ORIGINAL ARTICLE Journal of
Cytoplasm-Localized SIRT1 Enhances Apoptosis
Cellular Physiology
QIHUANG JIN, TINGTING YAN, XINJIAN GE, CHENG SUN, XIANGLIN SHI, AND QIWEI ZHAI* Institute for Nutritional Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, Graduate School of the Chinese Academy of Sciences, Shanghai, China In general, SIRT1 is localized in nuclei. Here, we showed that endogenous and exogenous SIRT1 were both able to partially localize in cytoplasm in certain cell lines, and cytoplasm-localized SIRT1 was associated with apoptosis and led to increased sensitivity to apoptosis. Furthermore, we demonstrated that translocation of nucleus-localized SIRT1 from nuclei to cytoplasm was the main pathway leading to localization of SIRT1 in cytoplasm. In HeLa cells, wild type SIRT1 was completely localized in nuclei. By truncation of two predicted nuclear localization signals or fusion with an exogenous nuclear export signal, SIRT1 was partially localized in cytoplasm of HeLa cells and resulted in increased sensitivity to apoptosis. The apoptosis enhanced by cytoplasm-localized SIRT1 was independent of its deacetylase activity, but dependent on caspases. SIRT1 was distributed in cytoplasm at metaphase during mitosis, and overexpression of SIRT1 significantly augmented apoptosis for cells at metaphase. In summary, we found SIRT1 is able to localize in cytoplasm, and cytoplasm-localized SIRT1 enhances apoptosis. J. Cell. Physiol. 213: 88–97, 2007. ß 2007 Wiley-Liss, Inc.
Silent information regulator 2 (Sir2) proteins are an evolutionarily conserved family of NAD-dependent protein deacetylases (Imai et al., 2000). In yeast, Sir2 is involved in the regulation of transcriptional silencing, DNA damage responses, senescence, and is required for lengthening of lifespan following caloric restriction (Guarente, 2000; Lin et al., 2002; Denu, 2003). Seven mammalian Sir2 homologs, referred as SIRT1–7, have been identified, and SIRT1 is the closest structural ortholog of the yeast Sir2 protein (Frye, 2000). SIRT1 plays a role in a wide variety of processes including stress resistance, metabolism, differentiation and aging (Blander and Guarente, 2004). SIRT1-deficient mice mostly die during the early postnatal period, and exhibit defects in spermatogenesis and in heart and retina development (Cheng et al., 2003; McBurney et al., 2003b). Overexpression of SIRT1 prevents adipogenesis in 3T3-L1 cells and myogenesis in C2C12 cells (Fulco et al., 2003; Picard et al., 2004). SIRT1 controls glucose homeostasis through modulating PGC-1a and regulates insulin secretion by repressing UCP2 in pancreatic b cells (Bordone et al., 2005; Rodgers et al., 2005). SIRT1 is also involved in both anti-apoptosis and proapoptosis. It has been reported that SIRT1 prevented apoptosis by three different mechanisms. The first one is that SIRT1 deacetylates the p53 tumor suppressor protein and attenuates its ability to trans-activate its downstream target genes (Luo et al., 2001; Vaziri et al., 2001). The second one is that SIRT1 deacetylases the DNA repair factor Ku70, causing deacetylated form of Ku70 to sequester Bax away from mitochondria and thereby inhibiting apoptosis (Cohen et al., 2004b). The third one is that SIRT1 deacetylates FOXO family proteins, resulting in dual effects on transcriptional repression of the downstream pro-apoptotic target gene Bim and upregulation of the stressresistance gene GADD45 (Brunet et al., 2004; Motta et al., 2004). However, SIRT1 was also demonstrated to augment apoptosis in response to TNFa by deacetylating RelA/p65 and inhibiting its transcriptional activity (Yeung et al., 2004). The seven mammalian Sir2 homologs have different subcellular localization, and SIRT1, SIRT6 and SIRT7 were reported to be localized in nuclei (Michishita et al., 2005). However, it was reported that SIRT1 was localized in both nucleoplasm and chromatin/matrix, and SIRT6 and SIRT7 were mainly associated with chromatin/matrix (Michishita et al., 2005; Mostoslavsky et al., 2006). During mitosis, SIRT1 was shown to be distributed over the whole cell (McBurney et al., 2003a; Michishita et al., 2005), distinguished from SIRT6 and ß 2 0 0 7 W I L E Y - L I S S , I N C .
SIRT7, which were associated with chromosomes (Michishita et al., 2005). In islet a and b cells, SIRT1 was appeared in cytoplasm (Moynihan et al., 2005). Recently Ohsawa et al. reported that localization of SIRT1 in cytoplasm is induced in the process of apoptosis (Ohsawa and Miura, 2006). Hallows et al. reported that in cytoplasm SIRT1 catalyzes the deacetylation of AceCS1 (Hallows et al., 2006). But the function of cytoplasm-localized SIRT1 in apoptosis is still unclear. In this paper, we found that SIRT1 was able to partially localize in cytoplasm in certain cell lines. Further investigation showed that cells with cytoplasm-localized SIRT1 were related to apoptosis. Furthermore, we demonstrated that localization of SIRT1 in cytoplasm led to increased cell sensitivity to apoptosis. Materials and Methods Materials
LMB, polybrene, H2O2, nocodazole, digitonin, etoposide, DAPI, mouse anti-tubulin and anti-FLAG monoclonal antibodies were purchased from Sigma (Saint Louis, MO). Pan-caspase inhibitor zVAD and mouse anti-cytochrome c monoclonal antibody for
Abbreviations: Sir2, silent information regulator 2; GFP, green fluorescent protein; NLS, nuclear localization signal; NES, nuclear export signal; CHX, cycloheximide; LMB, leptomycin B; zVAD, Z-VAD-FMK; HSV,herpes simplex virus; PARP, poly-ADP ribose polymerase. Contract grant sponsor: National Natural Science Foundation of China; Contract grant number: 30400083, 30570558. Contract grant sponsor: The Chinese Academy of Sciences; Contract grant number: KSCX2-2-25. Contract grant sponsor: Science and Technology Commission of Shanghai Municipality; Contract grant number: 04dz14007. *Correspondence to: Qiwei Zhai, Institute for Nutritional Sciences, Shanghai Institutes for Biological Sciences, Chinese Academy of Sciences, 294 Taiyuan Road, Shanghai 200031, China. E-mail:
[email protected] Received 31 October 2006; Accepted 26 February 2007 DOI: 10.1002/jcp.21091
88
CYTOPLASM-LOCALIZED SIRT1 ENHANCES APOPTOSIS
immunofluorescence were from Promega Life Science (Madison, WI). Mouse anti-cytochrome c for Western blot was from Santa Cruz Biotechnology (Santa Cruz, CA). Rabbit anti-SIRT1 and antiacetylated p53 (Lys373) antibodies were from Upstate (Chicago, IL). Puromycin was from Calbiochem (La Jolla, CA). Cell culture medium DMEM, F12 and lipofectamine 2000 were from Invitrogen (Carlsbad, CA). Cy3 conjugated anti-rabbit IgG or anti-mouse IgG, horseradish peroxidase conjugated anti-rabbit IgG or anti-mouse IgG secondary antibodies were from Jackson Immunoresearch (West Grove, PA). Rabbit anti-GFP was from Cell Signaling Technology (Beverly, MA). Mouse anti-PARP monoclonal antibody was from BD Biosciences (San Diego, CA). SuperSignal west pico chemiluminescent substrate was from Pierce Biotechnology (Rockford, IL).
Construction and use of recombinant herpes simplex virus (HSV)
Cell culture and transfection
Construction and use of recombinant lentivirus
Cells were grown in the presence of 7.5% CO2 at 37 8C in humidified chambers. HeLa, HEK293T, HEK293, C2C12 and 2-2 cells were grown in DMEM with 10% newborn calf serum, LoVo cells were in F12 with 10% FBS. The cells were transfected using lipofectamine 2000 according to the manufacturer’s instruction.
