Portal de Periódicos da CAPES

FOXO4 Is Acetylated upon Peroxide Stress and Deacetylated by the Longevity Protein hSir2

2004; Elsevier BV; Volume: 279; Issue: 28 Linguagem: Inglês

10.1074/jbc.m401138200

1083-351X

Armando van der Horst, Leon G.J. Tertoolen, Lydia M. M. de Vries-Smits, Roy A. Frye, René H. Medema, Boudewijn Burgering,

Sirtuins and Resveratrol in Medicine

FOXO transcription factors have important roles in metabolism, cellular proliferation, stress tolerance, and aging. FOXOs are negatively regulated by protein kinase B/c-Akt-mediated phosphorylation. Here we show that FOXO factors are also subject to regulation by reversible acetylation. We provide evidence that the acetyltransferase CREB-binding protein (CBP) binds FOXO resulting in acetylation of FOXO. This acetylation inhibits FOXO transcriptional activity. Binding of CBP and acetylation are induced after treatment of cells with peroxide stress. Deacetylation of FOXOs involves binding of the NAD-dependent deacetylase hSir2SIRT1. Accordingly, hSir2SIRT1-mediated deacetylation precludes FOXO inhibition through acetylation and thereby prolongs FOXO-dependent transcription of stress-regulating genes. These data demonstrate that acetylation functions in a second pathway of negative control for FOXO factors and provides a novel mechanism whereby hSir2SIRT1 can promote cellular survival and increase lifespan. FOXO transcription factors have important roles in metabolism, cellular proliferation, stress tolerance, and aging. FOXOs are negatively regulated by protein kinase B/c-Akt-mediated phosphorylation. Here we show that FOXO factors are also subject to regulation by reversible acetylation. We provide evidence that the acetyltransferase CREB-binding protein (CBP) binds FOXO resulting in acetylation of FOXO. This acetylation inhibits FOXO transcriptional activity. Binding of CBP and acetylation are induced after treatment of cells with peroxide stress. Deacetylation of FOXOs involves binding of the NAD-dependent deacetylase hSir2SIRT1. Accordingly, hSir2SIRT1-mediated deacetylation precludes FOXO inhibition through acetylation and thereby prolongs FOXO-dependent transcription of stress-regulating genes. These data demonstrate that acetylation functions in a second pathway of negative control for FOXO factors and provides a novel mechanism whereby hSir2SIRT1 can promote cellular survival and increase lifespan. The Forkhead box, class O subfamily of forkhead transcription factors (FOXO) 1The abbreviations used are: FOXO, Forkhead box, class O subfamily of forkhead transcription factors; Nam, nicotinamide; EGF, epidermal growth factor; 4-OHT, 4-hydroxy-tamoxifen; CREB, cAMP-response element-binding protein; CBP, CREB-binding protein; MnSOD, manganese superoxide dismutase; HA, hemagglutinin; PKB, protein kinase B; GAPDH, glyceraldehyde-3-phosphate dehydrogenase; siRNA, short interfering RNA; DBE, Daf-16 family protein binding element; RIPA, radioimmune precipitation assay buffer. consists of the functionally related proteins FOXO1, FOXO3a, and FOXO4 (also known as FKHR, FKHRL1, and AFX, respectively; Ref. 1Kaestner K.H. Knochel W. Martinez D.E. Genes Dev. 2000; 14: 142-146PubMed Google Scholar). The growth factor-stimulated phosphatidylinositol 3-kinase-protein kinase B (PKB)/c-Akt pathway negatively regulates FOXO factors by phosphorylation-mediated nuclear exclusion (2Brunet A. Bonni A. Zigmond M.J. Lin M.Z. Juo P. Hu L.S. Anderson M.J. Arden K.C. Blenis J. Greenberg M.E. Cell. 1999; 96: 857-868Abstract Full Text Full Text PDF PubMed Scopus (5454) Google Scholar, 3Kops G.J. de Ruiter N.D. De Vries-Smits A.M. Powell D.R. Bos J.L. Burgering B.M. Nature. 1999; 398: 630-634Crossref PubMed Scopus (953) Google Scholar, 4Tang E.D. Nunez G. Barr F.G. Guan K.L. J. Biol. Chem. 1999; 274: 16741-16746Abstract Full Text Full Text PDF PubMed Scopus (663) Google Scholar). This pathway is evolutionarily conserved between Caenorhabditis elegans and humans. DAF-16, the C. elegans homologue of mammalian FOXO, is also controlled by phosphatidylinositol 3-kinase/PKB signaling. DAF-16 regulates daver formation in larvae, and responses to various environmental stresses and longevity in adult worms (5Gottlieb S. Ruvkun G. Genetics. 1994; 137: 107-120Crossref PubMed Google Scholar, 6Kenyon C. Chang J. Gensch E. Rudner A. Tabtiang R. Nature. 1993; 366: 461-464Crossref PubMed Scopus (2523) Google Scholar, 7Murakami S. Johnson T.E. Genetics. 1996; 143: 1207-1218Crossref PubMed Google Scholar, 8Honda Y. Honda S. FASEB J. 1999; 13: 1385-1393Crossref PubMed Scopus (599) Google Scholar). In parallel, mammalian FOXO transcription factors have been implicated in regulating metabolism, cell cycle progression, and stress tolerance (9Tran H. Brunet A. Grenier J.M. Datta S.R. Fornace Jr., A.J. DiStefano P.S. Chiang L.W. Greenberg M.E. Science. 