Portal de Periódicos da CAPES

Primary Structure, Tissue Distribution, and Expression of Mouse Phosphoinositide-dependent Protein Kinase-1, a Protein Kinase That Phosphorylates and Activates Protein Kinase Cζ

1999; Elsevier BV; Volume: 274; Issue: 12 Linguagem: Inglês

10.1074/jbc.274.12.8117

1083-351X

Lily Dong, Ruo-bo Zhang, Paul R. Langlais, Huili He, Matthew Clark, Li Zhu, Feng Liu,

Mast cells and histamine

Phosphoinositide-dependent protein kinase-1 (PDK1) is a recently identified serine/threonine kinase that phosphorylates and activates Akt and p70S6K, two downstream kinases of phosphatidylinositol 3-kinase. To further study the potential role of PDK1, we have screened a mouse liver cDNA library and identified a cDNA encoding the enzyme. The predicted mouse PDK1 (mPDK1) protein contained 559 amino acids and a COOH-terminal pleckstrin homology domain. A 7-kilobase mPDK1 mRNA was broadly expressed in mouse tissues and in embryonic cells. In the testis, a high level expression of a tissue-specific 2-kilobase transcript was also detected. Anti-mPDK1 antibody recognized multiple proteins in mouse tissues with molecular masses ranging from 60 to 180 kDa. mPDK1 phosphorylated the conserved threonine residue (Thr402) in the activation loop of protein kinase C-ζ and activated the enzyme in vitro and in cells. Our findings suggest that there may be different isoforms of mPDK1 and that the protein is an upstream kinase that activates divergent pathways downstream of phosphatidylinositol 3-kinase. Phosphoinositide-dependent protein kinase-1 (PDK1) is a recently identified serine/threonine kinase that phosphorylates and activates Akt and p70S6K, two downstream kinases of phosphatidylinositol 3-kinase. To further study the potential role of PDK1, we have screened a mouse liver cDNA library and identified a cDNA encoding the enzyme. The predicted mouse PDK1 (mPDK1) protein contained 559 amino acids and a COOH-terminal pleckstrin homology domain. A 7-kilobase mPDK1 mRNA was broadly expressed in mouse tissues and in embryonic cells. In the testis, a high level expression of a tissue-specific 2-kilobase transcript was also detected. Anti-mPDK1 antibody recognized multiple proteins in mouse tissues with molecular masses ranging from 60 to 180 kDa. mPDK1 phosphorylated the conserved threonine residue (Thr402) in the activation loop of protein kinase C-ζ and activated the enzyme in vitro and in cells. Our findings suggest that there may be different isoforms of mPDK1 and that the protein is an upstream kinase that activates divergent pathways downstream of phosphatidylinositol 3-kinase. phosphatidylinositol 3-kinase Chinese hamster ovary expressed sequence tag glutathione S-transferase insulin receptor polyacrylamide gel electrophoresis protein kinase C protein kinase A phosphoinositide-dependent protein kinase-1 human PDK1 mouse PDK1 hemagglutinin kilobase(s) Activation of Akt (also called protein kinase B or PKB) by growth factors has been shown to be necessary for various cellular processes including cell growth, differentiation, metabolism, and apoptosis (1Bos J.L. Trends Biochem. Sci. 1995; 20: 441-442Abstract Full Text PDF PubMed Scopus (74) Google Scholar, 2Datta S.R. Dudek H. Tao X. Masters S. Fu H. Gotoh Y. Greenberg M.E. Cell. 1997; 91: 231-241Abstract Full Text Full Text PDF PubMed Scopus (4985) Google Scholar, 3Franke T.F. Kaplan D.R. Cantley L.C. Cell. 1997; 88: 435-437Abstract Full Text Full Text PDF PubMed Scopus (1529) Google Scholar, 4Magun R. Burgering B.M. Coffer P.J. Pardasani D. Lin Y. Chabot J. Sorisky A. Endocrinology. 1996; 137: 3590-3593Crossref PubMed Scopus (136) Google Scholar, 5Summers S.A. Birnbaum M.J. Biochem. Soc. Trans. 1997; 25: 981-988Crossref PubMed Scopus (59) Google Scholar, 6Ueki K. Yamamoto-Honda R. Kaburagi Y. Yamauchi T. Tobe K. Burgering B.M. Th. Coffer P.J. Komuro I. Akanuma Y. Yazaki Y. Kadowaki T. J. Biol. Chem. 1998; 273: 5315-5322Abstract Full Text Full Text PDF PubMed Scopus (330) Google Scholar, 7Kohn A.D. Summers S.A. Birnbaum M.J. Roth R.A. J. Biol. Chem. 1996; 271: 31372-31378Abstract Full Text Full Text PDF PubMed Scopus (1109) Google Scholar). However, the physiological roles of the enzyme and exact mechanisms by which the enzyme is regulated in the signaling processes remain elusive. Although some evidence suggests that Akt may be activated by products of phosphatidylinositol (PI) 3-kinase1 such as phosphotidylinositol 3, 4-diphosphate (8Franke T.F. Kaplan D.R. Cantley L.C. Toker A. Science. 1997; 275: 665-668Crossref PubMed Scopus (1317) Google Scholar, 9Klippel A. Kavanaugh W.M. Pot D. Williams L.T. Mol. Cell. Biol. 1997; 17: 338-344Crossref PubMed Scopus (449) Google Scholar), other studies have shown that the activation of the enzyme is primarily caused by phosphorylation (10Burgering B.M. Coffer P.J. Nature. 