For construction of lentivirus expressing SIRT1 fused with GFP protein, the GFP gene in pLentiLox 3.7 was cut off by NheI and EcoRI and replaced by a linker, which had multiple cloning sites including NheI and BamHI. The linker was gained by annealing the following two oligonucleotides:
Plasmid construction
pCMV-SIRT1. Expressing SIRT1 with myc tag. SIRT1 cDNA fragment in pBabepuro-SIRT1 (Luo et al., 2001) was cut out with BamHI and inserted into pCMV-Tag 3A (Stratagene, La Jolla, CA) at the BamHI site. pEGFP-SIRT1. Expressing GFP fused with SIRT1. SIRT1 cDNA fragment in pBabepuro-SIRT1 was cut out with BamHI and inserted into pEGFP-C3 (Clontech, Palo Alto, CA) at the BamHI site. pEGFP-SIRT1D238. Expressing GFP fused with a fragment of SIRT1 absent of 238 amino acids at N-terminal. The cDNA coding the fragment of SIRT1 was obtained by PCR using the following primer pair: 50 -GGCTCGAGGATATTAATACAATTGAAGATG-30 50 -CCGGATCCCTATGATTTGTTTGATGGATAGTTC-30
The fragment of gene was inserted into pEGFP-C3 at XhoI and BamHI sites. pEGFP-SIRT1D238 HY. Expressing GFP-SIRT1D238 with mutation H363Y. The process of constructing the plasmid was the same as constructing pEGFP-SIRT1D238 except using another PCR template SIRT1 H363Y (Luo et al., 2001), which has no deacetylase activity. pEGFP-SIRT1D517. Expressing GFP fused with a fragment of SIRT1 absent of 517 amino acids at N-terminal. The cDNA coding the fragment of SIRT1 was obtained by PCR using the following primer pair: 50 -GGCTCGAGCAAAAAGAATTGGCTTATTTGTCAG-30 50 -CCGGATCCCTATGATTTGTTTGATGGATAGTTC-30
The cDNA was inserted into pEGFP-C3 at XhoI and BamHI sites. pEGFP-NES. Expressing GFP fused with a nuclear export signal (NES) peptide (LQLPPLERLTLDC) from HIV-1 Rev protein (Fischer et al., 1995). The cDNA sequences coding the NES peptide were synthesized, annealed, and then inserted into pEGFP-C3 at sites BglII and ApaI. The NES cDNA sequences were: 50 -GATCTTCAGCTACCACCGCTTGAGAGACTTACTCTTGATTGTCGGGCC-30 50 -CGACAATCAAGAGTAAGTCTCTCAAGCGGTGGTAGCTGAA-30
pEGFP-NES-SIRT1. Expressing GFP fused with a NES peptide and SIRT1. SIRT1 cDNA fragment from pBabepuro-SIRT1 was cut out with BamHI and inserted into pEGFP-NES at the BamHI site. JOURNAL OF CELLULAR PHYSIOLOGY DOI 10.1002/JCP
SIRT1 gene was from pBabepuro-SIRT1 by BamHI digestion and subcloned into reconstructed pHSVPrPUC-myc vector at the BamHI site. After transfected with plasmids by lipofectamine 2000 for 24 h, 2-2 cells were super-infected with 5dl5 HSV Helper viruses. The recombinant HSV-SIRT1 virus particles were harvested, amplified as described previously (Wang et al., 2005). After C2C12 cells were infected with the HSV-SIRT1 virus particles for 24 h, the medium was replaced with fresh medium and the cells were incubated for another 24 h. Then the cells were fixed with 4% paraformdehyde and immunostained with anti-SIRT1 antibody (1:10,000).
50 -CTAGCTATGGATCCTGTTCTAGAGCAGAATTCCTCGAGAAAGGGCCCAGGCCTTGA-30 50 -AATTTCAAGGCCTGGGCCCTTTCTCGAGGAATTCTGCTCTAGAACAGGATCCATAG3-0
The cDNA coding GFP was re-inserted into the re-constructed pLentiLox 3.7 at NheI and BamHI sites, and the GFP cDNA was gained by PCR with the following primer pair: 50 -AAGCTAGCCGCCACCATGGTGAGCAAGG-30 50 -TTTGGATCCTTGTACAGCTCGTCCATGC-30
Finally, SIRT1 cDNA fragment in pBabepuro-SIRT1 was cut out with BamHI and inserted into the re-constructed pLentiLox 3.7 with GFP at the BamHI site. The recombinant pLentiLox 3.7 was cotransfected with VSV-G and D8.9 in HEK293T cells. Starting at 48 h, virus particles were harvested every 12 h. HeLa cells were infected with virus in the presence of 8 mg/ml polybrene for 12 h each time and totally three times. The infected cells were then subcultured and treated with nocodazole. Immunofluorescence
The cells were fixed with 4% paraformaldehyde for 40 min at 48C and permeabilized with permeabilization solution (0.1% Triton X-100, 0.1% sodium citrate in PBS). After blocked with 3% BSA, the cells were stained by primary antibodies, rabbit anti-SIRT1 antibody (1:10,000 or 1:1,000), or mouse anti-cytochrome c monoclonal antibody (1:1,000), and followed incubating with corresponding secondary antibodies, Cy3 conjugated anti-rabbit IgG or anti-mouse IgG (1:1,000). Then the cells were stained with DAPI (0.5 mg/ml) and observed under a fluorescence microscope (Olympus IX71) or a confocal fluorescence microscope (Carl Zeiss, LSM 510 META). Western blot
Western blot was carried out as previously described (Jin et al., 2004). Primary antibodies included rabbit anti-SIRT1 antibody (1:10,000), mouse anti-tubulin monoclonal antibody (1:10,000), rabbit anti-GFP antibody (1:1,000), mouse anti-PARP monoclonal antibody (1:1,000), and secondary antibodies linked to horseradish peroxidase including anti-rabbit IgG and anti-mouse IgG (1:1,000). Visualization of bound antibodies was by the chemiluminescence method.
89
90
JIN ET AL.
Subcellular fraction
LoVo cells treated with or without 200 mM H2O2 for 24 h were digested with trypsin and washed once with ice-cold PBS, and the cytosolic and pellet (mitochondrial and nuclear) fractions were generated using digitonin-permeabilized method as previously described (Adrain et al., 2001). Statistical analysis
All values are expressed as mean standard deviation (SD), and statistical analyses were carried out using a Student’s t-test, unless otherwise indicated. Differences were considered to be statistically significant when P < 0.05. Results SIRT1 is able to partially localize in cytoplasm in certain cell lines
In general SIRT1 is localized in nuclei. In HEK293 and HeLa cells except for those mitotic cells, endogenous SIRT1 was absolutely localized in nuclei (Fig. 1, middle and lower parts). However, in LoVo cells, endogenous SIRT1 was partial cytoplasmic localization in 3.05 0.7% LoVo cells excluding mitotic cells (Fig. 1, upper part). The staining of endogenous SIRT1 was very weak, to confirm the localization of SIRT1 in cytoplasm of LoVo cells, exogenous SIRT1 in expression vector pCMV-SIRT1 was transfected into LoVo and HEK293 cells. Exogenous SIRT1 was also partial localization in cytoplasm of LoVo cells (Fig. 2A, upper part). In HEK293 cells, nearly all cells presented SIRT1 in nuclei (Fig. 2A, lower part), and the percentage of cells with cytoplasm-localized SIRT1 were 33.6 2.0% for LoVo cells and 0.8 0.2% for HEK293 cells after transfection for 48 h (Fig. 2C). Cytoplasm-localized SIRT1 also appeared in C2C12 cells transfected with pCMV-SIRT1 (data not shown). C2C12 infected with recombinant herpes simplex virus (HSV) expressing SIRT1 (HSV-SIRT1) also presented SIRT1 in cytoplasm (Fig. 2A, middle part), and the percentage of cells with cytoplasm-localized SIRT1 was 26.0 4.5% after infection for 24 h (Fig. 2C). This result excluded the possibility that the methods to express exogenous SIRT1 led to the localization of SIRT1 in cytoplasm. The expression of SIRT1 by pCMV-SIRT1 and HSV-SIRT1 was confirmed by Western blot (Fig. 2B).