2002; 296: 530-534Crossref PubMed Scopus (711) Google Scholar, 10Kops G.J. Dansen T.B. Polderman P.E. Saarloos I. Wirtz K.W. Coffer P.J. Huang T.T. Bos J.L. Medema R.H. Burgering B.M. Nature. 2002; 419: 316-321Crossref PubMed Scopus (1282) Google Scholar; reviewed in Ref. 11Burgering B.M. Kops G.J. Trends Biochem. Sci. 2002; 27: 352-360Abstract Full Text Full Text PDF PubMed Scopus (624) Google Scholar). In C. elegans, overexpression of the NAD-dependent deacetylase Sir2 (silent information regulator 2) increases lifespan, which requires DAF-16 (12Tissenbaum H.A. Guarente L. Nature. 2001; 410: 227-230Crossref PubMed Scopus (1590) Google Scholar). The Sir2 family of genes is a highly conserved group of genes with seven human homologues, of which the SIRT1 gene encodes the closest homologue of yeast and C. elegans Sir2, hence named hSir2SIRT1 (13Frye R.A. Biochem. Biophys. Res. Commun. 2000; 273: 793-798Crossref PubMed Scopus (1174) Google Scholar). Recently, deacetylation of p53 by hSir2SIRT1 has been demonstrated, and it has been suggested that this functions in increasing cellular resistance against stress. However, subsequent studies (14Howitz K.T. Bitterman K.J. Cohen H.Y. Lamming D.W. Lavu S. Wood J.G. Zipkin R.E. Chung P. Kisielewski A. Zhang L.L. Scherer B. Sinclair D.A. Nature. 2003; 425: 191-196Crossref PubMed Scopus (3202) Google Scholar) showed that in HEK293T cells, which lack functional p53, activation of hSir2SIRT1 by resveratrol treatment still increases cellular resistance against gamma-radiation, thus suggesting alternative pathways. As DAF-16 is necessary for lifespan extension by Sir2, FOXOs may well function in such an alternative pathway. Cell Culture, Transfection, and Treatment—HEK293T, DL23 (DLD-1 human colon carcinoma cells expressing a conditionally active version of FOXO3a; Ref. 10Kops G.J. Dansen T.B. Polderman P.E. Saarloos I. Wirtz K.W. Coffer P.J. Huang T.T. Bos J.L. Medema R.H. Burgering B.M. Nature. 2002; 419: 316-321Crossref PubMed Scopus (1282) Google Scholar), and A14 cells (human insulin receptor overexpressing mouse NIH3T3 cells (15Burgering B.M. Coffer P.J. Nature. 1995; 376: 599-602Crossref PubMed Scopus (1884) Google Scholar) and C2C12 mouse myoblasts were maintained in Dulbecco's modified Eagle's medium supplemented with glutamine, penicillin/streptomycin, and 10% fetal bovine serum. In all experiments, cells were cultured in the presence of 10% fetal bovine serum, unless stated otherwise. HEK293T cells were transiently transfected using FuGENE6 reagent according to the manufacturer's suggestions (Roche Applied Science). A14 cells were transfected using the calcium phosphate method. Total amounts of transfected DNA were equalized using pBluescript II KS(+). 1 μm trichostatin A (ICN), 5 mm nicotinamide (Nam, Supelco), 1 μg/ml insulin, 20 ng/ml epidermal growth factor (EGF), 500 nm 4-hydroxy-tamoxifen (4-OHT), and 20–500 μm hydrogen peroxide (H2O2, Merck) were added as indicated. All experiments were performed at least three times and representative results are shown. Plasmids, Oligonucleotides, and Recombinant Proteins—pMT2-HA-FOXO4, pRP261-GST-FOXO4-DB (3Kops G.J. de Ruiter N.D. De Vries-Smits A.M. Powell D.R. Bos J.L. Burgering B.M. Nature. 1999; 398: 630-634Crossref PubMed Scopus (953) Google Scholar), pcDNA3-HA-FOXO3a.A3 (10Kops G.J. Dansen T.B. Polderman P.E. Saarloos I. Wirtz K.W. Coffer P.J. Huang T.T. Bos J.L. Medema R.H. Burgering B.M. Nature. 2002; 419: 316-321Crossref PubMed Scopus (1282) Google Scholar), pcDNA3.1-hSir2SIRT1 (16Vaziri H. Dessain S.K. Ng Eaton E. Imai S.I. Frye R.A. Pandita T.K. Guarente L. Weinberg R.A. Cell. 2001; 107: 149-159Abstract Full Text Full Text PDF PubMed Scopus (2316) Google Scholar), pBabe-puro (17Medema R.H. Herrera R.E. Lam F. Weinberg R.A. Proc. Natl. Acad. Sci. U. S. A. 1995; 92: 6289-6293Crossref PubMed Scopus (411) Google Scholar), 6× DBE (18Furuyama T. Nakazawa T. Nakano I. Mori N. Biochem. J. 2000; 349: 629-634Crossref PubMed Scopus (558) Google Scholar), and p27GL-1609 (19Medema R.H. Kops G.J. Bos J.L. Burgering B.M. Nature. 2000; 404: 782-787Crossref PubMed Scopus (1231) Google Scholar) have been described before. pcDNA3.1-myc-FOXO4 was created by ligating a Klenow-blunted SalI/NotI fragment from pMT2-HA-FOXO4 into Klenow-blunted BamHI/NotI digested pcDNA3.1-myc. GLOFLAG3-FLAG-FOXO4 was created by ligating a Klenow-blunted SalI/NotI fragment from pMT2-HA-FOXO4 into mungbean-blunted BamHI/NotI-digested GLOFLAG3. pcDNA3.1-myc-hSir2SIRT1 and pRP265-GST-hSir2SIRT1 were created by ligating a BamHI/XbaI fragment from pcDNA3.1-hSir2SIRT1 into BamHI/XbaI-digested pcDNA3.1-myc and pRP265, respectively. pRc/RSV-CBP-HA and pRP265-GST-p300-HAT were kindly provided by R. Giles (20Chrivia J.C. Kwok R.P. Lamb N. Hagiwara M. Montminy M.R. Goodman R.H. Nature. 