1995; 376: 599-602Crossref PubMed Scopus (1893) Google Scholar, 11Kohn A.D. Takeuchi F. Roth R.A. J. Biol. Chem. 1996; 271: 21920-21926Abstract Full Text Full Text PDF PubMed Scopus (410) Google Scholar, 12Alessi D.R. Andjelkovic M. Caudwell B. Cron P. Morrice N. Cohen P. Hemmings B.A. EMBO. J. 1996; 15: 6541-6551Crossref PubMed Scopus (2550) Google Scholar). The hypothesis that Akt is activated by phosphorylation is further supported by the recent identification of a phosphoinositide-dependent protein kinase-1 or PDK1 (13Alessi D.R. Deak M. Casamayor A. Caudwell F.B. Morrice N. Norman D.G. Gaffney P. Reese C.B. MacDougall C.N. Harbison D. Ashworth A. Bownes M. Curr. Biol. 1997; 7: 776-789Abstract Full Text Full Text PDF PubMed Scopus (628) Google Scholar,14Stephens L. Anderson K. Stokoe D. Erdjument-Bromage H. Painter G.F. Holmes A.B. Gaffney P.R. Reese C.B. McCormick F. Tempst P. Coadwell J. Hawkins P.T. Science. 1998; 279: 710-714Crossref PubMed Scopus (925) Google Scholar). The 556-residue enzyme isolated from human tissues (hPDK1) contains a catalytic domain (residues 84–341) and a COOH-terminal pleckstrin homology or pleckstrin homology domain (residues 450–550). hPDK1 phosphorylates Akt at Thr308 and activates the enzyme in a PI 3-kinase-dependent manner. In addition to Akt, hPDK1 phosphorylates and activates another PI 3-kinase downstream target, p70S6K (15Pullen N. Dennis P.B. Andjelkovic M. Dufner A. Kozma S.C. Hemmings B.A. Thomas G. Science. 1998; 279: 707-710Crossref PubMed Scopus (732) Google Scholar). Although it is generally agreed upon that PI 3-kinase is necessary for a variety of insulin-mediated metabolic events, some evidence shows that Akt may not be necessary for insulin-mediated glycogen synthesis and glucose transport (6Ueki K. Yamamoto-Honda R. Kaburagi Y. Yamauchi T. Tobe K. Burgering B.M. Th. Coffer P.J. Komuro I. Akanuma Y. Yazaki Y. Kadowaki T. J. Biol. Chem. 1998; 273: 5315-5322Abstract Full Text Full Text PDF PubMed Scopus (330) Google Scholar, 16Kitamura T. Ogawa W. Sakaue H. Hino Y. Kuroda S. Takata M. Matsumoto M. Maeda T. Konishi H. Kikkawa U. Kasuga M. Mol. Cell. Biol. 1998; 18: 3708-3717Crossref PubMed Scopus (299) Google Scholar). These findings raise a very interesting question as to how PI 3-kinase is coupled to the downstream targets of the insulin receptor. Potential candidates that may mediate part of the phosphatidylinositol 3-kinase (PI 3-kinase) signaling are atypical isoforms of PKC such as PKCζ. PKCζ is activated in vitroby phosphatidylserine and polyphosphoinositides, products of PI 3-kinase (17Nakanishi H. Brewer K.A. Exton J.H. J. Biol. Chem. 1993; 268: 13-16Abstract Full Text PDF PubMed Google Scholar). In addition, insulin provokes a rapid increase in both the serine phosphorylation and the enzyme activity of PKCζ in rat adipocytes. Furthermore, this insulin-stimulated phosphorylation of PKCζ was inhibited by PI 3-kinase-specific inhibitors such as LY294002 and wortmannin (18Standaert M.L. Galloway L. Karnam P. Bandyopadhyay G. Moscat S. Farese R.V. J. Biol. Chem. 1997; 272: 30075-30082Crossref PubMed Scopus (410) Google Scholar). Although these data suggest that PKCζ may be activated by PI 3-kinase-dependent phosphorylation, the kinase(s) that phosphorylates PKCζ has not been identified. To further understand the mechanism of insulin receptor signal transduction and regulation, we attempted to clone the pdk1 gene from mouse tissues. Here we report the cloning and characterization of mouse PDK1 (mPDK1). We have found that mPDK1 mRNA was broadly expressed in various mouse tissues and in early development stages. Western blot analysis suggests there may be isoforms of the protein. We have also shown that mPDK1 is a kinase that phosphorylates and activates PKCζ. Our findings suggest that PDK1 may be a key enzyme that regulates divergent signaling pathways downstream of PI 3-kinase. Buffer A consisted of 50 mm Hepes, pH 7.6, 1% Triton X-100, 150 mm NaCl, 1 mmphenylmethanesulfonyl fluoride, 1 mm sodium orthovanadate, 1 mm sodium fluoride, 1 mm sodium pyrophosphate, 10 μg/ml aprotinin and 10 μg/ml leupeptin. Buffer B consisted of 50 mm Tris-HCl, pH 7.5, 5 mmMgCl2, 1 mm sodium fluoride, 1 mmsodium orthovanadate, and 1 mm phenylmethanesulfonyl fluoride. Buffer C consisted of 20 mm Hepes, pH 7.6, 3 mm CaCl2, 5 mm KCl, 7 mm NaHCO3, 107 mm NaCl, 1 mm MgSO4, 10 mm glucose, and 0.1% bovine serum albumin. Buffer D consisted of 20 mmNa2HPO4, pH 8.6, 0.5% Triton X-100, 0.1% SDS, 0.02% NaN3, and 1 m NaCl. The EST clone (AA117492) that contains an insert encoding the COOH terminus of mPDK1 (residues 265–559, see Fig.1) was obtained from ATCC. A mouse liver cDNA library (19Li H. Dong L. Whitlock Jr., J.P. J. Biol. Chem. 1994; 269: 28098-28105Abstract Full Text PDF PubMed Google Scholar) was screened using the partial mPDK1 cDNA fragment as a probe. Positive clones, contained within the pBK-CMV phagemid, were excised in vivo from the ZAP ExpressTM vector using