Heart 6-Phosphofructo-2-kinase Activation by Insulin Results from Ser-466 and Ser-483 Phosphorylation and Requires 3-Phosphoinositide-dependent Kinase-1, but Not Protein Kinase B*

Previous studies have shown that (i) the insulin-induced activation of heart 6-phosphofructo-2-kinase (PFK-2) is wortmannin-sensitive, but is insensitive to rapamycin, suggesting the involvement of phosphatidylinositol 3-kinase; and (ii) protein kinase B (PKB) activates PFK-2 in vitro by phosphorylating Ser-466 and Ser-483. In this work, we have studied the effects of phosphorylation of these residues on PFK-2 activity by replacing each or both residues with glutamate. Mutation of Ser-466 increased theV max of PFK-2, whereas mutation of Ser-483 decreased citrate inhibition. Mutation of both residues was required to decrease the K m for fructose 6-phosphate. We also studied the insulin-induced activation of heart PFK-2 in transfection experiments performed in human embryonic kidney 293 cells. Insulin activated transfected PFK-2 by phosphorylating Ser-466 and Ser-483. Kinase-dead (KD) PKB and KD 3-phosphoinositide-dependent kinase-1 (PDK-1) cotransfectants acted as dominant negatives because both prevented the insulin-induced activation of PKB as well as the inactivation of glycogen-synthase kinase-3, an established substrate of PKB. However, the insulin-induced activation of PFK-2 was prevented only by KD PDK-1, but not by KD PKB. These results indicate that the insulin-induced activation of heart PFK-2 is mediated by a PDK-1-activated protein kinase other than PKB.

sizes fructose 2,6-bisphosphate (1,2). Fructose 2,6-bisphosphate is itself a potent stimulator of 6-phosphofructo-1-kinase and hence of glycolysis. The insulin-induced stimulation of glucose transport and PFK-2 activation is mediated by phosphatidylinositol 3-kinase, a component of one of the insulin signaling pathways (3). Phosphatidylinositol 3,4,5-trisphosphate and phosphatidylinositol 3,4-bisphosphate, products of phosphatidylinositol 3-kinase, in turn, induce the phosphorylation of PKB, which leads to its activation. 3-Phosphoinositidedependent kinase-1 (PDK-1) phosphorylates PKB␣ on Thr-308, as well as on Ser-473, when complexed to a targeting subunit (4,5). PKB has been proposed to mediate one of the metabolic effects of insulin, namely the inactivation of glycogen synthase kinase-3 (GSK-3), leading to stimulation of glycogen synthesis in skeletal muscle (6). Heart PFK-2 was the second substrate of PKB to be recognized, and PKB activates heart PFK-2 in vitro by phosphorylating Ser-466 and Ser-483 (7). The phosphorylation site sequences surrounding Ser-466 and Ser-483 resemble those found in the other PKB targets, namely GSK-3 and BAD (4). The activation of heart PFK-2 due to phosphorylation by PKB resulted from a 2-fold increase in both V max and affinity for Fru-6-P, one of the substrates of PFK-2. One of the aims of this work was to study three heart PFK-2 mutants (S466E, S483E, and S466E/S483E) in which each or both serine residues in the PKB phosphorylation sites were replaced with Glu to mimic their phosphorylation. The relative contribution of each phosphorylation site to the changes in activity of PFK-2 was assessed by kinetic measurements.
In isolated cardiomyocytes, the insulin-induced activation of PFK-2 was insensitive to rapamycin and PD98059 (8), which inhibit p70 ribosomal S6 kinase and mitogen-activated protein kinase activation, respectively. However, the effect of insulin to activate PFK-2 was blocked by inhibitors of phosphatidylinositol 3-kinase (wortmannin and LY294002) (8). To test the hypothesis that PKB is required for the insulin-induced activation of heart PFK-2, we carried out transfection experiments in human embryonic kidney (HEK)-293 cells, which contain the components of the insulin signaling pathways and have been used in mechanistic studies (9 -11). HEK-293 cells were transfected with cDNA constructs coding for recombinant wild-type bovine heart PFK-2 (BH1) and mutants. In these cells, we studied the effect of insulin on the extent of phosphorylation and changes in PFK-2 activity, and we identified the phosphorylation sites in the wild-type and the three mutants. Finally, cotransfection of heart PFK-2 together with different PKB and PDK-1 constructs (wild-type or dominant negatives) was performed to evaluate the role of PKB in the insulin-induced activation of heart PFK-2. Our findings suggest the possible participation of PDK-1-activated protein kinase(s), other than PKB, in the insulin signaling pathway that activates PFK-2.

