Protein Kinase B/Akt Is Activated by Polyomavirus Middle-T Antigen Via a Phosphatidylinositol 3-Kinase-Dependent Mechanism

Protein kinase B/Akt is activated by polyomavirus middle-T antigen via a phosphatidylinositol 3-kinase-dependent mechanism

R Meili, P Cron, BA Hemmings and K Ballmer-Hofer1

Friedrich Miescher Institute, PO Box 2543, CH-4002 Basel, Switzerland

The middle tumor antigen (middle-T) of mouse poly- threonine protein phosphatase 2A (PP2A) (Pallas et al., omavirus is responsible for the transforming potential of 1990; Walter et al., 1990), Src family tyrosine kinases this virus. Middle-T has been shown to interact with a (Courtneidge and Smith, 1983 and reviewed in Kiefer variety of cellular proteins known to mediate mitogenic et al., 1994), the adaptor protein SHC which binds to signaling, like phosphatase-2A, Src family kinases, phosphotyrosine 250 of middle-T via its phosphotyr- phosphatidylinositol 3-kinase (PI 3-kinase), the adapter osine binding (PTB) domain (Dilworth et al., 1994; protein SHC, phospholipase Cg-1 and 14-3-3 family Campbell et al., 1994), PI 3-kinase (Whitman et al., proteins. Association with SHC and PI 3-kinase, 1985; Talmage et al., 1989; Courtneidge et al., 1989) respectively, stimulates two independent signaling path- and phospholipase Cg-1 (Su et al., 1995). When ways that are indispensible for viral oncogenicity. SHC associated with middle-T, SHC gets phosphorylated activates the Ras/MAPK pathway via Grb2/SOS at tyrosine residues. This renders this protein resulting in changes in early gene expression. The competent to initiate downstream signaling through downstream targets of PI 3-kinase are less well studied Grb2 and SOS resulting in increased GTP loading of but seem to impinge on serum response factor (SRF) Ras. Activated Ras propagates and presumably which is also involved in regulating early gene expres- ampli®es incoming signals and initiates the MAP sion. Recently, the protein kinase B/Akt (PKB/Akt) has kinase cascade (Robinson and Cobb, 1997). Associa- been identi®ed as a target of PI 3-kinase in receptor tion of PI 3-kinase with tyrosine kinase receptors or tyrosine kinase signaling. Here we show that PKB/Akt is middle-T leads to activation of this enzyme, resulting a target of wild type middle-T, but not of mutants unable in increased production of D3 phosphatidylinositides. to activate PI 3-kinase. These data were con®rmed using These lipids have been implicated in the regulation of a inhibitors of PI 3-kinase as well as dominant-negative plethora of cellular phenomena such as cell growth, alleles of the catalytic subunit of this lipid kinase. In rearrangement of cytoskeletal structures, cellular addition, mutants of PKB/Akt lacking a pleckstrin motility, receptor tracking, cell survival and chemo- homology domain and therefore unable to bind to D3 taxis (reviewed in Toker and Cantley, 1997). Recent phospatidylinositides were not activated by middle-T. work shows that these lipids undergo speci®c interac- Taken together these data suggest that middle-T tions with target proteins having SH2- (Rameh et al., activates PKB/Akt in a PI 3-kinase-dependent manner. 1995) or PH-domains (Klarlund et al., 1997; Franke et Furthermore, direct association with D3 phosphatidyli- al., 1997). Protein kinase B/Akt (PKB/Akt), also nositides seems to be essential for activation of PKB/ known as RAC-PK, is one of these targets and is Akt. activated in response to growth stimulation (Franke et al., 1995). The phospholipid products of PI 3-kinase Keywords: middle-T; PI 3-kinase; protein kinase B/Akt; bind to the PH-domain of PKB/Akt and recruit this polyomavirus protein to the plasma membrane. This leads to activation upon phosphorylation by a not yet identi®ed cellular kinase, presumably also located at cellular membranes (Alessi et al., 1996). It has also Introduction been shown that PI 3-kinase is required for full activation of the Ras/MAPK pathway by middle-T as Mouse polyomavirus induces a broad spectrum of well as for signaling to serum response factor (Urich et tumors in mice and transforms cells in culture al., 1997). Mutants of middle-T still capable of (Treisman et al., 1981; Rassoulzadegan et al., 1982). activating PI 3-kinase, but lacking a binding site for The viral proteins mediating this function are the SHC, trigger stimulation of p70S6-kinase (Dahl et al., tumor antigens, in particular the middle-T antigen. 