A novel protein kinase B (PKB)/AKT-binding protein enhances PKB kinase activity and regulates DNA synthesis.

Protein kinase B (PKB)/Akt reportedly plays a role in the survival and/or proliferation of cells. We identified a novel protein, which binds to PKB, using a yeast two-hybrid screening system. This association was demonstrated not only in vivo by overexpressing both proteins or by coimmunoprecipitation of the endogenous proteins, but also in vitro using glutathione S-transferase fusion proteins. Importantly, this protein specifically associates with the C terminus of PKB but not with other AGC kinases and enhances PKB phosphorylation and kinase activation without growth factor stimulation. Thus, we termed this Akt-specific binding protein APE (Akt-phosphorylation enhancer). Since APE-induced phosphorylation of PKB did not occur in cells treated with wortmannin or LY294002, APE itself is not a kinase but seems to enhance or prolong the phosphoinositide 3-kinase-dependent phosphorylation of PKB. In cells in which APE was suppressed by small interfering RNA, DNA synthesis was significantly reduced with suppression of PKB phosphorylation, suggesting a synergistic role of APE in PKB-induced proliferation. On the other hand, in cells overexpressing both PKB and APE, despite markedly increased basal phosphorylation of PKB, both DNA rereplication and subsequent Chk2 phosphorylation and apoptosis were seen, suggesting the involvement of APE in the regulation of cell cycling replication licensing. Taking these observations together, APE appears to be a novel regulator of PKB phosphorylation. Furthermore, the interaction between APE and PKB, possibly dependent on the expression levels of both proteins, may be a novel molecular mechanism leading to proliferation and/or apoptosis.


Protein kinase B (PKB)/Akt reportedly plays a role in the survival and/or proliferation of cells.
We identified a novel protein, which binds to PKB, using a yeast twohybrid screening system. This association was demonstrated not only in vivo by overexpressing both proteins or by coimmunoprecipitation of the endogenous proteins, but also in vitro using glutathione S-transferase fusion proteins. Importantly, this protein specifically associates with the C terminus of PKB but not with other AGC kinases and enhances PKB phosphorylation and kinase activation without growth factor stimulation. Thus, we termed this Akt-specific binding protein APE (Akt-phosphorylation enhancer). Since APEinduced phosphorylation of PKB did not occur in cells treated with wortmannin or LY294002, APE itself is not a kinase but seems to enhance or prolong the phosphoinositide 3-kinase-dependent phosphorylation of PKB. In cells in which APE was suppressed by small interfering RNA, DNA synthesis was significantly reduced with suppression of PKB phosphorylation, suggesting a synergistic role of APE in PKB-induced proliferation. On the other hand, in cells overexpressing both PKB and APE, despite markedly increased basal phosphorylation of PKB, both DNA rereplication and subsequent Chk2 phosphorylation and apoptosis were seen, suggesting the involvement of APE in the regulation of cell cycling replication licensing. Taking these observations together, APE appears to be a novel regulator of PKB phosphorylation. Furthermore, the interaction between APE and PKB, possibly dependent on the expression levels of both proteins, may be a novel molecular mechanism leading to proliferation and/or apoptosis.
The serine/threonine protein kinase PKB 1 (also called Akt) is thought to be a key mediator of signal transduction. Upon growth factor stimulation, a family of lipid kinases known as class 1 phosphoinositide 3-kinases (PI 3-kinases) is recruited to the plasma membrane. PI 3-kinases phosphorylate phosphatidylinositol 4,5-bisphosphate at the D-3 position of the inositol ring, converting it to phosphatidylinositol 3,4,5-trisphosphate. Following the activation of PI 3-kinase, PKBs are recruited to the plasma membrane through direct contact of the pleckstrin homology (PH) domain with phosphatidylinositol 3,4,5trisphosphate and are phosphorylated at Thr 308 by PDK1 and at Ser 473 by PDK2, a kinase of which the molecular structure has not yet been identified (1,2). AGC kinases other than PKB are also known to be regulated by PI 3-kinase, and PKB acts downstream from PI 3-kinase to regulate numerous biological processes, such as proliferation, antiapoptosis, cell growth, and glucose metabolism (1,2).
PKB has a wide range of substrates, including GSK-3, FKHR (FoxO1), FKHR-L1 (FoxO3), AFX (FoxO4), and eNOS, all of which have the consensus motif RXRXX(S/T) (3,4). Protein kinases do not generally form stable complexes with their substrates, although PKB has been shown to exist in a stable complex with several of its substrates including MDM2, p21 Cip1 /WAF1, and TSC2 (5)(6)(7)(8). It was recently shown that several proteins interact with PKB as function modulators rather than as substrates. In a specific subset of T and B cells, TCL1 interacts with the PH domain of PKB and increases its kinase activity (9). Heat shock protein 90 (Hsp90) was shown to form complexes with Cdc37 and PKB, and PKB was stabilized and protected from dephosphorylation and degradation, resulting in increased kinase activity (10). Carboxyl-terminal modulator protein binds to the carboxyl terminus of PKB␣. Carboxyl-terminal modulator protein binding reportedly inhibits the phosphorylation and kinase activity of PKB, and stable expression of carboxylterminal modulator protein leads to phenotypic regression of a v-Akt transformed lung epithelial cell line to wild type (11). TRB3 was also identified as a negative modulator of the PKB type (12), although a contradictory report was very recently published (13). These results indicate that understanding PKB modulation is important for elucidating the mechanism of PKB activation and its regulation of cellular functions.