ExAssist-XLOLR system (Stratagene, CA), following the protocol recommended by the manufacturer. DNA was sequenced from both directions by the Institutional DNA Sequencing Facility. A mouse multiple Northern blot containing poly(A) RNA from different tissues of mice aged from 8 to 12 weeks old and an embryonic Northern blot (CLONTECH) were hybridized under conditions recommended by the manufacturer, using the radiolabeled EST mPDK1 cDNA fragment as a probe. The full-length mPDK1 cDNA (except for the last four amino acids) was amplified by polymerase chain reaction and subcloned into the mammalian expression vector, pBEX (20Bram R.J. Hung D.T. Martin P.K. Schreiber S.L. Crabtree G.R. Mol. Cell. Biol. 1993; 13: 4760-4769Crossref PubMed Scopus (251) Google Scholar), in-frame at its COOH terminus with a sequence encoding the hemagglutinin (HA)-tag (pBEX/mPDK1). A kinase-inactive mPDK1 (mPDK1K114G) was generated by replacing the conserved ATP binding site lysine residue with glycine with the use of a Quick-change site-directed mutagenesis kit (Stratagene, CA). The full-length human PKCζ, except the last valine residue, was subcloned into pcDNA3.1 (Invitrogene, CA) in-frame at its COOH terminus with the Myc-tag (pcDNA/PKCζ) and was provided by Dr. R. Lin (UTHSCSA). Mutant forms of PKCζ in which the threonine residue at position 402 was changed either to alanine (PKCT402A) or glutamate (PKCT402E) were also generated by site-directed mutagenesis. To establish cell lines expressing both the IR and mPDK1, the recombinant pBEX/mPDK1 plasmid was transfected into CHO/IR cells with LipofectAMINE (Life Technologies, Inc.). Stable cell lines expressing high levels of mPDK1 (CHO/IR/mPDK1) were selected with 10 μg/ml puromycin. Positives were identified by Western blot using the antibody to the HA-tag (Babco, CA) and cloned by limiting dilution. The cDNA encoding the COOH terminus of the enzyme (mPDK1CT, amino acid residues 285–559 of mPDK1, see Fig. 1) was amplified by polymerase chain reaction and subcloned into a bacterial expression vector pGEX-4T-1 (Amersham Pharmacia Biotech). Overexpression and purification of the fusion protein was carried out according to the manufacturer's protocol. Polyclonal antibodies to the GST/mPDK1CT fusion protein were generated by immunizing rabbits. Immobilization of GST and GST/mPDK1CT to Affi-Gel 10 beads was performed according to the protocol provided by the manufacturer (Bio-Rad). Anti-mPDK1 antibody was affinity-purified by, first, absorption of the antiserum with GST-immobilized Affi-Gel 10 beads to remove the antibody to GST. The resulting supernatant was then incubated with GST/mPDK1CT-immobilized Affi-Gel 10 beads. The beads were washed extensively with a buffer containing 50 mm Tris-HCl, pH 7.5, and 0.5 m NaCl. After further washing with 50 mm Tris-HCl, pH 7.5, the anti-mPDK1 antibody bound to the beads was eluted with a buffer containing 0.1 m glycine, pH 2.8, 0.1% Triton X-100, and 150 mm NaCl, neutralized with 1 m Tris-HCl, pH 8.5, and dialyzed against a buffer containing 50 mm Hepes, pH 7.6, and 150 mmNaCl. Tissues from mice aged from 3 to 6 months old were homogenized in ice-cold Buffer A. Tissue homogenates were clarified by centrifugation at 18,000 × g for 2 × 30 min. The concentration of the proteins in the supernatant was determined by the Bradford assay. Proteins were separated by SDS-PAGE, transferred to a nitrocellulose membrane, and detected with the affinity-purified polyclonal antibody to mPDK1. A peptide (TFCGTPNYIAPEI) corresponding to the consensus phosphorylation sequence in the activation loop of PKCζ (residues 402 to 414) was chemically synthesized and used as a substrate for mPDK1. To facilitate the in vitro assays, three arginine residues were added to the ends of the peptide (one at the NH2 terminus and two at the COOH terminus). To identify the PDK1 phosphorylation site in PKCζ, we synthesized a peptide with the potential phosphorylation site threonine (Thr402) changed to alanine (T402A). Two positive control peptides with the sequences derived from the activation loop of Akt (KTFCGTPEYLAPEVRR) and the cAMP-dependent protein kinase (PKA, RTLCGTPEYLAPEIRR) were also synthesized. For in vitro kinase assays, the HA-tagged mPDK1 was immunoprecipitated from CHO/IR/mPDK1 cells by the anti-HA antibody (BABCO, CA) or control normal mouse immunoglobulin bound to protein-G-agarose beads. Kinase reactions were initiated by the addition of 30 μl of Buffer B containing 4 μCi of [γ-32P]ATP and 10 μl of 0.4 mm substrate or control peptides. After incubation for 30 min at 30 °C, the incorporation of 32P into peptides was determined by absorption of the positively charged peptides to P81 phosphocellulose membrane and checked by Cerenkov counting. Phosphorylation of PKCζ by mPDK1 was performed using a protocol similar to that used in the study of PDK1-catalyzed p70S6K phosphorylation (14Stephens L. Anderson K. Stokoe D. Erdjument-Bromage H. Painter G.F. Holmes A.B. Gaffney P.R. Reese C.B. McCormick F. Tempst P. Coadwell J. Hawkins P.T. Science. 