EXPERIMENTAL PROCEDURES
Materials-C-terminally polyhistidine-tagged bovine heart PFK-2/ fructose-2,6-bisphosphatase (BH1(His) 6 ) cDNA, cloned in pBluescript II KS ϩ phagemid (7), was used to create two single mutations (S466E and S483E) by polymerase chain reaction with the mutant oligonucleotides 5Ј-GATGAGAAGGAACGAATTCACGCCTCTG-3Ј (for S466E) and 5Ј-CGAATACAATCAGACGTCCAAGAAATTACGAAGTTGGGAGC-3Ј (for S483E). The mutated polymerase chain reaction fragments were introduced into the original phagemid to create pBluescript II KS ϩ / BH1(His) 6 /S466E and pBluescript II KS ϩ /BH1(His) 6 /S483E. One unique restriction site (BstBI) is located between the two mutation sites, and the other (BamHI) is located at the end of the cDNA. The BstBI/BamHI fragment from pBluescript II KS ϩ /BH1(His) 6 /S483E, containing the S483E mutation, was introduced into pBluescript II KS ϩ /BH1(His) 6 /S466E to create the double mutant pBluescript II KS ϩ / BH1(His) 6 /S466E/S483E. The different mutations were verified by sequencing and introduced into the bacterial expression vector pET3a as described (7). The wild-type and mutant BH1(His) 6 cDNAs were also introduced into the pCMV5 eukaryotic expression vector (12) as follows. An optimal Kozak consensus sequence (TCCACC) was first added at the beginning of the BH1(His) 6 cDNA by polymerase chain reaction using pBluescript II KS ϩ /BH1(His) 6 as a template (the oligonucleotides were 5Ј-CCATCGATTCCACCATGTCCGGGAATCCTGCC-3Ј and 5Ј-CG-GAATTCTCCTCGGCAAAG-3Ј). The polymerase chain reaction product was finally reintroduced into the same phagemid using the ClaI and EcoRI unique restriction sites to form a complete BH1(His) 6 cDNA including the Kozak consensus sequence. The ClaI/BamHI fragment, containing this complete BH1(His) 6 cDNA, was then introduced into the pCMV5 vector using the same sites to form the pCMV5/BH1(His) 6 vector. The different mutations were introduced into this vector at the StuI and BamHI unique restriction sites.
Activated PKB, used for the in vitro phosphorylation experiments, was prepared as follows. N-terminally His-tagged PKB␣ (provided by A. Paterson, Dundee University) was activated in vitro by incubation with PDK-1, MgATP, and phosphatidylinositol 3,4,5-trisphosphate as described previously for His-tagged PKB␤ (13). PDK-1 was then removed from His-tagged PKB␣ by chromatography on heparin-Sepharose. pCMV5 vectors containing the cDNA coding human GSK-3␤ with the EFMPME epitope tag at the N terminus (EE-GSK-3) and hemagglutinin-tagged PKB (HA-PKB) have been described (14). pCMV5 vectors containing the cDNA coding the triple mutant kinase-dead (KD) PKB (called AAA-PKB, in which Lys-179 involved in ATP binding and Thr-308 and Ser-473 required for phosphorylation-induced activation have been mutated to Ala) and the double mutant KD PDK-1 (in which Lys-111 and Asp-223 involved in catalysis have been mutated to Ala) were obtained using the Quickchange kit (Stratagene) following instructions provided by the manufacturer. Anti-His antibody (CLON-TECH), anti-glutathione S-transferase (GST) antibody (Pharmingen), and anti-HA and anti-c-Myc antibodies (Roche Molecular Biochemicals) were from the indicated sources. A rabbit polyclonal antibody (which we call BAK) was raised against the C-terminal peptide of PKB (FPQF-SYSASSTA) and purified on protein A-Sepharose. This antibody was used for endogenous PKB assay. All other materials were from sources previously cited (7,(13)(14)(15).