In this study, we identified a novel protein that interacts with PKB (in vivo and in vitro). Without growth factor stimulation, overexpression of this protein markedly enhances phosphorylation of Thr 308 and Ser 473 in PKB, leading to its kinase activation and phosphorylation of its downstream substrates such as GSK-3 and FKHR. In addition, suppression of APE using RNA interference significantly reduces PKB phosphorylation and PKB kinase activity. Therefore, we termed this protein APE (Akt-phosphorylation enhancer) and herein demonstrate the possible role of APE in DNA synthesis and apoptosis in cooperation with PKB.

MATERIALS AND METHODS
Yeast Two-hybrid System-The DupLEX-A two-hybrid system (Ori-Gene) was used for screening. We screened a mouse embryonic cDNA library with the pJG4-5 vector with a bait protein corresponding to the full length of mouse PKB␣ using a pEG202 vector and yeast strain EGY48. The positive clones were selected and assayed for ␤-galactosidase activity. Plasmid DNAs were isolated from positive clones and co-transformed with bait cDNA or negative control cDNA back into yeast to reconfirm the interaction. A yeast ␤-galactosidase assay kit (Pierce) was used to measure the protein-protein in vivo interaction according to the manufacturer's instructions.
Northern Blotting-Mouse Multiple Tissue Northern blot (Clontech) was used for Northern blotting. APE cDNA corresponding to the 600 bp of the coding region of the C-terminal and 400 bp of the untranslated region was used as a probe.
Antibody against APE-Fragments of the cDNA clone were subcloned into a glutathione S-transferase (GST) expression vector (Amersham Biosciences), expressed in BL21 and purified using glutathionecoupled Sepharose beads. Purified GST fusion proteins were injected into rabbits, and antisera were affinity-purified using the respective antigens. APE-C was generated to a fragment of APE encompassing amino acids 1646 -1845. APE-N1 corresponded to amino acids 172-372, and APE-N3 to corresponded to 501-601. Polyclonal antibodies to each antigen were affinity-purified, using each GST fusion protein, after removal of GST-specific antibody.
Gene Constructions and Expression System in Yeast and Mammalian Cells-Full-length APE cDNAs were cloned into the pShuttle vector to express these proteins with an adenovirus expression system (Clontech). The expression cassette was excited and subcloned into pAdeno-X vector (Clontech). Adenovirus was cloned, and large scale virus purification from 293T cell lysates was achieved by performing CsCl density gradient centrifugation twice, followed by overnight dialysis as previously described. An adenovirus expression vector for PKB␣ with a Myc tag at its C terminus had previously been generated (14). PKB␤, the PKB␣ PH domain (residues 1-106), the PKB␣ kinase domain (residues 138 -418), PKB␣ kinase and its hydrophobic domain (residues 148 -480), and the PKB␣ hydrophobic domain (residues 418 -480) were generated by PCR using PKB␣ or PKC␤ cDNA as described previously (14). SGK1 (residues 98 -431), SGK2 (residues 35-333), PKC␤2 (residues 342-673), and PKC⑀ (residues 408 -737) were generated by PCR using a mouse testis cDNA template. APE fragments were generated by PCR using full-length cDNAs as templates. The PCR products were cloned into pEG202 or pJG4-5 vectors.
Cell Cultures-HepG2, COS-7, and HeLa cells were from the RIKEN Cell Bank. Cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal bovine serum under a 5% CO 2 atmosphere at 37°C.
In Vivo Association of APE and PKB-HeLa cells were resuspended (4 ϫ 10 7 cells/ml) in buffer A (10 mM HEPES (pH 7.9), 10 mM KCl, 1.5 mM MgCl 2 , 0.34 M sucrose, 10% glycerol, 1 mM dithiothreitol, 5 g/ml aprotinin, 5 g/ml leupeptin, 0.5 g/ml pepstatin A, 0.1 mM phenylmethylsulfonyl fluoride). Mouse testis was homogenized with 50 strokes, using a Teflon/glass homogenizer, in 10 volumes of ice-cold buffer A. Triton X-100 (0.1%) was added, and the cells were incubated for 5 min on ice. The supernatant was clarified by high speed centrifugation (15 min, 20,000 ϫ g, 4°C) to remove nuclei, cell debris, and insoluble aggregates. Endogenous PKB and APE in the HeLa cell lysate or mouse testis lysates were immunoprecipitated with 100 g of immobilized anti-PKB and anti-APE-C antibodies, respectively. One hundred micrograms of immobilized Rabbit IgG were used as a control. The immunoprecipitates were washed four times in buffer A with 0.1% Triton X-100, eluted with elution buffer, electrophoresed, and transferred to nitrocellulose membranes. These filters were subjected to inmmunoblotting using the antibodies against APE and PKB.