Fig. 1. Endogenous SIRT1 is partially localized in cytoplasm of LoVo cells but not in HEK293 and HeLa cells. SIRT1 was immunostained with anti-SIRT1 antibody diluted 1:1,000. The numbers at right of pictures are the percentage of cells with cytoplasm-localized SIRT1. Scale bar, 20 mm. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]
JOURNAL OF CELLULAR PHYSIOLOGY DOI 10.1002/JCP
Fig. 2. Exogenous SIRT1 is able to partially localize in cytoplasm in certain cell lines. A: SIRT1 was able to partially localize in cytoplasm in LoVo and C2C12 cells, but not in HEK293 cells. SIRT1 was immunostained with anti-SIRT1 antibody in LoVo and HEK293 cells transfected with pCMV-SIRT1 or in C2C12 cells infected with HSVSIRT1. Scale bar, 20 mm. B: The expression of SIRT1 by pCMV-SIRT1 and HSV-SIRT1 was confirmed by Western blot using anti-SIRT1 antibody. C: The percentage of cells with cytoplasm-localized SIRT1 in positively transfected or infected cells. Quantitation was done by immunostaining with anti-SIRT1 antibody after LoVo and HEK293 cells were transfected with pCMV-SIRT1 for 48 h and C2C12 cells were infected with HSV-SIRT1 for 24 h. Error bars indicate the WSD of three independent experiments. MP < 0.01 compared with HEK293 cells. D: GFP-SIRT1 could also partially localize in cytoplasm in LoVo cells, but not in HEK293 cells. Scale bar, 20 mm. E: The expression of GFP-SIRT1 was confirmed by Western blot with anti-SIRT1 or anti-GFP antibody. F: The percentage of cells with cytoplasmlocalized GFP-SIRT1 in LoVo or HEK293 cells expressing GFP-SIRT1, after the cells were transfected with pEGFP-SIRT1 for 48 h. Error bars indicate the WSD of three independent experiments. MP < 0.01. [Color figure can be viewed in the online issue, which is available at www. interscience.wiley.com.]
To exclude the possibility that immunostaining led to the artificial localization of SIRT1 in cytoplasm, we constructed the plasmid expressing SIRT1 protein fused with GFP. The expression of GFP-SIRT1 by this plasmid was confirmed by Western blot (Fig. 2E). GFP-SIRT1 was also able to localize in cytoplasm in LoVo cells, but not in HEK293 cells (Fig. 2D). After transfection for 48 h, the percentage of cells with cytoplasmlocalized GFP-SIRT1 was 40.1 7.8% for LoVo cells and 1.5 0.4% for HEK293 cells (Fig. 2F). These results showed that SIRT1 is able to partially localize in cytoplasm in certain cell lines.
CYTOPLASM-LOCALIZED SIRT1 ENHANCES APOPTOSIS
Cytoplasm-localized SIRT1 is associated with apoptosis
The percentage of cells with cytoplasm-localized SIRT1 in positively transfected cells decreased continuously along with culture time prolonging (data not shown). Therefore, it is possible that the cells with cytoplasm-localized SIRT1 are sensitivity to apoptosis. To investigate this possibility, after transfected with GFP or GFP-SIRT1 for 48 h, LoVo cells were treated with 200 mM H2O2 for 20 h and then stained by anticytochrome c antibody and DAPI. Apoptotic cells exhibited release of cytochrome c from mitochondria and condensed nuclei by DAPI staining (Fig. 3A). In the cells without cytochrome c release, cytochrome c staining was punctuated (Fig. 3A, upper and middle parts); the cells with cytochrome c release exhibited diffusive cytosolic cytochrome c staining (Fig. 3A, lower part). The cells with cytoplasm-localized GFPSIRT1 had a higher percentage of apoptosis than the cells only with nucleus-localized GFP-SIRT1 or the cells transfected with GFP either according to cytochrome c release (Fig. 3B) or according to the condensed nuclei (Fig. 3C). After the LoVo cells were treated with H2O2 for 24 h, cytosolic fraction was isolated. In the cytosolic fraction, cytochrome c was appeared concomitantly with increased SIRT1 protein level (Fig. 3D). These data demonstrated that cytoplasm-localized SIRT1 is associated with apoptosis. Cytoplasm-localized SIRT1 is mainly from nuclei
Next, we investigated possible mechanisms underlying the origin of cytoplasm-localized SIRT1. Since cytoplasm-localized SIRT1 was associated with apoptosis, we investigated the alternation of SIRT1 localization with apoptotic stimuli. The LoVo cells transfected with GFP-SIRT1 were treated with apoptosis inhibitor Z-VAD-FMK (zVAD) and cycloheximide (CHX), which inhibits protein synthesis and induces apoptosis. GFP-SIRT1 was found to partially localize in cytoplasm after treatment for 8 h in some LoVo cells, which completely presented SIRT1 in nuclei before treatment (Fig. 4A). Because protein synthesis was inhibited by CHX, these data demonstrated that cytoplasm-localized SIRT1 was exported from nuclei. After transfected with GFP-SIRT1, LoVo cells were immediately treated with H2O2 for 24 h, and more than 60% of positive transfected cells presented cytoplasm-localized SIRT1; while leptomycin B (LMB), an inhibitor of nuclear export, significantly suppressed localization of SIRT1 in cytoplasm induced by H2O2 (Fig. 4B and C). These results confirmed that cytoplasm-localized SIRT1 was mainly from nuclei. Endogenous SIRT1 was detectable in LoVo cells when antiSIRT1 antibody was diluted to 1/1,000. LoVo cells were treated with protein synthesis inhibitor CHX, which is also an apoptosis inducer, and then stained with anti-SIRT1 antibody. As shown in Figure 4D and E, under induction of CHX, the percentage of cells with cytoplasm-localized SIRT1 dramatically increased. Other means of apoptotic induction including serum deprivation and puromycin treatment also led to increase the percentage of cells with cytoplasm-localized SIRT1 (Fig. 4E). These data showed that cytoplasm-localized endogenous SIRT1 was detectable, and endogenous SIRT1 was also able to be exported from nuclei in LoVo cells. SIRT1 contains multiple NLSs and one or more NESs
SIRT1 protein might contain one or more nuclear export signals (NES), as cytoplasm-localized SIRT1 was mainly from nuclei. The potential NES was not found by software prediction with NetNES 1.1 (la Cour et al., 2004). To further search the NES in SIRT1, we constructed GFP-SIRT1D238 plasmid, expressing GFP fused with SIRT1 absent of 238 amino acids peptide at N-terminal, which contains two nuclear localization signals (NLS) predicted by PSORT II (Horton and Nakai, 1997), as shown in Figure 5A. The SIRT1 fragment absent of these two JOURNAL OF CELLULAR PHYSIOLOGY DOI 10.1002/JCP
Fig. 3. Cytoplasm-localized SIRT1 is associated with apoptosis. A: Cytoplasm-localized SIRT1 was associated with cytochrome c release and condensed nuclei. LoVo cells were transfected with GFP or GFP-SIRT1 for 48 h, and subsequently treated with 200 mM H2O2 for 20 h. The cells were immunostained with anti-cytochrome c antibody and observed under a confocal microscope. Scale bar, 10 mm. The percentage of apoptotic cells in positively transfected cells was determined according to cytochrome c release (B) or condensed nuclei by DAPI staining (C). Error bars indicate the WSD of three independent experiments. MP < 0.05, MMP < 0.01 compared with the cells transfected with GFP. D: Western blot analyzed the protein level of SIRT1 and cytochrome c in cytosolic and pellet fractions isolated from LoVo cells treated with or without 200 mM H2O2 for 24 h. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]
predicted NLSs was partially localized in cytoplasm, but still mainly localized in nuclei (Fig. 5B). These data suggested that there could exist other NLSs besides these two predicted NLSs. SIRT1 absent of 517 amino acids at N-terminal was also mainly localized in nuclei, while much less distributed in cytoplasm than GFP-SIRT1D238 did (Fig. 5B). These data suggested that there could exist one or more NESs in SIRT1D238 and other NLSs in SIRT1D517. To further confirm that SIRT1 absent of 238 amino acids was able to be exported out of nuclei, HeLa cells transfected with GFP-SIRT1D238 were
91
92
JIN ET AL.