In Vitro Phosphorylation of Recombinant Heart PFK-2 by PKB-Wild-type BH1(His) 6 and the S466E and S466E/S483E mutants were expressed in Escherichia coli strain BL21(DE3) pLysE, and the S483E mutant was expressed in BL21(DE3) pLysS. Culture, lysis, and purification were carried out as described (15). For measurement of the changes in kinetic properties induced by phosphorylation, the wild-type and mutant BH1(His) 6 preparations were incubated with PKB␣ (see table legends for details) in phosphorylation buffer containing 25 mM Tris-HCl (pH 7.5), 1 mM MgATP, 0.05% (v/v) ␤-mercaptoethanol, 5 mM magnesium acetate, 1 M microcystin, 1 M cAMP-dependent protein kinase inhibitor peptide, and 50 M EGTA. After 30 min, the reactions  6 preparations by PKB For measurements of the stoichiometry of 32 P incorporation, wildtype and mutant BH1(His) 6 preparations (all at 0.1 mg/ml) were phosphorylated with [␥-32 P]MgATP (0.1 mM) and PKB (100 milliunits/ml) in a final volume of 25 l until a plateau was reached (80 min). 32 P incorporation was measured by phosphoimaging after SDS polyacrylamide gel electrophoresis. To study the kinetic properties of phosphorylation, wild-type and mutant BH1(His) 6 preparations (10 nM to 3 M) were phosphorylated with MgATP (1 mM) and PKB (5 milliunits/ml) in a final volume of 40 l for 5 min. Following SDS polyacrylamide gel electrophoresis 32 P incorporation was also measured by phosphoimaging. The results are the means Ϯ S.E. for the number of determinations shown in parentheses. BH1(His) 6 6 preparations induced by mutation or phosphorylation by PKB BH1(His) 6 preparations (0.1 mg/ml) were incubated with 500 milliunits/ml PKB at 30°C. After 30 min, the reactions were stopped, and aliquots were taken for the measurement of PFK-2 activity. For the determination of K m values, the concentration of the substrate understudy was varied up to 10 times the K m , whereas the concentration of the other substrate was saturating (5 mM for MgATP and 2 mM for Fru-6-P). For the IC 50 for magnesium citrate, the concentrations of Fru-6-P and MgATP were 100 M and 5 mM, respectively. were stopped by diluting 10-fold in 20 mM Hepes (pH 7.5), 50 mM KCl, 0.5 mM EGTA, 5 mM EDTA, 1 mM potassium phosphate, 20% (v/v) glycerol, and 0.1% (v/v) ␤-mercaptoethanol and chilled in ice. Aliquots were taken for PFK-2 activity. Kinetic measurements were made when phosphorylation had reached a maximum. In vitro measurements of 32 P incorporation into purified PFK-2 were performed as described (7). Transfection of HEK-293 Cells-HEK-293 cells were cultured in 10-cm diameter dishes in Dulbecco's minimal essential medium containing 10% (v/v) fetal calf serum and transfected using a modified calcium phosphate procedure (11) with pCMV5 DNA constructs (wildtype or mutant BH1(His) 6 , EE-GSK-3, wild-type HA-PKB, GST-AAA-PKB, and mutant PDK-1) and empty vector to reach a total of 10 g of added DNA. In cotransfection experiments, the amount of cotransfected DNA was in ϳ5-10-fold excess with respect to the transfected vector. Transfection was performed overnight at 37°C, and then the medium was aspirated and replaced with fresh Dulbecco's minimal essential medium containing 10% (v/v) fetal calf serum. After 10 h, the cells were  6 from 32 P-labeled HEK-293 cells, which were incubated with rapamycin and PD98059 and with or without insulin (10 min) and wortmannin as indicated, was immunoprecipitated and subjected to SDS-polyacrylamide gel electrophoresis for autoradiography (11).