1993; 365: 855-859Crossref PubMed Scopus (1770) Google Scholar) and R. Vries (21Vries R.G. Prudenziati M. Zwartjes C. Verlaan M. Kalkhoven E. Zantema A. EMBO J. 2001; 20: 6095-6103Crossref PubMed Scopus (72) Google Scholar), respectively. Gal4-DBD-CBP-HA constructs were a kind gift of E. Kalkhoven (22Kalkhoven E. Teunissen H. Houweling A. Verrijzer C.P. Zantema A. Mol. Cell. Biol. 2002; 22: 1961-1970Crossref PubMed Scopus (87) Google Scholar). pRL-TK (Tk-Renilla luciferase) was purchased from Promega. siRNA duplexes to down-regulate hSir2SIRT1 expression (duplex 1, sense sequence, CAACUUGUACGACGAAGACdTdT; duplex 2, sense sequence, GCUGAUGAACCGCUUGCUAdTdT; duplex 3, sense sequence, CGUCUUAUCCUCUAGUUCUdTdT) were purchased from Dharmacon RNA Technologies and cotransfected using OligofectAMINE according to the manufacturer (Invitrogen). GST-FOXO4-DB, GST-p300-HAT, and GST-tagged hSir2SIRT1 were purified from bacteria using a standard GST-fusion protein-purification protocol (23O'Brien R.M. Noisin E.L. Suwanichkul A. Yamasaki T. Lucas P.C. Wang J.C. Powell D.R. Granner D.K. Mol. Cell. Biol. 1995; 15: 1747-1758Crossref PubMed Google Scholar). Antibodies—Monoclonal 12CA5 and 9E10 antibodies were produced using hybridoma cell lines. Monoclonal antibodies recognizing the FLAG-M2 epitope, p27kip1, and GAPDH were obtained from Sigma, Transduction Laboratories, and Chemicon, respectively. Polyclonal antibodies recognizing acetylated lysine residues (α-AcLys), actin, CBP, Gal4-DBD, MnSOD, and pS473-PKB were obtained from Cell Signaling Technologies, Santa Cruz Biotechnology (3×), Stressgen BioReagents, and New England Biolabs, respectively. Polyclonal antibodies recognizing FOXO3a were purchased from Sigma and Santa Cruz Biotechnology. Polyclonal antibodies recognizing FOXO4 (19Medema R.H. Kops G.J. Bos J.L. Burgering B.M. Nature. 2000; 404: 782-787Crossref PubMed Scopus (1231) Google Scholar) and PKB (15Burgering B.M. Coffer P.J. Nature. 1995; 376: 599-602Crossref PubMed Scopus (1884) Google Scholar) have been described before. Polyclonal antibody recognizing hSir2SIRT1 was raised in a rabbit using an N-terminally 6His-tagged fragment consisting of amino acids 506–747 of hSir2SIRT1. Western Blot Analysis—Protein samples were separated by SDS-PAGE and transferred to polyvinylidene difluoride membrane (Immobilon). Western blot analysis was performed under standard conditions using the indicated antibodies. FOXO Activity—To determine the expression of endogenous p27kip1 and MnSOD, HEK293T cells were transfected with empty vector or HA-FOXO4 together with pBabe-puro. Either constructs encoding Gal4-DBD-CBP-HA or myc-hSirSIRT1 were cotransfected or cells were treated using hydrogen peroxide or Nam for twenty-four hours. Twenty-four hours before harvest, 1 μg/ml puromycin was added to the cultures to select for transfected cells. Cells were lysed in 1× Laemmli sample buffer, and samples were analyzed by Western blot analysis. If indicated, DL23 cells treated for twenty-four hours with 4-OHT and/or Nam were used. For luciferase assays, either A14 cells transiently transfected with a reporter construct bearing six canonical FOXO binding sites (6× DBE-luciferase) or a reporter cell line (A14-p27luc) stably expressing the human p27kip1-promoter linked to luciferase (p27GL-1609) were used. Cells were treated as above. Luciferase counts were normalized using Tk-Renilla-luciferase. Samples were analyzed according to the manufacturer's instructions (Promega). To control for effects on basal transcription, the fold induction of luciferase expression was determined by dividing luciferase ratios from samples transfected with FOXO4 by the ratios from parallel samples transfected with empty vector. Statistical analysis was performed using a Student's t test. p < 0.05 was considered significant. Flow cytometric analysis to determine the percentage of cells in the G1-phase or S-phase of the cell cycle was essentially done as described (3Kops G.J. de Ruiter N.D. De Vries-Smits A.M. Powell D.R. Bos J.L. Burgering B.M. Nature. 1999; 398: 630-634Crossref PubMed Scopus (953) Google Scholar). Cells were either transfected with HA-FOXO3a.A3 and hSir2SIRT1 or treated with 4-OHT (500 nm) and/or Nam (5 mm). Statistical analysis was performed using a Student's t test. p < 0.05 was considered significant. For the measurement of reactive oxygen species, DL23 and A14 cells were loaded with 10 μg/ml CM-H2DCFDA probe (Molecular Probes) in HEPES buffered saline for 5 min at 37 °C. Cells were incubated in Dulbecco's modified Eagle's medium containing 10 mm HEPES, 10 mm bicarbonate, pH 7.2, without phenol red at 37 °C and treated with 150 μm H2O2. Fluorescence was collected with a Leica DMR microscope. Excitation was done with a monochromator (SPEX Industries) at 490 nm (slit = 8 nm), >515 nm dichroic mirror, and