Fig. 4. Cytoplasm-localized SIRT1 is mainly from nuclei. A: SIRT1 was partially exported out of nuclei in LoVo cells only with nucleus-localized SIRT1 after treatment with the protein synthesis inhibitor and apoptosis inducer CHX (100 mg/ml), and the apoptosis inhibitor zVAD (100 mM). Representative pictures were taken from the same cells at the indicated times. Scale bar, 10 mm. B: Exportation of SIRT1 out of nuclei was sensitive to nuclear export inhibitor LMB. Scale bar, 10 mm. C: The percentage of cells with cytoplasm-localized SIRT1 in positively transfected cells. LoVo cells transfected with GFP-SIRT1 were treated with or without 12.5 ng/ml LMB in the presence of 200 mM H2O2 for 24 h, and error bars indicate the WSD of three independent experiments. MP < 0.01. D: Endogenous SIRT1 translocated to cytoplasm in the presence of protein synthesis inhibitor and apoptosis inducer CHX (100 mg/ml) for 24 h in LoVo cells. The endogenous SIRT1 was stained with anti-SIRT1 antibody at high concentration. Scale bar, 10 mm. E: Quantitation of LoVo cells with cytoplasm-localized endogenous SIRT1 after induced by 100 mg/ml CHX, serum deprivation or 5 mg/ml puromycin for 24 h. Error bars indicate the WSD of three independent experiments. MP < 0.01. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]
treated with or without LMB in the presence of CHX and zVAD. The same cells were observed after 2 h. GFP-SIRT1D238 in cytoplasm obviously decreased in HeLa cells treated with LMB but not in cells treated without LMB (Fig. 5D, b). The intracellular distribution of GFP and GFP-SIRT1D517 proteins was not changed under treatment with LMB (Fig. 5D, a and d). These data suggested that SIRT1 possessed one or more NESs in the fragment of SIRT1 from D239 to T517 but not in GFP-SIRT1D517. But the results also might be resulted from degradation of cytoplasm-localized GFP-SIRT1D238, since nucleus-localized SIRT1 was not dramatically increased (Fig. 5D, b). To exclude this possibility, GFP-SIRT1D238 was expressed in HEK293 cells, where GFP-SIRT1D238 was mainly localized in cytoplasm in a lot of cells. After treated with LMB in the presence of zVAD and CHX for 2 h, the same cells were observed. We found GFP-SIRT1D238 in nuclei was dramatically increased (Fig. 5D, e). Interestingly, GFP-SIRT1D238 HY JOURNAL OF CELLULAR PHYSIOLOGY DOI 10.1002/JCP
(GFP-SIRT1D238 with mutation H363Y) was mainly localized in cytoplasm of HeLa cells (Fig. 5B). The HeLa cells transfected with GFP-SIRT1D238 HY were treated with LMB. After 2 h, GFP-SIRT1D238 HY was obviously accumulated in nuclei (Fig. 5D, c). These data further indicated that SIRT1 possessed one or more NESs in the fragment of SIRT1 from D239 to T517. It has been reported that etoposide can induce the acetylation of p53 (Luo et al., 2001). In the presence of etoposide, the acetylation of p53 was detected in the cells cotransfected with p53 and GFP, GFP-SIRT1D517 or GFP-SIRT1D238 HY. When the cells were cotransfected with p53 and GFP-SIRT1 or GFP-SIRT1D238, the acetylation level of p53 was significantly reduced. These data indicated that GFP-SIRT1 and GFP-SIRT1D238 still reserved the deacetylase activity, but not GFP-SIRT1D238 HY and GFP-SIRT1D517 (Fig. 5C). All these data suggested that SIRT1 should contain other NLSs besides the two predicted main NLSs and one or more NESs.
CYTOPLASM-LOCALIZED SIRT1 ENHANCES APOPTOSIS
Fig. 5. SIRT1 contains multiple NLSs and one or more NESs. A: The schematic composition of constructs. GFP-SIRT1, GFP fused SIRT1. GFP-SIRT1D238, GFP fused SIRT1 fragment absent of N-terminal 238 amino acids, which include two predicted NLSs. GFP-SIRT1D238 HY, GFP-SIRT1D238 with mutation H363Y. GFP-SIRT1D517, GFP fused SIRT1 fragment absent of N-terminal 517 amino acids, which include two predicted NLSs and the SIRT1 enzyme activity domain. B: The intracellular distribution of GFP and GFP fused SIRT1 or SIRT1 fragments in HeLa cells. GFP-SIRT1 was localized in nuclei. GFP-SIRT1D238 and GFP-SIRT1D517 were distributed over the whole cell. GFP-SIRT1D238 HY was mainly distributed in cytoplasm. Scale bar, 20 mm. C: GFP-SIRT1 and GFP-SIRT1D238 reserved the deacetylase activity, and GFP-SIRT1D238 HY and GFP-SIRT1D517 almost completely lost the deacetylase activity. p53 tagged with FLAG was transfected into HEK293T cells combined with GFP, GFPSIRT1, GFP-SIRT1D238, GFP-SIRT1D238 HY or GFP-SIRT1D517. After transfection for 18 h, the cells were treated with 20 mM etoposide for 6 h and then harvested for Western blot with antiacetylated p53 (Lys373), FLAG or GFP antibodies. D: GFP-SIRT1D238 and GFP-SIRT1D238 HY was able to shuttle between the cytoplasm and nucleus, but GFP-SIRT1D517 was not. The transfected cells were treated with or without nuclear export inhibitor LMB (12.5 ng/ml) in the presence of the protein synthesis inhibitor CHX (100 mg/ml) and the apoptosis inhibitor zVAD (100 mM) for indicated times, and then pictures were taken from the same cells. In the case of treatment without LMB, the localization of GFP (a, left part), GFP-SIRT1D238 (b and e, left part), GFP-SIRT1D238 HY (c, left part), GFP-SIRT1D517 (d, left part) was not changed. In the case of treatment with LMB, the localization of GFP (a, right part) and GFP-SIRT1D517 (d, right part) was not obviously changed; GFP-SIRT1D238 was reduced in cytoplasm of HeLa (b, right part) and accumulated in nuclei of HEK293 cells (e, right part); GFP-SIRT1D238 HY was significantly accumulated in nuclei of HeLa cells (c, right part). Scale bar, 20 mm. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]
JOURNAL OF CELLULAR PHYSIOLOGY DOI 10.1002/JCP
Fig. 6. SIRT1D238 enhanced apoptosis independent of its deacetylase activity. A: Both GFP-SIRT1D238 and GFP-SIRT1D238 HY enhanced apoptosis induced by H2O2 in HeLa cells, and GFPSIRT1D517 did not. After transfected for 24 h, the HeLa cells were treated with 200 mM H2O2 for another 24 h. The apoptotic cells presented shrinkage under phase contrast microscopy. Arrows indicate apoptotic cells. Scale bar, 20 mm. B: The percentage of apoptotic cells in positively transfected cells. Apoptotic cells were determined according to their morphological changes. Error bars indicate the WSD of three independent experiments. MP < 0.01 compared with the GFP-SIRT1 transfected cells treated with H2O2. C: Apoptotic cells judged by morphological changes were further confirmed by cytochrome c release and condensed nuclei. Arrows indicate apoptotic cells. Scale bar, 20 mm. D: The pan-caspase inhibitor zVAD (50 mM) prevented cells from apoptosis enhanced by GFP-SIRT1D238 and GFP-SIRT1D238 HY. Apoptotic cells were determined according to their morphological changes. Error bars indicate the WSD of three independent experiments. MP < 0.01 compared with the cells transfected with GFP-SIRT1D238 or GFP-SIRT1D238 HY in the absence of zVAD. [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]
SIRT1D238 enhances apoptosis independent of its deacetylase activity
Cytoplasm-localized SIRT1 was associated with apoptosis in LoVo cells (Fig. 3). However, whether cytoplasm-localized SIRT1 had a direct function in apoptosis is yet to be confirmed. In HeLa cells, SIRT1D238 was partially localized in cytoplasm, and SIRT1 was localized in nuclei (Fig. 5B). To examine the function of cytoplasm-localized SIRT1 in apoptosis, H2O2 was used to induce apoptosis. As shown in Figure 6A, under
93
94
JIN ET AL.