1998; 279: 710-714Crossref PubMed Scopus (925) Google Scholar). In brief, the Myc-tagged PKCζ was transiently expressed in CHO/IR cells. The extracts of these cells were mixed with those of insulin-treated or -nontreated CHO/IR cells expressing also the wild-type or mutant (K114G) mPDK1. The Myc-tagged PKCζ and the HA-tagged mPDK1 were co-immunoprecipitated by appropriate antibodies to the tags. In vitro phosphorylation was initiated with the addition of 30 μl of Buffer B containing 2 μCi of [γ-32P]ATP and incubated for 30 min at 30 °C. After washing with an ice-cold buffer containing 50 mm Hepes, pH 7.4, 150 mm NaCl, and 0.1% Triton X-100, the immunoprecipitated proteins were heated at 95 °C for 3 min, separated by SDS-PAGE, and blotted to a nitrocellulose membrane. Phosphorylation of PKCζ by PDK1 was visualized by autoradiography and quantified by PhosphoImager analysis. In vivolabeling was carried out as described previously with some modifications (21Dong L.Q. Du H. Porter S. Kolakowski Jr., L.F. Lee A.V. Mandarino J. Fan J. Yee D. Liu F. J. Biol. Chem. 1997; 272: 29104-29112Abstract Full Text Full Text PDF PubMed Scopus (66) Google Scholar). In brief, cDNAs encoding the wild-type or T402A mutant PKCζ were transiently transfected into either CHO/IR or CHO/IR/mPDK1 cells. After transfection (36 h), cells in 10-cm diameter plates were incubated in 4 ml of phosphate-free Buffer C for 30 min at 37 °C and then radiolabeled with 0.3 mCi of carrier-free [32P]orthophosphate/plate for 4 h at 37 °C. After treatment with or without insulin (10 nm) for an additional 5 min, cells were washed twice with ice-cold Buffer C and lysed with 0.4 ml/plate lysis Buffer A. After centrifugation at 12,000 ×g for 10 min at 4 °C, the 32P-labeled PKCζ was immunoprecipitated with anti-Myc antibody bound to protein G-agarose beads (Amersham Pharmacia Biotech). The immunoprecipitates were washed twice with Buffer D and then twice with the same buffer, except that the NaCl concentration in the buffer was changed to 0.15m. The radiolabeled PKCζ was separated by SDS-PAGE, blotted to Immobilon P membrane (Millipore), and visualized by autoradiography. Thein vitro assay of PKCζ activity was carried out by a two-step process. First, the immunoprecipitated PKCζ was activated by purified mPDK1 in the presence of Buffer B containing 25 μm ATP. After incubation for 30 min at 30 °C, the phosphorylated PKCζ/protein G beads were washed twice with an ice-cold buffer containing 50 mm Hepes, pH 7.4, 150 mm NaCl, and 0.1% Triton X-100 and once with Buffer B. The activity of PKCζ was then determined as described (18Standaert M.L. Galloway L. Karnam P. Bandyopadhyay G. Moscat S. Farese R.V. J. Biol. Chem. 1997; 272: 30075-30082Crossref PubMed Scopus (410) Google Scholar), using the Ser159-PKC-ε-(149–164) peptide (Quality Controlled Biochemicals, Inc., Hopkinton, MA) as a substrate. To study the effect of mPDK1 on PKCζ activation in cells, pcDNA/PKCζ plasmid was transfected into CHO/IR or CHO/IR/mPDK1 cells by electroporation. After transfection (48 h), the cells were serum-starved for 1 h, treated with or without 10−8m insulin for 5 min, and then lysed in Buffer A. The transiently expressed PKCζ was immunoprecipitated by the anti-Myc antibody (Santa Cruz, CA) bound to protein G beads. The activity of PKCζ was determined as described above. To study the functional role of PDK1, we searched the NCBI dbEST using the hPDK1 sequence (13Alessi D.R. Deak M. Casamayor A. Caudwell F.B. Morrice N. Norman D.G. Gaffney P. Reese C.B. MacDougall C.N. Harbison D. Ashworth A. Bownes M. Curr. Biol. 1997; 7: 776-789Abstract Full Text Full Text PDF PubMed Scopus (628) Google Scholar) as a probe. We identified a partial cDNA sequence (AA117492) highly homologous (85% identical) to the sequence encoding the COOH terminus of hPDK1 (amino acid residues 262–556) (13Alessi D.R. Deak M. Casamayor A. Caudwell F.B. Morrice N. Norman D.G. Gaffney P. Reese C.B. MacDougall C.N. Harbison D. Ashworth A. Bownes M. Curr. Biol. 1997; 7: 776-789Abstract Full Text Full Text PDF PubMed Scopus (628) Google Scholar), suggesting that the sequence is the mouse homologue of hPDK1. To clone the full-length cDNA for mPDK1, we screened a mouse liver cDNA library using the insert of the mPDK1 EST clone as a probe. Several positive clones were isolated and characterized. The longest sequence (clone L3, 1.9 kb) was predicted to encode a protein comprised of 559 amino acids with an estimated molecular mass of 64 kDa (Fig.1). The putative initiation codon at position 101 was identified based on the presence of a consensus Kozak sequence (22Kozak M. Nucleic Acids Res. 1987; 15: 8125-8148Crossref PubMed Scopus (4189) Google Scholar) and the putative assignment of the human and rat sequences (13Alessi D.R. Deak M. Casamayor A. Caudwell F.B. Morrice N. Norman D.G. Gaffney P. Reese C.B. MacDougall C.N. Harbison D. Ashworth A. Bownes M. Curr. Biol. 1997; 7: 776-789Abstract Full Text Full Text PDF PubMed Scopus (628) Google Scholar, 14Stephens L. Anderson K. Stokoe D. Erdjument-Bromage H. Painter G.F. Holmes A.B. Gaffney P.R. Reese C.B. McCormick F. Tempst P. Coadwell J. Hawkins P.T. Science. 