a >515 nm emission filter. Ultra-low light conditions were applied only during image collection, under which no photo-activation of the cellular H2DCFDA could be observed. Images were collected by integration of 20 frames at 10-s interval (fixed gain, intensifier/camera/grabber, during all experiments). Off-line analysis was done by quantification of the average pixel intensities of at least five individual cells/image during time in at least two separate experiments. Statistical analysis was performed using a Student's t test. p < 0.05 was considered significant. Co-immunoprecipitation—For exogenous proteins, HEK293T cells were transfected with the indicated constructs. For CBP/FOXO4 coimmunoprecipitations, cells were left untreated or were treated with insulin plus EGF or hydrogen peroxide only for 30 min (see Fig. 4A) or 60 min (see Fig. 4C) or with insulin plus EGF for 30 min before adding hydrogen peroxide for 30 min (see Fig. 4A, as indicated). For hSir2SIRT1/FOXO4 co-immunoprecipitations, cells were treated with Nam (5 mm) or hydrogen peroxide (200 μm) as indicated. Forty hours after transfection, cells were lysed in RIPA buffer. Immunoprecipitation samples were analyzed by Western blot analysis. For endogenous proteins, C2C12 or HEK293T cells were treated using 500 μm H2O2 for 1 h and lysed in lysis buffer (50 mm Tris-HCl, pH 7.5, 1% Triton X-100, 5% glycerol, 5 mm EDTA, 100 mm NaCl, protease inhibitors). Immunoprecipitations were carried out essentially as described for exogenous proteins. Acetylation—HEK293T cells were transfected with the indicated constructs. Cells were treated with Nam for 1 h or with hydrogen peroxide for the indicated times. Forty hours after transfection, cells were lysed in acetylation buffer (50 mm Tris-HCl, pH 7.5, 0.4% Triton X-100, 5 mm EDTA, 100 mm NaCl, protease inhibitors) or RIPA buffer. FOXO4 immunoprecipitation samples were analyzed by Western blot analysis. In Vitro and in Vivo Deacetylation—For in vitro deacetylation, 2.5 μg GST-FOXO4-DB were incubated with 0.05 μCi [14C]-acetyl-CoA (ICN) and 2.0 μg GST-p300-HAT in reaction buffer (50 mm Tris-HCl, pH 8.0, 5 mm EDTA, pH 8.0, 0.5% Nonidet P-40, 150 mm NaCl (21Vries R.G. Prudenziati M. Zwartjes C. Verlaan M. Kalkhoven E. Zantema A. EMBO J. 2001; 20: 6095-6103Crossref PubMed Scopus (72) Google Scholar) for 45 min at 30 °C. Subsequently, deacetylation buffer (50 mm Tris-HCl, pH 8.8, 4 mm MgCl2, 0.2 mm dithiothreitol, 2 μm trichostatin A plus 1 mm NAD+ if indicated (24Luo J. Nikolaev A.Y. Imai S. Chen D. Su F. Shiloh A. Guarente L. Gu W. Cell. 2001; 107: 137-148Abstract Full Text Full Text PDF PubMed Scopus (1902) Google Scholar) and 2.0 μg of GST-tagged wt-hSir2SIRT1 were added to a final volume of 100 μl. Samples were incubated for another 90 min at 37 °C, and the reaction was terminated using 5× Laemmli sample buffer. Samples were separated by SDS-PAGE and analyzed by autoradiography. For in vivo deacetylation, HEK293T cells were transfected with the indicated constructs. Forty hours posttransfection cells were lysed in RIPA buffer, and myc-FOXO4 was immunoprecipitated from cleared lysates using 9E10 antibody. Immunoprecipitates and total lysates were subjected to Western blotting. Nicotinamide Inhibits FOXO Transcriptional Activity—To investigate the possibility that hSir2SIRT1 controls FOXO function, we analyzed the effect of Nam, an inhibitor of the Sir2 family of deacetylases (25Bitterman K.J. Anderson R.M. Cohen H.Y. Latorre-Esteves M. Sinclair D.A. J. Biol. Chem. 2002; 277: 45099-45107Abstract Full Text Full Text PDF PubMed Scopus (822) Google Scholar), on FOXO4 transcriptional activity. Nam inhibited FOXO4-mediated transcription of a luciferase reporter construct bearing six canonical FOXO DNA-binding sites (Fig. 1A; Ref. 18Furuyama T. Nakazawa T. Nakano I. Mori N. Biochem. J. 2000; 349: 629-634Crossref PubMed Scopus (558) Google Scholar). Previously, we have shown that FOXO transcriptionally regulates the expression of the cell-cycle inhibitor p27kip1 (19Medema R.H. Kops G.J. Bos J.L. Burgering B.M. Nature. 2000; 404: 782-787Crossref PubMed Scopus (1231) Google Scholar). In support of this, expression of FOXO4 in HEK293T cells increased p27kip1 expression (Fig. 1B). Importantly, Nam inhibited the expression of the endogenous p27kip1 gene induced by FOXO4 (Fig. 1B). Inhibition of p27kip1 expression by Nam in these cells was not due to PKB activation, as Nam treatment under these conditions did not change phosphorylation of PKB on Ser-473, which is indicative of PKB activity. Also, in A14 cells stably expressing the luciferase gene driven by the p27kip1 promoter (A14-p27luc), FOXO4-induced luciferase expression was reduced by Nam (data not shown). These data demonstrate that the Sir2-inhibitor Nam inhibits FOXO4 transcriptional activity, possibly via hSir2SIRT1. Next we investigated whether Nam also affects the biological activity of FOXO transcription factors. Previously, we have employed DL23 cells, DLD-1 human colon carcinoma cells expressing an inducible conditionally active version of FOXO3a and lacking functional p53, and demonstrated in these cells transcriptional regulation of p27kip1 expression and induction ofaG1 cell-cycle arrest by FOXO (10Kops G.J. Dansen T.B. Polderman P.E. Saarloos I. Wirtz K.W. Coffer P.J. Huang T.T. Bos J.L. Medema R.H. Burgering B.M. Nature. 