FIG. 3. HPLC profiles of 32 P-labeled tryptic peptides of BH1(His) 6 from transfected HEK-293 cells (A-C) and of BH1(His) 6 phosphorylated in vitro by PKB (D).
Immunoprecipitated 32 P-labeled BH1(His) 6 was digested with trypsin. Peptides were purified by reverse-phase HPLC using a linear gradient of acetonitrile (dashed line). Tryptic peptide maps (A-C) are the means of two separate experiments. In all cases, HEK-293 cells were preincubated with rapamycin and PD98059. In vitro phosphorylation of purified recombinant BH1(His) 6 by PKB (D) had reached a plateau of 32 P incorporation before digestion with trypsin. In all experiments, Ͼ95% of the radioactivity was eluted between 10 and 30% of acetonitrile. deprived of serum in Dulbecco's minimal essential medium for 16 h. The cells were then treated with or without wortmannin (100 nM, 10 min), rapamycin (100 nM, 30 min) and PD98059 (50 M, 30 min) and further incubated with or without insulin (100 nM) for the indicated periods of time.
Measurements of PFK-2 and PKB Activities in Cell Extracts-After removing the incubation medium, the HEK-293 cells were frozen in liquid nitrogen. The cells were then lysed in 1 ml of ice-cold Buffer A (50 mM Hepes (pH 7.5), 50 mM KF, 1 mM potassium phosphate, 1 mM EDTA, 1 mM EGTA, 1 M microcystin, 1 mM Na 3 VO 4 , 0.1% (v/v) ␤-mercaptoethanol, 1 mM phenylmethanesulfonyl fluoride, 1 mM benzamidine hydrochloride, and 1 g/ml leupeptin) by passage through a syringe fitted with a fine needle. PFK-2 activity was measured in supernatants (15,000 ϫ g for 5 min) as described (16). In the experiments reported in this paper, transfected HEK-293 cells contained between 20 and 100 times more PFK-2 activity (0.1-0.5 milliunits/mg of protein) than the non-transfected control cells (5 microunits/mg of protein), depending on the amount of transfected PFK-2 DNA (1-10 g) and on transfection efficiency (30 -80% of cells transfected after 40 h). Endogenous PKB activity was measured in immunoprecipitates as follows. Cell extracts (0.3 ml), corresponding to ϳ3 mg of protein, were incubated for 2 h at 4°C with agitation in a final volume of 0.5 ml of Buffer A and 60 g of PKB antibody (BAK) coupled to 25 l of protein A-Sepharose. The immune complexes were washed three times with buffer containing 50 mM Tris-HCl (pH 7.5), 1 mM EDTA, 1 mM EGTA, 50 mM NaF, 10 mM sodium ␤-glycerophosphate, 5 mM sodium pyrophosphate, 0.1% (v/v) ␤-mercaptoethanol, 0.1% (w/v) Triton X-100, and 0.5 M NaCl and twice with buffer containing 50 mM Tris-HCl (pH 7.5), 0.03% (w/v) Brij-35, 0.1 mM EGTA, and 0.1% (v/v) ␤-mercaptoethanol. The beads were then resuspended in 20 l of PKB assay buffer (10 mM Mops (pH 7), 0.5 mM EDTA, 10 mM magnesium acetate, and 0.1% (v/v) ␤-mercaptoethanol). PKB activity was measured at 30°C in a final volume of 50 l of PKB assay buffer in the presence of 2.5 M cAMP-dependent protein kinase inhibitor peptide with 0.25 mM substrate peptide (RPRAATF (17)) and 0.1 mM [␥-32 P]MgATP (specific radioactivity of 1500 cpm/pmol). After 20 min, 20-l aliquots were removed for the measurement of 32 P incorporation (18). Transfected HA-PKB activity was measured by the same method on immune complexes obtained from 100 g of protein extract treated with 3 g of anti-HA antibody.