1998; 279: 710-714Crossref PubMed Scopus (925) Google Scholar). mPDK1 contains a catalytic domain located between residues 84 and 344 and a COOH-terminal pleckstrin homology domain located between residues 453 and 553. The overall sequence percent identity between mPDK1 and hPDK1 (13Alessi D.R. Deak M. Casamayor A. Caudwell F.B. Morrice N. Norman D.G. Gaffney P. Reese C.B. MacDougall C.N. Harbison D. Ashworth A. Bownes M. Curr. Biol. 1997; 7: 776-789Abstract Full Text Full Text PDF PubMed Scopus (628) Google Scholar) or rat PDK1 (14Stephens L. Anderson K. Stokoe D. Erdjument-Bromage H. Painter G.F. Holmes A.B. Gaffney P.R. Reese C.B. McCormick F. Tempst P. Coadwell J. Hawkins P.T. Science. 1998; 279: 710-714Crossref PubMed Scopus (925) Google Scholar) is 95 and 98, respectively. However, like both the human and rat PDK1 cDNA sequences, no in-frame upstream stop codons were observed in mPDK1 cDNA. The delineation of the bona fide open reading frame may await further investigation. To assess the tissue distribution of PDK1 mRNA expression, Northern blot analysis was performed. A 7-kb mPDK1 transcript was found to be expressed in all tissues examined, and the highest expression was detected in the heart, brain, liver, and testis (Fig.2 A). In the testis, a high level expression of an additional 2-kb transcript was also detected (Fig. 2 A). The high and specific expression of this transcript in testis is consistent with the hypothesis that there may be testis-specific isoforms of PDK1 that function in cell survival or cell death decisions during spermatogenesis (13Alessi D.R. Deak M. Casamayor A. Caudwell F.B. Morrice N. Norman D.G. Gaffney P. Reese C.B. MacDougall C.N. Harbison D. Ashworth A. Bownes M. Curr. Biol. 1997; 7: 776-789Abstract Full Text Full Text PDF PubMed Scopus (628) Google Scholar). The expression of PDK1 mRNA in mouse embryos was also examined. The 7-kb mRNA transcript could be detected as early as day 7 and maintained constant to day 17 post-conception (Fig. 2 B). The high expression of mPDK1 mRNA throughout the embryonic stages suggests that the protein may play important roles in embryonic development. Using the affinity-purified polyclonal anti-mPDK1 antibody, we examined endogenous mPDK1 expression in different mouse tissues (Fig.3). Protein bands with molecular masses of approximately of 62 to 64 kDa was detected in most of the tissues examined, with the highest expression in the testis (Fig. 3, lane 1), brain (Fig. 3, lane 2), heart (Fig. 3, lane 3), spleen (Fig. 3, lane 5), and adipose tissue (Fig.3, lane 6). Our unpublished data suggest that the heterogeneity of the protein bands could correspond to mPDK1 isoforms translated from different mRNAs or generated by alternative use of initiation sites of translation of the same mRNA. In addition to these isoforms, the antibody also detected two other protein bands in most of the tissues examined, with molecular masses of approximately 125 and 180 kDa, respectively (Fig. 3, indicated by arrows). Preincubation of the antibody with excess amounts of GST-mPDK1CT antigen eliminated these protein bands, suggesting that the immunoreactivity was specific. hPDK1 has been shown to phosphorylate Akt and p70S6Kat Thr308 (14Stephens L. Anderson K. Stokoe D. Erdjument-Bromage H. Painter G.F. Holmes A.B. Gaffney P.R. Reese C.B. McCormick F. Tempst P. Coadwell J. Hawkins P.T. Science. 1998; 279: 710-714Crossref PubMed Scopus (925) Google Scholar) and Thr229 (15Pullen N. Dennis P.B. Andjelkovic M. Dufner A. Kozma S.C. Hemmings B.A. Thomas G. Science. 1998; 279: 707-710Crossref PubMed Scopus (732) Google Scholar), respectively. Comparison of the sequences surrounding these phosphorylation sites revealed a high homology between these two enzymes (Fig.4 A). This finding suggests that PDK1 recognizes a consensus sequence that may also exist in other protein kinases. To test this hypothesis, we searched the NCBI Brookhaven protein data bank using the Akt phosphorylation consensus sequence TFCGTPEYLAPE (Fig. 4 A) as a probe. In addition to the two known PDK1 substrates Akt and P70S6K, we identified a number of kinases including the catalytic domain of cAMP-dependent protein kinase (PKA, protein data bank accession number P05206), PKC isoforms, cell cycle protein kinase CDC5 (G416768), cGMP-dependent protein kinase (P00516), the polo-like Ser/Thr protein kinase PLK (P53350), and spermatozoon-associated protein kinase (P21901) (Fig. 4 A). This finding was very interesting