In Vitro Association of APE and PKB-cDNAs encoding amino acids 418 -480 of mouse PKB␣ with the Myc tag and amino acids 1646 -1845 of mouse APE with the FLAG tag at their C termini were amplified by PCR. These cDNAs were cloned into pGEX-5X-1 and pET-28a vectors, and fusion protein expressions were induced in E. coli strain BL21 by the addition of 0.1 mM isopropyl ␤-D-thiogalactoside. The expressed proteins were purified using GST or His tag, according to the manufacturer's instructions. GST alone, GST-APE fragment fusion protein, and GST-PKB fragment fusion protein were incubated with glutathione-Sepharose 4B beads (Amersham Biosciences) for 3 h at 4°C followed by extensive washing in NETN buffer (20 mM Tris-HCl, pH 8.0, 100 mM NaCl, 1 mM EDTA, 0.5% Nonidet P-40). An aliquot containing 20 g of GST alone and GST-APE fragment fusion protein bound to beads were then incubated with the His tag PKB fragment. Similarly, an aliquot containing 20 g of GST alone and the GST-PKB fragment fusion protein bound to beads were then incubated with the His tag APE fragment. After a 4-h incubation at 4°C, the beads were washed five times with NETN buffer. The bound proteins were eluted by incubating the beads in SDS loading buffer containing 0.1 M dithiothreitol, electrophoresed, and then immunoblotted using anti-Myc and anti-FLAG antibodies for detection of the His tag PKB fragment and His tag APE fragment, respectively.
Immunoprecipitation and PKB Kinase Assay-Cells were lysed in Nonidet P-40 lysis buffer (50 mM Tris-HCl pH 7.5, 1% Nonidet P-40, 120 mM NaCl, 1 mM EDTA, 50 mM NaF, 40 mM ␤-glycerophosphate, 0.1 mM Na 3 VO 4 , 1 mM phenylmethylsulfonyl fluoride, 10 g/ml aprotinin, 50 g/ml leupeptin). For co-immunoprecipitation, we incubated protein lysates with primary antibodies overnight at 4°C followed by incubation with Protein A-Sepharose. Immunoprecipitates were washed three times with Nonidet P-40 lysis buffer. An Akt kinase assay kit (Cell Signaling Technology) was used to measure PKB kinase activity. PKB was immunoprecipitated with anti-Myc antibody and Protein G-Sepharose. PKB kinase activity was detected using GSK-3␤ recombinant protein as a substrate according to the manufacturer's instructions.
Gene Transduction and in Vivo Phosphorylation of PKB-Mammalian cell lines were infected with adenovirus the day after plating. Purified virus was added directly to the culture medium. Titers of adenovirus for protein overexpression were adjusted so that the expression levels of the Myc-tagged PKB␣ were similar, irrespective of APE co-expression. Likewise, APE expression levels were adjusted so as to be similar, irrespective of Myc-tagged PKB␣ co-expression. Experiments were performed 36 h later for in vivo phosphorylation. Cells were starved for 12 h with KRB-Hepes buffer (118.5 mM NaCl, 4.7 mM KCl, 2.5 mM CaCl 2 , 1.2 mM KH 2 PO 4 , 1.2 mM MgSO 4 , 24.9 mM NaHCO 3 , 30 mM HEPES, pH 7.4) containing 20 mg/ml bovine serum albumin. Cells were stimulated by the indicated stimulant in each experiment and, whenever indicated, 1 M wortmannin or 10 M LY294002 1 h prior to stimulation.
Gene Silencing by siRNA-Gene silencing was performed by an adenovirus-mediated siRNA method. For silencing of endogenous APE gene expression in HepG2 cells, a sense fragment (GGATCCGCATTA-ACACCCACCCGCTCTTCAAGAGAGAGCGGGTGGGTGTTAATGTT-TTTTCTAGAGAATTC) and an antisense fragment (GAATTCTCTAG-AAAAAACATTAACACCCACCCGCTCTCTCTTGAAGAGCGGGTGG-GTGTTAATGCGGATCC) were used for human APE. These two oligonucleotides were annealed in vitro, and the resultant doublestranded DNA fragments were subcloned into the BamHI-EcoRI site of a pSIREN-Shuttle vector. A negative control vector was supplied by Clontech. The expression cassette containing siRNA of APE or the negative control was excited and subcloned into pAdeno-X vector.