2002; 419: 316-321Crossref PubMed Scopus (1282) Google Scholar). Treatment of DL23 cells with 4-OHT induced p27kip1 expression (Fig. 1C, left panel). This induction was reduced by treatment of these cells with Nam (Fig. 1C, left panel). Furthermore, FOXO3a activation imposed a cell-cycle arrest, which was partially relieved by treatment with Nam, in keeping with its effect on p27kip1 expression (Fig. 1C, right panel). In addition to imposing a cell-cycle arrest, FOXO activation increases cellular resistance against oxidative stress by transcriptional regulation of the gene encoding manganese superoxide dismutase (MnSOD) (10Kops G.J. Dansen T.B. Polderman P.E. Saarloos I. Wirtz K.W. Coffer P.J. Huang T.T. Bos J.L. Medema R.H. Burgering B.M. Nature. 2002; 419: 316-321Crossref PubMed Scopus (1282) Google Scholar). Activation of FOXO in DL23 cells by 4-OHT treatment increased MnSOD expression, and, similar to FOXO-induced p27kip1 expression, MnSOD levels were reduced in cells treated with Nam (Fig. 1C, left panel). In keeping with MnSOD mediating FOXO-induced cellular resistance against oxidative stress, H2O2-induced oxidative stress was reduced in cells treated with 4-OHT, whereas Nam treatment inhibited the effect of FOXO3a (Fig. 1D, upper and lower left panels). In addition, A14 cells expressing active FOXO3a after retroviral transduction displayed increased resistance against H2O2-induced oxidative stress, and this was also reversed by Nam treatment (Fig. 1D, lower right panel). These results show that inhibition of Sir2 function by Nam treatment inhibits FOXO-induced gene expression of p27kip1 and MnSOD and, consequently, FOXO-induced cell-cycle arrest and resistance against oxidative stress. hSir2SIRT1 Stimulates FOXO Transcriptional Activity—Next we determined whether the deacetylase inhibitor Nam indeed exerts its inhibitory effect on FOXO-mediated transcription through hSir2SIRT1. We first tested the effect of hSir2SIRT1 expression on FOXO4 activity. To this end, hSir2SIRT1 and FOXO4 were transfected into HEK293T cells, and the expression of endogenous p27kip1 and MnSOD was determined. Both genes were up-regulated by hSir2SIRT1 in a FOXO-dependent manner (Fig. 2A). Also, in A14-p27luc cells, FOXO4-induced luciferase expression was stimulated by co-expression of hSir2SIRT1 (data not shown). To further demonstrate involvement of hSir2SIRT1 in regulating FOXO4 transcriptional activity, we used siRNA to hSir2SIRT1, which resulted in reduced levels of endogenous hSir2SIRT1 (Fig. 2B). In keeping with the inhibitory effect of Nam on FOXO activity, knock-down of hSir2SIRT1 impaired FOXO4-induced p27kip1 expression. Expression of MnSOD, another transcriptional target of FOXOs (10Kops G.J. Dansen T.B. Polderman P.E. Saarloos I. Wirtz K.W. Coffer P.J. Huang T.T. Bos J.L. Medema R.H. Burgering B.M. Nature. 2002; 419: 316-321Crossref PubMed Scopus (1282) Google Scholar), was also reduced by siRNA to hSir2SIRT1. As regulation of p27kip1 is important in mediating a FOXO-induced cell-cycle arrest, we also analyzed the effect of hSir2SIRT1 on FOXO-induced cell-cycle arrest. In keeping with the effect of Nam (see Fig. 1C), hSir2SIRT1 increased the percentage of cells arrested in G1 (Fig. 2C). Thus, these data indicate that hSir2SIRT1 enhances the transcriptional activity of FOXO and thereby modulates its biological activity. CBP Acetylates FOXO and Inhibits FOXO Transcriptional Activity—As Sir2 deacetylases are generally thought to act as repressors of transcription via deacetylating histones (26Moazed D. Mol. Cell. 2001; 8: 489-498Abstract Full Text Full Text PDF PubMed Scopus (238) Google Scholar) we hypothesized that hSir2SIRT1 directly deacetylates FOXO transcription factors. Acetylation of proteins is mediated by acetyltransferases such as the related proteins p300 and CBP; binding between p300/CBP and FOXO as well as FOXO acetylation have been reported (27So C.W. Cleary M.L. Mol. Cell. Biol. 2002; 22: 6542-6552Crossref PubMed Scopus (92) Google Scholar, 28Nasrin N. Ogg S. Cahill C.M. Biggs W. Nui S. Dore J. Calvo D. Shi Y. Ruvkun G. Alexander-Bridges M.C. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 10412-10417Crossref PubMed Scopus (141) Google Scholar, 29Mahmud D.L. Amlak M. Deb D.K. Platanias L.C. Uddin S. Wickrema A. Oncogene. 