1996) or p21Rac-mediated transcriptional regulation Oncogenic transformation by polyomavirus has been (Urich et al., 1997). used as a paradigm for the study of mitogenic signaling in mammalian cells since this protein activates many of the cellular pathways also targeted by tyrosine kinase Results growth factor receptors. Signaling by middle-T requires interactions with cellular proteins such as the serine/ Middle-T activates PKB/Akt To determine the eect of middle-T on PKB/Akt Correspondence: K Ballmer-Hofer activity, a HA-tagged version of this kinase was 1Current address: Institute for Medical Radiobiology, CH-5232 Villigen-PSI, Switzerland transiently expressed together with middle-T in Received 7 May 1997; revised 29 September 1997; accepted 29 NIH3T3 cells. Kinase activity was measured in September 1997 immunecomplex assays using histone 2B as substrate. PKB/Akt activation by middle-T RMeiliet al 904 In the absence of serum, wild type middle-T activated reverted middle-T- as well as PDGF-induced PKB/ PKB/Akt about 20-fold (Figure 1a, compare lanes 1 Akt activation (Figure 3, compare lane 2 with 5 and 3 and 2) whereas in the presence of serum activation was with 6). Similarly, coexpression of an enzymatically only twofold (Figure 1a, compare lanes 6 and 7) inactive mutant of the catalytic p110 subunit of PI 3- indicating that PKB/Akt is maintained at high kinase blocked PKB/Akt activation by middle-T enzymatic activity by serum growth factors. (Figure 3, lane 9). A constitutively active form of the p110 subunit, on the other hand, slightly increased PKB/Akt activity (Figure 3, lane 8). Taken together PI 3-kinase but not SHC is required for PKB/Akt these data indicate that middle-T-stimulated activation activation by middle-T of PI 3-kinase results in activation of PKB/Akt. Mutants of middle-T were used to further de®ne the requirements for PKB/Akt activation. The mutant The PH-domain of PKB/Akt is required for activation Y315F, which does not bind p85 and therefore cannot by middle-T activate PI 3-kinase, was unable to activate PKB/Akt (Figure 1a, lanes 5 and 9 and 1b, bar 4). The mutant To investigate whether the phosphoinositides produced Y250F, unable to associate with SHC, activated PKB/ upon activation of PI 3-kinase by middle-T are Akt almost to the level of wild type middle-T (Figure required for PKB/Akt activation, we studied the 1a, lanes 4 and 8 and 1b, bar 3). The fact that this activity of a mutant of PKB/Akt, DPH-PKB/Akt, mutant was slightly less potent at activating PKB/Akt lacking a pleckstrin homology domain and therefore might be explained by the observation that it binds less unable to bind D3 phosphoinositides (Franke et al., p85 as shown earlier (Dilworth et al., 1994) (Figure 2, 1997; Alessi et al., 1996). This mutant was not compare lanes 2 and 3). Finally, the mutant dl1015 activated by middle-T (Figure 4, lanes 3 and 4). The which binds p85 (Figure 2, lane 4), but fails to activate fact that DPH-PKB/Akt was almost fully activated by PI 3-kinase in vivo (Ling et al., 1992), showed that the nonspeci®c tyrosine phosphatase inhibitor vanadyl interaction of PI 3-kinase with middle-T per se is not hydroperoxide as previously described (Andjelkovic et sucient for PKB/Akt activation (Figure 1b, bar 5). al., 1996), con®rms that the catalytic center of the Additional experiments using the PI 3-kinase kinase domain was unaected by the deletion (Figure inhibitor LY294002 were performed to demonstrate 4, lane 5). From these data we conclude that middle-T that lipid kinase activity is essential for PKB/Akt signaling to PKB/Akt is mediated by binding of D3 activation. The data show that LY294002 rapidly phosphoinositides to the PH-domain of PKB/Akt.