DNA Synthesis-HepG2 cells were maintained in Dulbecco's modified Eagle's medium supplemented with 10% fetal calf serum for 36 h and then transfected with an adenovirus encoding siRNA of APE or the negative control. Seventy-two hours after transfection, culture media were changed to Dulbecco's modified Eagle's medium supplemented with 0.2% bovine serum albumin, and cells were incubated for an additional 24 h. The cells were then incubated with BrdUrd labeling solution for 4 h. Incorporated BrdUrd was detected by cell proliferation ELISA, BrdUrd (colorimetric) (Roche Applied Science).
3-[4,5-Dimethylthiazol-2-yl]-2,5-diphenyltetrazolium Bromide (MTT) Assay-Cellular proliferation was measured by reduction of MTT, which corresponds to the living cell number and metabolic activity (15). Cells were plated at 5 ϫ 10 4 cells/well in 24-well plates and transfected with adenovirus. After incubation for the indicated time, MTT solution was added to each well. After 1 h of incubation, the reaction was stopped by adding 1 ml of isopropyl alcohol with 0.04 N HCl. The absorbance of each well was measured at 492 and 630 nm using a microplate reader.
Cell Viability Analysis-HepG2 cells and HeLa cells were infected with control GFP, PKB, APE, or both PKB and APE adenoviruses and incubated for the indicated times. Floating cells were recovered from culture medium by centrifugation at 1200 ϫ g for 1 min, and adherent cells were harvested by trypsinization. Both the floating and adherent cells were observed for morphologic changes with a light microscope at ϫ200 magnification. We combined the adherent and floating cells and measured their viability by using a trypan blue dye exclusion assay. Cell Cycle Analysis by Flow Cytometry-For cell cycle synchronization, cells were arrested at the G 1 -S phase transition separated by two subsequent thymidine blocks (2 mM thymidine) for 14 h, separated by a period of 10 h without thymidine. Both adherent and nonadherent cells were harvested by trypsinization, and an aliquot of 2 ϫ 10 6 cells was fixed in ice-cold ethanol for at least 1 h at 4°C. The cells were collected by centrifugation and resuspended in propidium iodine (10 g/ml) solution containing RNase for analysis of DNA content. Data were then collected on a BD Biosciences FACScan, 20,000 events/sample, using Cellquest software. DNA content analysis was performed with Verity ModFit software for the Macintosh computer.

Cloning of cDNAs Encoding the Protein Binding with PKB-
Using full-length mouse PKB as bait in a yeast two-hybrid screen of an embryonic mouse complementary DNA library, we isolated 31 clones displaying ␤-galactosidase activity. Sequencing analysis revealed 10 of the clones with the strongest ␤-galactosidase activity to be identical. In all cases, 606 bp of the coding region were followed by a 3Ј-untranslated region. Isolation of the full-length cDNA by phage screening of the mouse embryonic cDNA library and a series of 5Ј rapid amplifications of cDNA ends by PCR showed the largest open reading frame to be 5538 bp, which encode a 1845-amino acid protein with a predicted relative molecular mass (M r ) of 212,478 (Fig. 1A, accession number AB087827). By searching several data bases, we found that some mouse clones (BC037020, BC079895, AK129310) and this cDNA to be identical to a mouse homologue of the Kazusa DNA Research Institute clone KIAA1212. This clone is located on mouse chromosome 11 and on human chromosome 2. Although some cDNAs in the data base are presented as "full-length," it seems that they are not, judging from the size of the protein shown in this study. Our mouse cDNA is very likely to be full-length, and this protein was subsequently shown to enhance the phosphorylation of PKB, such that we designated the clone APE (Akt-phosphorylation enhancer). Protein analysis of APE revealed it to be a hydrophilic protein, and that its N terminus has a significant similarity with the putative coiled coil domain of the myosin heavy chain (Fig. 1B).
Tissue Distribution of APE-Northern blot analysis detected a 7.9-kb band of APE messenger RNA in the testis ( Fig. 2A). Longer exposure revealed moderate expression in the brain, and low expressions in the spleen and lungs ( Fig. 2A,   upper panel). Anti-APE antibodies were generated against the three different portions of APE (Fig. 1B). Immunoblotting with antibodies against APE, irrespective of the differences in the HeLa cells (B) or mouse testes (C) were homogenized and immunoprecipitated with the antibody indicated below the panels. APE specifically immunoprecipitated with anti-PKB antibody, and PKB specifically immunoprecipitated with APE antibody. D, interaction of APE and PKB was inhibited after insulin stimulation. HEK293 cells were serum starved for 24 h and treated with 100 nM insulin for 15 min. Endogenous PKB were immunoprecipitated with ␣-PKB monoclonal antibody or with control IgG, and the amount of APE associated with PKB was analyzed by Western blotting. epitopes of these antibodies (Fig. 1B), identified three bands of 220, 213, and 203 kDa in mouse tissues, whereas the control antibody against GST did not recognize any of these bands (Fig.  2B). The largest band of 220 kDa was observed in the lungs, testis, and fat. The 213-kDa band was detected in the brain, testis, heart, and fat. Finally, the smallest (203-kDa) band was detected in the lungs and spleen. These results were similarly obtained by immunoblotting of either immunoprecipitates of Extracts from E. coli BL21 cells expressing PKB␣ or APE with a pET system were used to test for APE or PKB␣ binding to the following bead matrices: GST beads coupled to either bacterially expressed GST or GST-PKB (amino acids 418 -480) or GST-APE (amino acids 1646 -1845). Extract bead complexes were washed three times to remove weakly bound protein prior to eluting off specifically bound proteins. The pulled down APE (FLAG-tagged) and PKB␣ (Myc-tagged) were resolved on an SDS-polyacrylamide gel and detected by ␣-FLAG or ␣-Myc antibody.