treatment with H2O2, some cells showed apoptotic morphological changes under phase contrast microscopy. Transfection with SIRT1D238 resulted in a much higher percentage of apoptotic cells than transfection with GFP or GFP-SIRT1. GFP-SIRT1D517 absent of enzyme activity domain was also partially localized in cytoplasm in HeLa cells, but did not increase the percentage of apoptotic cells compared with GFP or GFP-SIRT1. The percentage of apoptotic cells in the cells transfected with SIRT1D238 was nearly three times of that in the cells transfected with GFP or GFP-SIRT1, but the percentage of apoptotic cells in the cells transfected with GFPSIRT1D517 was similar with the cells transfected with GFP or GFP-SIRT1 (Fig. 6B). The apoptotic cells determined according to morphological changes were confirmed by cytochrome c release and condensed nuclei stained by DAPI (Fig. 6C). To further investigate whether the enhanced apoptosis by cytoplasm-localized SIRT1 is dependent on its deacetylase activity, HeLa cells transfected with GFP-SIRT1D238 HY were induced to apoptosis by H2O2. SIRT1 HY losses its deacetylase activity, and antagonize SIRT1 anti-apoptotic activity (Luo et al., 2001). However, the apoptosis was also enhanced by GFPSIRT1D238 HY (Fig. 6A and B). This result suggested that apoptosis enhanced by cytoplasm-localized SIRT1 was not dependent on its deacetylase activity. The apoptosis enhanced by SIRT1D238 or SIRT1D238 HY was caspase-dependent, as pan-caspase inhibitor zVAD prevented apoptosis enhanced by SIRT1D238 (Fig. 6D). Localization of SIRT1 partially in cytoplasm by fusion with an exogenous NES enhances apoptosis
To exclude the possibility that the deletion of 238 amino acids in SIRT1D238 leads to increased sensitivity to apoptosis, SIRT1 was fused with an exogenous NES (LQLPPLERLTLDC) from HIV-1 Rev protein (Fischer et al., 1995) (Fig. 7A), which resulted in partially cytoplasmic localization of SIRT1 (Fig. 7B). In the presence of etoposide, the acetylation level of p53 was lower in cells cotransfected with p53 and GFP-NES-SIRT1 than that in cells contransfected with p53 and GFP-NES, and suggested that GFP-NES-SIRT1 still reserved the deacetylase activity (Fig. 7C). Under treatment with H2O2, some cells showed apoptotic morphological changes under phase contrast microscopy. Transfection with GFP-NES-SIRT1 resulted in a much higher percentage of apoptotic cells than transfection with GFP-NES (Fig. 7D). The percentage of apoptotic cells in the cells with GFP-NES-SIRT1 was more than 3 times of that in the cells with GFP-SIRT1 or GFP-NES (Fig. 7E). The apoptotic cells determined according to morphological changes were confirmed by cytochrome c release and condensed nuclei stained by DAPI (Fig. 7F). Pan-caspase inhibitor zVAD significantly inhibited apoptosis enhanced by GFP-NES-SIRT1 (Fig. 7G). These results further confirmed that cytoplasmlocalized SIRT1 could enhance apoptosis, and this effect was caspase-dependent. SIRT1 localized in cytoplasm at metaphase during mitosis enhances apoptosis
SIRT1 was localized in cytoplasm at metaphase during mitosis (McBurney et al., 2003a; Michishita et al., 2005). Since our data demonstrated that cytoplasm-localized SIRT1 resulted in increased sensitivity to apoptosis, overexpression of SIRT1 should increase cell sensitivity to apoptosis inducers at metaphase. HeLa cells were highly infected with lentivirus expressing GFP or GFP-SIRT1 (Fig. 8A) and then treated for 24 h with 2 mg/ml nocodazole, which induces cell cycle arrest at metaphase and apoptosis. When mitosis was blocked by nocodazole at metaphase, SIRT1 was localized in cytoplasm (Fig. 8B). It was not feasible to distinguish apoptotic cells based JOURNAL OF CELLULAR PHYSIOLOGY DOI 10.1002/JCP
Fig. 7. SIRT1 fused with an exogenous NES was partially localized in cytoplasm and enhanced apoptosis. A: The schematic composition of constructs. GFP-NES, GFP fused with a NES, LQLPPLERLTLDC, from HIV-1 Rev protein. GFP-NES-SIRT1, GFP-NES further fused with SIRT1. B: The intracellular distribution of GFP-NES and GFP-NES-SIRT1. GFP-NES was mainly distributed in cytoplasm. GFP-NES-SIRT1 was partially localized in cytoplasm. After the green fluorescence pictures were taken, cells were fixed and stained with DAPI. Scale bar, 20 mm. C: GFP-NES-SIRT1 reserved the deacetylase activity. p53 tagged with FLAG was transfected into HEK293T cells combined with GFP-NES or GFP-NES-SIRT1. After 18 h, the cells were treated with 20 mM etoposide for 6 h and then harvested for Western blot with anti-acetylated p53 (Lys373), FLAG or GFP antibodies. D: GFP-NES-SIRT1 enhanced apoptosis induced by H2O2 in HeLa cells. After transfected for 24 h, the HeLa cells were treated with 200 mM H2O2 for another 24 h. The apoptotic cells presented shrinkage under phase contrast microscopy. Arrows indicate apoptotic cells. Scale bar, 20 mm. E: The percentage of apoptotic cells in positively transfected cells, and apoptotic cells were determined according to their morphological changes. Error bars indicate the WSD of three independent experiments. MP < 0.01, compared with the GFP-SIRT1 transfected cells treated with H2O2. F: Apoptotic cells judged by morphological changes were further confirmed by cytochrome c release and condensed nuclei. Arrows indicate apoptotic cells. Scale bar, 20 mm. G: The pan-caspase inhibitor zVAD (50 mM) prevented cells from apoptosis enhanced by GFP-NES-SIRT1. Apoptotic cells were determined according to their morphological changes. Error bars indicate the WSD of three M independent experiments. P < 0.01, compared with the cells transfected with GFP-NES-SIRT1 in the absence of zVAD. [Color figure can be viewed in the online issue, which is available at www. interscience.wiley.com.]