as it suggests that PDK1 may be a potential upstream kinase for these kinases, including PKCζ. As PKCζ has been shown to undergo PI 3-kinase-dependent phosphorylation and activation, we attempted to examine whether this enzyme was a substrate of mPDK1. We established a stable Chinese hamster ovary cell line expressing both the IR and an HA-tagged mPDK1 (CHO/IR/mPDK1). We also transiently expressed mutant mPDK1 (mPDK1K114G) and Myc-tagged PKCζ into CHO/IR cells. We then carried out in vitro phosphorylation studies using the immunoaffinity-purified mPDK1 and PKCζ. PKCζ was significantly phosphorylated in the presence of mPDK1 (Fig. 4 B,lanes 3 and 4) but not in the absence of the enzyme (Fig. 4B, lanes 1 and 2) or in the presence of a kinase-inactive mutant mPDK1 (mPDK1K114G, Fig. 4 B, lanes 7 and 8). The in vitro phosphorylation of PKCζ by mPDK1 was insulin-independent as a pretreatment of CHO/IR/mPDK1 cells with insulin had no significant effect on mPDK1 activity (Fig. 4 B), although under the same conditions the insulin-stimulated insulin receptor and IRS-1 tyrosine phosphorylation as well as Akt activation were all significantly increased (data not shown). PhosphoImager analysis showed that the amount of phosphate incorporated into PKCζ was 20-fold in the presence of mPDK1 over that in the absence of the enzyme. The wild-type mPDK1 also underwent significant autophosphorylation in vitro and in cells (data not shown). However, whether the phosphorylation plays a role in the regulation of the activity of the enzyme remain to be established. To identify the phosphorylation site on PKCζ, we carried out in vitro phosphorylation studies using synthetic peptides corresponding to the activation loop of PKCζ as substrates. As shown in Fig. 4 C, the peptide derived from the wild-type PKCζ activation loop sequence was readily phosphorylated by mPDK1 in vitro (lanes 5 and 6). Similar levels of phosphorylation were observed for the peptide substrates derived from Akt and PKA (Fig. 4C, lanes 7 and 8). Replacement of Thr402 with alanine in PKCζ peptide reduced the phosphorylation to the basal level (Fig. 4 C, lanes 1–4), suggesting that Thr402 was the site of phosphorylation by PDK1. Consistent with this observation, the phosphorylation of the mutant PKCζ (PKCζT402A) by mPDK1 was significantly decreased (Fig. 4 B, lanes 5 and6). Similar results were also obtained for the PKCT402E mutant (data not shown). Treatment of the cells with either wortmannin (50 nm, 1 h) or LY294002 (50 μm, 1 h) had no significant effect on the in vitro mPDK1 activity toward its peptide substrates nor did it affect PDK1-mediated phosphorylation of PKCζ in vitro(data not shown). Under the this same condition, the insulin-stimulated Akt in vivo phosphorylation was completely blocked (data not shown). To investigate whether PKCζ is an in vivo substrate for PDK1 and whether Thr402 is the site of phosphorylation by PDK1 in cells, we performed in vivo labeling experiments. As shown in Fig. 4 D, PKCζ was phosphorylated in CHO/IR cells, and insulin-treatment stimulated the phosphorylation of the enzyme (Fig. 4 D, lanes 1 and 2). Overexpressing mPDK1 increased the basal phosphorylation of the enzyme (Fig. 4 D, lanes 3 and 4). In agreement with the in vitro phosphorylation results, the phosphorylation of PKCζT402A mutant was significantly decreased by mPDK1 in the CHO/IR/mPDK1 cells (Fig. 4 D,lanes 5 and 6). These data provide further evidence that Thr402 in the activation loop of PKCζ is the site of phosphorylation by PDK1. To test whether phosphorylation of PKCζ affected its enzymatic activity, we transiently expressed the Myc-tagged enzyme in CHO/IR or CHO/IR/mPDK1 cells. In the parental CHO/IR cells, insulin stimulation resulted in an approximate 1.6-fold increase in PKCζ activity (Fig.5 A, lanes 1 and2). Overexpression of mPDK1 caused an approximate 2-fold increase in the basal PKCζ activity (Fig. 5 A, lanes 1 and 3), and insulin treatment further activated the enzyme (Fig. 4 D, lane 4). A Western blot showed similar levels of PKCζ in CHO/IR and CHO/IR/mPDK1 cells (data not shown). These results suggest that although mPDK1 was able to phosphorylate PKCζ in vitro in an insulin-independent manner, other effector(s) may also be necessary for maximum activation of PKCζ in cells. To examine whether PDK1 had a direct effect on PKCζ activity, we examined PKCζ activity in vitro before and after phosphorylation by PDK1. As shown in Fig. 5 B, phosphorylation of PKCζ by the wild-type PDK1 resulted in a 3-fold increase in the activity of PKCζ (Fig. 5, lanes 2 and3). No significant increase in PKCζ activity was observed when the enzyme was preincubated with the kinase-inactive mutant mPDK1K114G (Fig. 5 B, lane 5). Under this assay condition, neither mPDK1 itself (Fig. 5 B,lane 1) nor the mutant PKCζT402A (Fig.5 B, lane 4) had significant effect on the phosphorylation of the peptide substrates. PDK1 is a recently identified PI 3-kinase downstream protein kinase that phosphorylates and activates Akt, p70S6K and cAMP-dependent protein kinase (13Alessi D.R. Deak M. Casamayor A. Caudwell F.B. Morrice N. Norman D.G. Gaffney P. Reese C.B. MacDougall C.N. Harbison D. Ashworth A. Bownes M. Curr. Biol. 1997; 7: 776-789Abstract Full Text Full Text PDF PubMed Scopus (628) Google Scholar, 14Stephens L. Anderson K. Stokoe D. Erdjument-Bromage H. Painter G.F. Holmes A.B. Gaffney P.R. Reese C.B. McCormick F. Tempst P. Coadwell J. Hawkins P.T. Science. 