tissue lysates (Fig. 2B, upper panel) or nonimmunoprecipitated lysate (Fig. 2B, lower panel). These results suggest the existence of alternatively spliced protein products from the APE gene because a search of human expressed sequence tag databases indicated the existence of alternatively spliced forms of APE (data not shown).
In Vivo Association of PKB and APE-Next, to demonstrate in vivo association between APE and PKB, full-length APE and c-Myc-tagged PKB␣ were overexpressed in COS-7 cells. As shown in the upper panel of Fig. 3A, APE was detected in the immunoprecipitate by the anti-Myc antibody (Fig. 3A, upper  panel). Similarly, PKB was detected in the anti-APE immunoprecipitate (Fig. 3A, lower panel). This interaction between APE and PKB was demonstrated when both were overexpressed in Sf-9 insect cells or HepG2 cells (data not shown).
We also demonstrated an endogenous interaction between PKB and APE by coimmunoprecipitation of the endogenous proteins using specific antibodies in HeLa cells, and mouse testis. As shown in Fig. 3B, APE was coimmunoprecipitated by anti-PKB antibody in HeLa cells. PKB was also coimmu-noprecipitated by anti-APE-C antibody as shown in the lower panel. The PKB and APE interaction was reconfirmed by the same procedure using mouse testis homogenates (Fig. 3C), indicating that the PKB-APE interaction occurs under physiological conditions. APE Binds to Nonphosphorylated PKB More Efficiently than Phosphorylated PKB-The effect of PKB phosphorylation on the interaction between APE and PKB was assessed by measuring the amount of APE co-immunoprecipitated with PKB in the presence and absence of insulin stimulation (Fig. 3D). Insulin stimulation induced the phosphorylation of PKB on Thr 308 and Ser 473 . The amount of APE co-immunoprecipitated with PKB was revealed to be significantly lower in the insulinstimulated condition, compared with the unstimulated condition (Fig. 3D, bottom panel). This result suggests that APE has a higher affinity for nonphosphorylated than for phosphorylated PKB.
APE Binds to the C-terminal Portion of PKB but Not to Other AGC Kinases-To determine the region of PKB responsible for binding with APE, we generated four deletion mutants consist- Phosphorylations of Ser 9 of GSK-3␤ and Ser 256 of FKHR were detected by specific antibodies. D, APE enhanced PKB kinase activity without pervanadate stimulation. Myc-PKB␣ was transfected into COS-7 cells in the presence or absence of APE, as indicated. Cells were serum-starved and then stimulated with vehicle control or 100 M pervanadate for the indicated times. Myc-PKB was immunoprecipitated, and phosphorylation of Ser 473 of PKB was detected by a specific antibody. An in vitro kinase assay was performed using a GSK-3␤ fragment as the substrate for PKB kinase, and kinase activity was determined using the GSK-3␤ phospho-Ser 9 antibody. E, APE-induced basal phosphorylation of PKB␣ was inhibited by LY294002 and wortmannin. COS-7 cells were transfected with Myc-PKB␣, with or without APE. Cells were serum-starved for 12 h and then stimulated with 50 M epidermal growth factor for 15 min. The cells were incubated with 10 M LY294002 or 1 M wortmannin for 1 h prior to epidermal growth factor stimulation.
ing of a PH domain, kinase domain, kinase and hydrophobic domain, or the hydrophobic domain in the carboxyl terminus of PKB (16) (Fig. 4A). These mutants were subjected to baits in a yeast two hybrid screening with APE. It was revealed that the kinase domain with the hydrophobic motif or the hydrophobic motif alone binds with APE, whereas neither PH nor the kinase domain can bind with APE (Fig. 4B).
Subsequently, several deletion mutants of APE were produced to determine the portion responsible for the association with PKB. The C-terminal portion was shown to consist of 200 amino acids of APE, sufficient for the association with PKB. Since the deletion mutant amino acids 101-200 or 1-150 retain the ability to bind PKB, it is likely that the minimal portion necessary for the association with PKB is located within amino acid sequence 101-150 (Fig. 4C).