Phosphorylation Site Identification-HEK-293 cells were transfected with 10 g of plasmid containing wild-type or mutant BH1(His) 6 cDNAs. The cells were washed with phosphate-free Dulbecco's minimal essential medium and incubated for 4 h with [ 32 P]orthophosphate (1 mCi/ml). The cells were pretreated with rapamycin (100 nM) and PD98059 (50 M) for 30 min with or without wortmannin (100 nM) for 10 min and then incubated with or without insulin (100 nM) for 10 min. The cells were lysed (11), and PFK-2 was immunoprecipitated using 160 g of anti-His monoclonal antibody/10-cm dish of cells. Cell extracts were incubated for 30 min at 4°C on a shaker with antibody coupled to 50 l of protein G-Sepharose. The immune complexes were washed six times with lysis buffer (11)  was hydrolyzed and subjected to thinlayer cellulose electrophoresis prior to autoradiography (11). This resolved orthophosphate (Pi), phosphoserine (pS), phosphothreonine (pT), and phosphotyrosine (pY). B, each radioactive peak (500 -1500 cpm) from Fig. 3B was also analyzed by burst sequencing and MALDI-MS to identify the phosphorylated residues (11). The numbers in parentheses correspond to the peptide masses measured by MALDI-MS and calculated from the sequence, respectively. The results are representative of two experiments. (11). Following SDS-polyacrylamide gel electrophoresis, the PFK-2 band was excised and digested with trypsin (11). Peptides were separated by reverse-phase chromatography on a C 18 column (0.46 ϫ 25 cm, Vydac) in an acetonitrile gradient, and the radioactive peaks were identified by Cerenkov counting (11).

4-vinylpyridine
Other Methods-Identification of phosphorylation sites by phosphoamino acid analysis, solid-phase sequencing, and matrix-assisted laser desorption-ionization mass spectrometry (MALDI-MS) were carried out as described (7,11). Proteins were measured (19) using ␥-globulin as a standard or by the reaction with ninhydrin, after total alkaline hydrolysis (20), using bovine serum albumin as a standard. PFK-2 was assayed (16) under the conditions described in the figure and table legends. GSK-3 activity ratio (with and without protein phosphatase 2A treatment) was measured in immunoprecipitates as described (14). Kinetic constants were calculated by computer fitting of the data to a hyperbola describing the Michaelis-Menten equation by nonlinear least-squares regression. One unit of enzyme activity corresponds to the formation of 1 mol (PFK-2) or 1 nmol (protein kinases) of product/min under the assay conditions.

In Vitro Studies
The contribution of the C-terminal residues (Ser-466 and Ser-483) to the changes in the kinetic properties of PFK-2 induced by phosphorylation was studied by site-directed mu-tagenesis. Wild-type BH1(His) 6 and the three mutants (S466E, S483E, and S466E/S483E) were expressed in bacteria and purified by a two-step procedure including anion-exchange and metal affinity chromatography (7). The mutant preparations had the same chromatographic behavior as the wild-type. After SDS-polyacrylamide gel electrophoresis and Coomassie Blue staining, a single band was observed for each preparation, which migrated with the expected M r of 61,000 (data not shown).
The stoichiometry of phosphorylation of recombinant wildtype BH1(His) 6 by PKB (Table I) was close to 1 mol of phosphate incorporated per mol of enzyme subunit (measured by the ninhydrin method), in agreement with previous work (7). PKB phosphorylated S466E and S483E with a stoichiometry corresponding to about half that of the wild-type. The double mutant (S466E/S483E) was not phosphorylated by PKB. Moreover, the V max of PKB for wild-type PFK-2 was twice that of the S466E and S483E mutants, whereas the affinity of PKB for all the preparations was the same ( Table I). The difference in the rate of phosphorylation could be due to the incorporation of a negative charge, which could then act as a product inhibitor in the PKB reaction.