2002; 21: 1556-1562Crossref PubMed Scopus (94) Google Scholar). To confirm these findings, HA-CBP and FLAG-FOXO4 were co-expressed in HEK293T cells. Indeed FLAG-FOXO4 was found to bind to HA-CBP (Fig. 3A). More importantly, binding of CBP to FOXO4 induced acetylation of FOXO4 (Fig. 3B). Also, co-expression of CBP and FOXO3a resulted in acetylation of FOXO3a (data not shown), suggesting that CBP is able to mediate acetylation of other FOXO transcription factors. Similarly, co-expression of p300 and FOXO4 induced acetylation of FOXO4 (data not shown), indicating that CBP and p300 are functionally equivalent in FOXO4 acetylation. Because inhibiting hSir2SIRT1 activity by either Nam or siRNA-mediated knockdown of hSir2SIRT1 inhibited transcriptional activity of FOXO, we next determined the effect of CBP-mediated acetylation of FOXO4 on its transcriptional activity. In A14-p27luc cells expressing wt-CBP, FOXO4 activity was considerably impaired as compared with control cells expressing FOXO4 only (Fig. 3D). This repression was due to the acetyltransferase activity of CBP as co-expression of the CBP-mutant F1541A, which possesses reduced but not absent acetyltransferase activity (Fig. 3C; Ref. 30Martinez-Balbas M.A. Bannister A.J. Martin K. Haus-Seuffert P. Meisterernst M. Kouzarides T. EMBO J. 1998; 17: 2886-2893Crossref PubMed Scopus (228) Google Scholar), impaired FOXO4 transcriptional activity to a smaller extent. We also analyzed the effect of CBP expression on the endogenous FOXO target gene p27kip1. Consistent with the reporter assays, wt-CBP inhibited the induction of p27kip1 expression by FOXO4, whereas the F1541A mutant displayed a smaller effect because of residual activity (Fig. 3E). Taken together, these data indicate that CBP inhibits FOXO4 transcriptional activity by acetylation, in agreement with the observed inhibition of transcriptional activity of FOXO4 by the deacetylase inhibitor Nam. hSir2SIRT1 Deacetylates FOXO4 —As CBP and p300 can acetylate FOXO transcription factors, we tested whether hSir2SIRT1 could deacetylate FOXO4 in vitro. To this end, bacterially expressed FOXO4-DNA-binding domain (GST-FOXO4-DB) was in vitro acetylated by a GST-tagged p300 acetyltransferase domain in the presence of [14C]acetyl-CoA (21Vries R.G. Prudenziati M. Zwartjes C. Verlaan M. Kalkhoven E. Zantema A. EMBO J. 2001; 20: 6095-6103Crossref PubMed Scopus (72) Google Scholar). Subsequently, the reactions were incubated with GST-tagged hSir2SIRT1 in the absence or presence of the essential cofactor NAD+ (24Luo J. Nikolaev A.Y. Imai S. Chen D. Su F. Shiloh A. Guarente L. Gu W. Cell. 2001; 107: 137-148Abstract Full Text Full Text PDF PubMed Scopus (1902) Google Scholar). GST-hSir2SIRT1 completely deacetylated GST-FOXO4-DB in an NAD-dependent manner (Fig. 3F). The incomplete inhibition of deacetylase activity by omitting NAD+ was possibly due to traces of NAD+ in the GST-hSir2SIRT1 purification. Next, we tested deacetylation by hSir2SIRT1 in vivo. In HEK293T cells, a clear deacetylation of CBP-acetylated FOXO4 was observed upon cotransfection with hSir2SIRT1 (Fig. 3G). Taken together, these results show that hSir2SIRT1 can directly deacetylate FOXO4 both in vitro and in vivo and thereby stimulate FOXO transcriptional and biological activity. Hydrogen Peroxide Induces Acetylation of FOXO and Inhibits FOXO Activity—To investigate whether extracellular signaling could regulate acetylation/deacetylation of FOXO4, we tested the effects of insulin/insulin-like growth factor-1, a known regulator of FOXO function, and hydrogen peroxide. Hydrogen peroxide is known to promote acetylation of proteins like p53 and histones (31Ito A. Lai C.H. Zhao X. Saito S. Hamilton M.H. Appella E. Yao T.P. EMBO J. 2001; 20: 1331-1340Crossref PubMed Scopus (441) Google Scholar, 32Rahman I. Gilmour P.S. Jimenez L.A. MacNee W. Mol. Cell Biochem. 2002; 234–235: 239-248Crossref PubMed Scopus (283) Google Scholar), and FOXO transcription factors have been shown to be important in regulating stress responses (10Kops G.J. Dansen T.B. Polderman P.E. Saarloos I. Wirtz K.W. Coffer P.J. Huang T.T. Bos J.L. Medema R.H. Burgering B.M. Nature. 