PKB belongs to a family of protein kinases, originally including protein kinase A, cGMP-dependent protein kinase and protein kinase C, termed the AGC family. Proteins in this family contain regions of high homology in their kinase domains (1). Since AGC kinases contain regions of high homology with the hydrophobic motif in PKB, we further examined whether APE interacted with AGC kinases other than PKB, using a yeast two-hybrid system. SGK1, SGK2, PKC␤2, PKC⑀, and PKB␤/Akt2 have a kinase domain and a hydrophobic motif highly homologous to those of PKB␣. As a result, PKB␤ and PKB␣ bind efficiently to APE in yeast (Fig. 4D). Conversely, very little interaction with APE was observed for SGK1, SGK2, PKC␤2, or PKC⑀. These results indicate that APE is not a common AGC kinase-binding protein but, rather, a PKB-specific binding protein.
In Vitro Association between Amino Acids 418 -480 of PKB and Amino Acids 1646 -1845 of APE-To examine whether the association of APE and PKB occurs in vitro, amino acids 418 -480 of PKB and 1646 -1845 of APE were expressed using E. coli and then purified. As shown in the left panel of Fig. 4E, GSTamino acids 418 -480 of mouse PKB␣ fusion protein bound to His-tagged amino acids 1646 -1845 of mouse APE protein, whereas GST alone did not. Similarly, GST-amino acids 1646 -1845 of mouse APE fusion protein, but not GST alone, bound to His-tagged amino acids 418 -480 of mouse PKB␣ (Fig. 4E, right  panel). These results indicate that the interaction between APE and PKB is direct.
APE Markedly Enhances Basal Phosphorylation of PKB-PKB␣ is activated via phosphorylation of Thr 308 in the activation loop of the kinase domain and of Ser 473 in the hydrophobic motif of the carboxyl terminus (17)(18)(19)(20)(21). To test the effect of APE binding on phosphorylation of PKB, basal phosphorylation of endogenous PKB␣ in COS-7 cells transfected with GFP adenovirus or APE adenovirus was analyzed. As shown in Fig. 5A, there was no significant phosphorylation on Thr 308 or Ser 473 of PKB␣ after 12-h serum starvation (left lane of Fig. 5A), but Thr 308 and Ser 473 of endogenous PKB were apparently phosphorylated in cells overexpressing APE. A similar result was obtained for HepG2 cells (data not shown).
Pretreatment with pervanadate increased PKB phosphorylation, time-dependently, as reported previously (11). To explore the effect of APE on enhanced PKB␣ phosphorylation, we treated COS-7 cells with adenoviruses expressing PKB␣ and various amounts of APE. APE overexpression increased PKB phosphorylation, in a titer-dependent manner, and the maximal phosphorylation of PKB obtained by APE overexpression was comparable with that achieved by long term pervanadate stimulation (Fig. 5B). These results suggest that APE overexpression can induce essentially maximal phosphorylation of PKB on Thr 308 and Ser 473 , which indicates that APE is an enhancer of PKB in vivo.
Phosphorylation of PKB by APE Induces the Phosphorylation of GSK-3␣/␤ and FKHR-PKB reportedly phosphorylates several downstream molecules such as GSK-3␣/␤ and FKHR (17,22). As a positive control, we confirmed that insulin stimulation induced PKB phsphorylation as well as downstream phosphorylation of Ser 256 of FKHR and Ser 9 of GSK-3␤ in HepG2 cells. Then we examined whether PKB phosphorylated by the overexpressed APE can induce the phosphorylations of GSK-3␤ and FKHR without growth factor stimulation. As shown in the right two lanes of Fig. 5C, overexpressed APE markedly enhanced phosphorylation of Ser 256 of FKHR and Ser 9 of GSK-3␤, to degrees similar to those seen with insulin stimulation.
FIG. 6. APE siRNA reduced insulin-stimulated endogenous PKB phosphorylation and kinase activity in HepG2 cells. HepG2 cells were transfected with adenovirus expressing negative control siRNA or APE siRNA for 60 h, starved for 12 h, and treated with 100 nM insulin for 15 min. A, endogenous PKB were immunoprecipitated with ␣-PKB monoclonal antibody, and PKB phosphorylation was analyzed by Western blotting using phosphospecific antibodies of Thr 308 or Ser 473 . PKB and APE expression levels were analyzed using cell extracts. B, endogenous PKB immunoprecipitated by ␣-PKB monoclonal antibody from control or APE-depleted HepG2 cells was also used for an in vitro kinase assay. In Vitro Kinase Activity of PKB Enhanced by APE-To test the influence of APE binding on PKB kinase activity, we assayed kinase activity in immune complexes from transfected COS-7 cells treated with pervanadate. Pervanadate-stimulated PKB activity was time-dependently increased when COS-7 cells were transfected with PKB alone, and kinase activity paralleled the phosphorylations of Thr 308 and Ser 473 . APE enhanced the basal phosphorylations of Thr 308 and Ser 473 , and in vitro kinase activity was also maximally enhanced and paralleled these phosphorylations. These results indicate that APE induces maximal basal phosphorylation of PKB, thereby maximally enhancing its kinase activity (Fig. 5D).