In agreement with previous results (7), phosphorylation of recombinant wild-type BH1(His) 6 by PKB increased the V max (2-fold) and decreased the K m of PFK-2 for Fru-6-P (2-fold) ( Table II). In addition, phosphorylation of the wild type by PKB decreased the sensitivity toward magnesium citrate inhibition (21), without affecting the K m for ATP. The double mutation (S466E/S483E) had the same effects on the kinetic properties as those observed in the phosphorylated wild-type, suggesting that the double mutation indeed mimicked the effects of phosphorylation on activity. As expected, incubation of the S466E/ S483E mutant with PKB had no further effect on its kinetic properties. In summary, our results show that introduction of a negative charge (by phosphorylation or mutation) at Ser-466 or Ser-483 doubled the V max or decreased the sensitivity toward magnesium citrate inhibition, respectively, whereas a negative charge at both Ser-466 and Ser-483 was required to decrease the K m for Fru-6-P (Table II).

Studies in HEK-293 Cells
Activation and Phosphorylation of Transfected Heart PFK-2 by Insulin in HEK-293 Cells-Incubation of HEK-293 cells with insulin for 5-20 min had no effect on the activity of endogenous PFK-2 in non-transfected cells (data not shown), whereas it doubled the activity of transfected PFK-2 (wild-type BH1(His) 6 ) (Fig. 1A). The effect of insulin to double PFK-2 activity was consistently observed when the activity of transfected PFK-2 ranged between 100 and 500 microunits/mg of protein in unstimulated cells (Fig. 1A, inset). The insulin-induced PFK-2 activation was sensitive to LY294002 (data not shown) and wortmannin and was insensitive to rapamycin and PD98059 (Fig. 1A). Endogenous PKB activation correlated with the insulin-induced increase in PFK-2 activity (Fig. 1B).
To identify the phosphorylated residue(s) in transfected PFK-2 following insulin stimulation, HEK-293 cells expressing wild-type heart PFK-2 were 32 P-labeled and incubated with insulin for 10 min in the presence or absence of wortmannin. Rapamycin and PD98059 were present under all conditions. Immunoprecipitation of transfected PFK-2 and autoradiography after SDS-polyacrylamide gel electrophoresis revealed a single 32 P-labeled band with an expected M r of 61,000. The labeling of this band was increased 2-fold by insulin and blocked by wortmannin (Fig. 2). The 32 P-labeled bands were excised and digested with trypsin, and peptides were separated by reverse-phase HPLC. For wild-type PFK-2 from unstimu- lated cells, two major radioactive peaks (peaks 1 and 2) and four minor radioactive peaks (peaks 3-6) were observed (Fig.  3A). Insulin increased only the labeling of peaks 3 and 6, and the insulin-induced labeling of these peaks was blocked by wortmannin (Fig. 3, B and C).
Phosphoamino acid analysis showed that all the peptides were phosphorylated only on serine residues (Fig. 4A). To identify the phosphorylation sites in the 32 P-labeled peptides, each peak was analyzed by MALDI-MS and solid-phase sequencing (Fig. 4B). Five of the six 32 P-labeled peptides were identified. The two insulin-sensitive peaks (peaks 3 and 6) were those that were also labeled in vitro when recombinant BH1(His) 6 was phosphorylated with [␥-32 P]MgATP and PKB (Fig. 3D). These peaks contain phosphorylated Ser-466 (peak 3) and phosphorylated Ser-483 (peak 6) (Fig. 4), in agreement with previous work (7). Among the peaks insensitive to insulin, peak 1 corresponds to the C-terminal peptide phosphorylated on both Ser-522 and Ser-528; peak 2 would correspond to the same peptide phosphorylated on either Ser-522 or Ser-528; and peak 5 contains phosphorylated Ser-493. These proline-rich peptides could correspond to phosphorylation sites for the mitogen-activated protein kinase family. We were unable to identify the phosphorylation sites in peak 4, whose labeling was unaffected by insulin.
