glucagon increase the activity and degree of phosphorylation of Fru-1,6-P2:ase but in different ways. This hypothesis is further supported by the observation that the increases in phosphorylation caused by epinephrine and glucagon are additive. The method of sequentially digesting peptides containing the phosphorylatable serine residues is a suitable tool for such studies.
One of the effects of insulin is to lower the degree of phosphorylation, accompanied by a decrease in activity of Fru-l,6-P2:ase. An explanation of this effect could be the activation of a phosphoprotein phosphatase which dephosphorylates Fru-1,6-P2:ase. The decrease in activity is not thought to be a consequence of the decrease in the level of CAMP (28).
Another question that might be raised by the occurrence of three different phosphorylation sites is whether phosphorylation of all the sites occurs in v i v a The simultaneous phosphorylation in vitro of three different sites could be an artefact due to dephosphorylation of the enzyme during the course of preparation, most probably before the heating step when most other cellular proteins are denatured.
It was observed that native Fru-1,6-P2:ases from pig and rabbit liver were phosphorylated by CAMP-PK in vitr2, with a concomittant change in apparent K, for Fru-1,6-P2 (19). This would mean that regulation of Fru-1,6-P2:ase activity by phosphorylation is not a feature unique for the rat liver enzyme. However in contrast to the rat liver enzyme, phosphorylation of Fru-1,6-P2:ase from pig and rabbit had no effect on the inhibition by AMP and Fru-2,6-P2.
The capacity of glucagon and related hormones is to stimulate CAMP-PK to phosphorylate among many other proteins pyruvate kinase, Fru-1,6-P2:ase and 6-phosphofructo-2-kinase/fructose-2,6-bisphosphatase. These enzymatic modifications lead to a decrease in the rate of glycolysis by inactivating pyruvate kinase and to an increase in the rate of gluconeogenesis by activating Fru-1,6-P2:ase and decreasing the amount of the inhibitor Fru-2,6-P2. The overall effect of glucagon and other hormones related to starvation is to induce gluconeogenesis and thus to maintain the glucose homeostasis.