a 0.1% CS 14h 10% CS b Relative activation of HA PKB — mT wt mT Y250F mT Y315F mT dl1015 — 20% CS mT wt mT Y250F mT Y315F

HA PKB

H2B

Figure 1 Activation of PKB/Akt by middle-T. (a) Cells were transfected as described and serum deprived (lanes 1 ± 5) for 14 h. Cells in lane 2 were treated with 20% calf serum for 30 min prior to cell lysis. Cells were lysed and HA-tagged PKB/Akt was immunoprecipitated with the monoclonal antibody 12CA5. The immunoprecipitates were assayed for kinase activity using histone 2B as substrate and analysed by SDS ± PAGE (bottom panel). To determine the amount of kinase in the immunoprecipitates, a Western blot was decorated with the anti-HA- tag antibody (top panel). The bar graph at the bottom shows relative kinase activities of one representative experiment. (b) Shows the mean relative activation of PKB in serum-deprived NIH3T3 cells by various middle-T mutants calculated from ®ve separate experiments (except for dl1015: n=2). Activation by wild type middle-T was taken as 100% PKB/Akt activation by middle-T RMeiliet al 905 Anti-middle-T immunoprecipitates

Blot HA PKB + mT wt + mT Y250F + mT dl1015 + mT Y315F HA PKB + mT wt + mT Y250F + mT dl1015 + mT Y315F α p85

p85

α mT

Src

p43

α p36 p36

1 2 3 4 5 1 2 3 4 5 Westernblot in vitro kinase Figure 2 Analysis of middle-T associated proteins. Middle-T was immunoprecipitated from lysates by the anti-middle-T monoclonal antibody Pab 762. Immunoprecipitates were subjected to an in vitro kinase reaction allowing radiolabeling of p85, Src, middle-T and an unidenti®ed band, p43, as shown in the right panel. The Pab 762 anti-middle-T blot and the anti-p36 blot detecting the catalytic subunit of protein phosphatase 2A show that the amounts of complex present in the individual immuneprecipitates are within twofold

mT wt plus This study identi®es PKB/Akt as a target of polyomavirus middle-T. It also shows that activation of PKB/Akt strictly depends on T antigen-associated — mT wt PDGF — mT wt PDGF — p110 p110K- PI 3-kinase. Our data are in agreement with the earlier LY294002 – – – + + + – – – proposed concept that the pleckstrin homology domain of PKB/Akt associates with D3 phosphoinositides HA PKB produced by PI 3-kinase (Alessi et al., 1996; Franke et al., 1995, 1997; Bos, 1995). PKB/Akt has been shown to assume at least two functions in tissue homeostasis (Bos, 1995; Franke et al., 1997). First, it is involved in mitogenic signaling and regulation of cellular kinases such as glycogen H2B synthase kinase 3 (Cross et al., 1995). When glycogen synthase kinase 3 is inactivated by phosphorylation, anabolic pathways are upregulated (Welsh et al., 1996). Another kinase associated with growth factor stimulation, p70S6-kinase, is activated by constitutively active mutants of PKB/Akt, but is not a direct substrate (Burgering and Coer, 1995). p70S6-kinase activation has also been observed as a consequence of polyomavirus infection (Talmage et al., 1988; Dahl et al., 1996). The involvement of PKB/Akt in growth regulation is further supported by data obtained with an oncogenic form of PKB/Akt expressed as a Gag-PKB/Akt fusion protein by the acute retrovirus AKT8 (Bellacosa et al., 1991). The viral Gag portion leads to membrane Figure 3 PI 3-kinase activity is required for PKB activation. Cells association and constitutive activation of this mutant were transfected as described. In lane 3, 30 ng/ml PDGF BB were kinase. added for 15 min. In lanes 4 ± 6, the inhibitor was added for 15 min The second role of PKB/Akt is to block apoptosis, prior to cell lysis at a concentration of 15 mM. Cells in lane 6 were as e.g. demonstrated in cells overexpressing PKB/Akt treated with 30 ng/ml PDGF BB for 5 min before the inhibitor was (Dudek et al., 1997; Kennedy et al., 1997; Kulik et al., added. In lanes 8,9, cells were cotransfected with a wt (p110) and an enzymatically inactive mutant form of the p110 subunit of PI 3- 1997). In agreement with this concept, dominant- kinase (p110K-), respectively. The bar graph at the bottom shows negative alleles of PKB/Akt block survival of starved the relative activities in one representative experiment PKB/Akt activation by middle-T RMeiliet al 906 (Didichenko et al., 1996) and the antibody used to detect phosphatase 2A was raised against a 20 amino acid peptide derived from the amino terminus of the catalytic subunit of phosphatase 2A (B Hemmings, FMI).