The three PFK-2 mutants (S466E, S483E, and S466E/ S483E) were transfected in HEK-293 cells, which were then stimulated with insulin to study their phosphorylation (Fig. 5). As expected, peaks 3 (Ser-466) and 6 (Ser-483) were absent in the S466E/S483E mutant (Fig. 5C). In this double mutant, no other radioactive peaks were observed after insulin treatment, indicating that the insulin-induced activation of PFK-2 is solely mediated by the phosphorylation of Ser-466 and Ser-483. In the single mutants (S466E and S483E), the phosphorylation of the unmodified Ser residue was either decreased or abolished (compare Fig. 5 with Fig. 3). This difference in behavior, which contrasts with the rate of phosphorylation of these mutants by PKB in vitro, could be explained by the fact that a protein kinase other than PKB is involved in PFK-2 activation by insulin.

Effects of Cotransfected PKB or PDK-1 on PFK-2 Activation by Insulin in HEK-293 Cells-
The potential role of PKB and PDK-1 in the insulin-induced activation of PFK-2 was studied in HEK-293 cells that had been cotransfected with wild-type BH1(His) 6 PFK-2 and wild-type HA-PKB, dominant-negative GST-AAA-PKB, or dominant-negative KD Myc-PDK-1. Under all conditions, rapamycin and PD98059 were present. The effect of insulin was tested in experiments that lasted only 10 min and were therefore too short to affect the level of expression of PFK-2. This lack of change in enzyme content and expression was verified by immunoblotting ( Figs. 6 and 7).
We first verified that GST-AAA-PKB and KD PDK-1 indeed behave as dominant negatives by measuring the effect of insulin on the activity of cotransfected PKB or GSK-3, two downstream components in insulin signaling (14). When wild-type HA-PKB was used as a reporter (Fig. 6A), overexpression of either GST-AAA-PKB or KD Myc-PDK-1 prevented the insulininduced activation of HA-PKB. Moreover, and in agreement with previous reports (14), insulin inactivated EE-GSK-3 as indicated by the decrease in its activity ratio (Fig. 6B). This ratio is a measure of the activation state of GSK-3 because it relates its activity to that obtained after full activation by treatment with protein phosphatase 2A. This index has been used to assess the extent of inactivation of GSK-3 by insulin or insulin-like growth factor-1 (14). The overexpression of wildtype HA-PKB inactivated EE-GSK-3, and as reported (14), this inactivation was even more pronounced than with insulin. Overexpression of GST-AAA-PKB or KD Myc-PDK-1 prevented the insulin-induced inactivation of EE-GSK-3 (Fig. 6B). Therefore, GST-AAA-PKB and KD Myc-PDK-1 acted as dominantnegative mutants, as described (22)(23)(24).
We then tested whether GST-AAA-PKB and KD Myc-PDK-1 could also prevent the insulin-induced activation of transfected PFK-2. As for GSK-3, we determined the activity ratio as an index of PFK-2 activation. Insulin activated endogenous PKB and transfected PFK-2 (Fig. 7), in agreement with the results in Fig. 1. Cotransfection of wild-type PKB resulted in little, if any, effect on PFK-2 activity ratio in control or insulin-treated cells. As expected, the activation of endogenous PKB was di- In conclusion, insulin activates heart PFK-2 by inducing the phosphorylation of Ser-466 and Ser-483 in intact cells. Although these sites are known to be phosphorylated by PKB in vitro, our results indicate that, in intact cells, PFK-2 activation by insulin is mediated by PDK-1 and that PKB may not be essential. Therefore, other protein kinase(s), downstream of PDK-1, might mediate PFK-2 activation by phosphorylation. Examples of insulin-sensitive protein kinases, whose activation is dependent upon phosphatidylinositol 3-kinase and PDK-1, include protein kinase C and serum-and glucocorticoid-regulated protein kinase (25,26). Future work will focus on whether these kinases or a novel protein kinase mediates the activation of PFK-2 by insulin in intact cells.