CYTOPLASM-LOCALIZED SIRT1 ENHANCES APOPTOSIS
Fig. 8. SIRT1 localized in cytoplasm at metaphase during mitosis enhanced apoptosis. A: HeLa cells were highly infected with lentivirus expressing GFP (lenti-GFP) or expressing GFP-SIRT1 (lenti-GFP-SIRT1). Scale bar, 20 mm. B: GFP-SIRT1 was distributed over the whole cell after treatment with 2 mg/ml nocodazole, which inhibits mitosis at metaphase and induces apoptosis. Scale bar, 20 mm. C: Cells overexpressing SIRT1 was sensitive to apoptosis induced by nocodazole. The HeLa cells infected with lenti-GFP or lenti-GFPSIRT1 were incubated with or without 2 mg/ml nocodazole for 24 h. Then the cells were collected for Western blot with anti-SIRT1, anti-PARP or anti-tubulin antibody. PARP cleavage is a marker of apoptosis. D: The relative quantity of cleaved PARP in the cells infected with lenti-GFP or lenti-GFP-SIRT1 after treatment with nocodazole for 24 h. Error bars indicate the WSD of three independent experiments. MP < 0.05 (paired Student’s t-test). [Color figure can be viewed in the online issue, which is available at www.interscience.wiley.com.]
on morphological changes or condensed nuclei by DAPI staining at metaphase, so the degree of apoptosis was determined by the cleavage of poly-ADP ribose polymerase (PARP), which is another apoptotic marker. As shown in Figure 8C, after treatment with nocodazole for 24 h, PARP cleavage significantly increased in HeLa cells overexpressing SIRT1 compared with the cells expressing GFP. The quantity of cleaved PARP in HeLa cells overexpressing SIRT1 was more than two times of that in HeLa cells expressing GFP (Fig. 8D). These results further confirmed that cytoplasm-localized SIRT1 resulted in increased sensitivity to apoptosis, and suggested that under physiological conditions SIRT1 might have a role in getting rid of abnormal dividing cells at metaphase. Discussion
In this paper, we found SIRT1 was able to localize in cytoplasm in certain cell lines, such as LoVo and C2C12 cells. To confirm this conclusion, we excluded several possibilities to get artificial results. First, SIRT1 expressed by different plasmids including pCMV-SIRT1 (myc tag), pECE-SIRT1 (flag tag, data not shown) was all able to localize in cytoplasm of LoVo cells, and it excluded the possibility that plasmids themselves affected the localization of SIRT1 in cytoplasm. Second, cytoplasm-localized SIRT1 was presented in C2C12 cells transfected pCMV-SIRT1 using lipofectamine 2000 or infected with HSV-SIRT1. The results demonstrated that localization of SIRT1 in cytoplasm was not attributed to the methods for expressing exogenous JOURNAL OF CELLULAR PHYSIOLOGY DOI 10.1002/JCP
SIRT1. Third, cytoplasm-localized SIRT1 was detected by immunostaining with either anti-SIRT1 antibody or anti-tag (myc and flag tags, data not shown) antibodies in LoVo cells. Moreover, GFP-SIRT1 was observed localizing in cytoplasm in living LoVo cells. These data excluded the possibility that cytoplasm-localized SIRT1 was resulted from nonspecific binding of antibodies in cytoplasm. In addition, endogenous SIRT1 was also able to localize in cytoplasm in LoVo cells (Fig. 4D), and it showed the localization of SIRT1 in cytoplasm was not due to overexpression of exogenous SIRT1. These data strongly demonstrated that the localization of SIRT1 in cytoplasm is not an artificial result, and is a natural characteristic of SIRT1. Cytoplasm-localized SIRT1 was mainly from nuclei. But the mechanisms of translocation of SIRT1 from the nucleus to cytoplasm are yet to be further elucidated. Wilson et al. reported that yeast Sir2 homolog Hst2 shuttles between the nucleus and cytoplasm and contains an identified NES, and also reported that Hst2 mammalian orthologue SIRT2 has an NES (Wilson et al., 2006). So we conjectured that SIRT1 might export out of the nuclei by a similar way. We aligned the NES in Hst2 with SIRT1, and found the Hst2 NES homolog in SIRT1 was in the region of SIRT1D517. While SIRT1D517 could not shuttling between cytoplasm and nucleus as shown in Figure 5D, d, which suggested that there was no potential NES in SIRT1D517. Our results showed that there could exist one or more NESs in the fragment of SIRT1 from D239 to T517 (Fig. 4D). The SIRT1 homologue in Drosophila, dSIR2, is appeared in both nuclei and cytoplasm or only in cytoplasm at the different nuclear cycle during early embryonic development (Rosenberg and Parkhurst, 2002). These data also suggested that dSIR2 contained NESs. A potential NES site from L456 to L465 in dSIR2 was found by NetNES 1.1. The homologous NES fragment of dSIR2 in SIRT1 is L469PHLHFDVEL478, which is localized between D239 to T517. It supported our result that there could exist one or more NESs in the fragment of SIRT1 between D239 to T517. But the predicted NES might be not responsible for SIRT1 exportation out of nuclei, as GFP fusion with the predicted NES L469PHLHFDVEL478 in SIRT1 did not significantly export out of nuclei (data not shown). GFPSIRT1D238 HY containing NESs is mainly localized in cytoplasm and could shuttle between the nucleus and cytoplasm (Fig. 5B and D c). When Val476 and Leu478 were mutated to disrupt the predicted NES, GFP-SIRT1D238 HY was not accumulated in the nuclei of HeLa cells (data not shown). The above data demonstrated that the predicted NES by NetNES 1.1 is not the real NES for SIRT1. Recently two NESs were identified in SIRT1 (Tanno et al., 2007). The major one was contained in SIRT1D238, as we suggested. It is interesting to investigate how SIRT1 is able to overcome the strong NLSs and localized in cytoplasm in LoVo cells, as well as in islet a and b cells (Moynihan et al., 2005). There might exist SIRT1 anchor proteins in cytoplasm to sequestrate SIRT1, as Parc sequestrates p53 in cytoplasm (Nikolaev et al., 2003). SIRT1 prevents apoptosis by deacetylating p53, FOXO and Ku70 (Luo et al., 2001; Vaziri et al., 2001; Brunet et al., 2004; Motta et al., 2004; van der Horst et al., 2004; Cohen et al., 2004b; Yang et al., 2005). Here, we reported that localization of SIRT1 in cytoplasm enhanced apoptosis, which was supported by two evidences. One is that truncation of 238 amino acids or fusion with an exogenous NES at SIRT1 N-terminal induced partial localization of SIRT1 in cytoplasm and increased cell sensitivity to apoptosis. The other is that SIRT1 localized in cytoplasm at metaphase during mitosis and overexpression of SIRT1 led to increased sensitivity to apoptosis induced by nocodazole, which inhibited mitosis at metaphase. As shown in Figures 6B and 7E, localization of SIRT1 in cytoplasm might slightly induce apoptosis. While we found that there were still some living cells with GFP-SIRT1D238 or GFP-NES-SIRT1 after
95
96
JIN ET AL.