2002; 419: 316-321Crossref PubMed Scopus (1282) Google Scholar). First, we investigated whether hydrogen peroxide could affect binding of CBP to FOXO4. Therefore, HA-CBP was immunoprecipitated from HEK293T cells co-expressing myc-FOXO4. Increasing amounts of hydrogen peroxide enhanced the interaction between HA-CBP and myc-FOXO4 (Fig. 4A), suggesting that peroxide stress may induce acetylation of FOXO4 by CBP. Furthermore, activation of phosphatidylinositol 3-kinase/PKB signaling by treating cells with insulin and epidermal growth factor decreased basal binding between HA-CBP and myc-FOXO4. As CBP is localized exclusively within the nucleus, the decreased binding observed is likely due to the fact that phosphatidylinositol 3-kinase/PKB signaling induces relocalization of myc-FOXO4 to the cytosol. Importantly, activation of phosphatidylinositol 3-kinase/PKB signaling could not prevent the increase of HA-CBP/myc-FOXO4 complex formation induced by hydrogen peroxide treatment, indicating that the effect of peroxide is dominant over growth factor stimulation in this respect. Next, we determined whether endogenous FOXO and CBP could interact. We detected endogenous interaction between CBP and both FOXO4 and FOXO3a. Similar to the transient expression studies, H2O2 treatment increased the interaction between FOXO and CBP, although the magnitude differed (Fig. 4B). Finally, we tested whether enhanced binding of CBP to FOXO4 after treatment of cells with hydrogen peroxide resulted in increased acetylation of FOXO4. Thus, HEK293T cells transfected with a myc-FOXO4 construct were treated with different concentrations of peroxide and followed up for several periods of time. FOXO4 was immunoprecipitated and acetylation was determined by Western blotting. Hydrogen peroxide treatment indeed induced acetylation of FOXO4 (Fig. 4C). Also, treatment of cells cotransfected with FLAG-FOXO4 and HA-CBP with hydrogen peroxide resulted in increased acetylation in the fraction of FLAG-FOXO4 bound to HA-CBP (Fig. 4D). As noted, we detected acetylation by endogenous acetyltransferases only at later timepoints than would be suggested by the binding of CBP to FOXO4 (Fig. 4, A and B). As we used a pan-acetyllysine-antibody, we attribute this result to the low sensitivity of the antibody toward FOXO4, precluding detection of a low stoichiometry. This is further suggested by the observation that upon overexpression of CBP, we detected acetylation at shorter timepoints after peroxide treatment. Therefore, increased binding of FOXO4 to CBP after hydrogen peroxide treatment likely leads to increased acetylation. As hydrogen peroxide induces acetylation of FOXO4, we expected that peroxide would have the same effect as CBP or Nam on FOXO4 transcriptional activity. To study this, FOXO4 was transfected in HEK293T cells, and the expression of endogenous p27kip1 was determined upon treatment of these cells with various concentrations of hydrogen peroxide for twenty-four hours. Indeed, hydrogen peroxide inhibited p27kip1 expression (Fig. 4E). Inhibition of p27kip1 expression was not due to activation of PKB by hydrogen peroxide, as under these conditions we could not detect any changes in the Ser-473 phosphorylation of PKB. Taken together, these data indicate that hydrogen peroxide inhibits FOXO4 transcriptional activity through induction of FOXO4 acetylation. Hydrogen Peroxide Stimulates Binding of hSir2SIRT1 to FOXO—Finally, because acetylation of FOXO4 occurs in conjunction with binding of the acetyltransferase CBP, we also investigated whether deacetylation by hSir2SIRT1 involves direct binding. Therefore, we expressed myc-hSir2SIRT1, HA-FOXO4, and treated cells with Nam or hydrogen peroxide for one hour to induce acetylation of HA-FOXO4. Complex formation between hSir2SIRT1 and FOXO4 could only be detected after peroxide treatment (Fig. 4F), suggesting that hSir2SIRT1 binds to acetylated FOXO4. On the contrary, Nam was not able to stabilize the interaction of hSir2SIRT1 with FOXO4, indicating that the effect of Nam on FOXO transcriptional activity can not be accounted for by mere binding of hSirSIRT1 to FOXO. We also tested whether FOXO and hSir2SIRT1 interacted endogenously. Antibody quality precluded the detection of a clear interaction between FOXO4 and hSir2SIRT1, but in HEK293T cells, an interaction between endogenous FOXO3a and hSir2SIRT1 was detected, and this interaction was weakly induced by peroxide treatment (Fig. 4G). Taken together, these data show that hSir2SIRT1 directly binds to FOXO, which is induced by peroxide stress. Our results suggest a model in which FOXO transcription factors are subject to regulation by reversible acetylation. Acetylation of FOXOs results in inhibition of their transcriptional and biological activities, which is reversed by the longevity protein hSir2SIRT1. Inhibition through acetylation of FOXO transcription factors is functionally equivalent to the previously described inhibition by PKB-mediated phosphorylation. We show that the acetyltransferase CBP binds to and acetylates FOXO4, thereby inhibiting its transcriptional activity. Previous reports on CBP/p300 binding to FOXO (27So C.W. Cleary M.L. Mol. Cell. Biol. 2002; 22: 6542-6552Crossref PubMed Scopus (92) Google Scholar, 28Nasrin N. Ogg S. Cahill C.M. Biggs W. Nui S. Dore J. Calvo D. Shi Y. Ruvkun G. Alexander-Bridges M.C. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 10412-10417Crossref PubMed Scopus (141) Google Scholar) suggested that CBP/p300 binding is required for activation of FOXO-dependent transcription. However, there are several important differences between these and our studies that may explain this apparent discrepancy. First, Nasrin et al. (28Nasrin N. Ogg S. Cahill C.M. Biggs W. Nui S. Dore J. Calvo D. Shi Y. Ruvkun G. Alexander-Bridges M.C. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 10412-10417Crossref PubMed Scopus (141) Google Scholar) inferred a role for CBP/p300 in the activation of FOXO because they observed that the viral protein E1A, which can inhibit CBP co-activator function, also inhibits FOXO transcriptional activity on an insulin-like growth factor BP1 promoter. However, E1A is known to have a plethora of targets within the transcriptional machinery and thus, by itself, a block by E1A is not sufficient proof of the involvement of CBP/p300 (for a review, see Ref. 33Goodman R.H. Smolik S. Genes Dev. 2000; 14: 1553-1577Crossref PubMed Google Scholar). Second, whereas we have also analyzed the effect of CBP on the regulation of endogenous gene transcription, other studies have relied on transient reporter assays in which a proper chromatin context is likely to be lacking. Third, the role of CBP/p300 can be promoter-context-dependent, and accordingly, different FOXO-responsive genes may respond differently to the presence of CBP/p300. Differential responsiveness could relate to the fact that we have to discriminate between the effect of CBP/p300 as histone acetyltransferase and as acetyltransferase acting on FOXO itself. It could be possible that, with respect to different promoters, these activities of CBP/p300 may synergize or antagonize. Because CBP/p300-mediated histone acetylation is generally considered to contribute to transcriptional activation, and as we observe inhibition of FOXO transcriptional activity, we conclude that in the context of the promoters analyzed here, the direct acetylation of FOXO acts as an inhibitory signal. The binding of CBP to FOXO4 as well as the acetylation of FOXO4 is induced upon treatment of cells with hydrogen peroxide. Furthermore, hSir2SIRT1 can bind to acetylated FOXO4 and consequently deacetylate FOXO4. Thus, hSir2SIRT1 can counteract the inhibition of FOXO induced by acetylation. Recently, regulation of p53 acetylation by hSir2SIRT1 has been described (16Vaziri H. Dessain S.K. Ng Eaton E. Imai S.I. Frye R.A. Pandita T.K. Guarente L. Weinberg R.A. Cell. 2001; 107: 149-159Abstract Full Text Full Text PDF PubMed Scopus (2316) Google Scholar, 24Luo J. Nikolaev A.Y. Imai S. Chen D. Su F. Shiloh A. Guarente L. Gu W. Cell. 2001; 107: 137-148Abstract Full Text Full Text PDF PubMed Scopus (1902) Google Scholar), and this has been implicated in mediating cellular protection against genotoxic stresses. However, apoptosis induced by loss of hSir2SIRT1 is only partially rescued by dominant-negative p53 (16Vaziri H. Dessain S.K. Ng Eaton E. Imai S.I. Frye R.A. Pandita T.K. Guarente L. Weinberg R.A. Cell. 2001; 107: 149-159Abstract Full Text Full Text PDF PubMed Scopus (2316) Google Scholar). Furthermore, in cells lacking functional p53, resveratrol, a presumed activator of hSir2SIRT1, can still enhance protection against genotoxic stress (14Howitz K.T. Bitterman K.J. Cohen H.Y. Lamming D.W. Lavu S. Wood J.G. Zipkin R.E. Chung P. Kisielewski A. Zhang L.L. Scherer B. Sinclair D.A. Nature. 2003; 425: 191-196Crossref PubMed Scopus (3202) Google Scholar). These observations suggest alternative p53-independent pathways involved in cellular protection by hSir2SIRT1. Recently, we have shown that FOXO activation results in increased resistance against oxidative stress (10Kops G.J. Dansen T.B. Polderman P.E. Saarloos I. Wirtz K.W. Coffer P.J. Huang T.T. Bos J.L. Medema R.H. Burgering B.M. Nature. 2002; 419: 316-321Crossref PubMed Scopus (1282) Google Scholar), and here we show that this is regulated by hSir2SIRT1. We like to propose that FOXOs play an important role in the hSir2SIRT1-mediated, p53-independent protection against stress. Therefore, by preventing negative regulation of FOXO due to acetylation after increased oxidative stress, hSir2SIRT1 can enhance cellular defenses against oxidative stress provided for by FOXO. This may be especially relevant under conditions of endogenous oxidative stress occurring during normal metabolic activities. Oxidative stress is considered to be a prime parameter of aging in all living organisms. Therefore, the results presented in this report provide a rationale for the observed increase in longevity induced by C. elegans Sir2, which is FOXO-dependent. We thank Hans Bos, Leo Price, Arjan Brenkman, and Bas Winkler for critical reading of the manuscript and stimulating discussions, Anneke Steegh for technical assistance, and Rachel Giles, Robert Vries, and Eric Kalkhoven for providing DNA constructs and reagents.