— mT wt — mT wt Vanadyl hydroperoxide Plasmids The construction of the plasmid encoding a tagged form of HA PKB PKB/Akt (HA-tagged human PKB a) under the control of ∆PH a CMV promoter has been reported before (Alessi et al., 1996). DPH-PKB/Akt was subcloned from the correspond- ing pECE vector HA-DNRac a (Andjelkovic et al., 1996) into the same CMV based vector used throughout this study. Plasmids encoding the wild type (pcDNA3 HA- p110) and kinase inactive (pcDNA3 HA-p110 R916P) catalytic subunit of PI 3-kinase were obtained from M H2B Thelen, Bern (Didichenko et al., 1996).

Tissue culture NIH3T3 cells were maintained in Dulbecco's modi®ed Eagle medium (DMEM) supplemented with 10% calf serum (CS) (GIBCO BRL).

Transient transfection Transient transfections were performed with Lipofectamine (GIBCO BRL). 12 mg of plasmid DNA in 1 ml DMEM were mixed with 50 ml Lipofectamine in 1 ml DMEM for 15 min at room temperature. This mixture plus an additional 2 ml DMEM were added to one million cells in a 10 cm dish. Cells were incubated for 4 h and then the Figure 4 Activation of PKB/Akt by middle-T requires the PH- medium was changed to 10% CS DMEM for 12 h. domain. A PKB/Akt mutant lacking the PH-domain was transfected together with middle-T into NIH3T3 cells. This Immunoprecipitation mutant was only activated when stimulated with vanadyl hydroperoxide (lane 5). The bar graph shows relative activities A 10 cm dish was lysed on ice with 500 mllysisbuer from one representative experiment (25 mM Tris pH 7.6 at RT, 100 mM NaCl, 10 mM NaF, 1mM Vanadate, 25 mM b-glycerophosphate, 3 mM sodium pyrophosphate, 1 mM EDTA, 1 mM EGTA, 2 mg/ml leupeptin, 2 mg/ml aprotinin, 1% NP40, 10% Glycerol, cells treated with survival factors such as IGF-1 or IL- 1mM DTT). The lysate was precleared by adding 20 ml 2 (Ahmed et al., 1997; Dudek et al., 1997; Kulik et al., Pansorbin (Calbiochem) followed by centrifugation. 50 ml 1997). It has also been shown that PKB/Akt promotes of a 1 : 1 slurry of Protein A beads in lysis buer together with 1 mgantibodywereaddedto200ml supernatant and survival in cells destined to enter apoptosis upon rocked on ice for 1 h. The beads were washed 26 with overexpression of Myc (Kaumann-Zeh et al., 1997). lysis buer and 26 with kinase buer (25 mM MOPS Based on these data one might speculate that PKB/Akt pH 7.4 at RT, 25 mM b-glycerophosphate, 20 mM magne- also plays a role in the lytic cycle of many viruses sium chloride, 1 mM DTT). where inadequate growth stimulation and viral DNA replication have been shown to eciently trigger In vitro kinase