treated with 1 mg/ml G418 for 1 week (data not shown). These results suggested that localization of SIRT1 in cytoplasm might not significantly induce apoptosis. SIRT1 absent of 238 amino acids or fused with an exogenous NES was only partially localized in cytoplasm, and most SIRT1 was still in nuclei. Nucleus-localized SIRT1 was reported to prevent apoptosis (Luo et al., 2001; Vaziri et al., 2001; Brunet et al., 2004; Motta et al., 2004; van der Horst et al., 2004; Cohen et al., 2004b; Yang et al., 2005), so how apoptosis could be enhanced in these cells under induction of H2O2? One explanation is that the HeLa cells overexpressing SIRT1 did not resist to apoptosis induced by H2O2 (Fig. 6B), and it is consistent with the report that overexpression of SIRT1 in mammalian cells has no effect on cell viability after exposure to H2O2 (Solomon et al., 2006). The other explanation is that cytoplasm-localized SIRT1 enhancing apoptosis is more directly than nucleus-localized SIRT1 preventing apoptosis. In response to apoptotic stimuli, Ku70, p53 and FOXO3 are acetylated and then promote apoptosis (Sakaguchi et al., 1998; Abraham et al., 2000; Cohen et al., 2004a, 2004b). While SIRT1 deacetylates acetylated Ku70, p53 and FOXO and prevents apoptosis. Under apoptotic stimuli, cytoplasm-localized SIRT1 might lead damaged cells immediately to undergo apoptosis before deacetylation of Ku70, p53 and FOXO. Under 5 mg/ml puromycin stimulation, the treatment of zVAD increased the percentage of nonapoptotic cells with cytoplasm-localized SIRT1 (data not shown). This data suggested that SIRT1 exported out of nuclei before the onset of apoptosis, instead of passively permeation from the nucleus to cytoplasm. It was reported that depletion of NADþ level and the subsequent reduction of SIRT1 activity contributed to PARPmediated myocyte cell death (Pillai et al., 2005). Here we reported that overexpression of SIRT1 enhanced PARP cleavage in nocodazole induced apoptosis. It might be attributed to cytoplasm-localized SIRT1 enhanced apoptosis independent of its deacetylase activity, as GFP-SIRT1D238 HY still enhanced apoptosis induced by H2O2. Moreover, 5 mM nicotinamide, which effectively inhibits the deacetylase activity of SIRT1 (Luo et al., 2001), did not protect cells from apoptosis enhanced by GFP-SIRT1D238 or GFP-NES-SIRT1 (data not shown). Although LMB inhibited SIRT1 exportation out of nuclei under apoptotic stimuli and induced reduction of GFPSIRT1D238 or GFP-NES-SIRT1 in cytoplasm, but we could not clearly conclude whether LMB could attenuate this increased ability of GFP-SIRT1D238 or GFP-NES-SIRT1 to potentiate apoptosis, as LMB is toxic to cells and able to induce apoptosis (Jang et al., 2004). In this paper, we also showed cytoplasmlocalized SIRT1 enhanced apoptosis was caspase-dependent. The detailed mechanisms of how cytoplasm-localized SIRT1 enhanced apoptosis via activation of caspases need to be studied in the future. In mammals, caloric restriction delays the onset of numerous age-associated diseases such as cancer (Masoro, 2000). Depression of cancer by caloric restriction is associated with reduction of cell proliferation and enhanced rates of apoptosis (Hursting et al., 2003). SIRT1 increases in caloric restriction mice (Cohen et al., 2004b), and increased SIRT1 might mediate the reduction of cell proliferation by caloric restriction, as mammalian SIRT1 was reported to limit cell replication (Chua et al., 2005). Based on our findings, increased SIRT1 enhances apoptosis at metaphase during mitosis, which might result in that cancer cells are more sensitive to apoptosis than normal cells due to usually more cell division in cancer cells. The localization of SIRT1 in cytoplasm is cell selective and cytoplasm-localized SIRT1 was found in all three colon cancer cell lines we tested, including LoVo, SW480 and Ls-174-T (data not shown). As cytoplasm-localized SIRT1 enhances apoptosis, increased SIRT1 is benefit to prevent colon cancer. It is consistent with the report that SIRT1 is down-regulated in JOURNAL OF CELLULAR PHYSIOLOGY DOI 10.1002/JCP
colorectal tumors (Ozdag et al., 2006). All these data suggested that alteration of intracellular localization of SIRT1 from nuclei to cytoplasm might be a novel way for cancer therapy. In summary, SIRT1 is able to localize in cytoplasm mainly through translocation of SIRT1 from nuclei to cytoplasm, and cytoplasm-localized SIRT1 increases cell sensitivity to apoptosis. Acknowledgments
Q. Zhai and X. Shi are scholars of the Hundred Talents Program from Chinese Academy of Sciences. Q. Zhai is also a scholar of the Shanghai Rising-Star Program from Science and Technology Commission of Shanghai Municipality. Q. Jin is supported by a postdoctoral fellowship from K. C. Wong Education Foundation, Hong Kong. Literature Cited Abraham J, Kelly J, Thibault P, Benchimol S. 2000. Post-translational modification of p53 protein in response to ionizing radiation analyzed by mass spectrometry. J Mol Biol 295:853–864. Adrain C, Creagh EM, Martin SJ. 2001. Apoptosis-associated release of Smac/DIABLO from mitochondria requires active caspases and is blocked by Bcl-2. EMBO J 20:6627– 6636. Blander G, Guarente L. 2004. The Sir2 family of protein deacetylases. Annu Rev Biochem 73:417–435. Bordone L, Motta MC, Picard F, Robinson A, Jhala US, Apfeld J, McDonagh T, Lemieux M, McBurney M, Szilvasi A, Easlon EJ, Lin SJ, Guarente L. 2005. Sirt1 regulates insulin secretion by repressing UCP2 in pancreatic beta cells. PLoS Biol 4:e31. Brunet A, Sweeney LB, Sturgill JF, Chua KF, Greer PL, Lin Y, Tran H, Ross SE, Mostoslavsky R, Cohen HY, Hu LS, Cheng HL, Jedrychowski MP, Gygi SP, Sinclair DA, Alt FW, Greenberg ME. 2004. Stress-dependent regulation of FOXO transcription factors by the SIRT1 deacetylase. Science 303:2011–2015. Cheng HL, Mostoslavsky R, Saito S, Manis JP, Gu Y, Patel P, Bronson R, Appella E, Alt FW, Chua KF. 2003. Developmental defects and p53 hyperacetylation in Sir2 homolog (SIRT1)deficient mice. Proc Natl Acad Sci USA 100:10794–10799. Chua KF, Mostoslavsky R, Lombard DB, Pang WW, Saito S, Franco S, Kaushal D, Cheng HL, Fischer MR, Stokes N, Murphy MM, Appella E, Alt FW. 2005. Mammalian SIRT1 limits replicative life span in response to chronic genotoxic stress. Cell Metab 2:67–76. Cohen HY, Lavu S, Bitterman KJ, Hekking B, Imahiyerobo TA, Miller C, Frye R, Ploegh H, Kessler BM, Sinclair DA. 2004a. Acetylation of the C terminus of Ku70 by CBP and PCAF controls Bax-mediated apoptosis. Mol Cell 13:627–638. Cohen HY, Miller C, Bitterman KJ, Wall NR, Hekking B, Kessler B, Howitz KT, Gorospe M, de Cabo R, Sinclair DA. 2004b. Calorie restriction promotes mammalian cell survival by inducing the SIRT1 deacetylase. Science 305:390–392. Denu JM. 2003. Linking chromatin function with metabolic networks: Sir2 family of NAD(þ)dependent deacetylases. Trends Biochem Sci 28:41–48. Fischer U, Huber