PI 3-Kinase Activity Is Needed for APE-induced PKB Phosphorylation-To examine whether the APE-induced increase in PKB phosphorylation is mediated only by PI 3-kinase, we examined the effects of the PI 3-kinase specific inhibitors LY294002 and wortmannin on APE-induced PKB phosphorylation. As shown in Fig. 5E, both epidermal growth factorinduced and APE-induced phosphorylation of PKB were completely inhibited by LY294002 and wortmannin treatments. These results indicate PI 3-kinase activity to be essential for APE-induced phosphorylation of PKB.
APE siRNA Inhibits Insulin-stimulated PKB Phosphorylation and Activation-To verify the role of endogenous APE in PKB phosphorylation, HepG2 cells were transfected with the negative control or small interfering RNA (siRNA) mediated by the adenoviral expression system. Suppression of endogenous APE by APE siRNA overexpression markedly reduced the APE protein level (Fig. 6A, upper panel). Under these conditions, endogenous PKB phosphorylation of both Thr 308 and Ser 473 in response to insulin was apparently reduced (Fig. 6A, third and  fourth panels). Consistent with the PKB phosphorylation results, insulin-induced PKB kinase activity measured by in vivo kinase assay was also reduced in APE-deficient cells (Fig. 6B).
Knockdown of APE Reduces DNA Synthesis-To explore the effect of APE depletion on proliferation, DNA synthesis in HepG2 cells were measured by BrdUrd incorporation. It was shown that suppressed expression of endogenous APE by siRNA led to decreased DNA synthesis in an APE siRNA titerdependent manner (Fig. 7).
Cell Death Induced by Overexpression of Both APE and PKB-Recent investigations have shown that overexpression of constitutively activated PKB mutants in many cell types promotes cellular proliferation and inhibits apoptosis (23)(24)(25). On the contrary, several lines of evidence indicate that downregulation of PI 3-kinase/PKB is required to execute the mitotic program efficiently (26). To explore the effect of prolonged PKB activation induced by APE, we next analyzed the effect of APE on cellular proliferation using COS-7 cells (Fig. 8A). The expression of GFP protein by adenovirus had no effect on COS-7 proliferation. COS-7 cells expressing PKB␣ proliferated slightly more slowly than the control GFP-expressing cells. However, COS-7 cells expressing both PKB␣ and APE showed essentially no proliferation. Trypan blue exclusion was employed to assay cell viability in COS-7 cells and HepG2 cells overexpressing PKB and APE. COS-7 cells expressing both PKB␣ and APE showed reduced viability (i.e. these cells ultimately died) (Fig. 8C). Virtually the same observations were made in HepG2 cells (Fig. 8D).
Induction of Apoptosis by Overexpression of Both APE and PKB-To elucidate whether apoptosis is involved in the molecular mechanism of APE-induced inhibition of cellular proliferation and cell death, we analyzed the cleavage of caspase-3 and PARP (Fig. 8B). Caspase-3 and PARP are key mediators of apoptosis, and cleavage of these enzymes to their active form correlates with the onset of apoptosis (27,28). When COS-7 cells were treated with 1 M staurosporine, cleaved caspase-3 and cleaved PARP were detectable after 3-6 h, in a time-dependent manner. In COS-7 cells expressing GFP, PKB␣, or APE, using an adenovirus expression system, no cleavage of caspase-3 or PARP was detectable. However, in cells expressing both PKB␣ and APE, cleaved caspase-3 and PARP were FIG. 9. Overexpressions of PKB␣ and APE result in cells with greater than 4n DNA content. A, APE and PKB, alone or in combination, were expressed with an adenovirus expression system, 18 h prior to thymidine release. Cells were collected at the indicated time points after thymidine release. As a control, cells expressing GFP were used. The DNA content was analyzed by fluorescence-activated cell sorting analysis. 2n (diploid) and 4n (tetraploid) represent the cells containing 2n and 4n DNA content, respectively. B, APE plus PKB␣ induces phosphorylation of Thr 68 of Chk2. HeLa cells arrested in G 1 /S by a double-thymidine method were released into a synchronous cell cycle and sampled every 3 h. Phosphorylation of Chk2 was determined by immunoblotting with a specific antibody for phospho-Thr 68 of Chk2. detected. These findings indicate that apoptosis is not induced by PKB alone but rather by the interaction between PKB and APE accompanying prolonged activation of PKB␣.