1998; 279: 710-714Crossref PubMed Scopus (925) Google Scholar, 15Pullen N. Dennis P.B. Andjelkovic M. Dufner A. Kozma S.C. Hemmings B.A. Thomas G. Science. 1998; 279: 707-710Crossref PubMed Scopus (732) Google Scholar, 23Alessi D.R. Kozlowski M.T. Weng Q.-P. Morrice N. Avruch J. Curr. Biol. 1998; 8: 69-81Abstract Full Text Full Text PDF PubMed Scopus (521) Google Scholar, 24Cheng X. Ma Y. Moore M. Hemmings B.A. Taylor S.S. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 9849-9854Crossref PubMed Scopus (192) Google Scholar). Although these data suggest that PDK1 may play an important part in cell signaling processes, direct evidence of the role of PDK1 remains elusive. We report here the cloning of PDK1 in mouse tissue. Our results show that the mPDK1 mRNA is broadly expressed in various tissues and in early embryonic stages. These findings suggest that the enzyme may play a general role in signaling processes and in development. The high level expression of a testis-specific mPDK1 mRNA also suggests that the enzyme may be involved in sex differentiation processes. This hypothesis is consistent with the finding that PDK1 is homologous to the Drosophila protein kinase DSTPK61, an enzyme that is implicated in the regulation of sex differentiation, oogenesis, and spermatogenesis (13Alessi D.R. Deak M. Casamayor A. Caudwell F.B. Morrice N. Norman D.G. Gaffney P. Reese C.B. MacDougall C.N. Harbison D. Ashworth A. Bownes M. Curr. Biol. 1997; 7: 776-789Abstract Full Text Full Text PDF PubMed Scopus (628) Google Scholar). Western blot analysis using anti-mPDK1 antibody revealed multiple protein bands in various mouse tissues (Fig. 3), suggesting that there may be isoforms of mPDK1. This hypothesis is consistent with the following observations. First, it has been shown that in sheep brain there were four phosphatidylinositol 3,4,5-trisphosphate-binding proteins with Akt kinase activity (14Stephens L. Anderson K. Stokoe D. Erdjument-Bromage H. Painter G.F. Holmes A.B. Gaffney P.R. Reese C.B. McCormick F. Tempst P. Coadwell J. Hawkins P.T. Science. 1998; 279: 710-714Crossref PubMed Scopus (925) Google Scholar). Second, PDK1 cDNAs isolated from human (13Alessi D.R. Deak M. Casamayor A. Caudwell F.B. Morrice N. Norman D.G. Gaffney P. Reese C.B. MacDougall C.N. Harbison D. Ashworth A. Bownes M. Curr. Biol. 1997; 7: 776-789Abstract Full Text Full Text PDF PubMed Scopus (628) Google Scholar), rat (14Stephens L. Anderson K. Stokoe D. Erdjument-Bromage H. Painter G.F. Holmes A.B. Gaffney P.R. Reese C.B. McCormick F. Tempst P. Coadwell J. Hawkins P.T. Science. 1998; 279: 710-714Crossref PubMed Scopus (925) Google Scholar), and mouse (this study) all contain no 5′-upstream stop codon, raising the possibility that they may be partial DNA sequences of larger cDNAs. In agreement with this observation, protein bands with molecular weights larger than that encoded by the mPDK1 cDNA were also detected by the antibody (Fig.3). Because the polyclonal antibody recognizes the COOH terminus of mPDK1, the antibody-immunoreactive proteins, if they are indeed mPDK1 isoforms, may only differ at their amino termini. In agreement with this hypothesis, we have cloned a cDNA from mouse testis. 2L. Q. Dong and F. Liu, unpublished data. This cDNA encodes a 62-kDa mPDK1 isoform with an amino terminus different from that of mPDK1 presented in this report (Fig. 1). It is interesting to notice that multiple isoforms have also been found in many other protein kinases including the PDK1 downstream effectors such as protein kinase C (25Newton A.C. J. Biol. Chem. 1995; 270: 28495-28498Abstract Full Text Full Text PDF PubMed Scopus (1480) Google Scholar, 26Mellor H. Parker P.J. Biochem. J. 1998; 332: 281-292Crossref PubMed Scopus (1365) Google Scholar) and Akt (27Marte B.M. Downward J. Trends Biochem. Sci. 1997; 22: 355-358Abstract Full Text PDF PubMed Scopus (654) Google Scholar). The presence of PDK1 isoforms may thus have some physiological relevance. It is possible that the amino acid differences between these isoforms could lead to differences in specificity between different substrates of the enzyme. In addition, different isoforms may be differently regulated in cells. Thus, the existence of multiple and tissue-specific isoforms of PDK1 suggests an additional level of regulation may be involved. Studies are currently in progress to test these ideas. In agreement with the findings of others (13Alessi D.R. Deak M. Casamayor A. Caudwell F.B. Morrice N. Norman D.G. Gaffney P. Reese C.B. MacDougall C.N. Harbison D. Ashworth A. Bownes M. Curr. Biol. 1997; 7: 776-789Abstract Full Text Full Text PDF PubMed Scopus (628) Google Scholar, 15Pullen N. Dennis P.B. Andjelkovic M. Dufner A. Kozma S.C. Hemmings B.A. Thomas G. Science. 