J, Boelens WC, Mattaj IW, Luhrmann R. 1995. The HIV-1 Rev activation domain is a nuclear export signal that accesses an export pathway used by specific cellular RNAs. Cell 82:475–483. Frye RA. 2000. Phylogenetic classification of prokaryotic and eukaryotic Sir2-like proteins. Biochem Biophys Res Commun 273:793–798. Fulco M, Schiltz RL, Iezzi S, King MT, Zhao P, Kashiwaya Y, Hoffman E, Veech RL, Sartorelli V. 2003. Sir2 regulates skeletal muscle differentiation as a potential sensor of the redox state. Mol Cell 12:51–62. Guarente L. 2000. Sir2 links chromatin silencing, metabolism, and aging. Genes Dev 14:1021– 1026. Hallows WC, Lee S, Denu JM. 2006. Sirtuins deacetylate and activate mammalian acetyl-CoA synthetases. Proc Natl Acad Sci U S A 103:10230–10235. Horton P, Nakai K. 1997. Better prediction of protein cellular localization sites with the k nearest neighbors classifier. Proc Int Conf Intell Syst Mol Biol 5:147–152. Hursting SD, Lavigne JA, Berrigan D, Perkins SN, Barrett JC. 2003. Calorie restriction, aging, and cancer prevention: mechanisms of action and applicability to humans. Annu Rev Med 54:131–152. Imai S, Armstrong CM, Kaeberlein M, Guarente L. 2000. Transcriptional silencing and longevity protein Sir2 is an NAD-dependent histone deacetylase. Nature 403:795–800. Jang BC, Paik JH, Jeong HY, Oh HJ, Park JW, Kwon TK, Song DK, Park JG, Kim SP, Bae JH, Mun KC, Suh MH, Yoshida M, Suh SI. 2004. Leptomycin B-induced apoptosis is mediated through caspase activation and down-regulation of Mcl-1 and XIAP expression, but not through the generation of ROS in U937 leukemia cells. Biochem Pharmacol 68:263–274. Jin QH, Zhao B, Zhang XJ. 2004. Cytochrome c release and endoplasmic reticulum stress are involved in caspase-dependent apoptosis induced by G418. Cell Mol Life Sci 61:1816–1825. la Cour T, Kiemer L, Molgaard A, Gupta R, Skriver K, Brunak S. 2004. Analysis and prediction of leucine-rich nuclear export signals. Protein Eng Des Sel 17:527–536. Lin SJ, Kaeberlein M, Andalis AA, Sturtz LA, Defossez PA, Culotta VC, Fink GR, Guarente L. 2002. Calorie restriction extends Saccharomyces cerevisiae lifespan by increasing respiration. Nature 418:344–348. Luo J, Nikolaev AY, Imai S, Chen D, Su F, Shiloh A, Guarente L, Gu W. 2001. Negative control of p53 by Sir2alpha promotes cell survival under stress. Cell 107:137–148. Masoro EJ. 2000. Caloric restriction and aging: an update. Exp Gerontol 35:299– 305. McBurney MW, Yang X, Jardine K, Bieman M, Th’ng J, Lemieux M. 2003a. The absence of SIR2alpha protein has no effect on global gene silencing in mouse embryonic stem cells. Mol Cancer Res 1:402–409. McBurney MW, Yang X, Jardine K, Hixon M, Boekelheide K, Webb JR, Lansdorp PM, Lemieux M. 2003b. The mammalian SIR2alpha protein has a role in embryogenesis and gametogenesis. Mol Cell Biol 23:38–54. Michishita E, Park JY, Burneskis JM, Barrett JC, Horikawa I. 2005. Evolutionarily conserved and nonconserved cellular localizations and functions of human SIRT proteins. Mol Biol Cell 16:4623–4635.
CYTOPLASM-LOCALIZED SIRT1 ENHANCES APOPTOSIS
Mostoslavsky R, Chua KF, Lombard DB, Pang WW, Fischer MR, Gellon L, Liu P, Mostoslavsky G, Franco S, Murphy MM, Mills KD, Patel P, Hsu JT, Hong AL, Ford E, Cheng HL, Kennedy C, Nunez N, Bronson R, Frendewey D, Auerbach W, Valenzuela D, Karow M, Hottiger MO, Hursting S, Barrett JC, Guarente L, Mulligan R, Demple B, Yancopoulos GD, Alt FW. 2006. Genomic instability and aging-like phenotype in the absence of mammalian SI RT6. Cell 124:315–329. Motta MC, Divecha N, Lemieux M, Kamel C, Chen D, Gu W, Bultsma Y, McBurney M, Guarente L. 2004. Mammalian SIRT1 represses forkhead transcription factors. Cell 116:551–563. Moynihan KA, Grimm AA, Plueger MM, Bernal-Mizrachi E, Ford E, Cras-Meneur C, Permutt MA, Imai S. 2005. Increased dosage of mammalian Sir2 in pancreatic beta cells enhances glucose-stimulated insulin secretion in mice. Cell Metab 2:105–117. Nikolaev AY, Li M, Puskas N, Qin J, Gu W. 2003. Parc: a cytoplasmic anchor for p53. Cell 112:29–40. Ohsawa S, Miura M. 2006. Caspase-mediated changes in Sir2alpha during apoptosis. FEBS Lett 580:5875–5879. Ozdag H, Teschendorff AE, Ahmed AA, Hyland SJ, Blenkiron C, Bobrow L, Veerakumarasivam A, Burtt G, Subkhankulova T, Arends MJ, Collins VP, Bowtel D, Kouzarides T, Brenton JD, Caldas C. 2006. Differential expression of selected histone modifier genes in human solid cancers. BMC Genom 7:90. Picard F, Kurtev M, Chung N, Topark-Ngarm A, Senawong T, Machado De Oliveira R, Leid M, McBurney MW, Guarente L. 2004. Sirt1 promotes fat mobilization in white adipocytes by repressing PPAR-gamma. Nature 429:771–776. Pillai JB, Isbatan A, Imai S, Gupta MP. 2005. Poly(ADP-ribose) polymerase-1-dependent cardiac myocyte cell death during heart failure is mediated by NADþ depletion and reduced Sir2alpha deacetylase activity. J Biol Chem 280:43121–43130. Rodgers JT, Lerin C, Haas W, Gygi SP, Spiegelman BM, Puigserver P. 2005. Nutrient control of glucose homeostasis through a complex of PGC-1alpha and SIRT1. Nature 434:113–118.
JOURNAL OF CELLULAR PHYSIOLOGY DOI 10.1002/JCP
Rosenberg MI, Parkhurst SM. 2002. Drosophila Sir2 is required for heterochromatic silencing and by euchromatic Hairy/E(Spl) bHLH repressors in segmentation and sex determination. Cell 109:447–458. Sakaguchi K, Herrera JE, Saito S, Miki T, Bustin M, Vassilev A, Anderson CW, Appella E. 1998. DNA damage activates p53 through a phosphorylation–acetylation cascade. Genes Dev 12:2831–2841. Solomon JM, Pasupuleti R, Xu L, McDonagh T, Curtis R, Distefano PS, Huber LJ. 2006. Inhibition of SIRT1 catalytic activity increases p53 acetylation but does not alter cell survival following DNA damage. Mol Cell Biol 26:28–38. Tanno M, Sakamoto J, Miura T, Shimamoto K, Horio Y. 2007. Nucleo-cytoplasmic shuttling of NADþ-dependent histone deacetylase SI RT1. J Biol Chem 282:6823–6832. van der Horst A, Tertoolen LG, de Vries-Smits LM, Frye RA, Medema RH, Burgering BM. 2004. FOXO4 is acetylated upon peroxide stress and deacetylated by the longevity protein hSir2(SIRT1). J Biol Chem 279:28873–28879. Vaziri H, Dessain SK, Ng Eaton E, Imai SI, Frye RA, Pandita TK, Guarente L, Weinberg RA. 2001. hSIR2(SIRT1) functions as an NAD-dependent p53 deacetylase. Cell 107:149– 159. Wang J, Zhai Q, Chen Y, Lin E, Gu W, McBurney MW, He Z. 2005. A local mechanism mediates NAD-dependent protection of axon degeneration. J Cell Biol 170:349– 355. Wilson JM, Le VQ, Zimmerman C, Marmorstein R, Pillus L. 2006. Nuclear export modulates the cytoplasmic Sir2 homologue Hst2. EMBO Rep 7:1247–1251. Yang Y, Hou H, Haller EM, Nicosia SV, Bai W. 2005. Suppression of FOXO1 activity by FHL2 through SIRT1-mediated deacetylation. EMBO J 24:1021–1032. Yeung F, Hoberg JE, Ramsey CS, Keller MD, Jones DR, Frye RA, Mayo MW. 2004. Modulation of NF-kappaB-dependent transcription and cell survival by the SIRT1 deacetylase. EMBO J 23:2369–2380.
97