Effect of APE-PKB Interaction on the Cell Cycle-To study the effect of the APE-PKB interaction on cell cycle progression, cell cycle profiles were analyzed using flow cytometry (Fig. 9A). Overexpression of PKB or APE alone did not promote rereplication. However, co-expression of PKB and APE generated cells that had DNA contents greater than normal G 2 /M cells from 5 to 15 h after thymidine release, indicating that PKB and APE interact to induce DNA rereplication without mitosis.
APE-PKB Interaction Induces Chk2 Phosphorylation-Rereplication reportedly leads to DNA damage, and Vaziri et al. (29) demonstrated Chk2 phosphorylation in mammalian cells in which rereplication had been induced by overexpression of the replication license factors CDT1 and CDC6. The amount of Chk2 protein was not altered by PKB␣ or APE expression (data not shown). However, in HeLa cells overexpressing both PKB␣ and APE, Chk2 was apparently phosphorylated starting 6 h after thymidine block release, and peak phosphorylation was observed 12-21 h thereafter (Fig. 9B). In contrast, overexpression of neither PKB␣ nor APE induced apparent Chk2 phosphorylation.

DISCUSSION
In this study, we identified a novel PKB-binding protein using a yeast two-hybrid screening system and named it APE. The APE protein was detectable in many tissues including the brain, spleen, lung, fat, and heart, although APE mRNA was most abundant in the testis. In addition, from the immunoblotting results obtained using antibodies against different portions of APE, the presence of alternatively spliced protein products is likely.
The in vivo interaction of APE with PKB was clearly demonstrated by the overexpression of both proteins as well as by coimmunoprecipitation of the endogenous proteins. In vitro association was also demonstrated using bacterially expressed recombinant proteins. Notably, APE did not interact with any of the other AGC kinases tested in this study such as SGK1/2, PKC␤2, and PKC⑀, which have regions highly homologous to the hydrophobic motif of PKB. Thus, it is reasonable to consider APE a PKB-specific binding protein.
Subsequently, by overexpressing APE, we demonstrated that this protein markedly enhances the phosphorylation and kinase activity of PKB, whereas reducing endogenous APE expression using siRNA suppressed both. In addition, although APE binds to both phosphorylated and nonphosphorylated PKB, it seems that more PKB binds to APE when PKB is nonphosphorylated. Taking into consideration that APE-induced phosphorylation of PKB did not occur in cells treated with wortmannin or LY294002, APE itself is not a kinase and it is likely that APE enhances or prolongs the PI 3-kinase-dependent phosphorylation of PKB. In other words, we speculate that APE functions as a scaffold protein and facilitates Thr 308 and Ser 473 phosphorylation of PKB by PDK1/2. Alternatively, the APE⅐PKB complex may inhibit access of serine/threonine phosphatases such as protein phosphatase 2A.
Recent evidence indicates that PI 3-kinase and PKB play important roles in regulating cell proliferation. In this study, it was demonstrated that suppression of APE by siRNA reduced DNA synthesis, with decreased phosphorylation and kinase activity of PKB. This result agrees with those of previous reports showing the important role of PKB in proliferation. Thus, when the level of PKB expression is limited, APE apparently enhances proliferation in cooperation with PKB.
On the other hand, interestingly, we demonstrated overexpressions of both APE and PKB to induce DNA rereplication rather than normal DNA synthesis, thereby proving that these overexpressions together increase the cellular DNA content more than 4n in the S phase within 10 h after initiation of the S phase. Similar rereplication was reported with overexpression of the DNA replication factors Cdt1 and Cdc6 in either yeast or mammalian cells (29). In such cell systems, rereplication induced DNA damage, and the checkpoint pathway including Chk2 was activated. Chk2 activation is involved in the p53-dependent apoptotic response observed with DNA damage (30). In good agreement with these previous reports, we observed Chk2 phosphorylation and subsequent apoptosis in PKB and APE-expressing cells, after DNA rereplication. Thus, although the overexpression of both APE and PKB observed in this study may not be physiological, it is likely that the prolonged PKB phosphorylation induced by the association with APE does not lead to normal cell proliferation but rather to rereplication and the ensuing apoptosis. In other words, in the cells with high PKB expression, increased APE expression could lead to apoptosis after DNA rereplication.
Although we cannot explain how PKB and APE induce rereplication in human cells, we observed APE-induced PKB phosphorylation to be markedly enhanced not only in the cytoplasm but also the nucleus (data not shown). Thus, we speculate that overexpressed PKB and APE might phosphorylate some unidentified proteins in the nucleus such that normal replication licensing is blunted. However, this phenomenon was observed only with the overexpression system, and further study is needed to clarify whether this phenomenon is physiological.
In summary, we identified a novel PKB-binding protein, which enhances the phosphorylation of PKB and termed it APE. APE plays a role in regulating the phosphorylation state of PKB and the resultant DNA synthesis. In addition, DNA rereplication and the resultant apoptosis might be a novel mechanism that is induced by the enhanced interaction between APE and PKB, the physiological significance of which merits further investigation.