1998; 279: 707-710Crossref PubMed Scopus (732) Google Scholar), we have found that PDK1 was constitutively activated, and insulin treatment did not further increase its activity in vitro (Fig. 4, Band C). However, we also observed an additive effect of PDK1 and insulin in PKCζ activity in cells (Fig. 5 A). One possible explanation for this discrepancy is that although PDK1 was constitutively active and able to phosphorylate and activate PKCζin vitro and in cells, other insulin-dependent mechanism(s), such as phosphorylation by other kinase(s) or translocation to a specific cellular compartment, may also be required for the full activation of PKCζ. Consistent with this idea, our unpublished data showed that overexpression of mPDK1 in cells resulted in an increase in Akt activity but did not stimulate its phosphorylation at Ser473 in cells, whose phosphorylation (by an unknown kinase) has been shown to be necessary for the full activation of the enzyme (24Cheng X. Ma Y. Moore M. Hemmings B.A. Taylor S.S. Proc. Natl. Acad. Sci. U. S. A. 1998; 95: 9849-9854Crossref PubMed Scopus (192) Google Scholar, 27Marte B.M. Downward J. Trends Biochem. Sci. 1997; 22: 355-358Abstract Full Text PDF PubMed Scopus (654) Google Scholar). Our results are also consistent with those of Alessi et al. (13Alessi D.R. Deak M. Casamayor A. Caudwell F.B. Morrice N. Norman D.G. Gaffney P. Reese C.B. MacDougall C.N. Harbison D. Ashworth A. Bownes M. Curr. Biol. 1997; 7: 776-789Abstract Full Text Full Text PDF PubMed Scopus (628) Google Scholar) who found that overexpression of PDK1 activated Akt and potentiated the IGF-1-induced increase of Akt activity in cells. In addition, previous studies have shown that phosphorylation of the conserved threonine residue in the activation loop of PKC isoforms resulted in a conformational change of the enzymes and rendered the inactive PKCs to become the cofactor-activable, mature form (28Cazaubon S.M. Parker P.J. J. Biol. Chem. 1993; 268: 17559-17563Abstract Full Text PDF PubMed Google Scholar, 29Orr J.W. Newton A.C. J. Biol. Chem. 1994; 269: 27715-27718Abstract Full Text PDF PubMed Google Scholar). Our studies provide evidence that PDK1 is the kinase that phosphorylates the conserved threonine residue in PKCζ and initiates the activation process. During the reviewing process of our manuscript, two groups reported their findings on the activation of PKCζ by PDK1 (30Chou M.M. Hou W. Johnson J. Graham L.K. Lee M.H. Chen C.-S. Newton A.C. Schaffhausen B.S. Toker A. Curr. Biol. 1998; 8: 1069-1077Abstract Full Text Full Text PDF PubMed Google Scholar, 31Good J.A. Ziegler W.H. Parekh D.B. Alessi D.R. Cohen P. Parker P.J. Science. 1998; 281: 2042-2045Crossref PubMed Scopus (977) Google Scholar). By using antibodies specific to the phosphothreonine residue in the activation loop of PKC, both groups showed that Thr410 in rat PKCζ (equivalent to Thr402 in the human version enzyme) was indeed the in vivo phosphorylation site of PDK1. Our results are in full agreement with these findings. In summary, available data have shown that PDK1 is an upstream kinase for different effectors including Akt, p70S6K, cAMP-dependent protein kinase, and PKCζ. These findings suggest that PDK1 may be a site for divergence of different signaling pathways downstream of PI 3-kinase. There is evidence suggesting that PKCζ is downstream of PI 3-kinase and may contribute to insulin-stimulated glucose transport in 3T3-L1 adipocytes (17Nakanishi H. Brewer K.A. Exton J.H. J. Biol. Chem. 1993; 268: 13-16Abstract Full Text PDF PubMed Google Scholar, 18Standaert M.L. Galloway L. Karnam P. Bandyopadhyay G. Moscat S. Farese R.V. J. Biol. Chem. 1997; 272: 30075-30082Crossref PubMed Scopus (410) Google Scholar, 32Bandyopadhyay G. Standaert M.L. Zhao L. Yu B. Avignon A. Galloway L. Karnam P. Moscat J. Farese R.V. J. Biol. Chem. 1997; 272: 2551-2558Crossref PubMed Scopus (276) Google Scholar). Activation of PKC isoforms by PDK1 may thus provide an alternative pathway to mediate some PI 3-kinase-dependent downstream events such as glycogen synthesis and GLUT4 translocation in cells. The finding that PDK1 recognizes the consensus activation loop sequence in various protein kinases suggests that other PKC isoforms may also be substrates of PDK1. Further studies will help us to define the importance of the PDK1/PKC pathway in insulin and other growth factor-mediated signaling processes. We thank Dr. H. Li for the mouse liver cDNA library and Dr. R. Lin for pcDNA/PKCζ.