assay intracellular defense mechanisms like G1 cell cycle arrest and apoptosis (reviewed in Teodoro and Pab 762 middle-T immunoprecipitates collected on beads Branton, 1997). One of the mediators of apoptosis is were incubated with 75 ml kinase buer containing 1 mCi 32 p53 and several viruses produce proteins, such as the [g- P]ATP for 10 min at 308C. The reaction was stopped by the addition of 25 ml46sample buer and boiled for papovavirus large-T antigens, that are capable to 5 min. PKB/Akt activity was measured in immunecomplex subvert the function of p53. Since there is no evidence kinase assays following immunoprecipitation with the anti- that polyomavirus large-T directly interacts with p53, it HA-tag antibody. The beads were incubated with 75 ml is tempting to speculate that in this virus middle-T kinase buer containing 5 mCi [g-32P]ATP, 5 mM cold ATP ful®lls the role of an anti-apoptotic protein. and 5 mg histone 2B as substrate. The reaction was stopped after 10 min by the addition of 25 ml46sample buer and boiled for 5 min. The samples were separated by SDS ± PAGE (12.5%), blotted to a PVDF membrane (Millipore) and exposed to a phosphorimager cassette (Molecular Materials and methods Dynamics) for quantitation of the radioactivity incorpo- rated into histone 2B. PKB/Akt was detected by Western blotting with the anti-HA-tag antibody and quantitated Antibodies after scanning. To calculate the relative activity of PKB/ Mouse monoclonal anti-middle-T Pab 762 was kindly Akt, the quanti®ed kinase activities were divided by the provided by S Dilworth, London (Dilworth and Horner, amount of PKB/Akt determined using Western analysis 1993), antibodies against p85 were from Dr M Thelen and normalized to the value for unstimulated cells. PKB/Akt activation by middle-T RMeiliet al 907 Western analysis or the chemiluminescent substrate CDP-Star manufactured by TROPIX. PVDF membranes were immunodecorated following standard Western blotting protocols. For detection of the primary antibody either an alkaline phosphatase-linked Acknowledgements secondary antibody (Southern Biotechnology Associates) We wish to thank Mirjana Andjelkovic, Nicolas R Dunant, or an alkaline phosphatase protein A conjugate (Calbio- Nicholas Pullen and Mahmoud El-Shemerly for helpful chem) was used. The membranes were either developed discussions and for their advice on this manuscript. We using a colorimetric assay with 5-bromo-4-chloro-3-indolyl would also like to thank Drs Dilworth and Thelen for phosphate/nitro blue tetrazolium (Boehringer Mannheim) supplying reagents.

References

Ahmed NN, Grimes HL, Bellacosa A, Chan TO and Tsichlis Kennedy SG, Wagner AJ, Conzen SD, Jordan J, Bellacosa PN. (1997). Proc. Natl. Acad. Sci. USA, 94, 3627 ± 3632. A, Tsichlis PN and Hay N. (1997). Genes Dev., 11, 701 ± Alessi DR, Andjelkovic M, Caudwell BF, Cron P, Morrice 713. N, Cohen P and Hemmings BA. (1996). EMBO J., 15, Kiefer F, Courtneidge SA and Wagner EF. (1994). Adv. 6541 ± 6551. Cancer Res., 64, 125 ± 157. Andjelkovic M, Jakubowicz T, Cron P, Ming XF, Han JW Klarlund JK, Guilherme A, Holik JJ, Virbasius JV, Chawla and Hemmings BA. (1996). Proc.Natl.Acad.Sci.USA, A and Czech MP. (1997). Science, 275, 1927 ± 1930. 93, 5699 ± 5704. Kulik G, Klippel A and Weber MJ. (1997). Mol. Cell. Biol., Bellacosa A, Testa JR, Staal SP and Tsichlis PN. (1991). 17, 1595 ± 1606. Science, 254, 274 ± 277. Ling LE, Druker BJ, Cantley LC and Roberts TM. (1992). J. Bos JL. (1995). Trends Biochem. Sci., 20, 441 ± 442. Virol., 66, 1702 ± 1708. Burgering BM and Coer PJ. (1995). Nature, 376, 599 ± 602. Pallas DC, Shahrik LK, Martin BL, Jaspers S, Miller TB, CampbellKS,OgrisE,BurkeB,SuW,AugerKR,Druker Brautigan DL and Roberts TM. (1990). Cell, 60, 167 ± 176. BJ, Schahausen BS, Roberts TM and Pallas DC. (1994). Rameh LE, Chen CS and Cantley LC. (1995). Cell, 83, 821 ± Proc. Natl. Acad. Sci. USA, 91, 6344 ± 6348. 830. Courtneidge SA, Read M, Wilson JB and Fried M. (1989). Rassoulzadegan M, Cowie A, Carr A, Glaichenhaus N, Oncogene Res., 4, 75 ± 80. Kamen R and Cuzin F. (1982). Nature, 300, 713 ± 718. Courtneidge SA and Smith AE. (1983). Nature, 303, 435 ± Robinson MJ and Cobb MH. (1997). Curr. Opin. Cell. Biol., 439. 9, 180 ± 186. Cross DA, Alessi DR, Cohen P, Andjelkovic M and Su W, Liu W, Schahausen BS and Roberts TM. (1995). J. Hemmings BA. (1995). Nature, 378, 785 ± 789. Biol. Chem., 270, 12331 ± 12334. Dahl J, Freund R, Blenis J and Benjamin TL. (1996). Mol. Talmage DA, Blenis J and Benjamin TL. (1988). Mol. Cell. Cell. Biol., 16, 2728 ± 2735. Biol., 8, 2309 ± 2315. Didichenko SA, Tilton B, Hemmings BA, Ballmer-Hofer K Talmage DA, Freund R, Young AT, Dahl J, Dawe CJ and and Thelen M. (1996). Curr. Biol., 6, 1271 ± 1278. Benjamin TL. (1989). Cell, 59, 55 ± 65. Dilworth SM, Brewster CE, Jones MD, Lanfrancone L, Teodoro JG and Branton PE. (1997). J. Virol., 71, 1739 ± Pelicci G and Pelicci PG. (1994). Nature, 367, 87 ± 90. 1746. Dilworth SM and Horner VP. (1993). J. Virol., 67, 2235 ± Toker A and Cantley LC. (1997). Nature, 387, 673 ± 676. 2244. Treisman R, Novak U, Favaloro J and Kamen R. (1981). Dudek H, Datta SR, Franke TF, Birnbaum MJ, Yao R, Nature, 292, 595 ± 600. Cooper GM, Segal RA, Kaplan DR and Greenberg ME. Urich M, Senften M, Shaw P and Ballmer-Hofer K. (1997). (1997). Science, 275, 661 ± 665. Oncogene, 14, 1235 ± 1241. Franke TF, Kaplan DR, Cantley LC and Toker A. (1997). Walter G, Ruediger R, Slaughter C and Mumby M. (1990). Science, 275, 665 ± 668. Proc.Natl.Acad.Sci.USA,87, 2521 ± 2525. Franke TF, Yang SI, Chan TO, Datta K, Kazlauskas A, Welsh GI, Wilson C and Proud CG. (1996). Trends Cell. Morrison DK, Kaplan DR and Tsichlis PN. (1995). Cell, Biol., 6, 274 ± 279. 81, 727 ± 736. Whitman M, Kaplan DR, Schahausen BS, Cantley L and Kaumann-Zeh A, Rodriguez-Viciana P, Ullrich E, Gilbert Roberts TM. (1985). Nature, 315, 239 ± 242. C, Coer P, Downward J and Evan G. (1997). Nature, 385, 544 ± 548.