Oncogene (2008) 27, 6707–6719 & 2008 Macmillan Publishers Limited All rights reserved 0950-9232/08 $32.00 www.nature.com/onc ORIGINAL ARTICLE Cooperative effect of p21Cip1/WAFÀ1 and 14-3-3r on arrest and induction byp14 ARF

PG Hemmati1,2,3, G Normand1,3, B Gillissen1,2, J Wendt1,BDo¨ rken1,2 and PT Daniel1,2

1Department of Hematology, Oncology and Tumor Immunology, University Medical Center Charite´, Campus Berlin-Buch, Germany and 2Clinical and Molecular Oncology, Max-Delbru¨ck-Center for Molecular Medicine, Berlin-Buch, Germany

P14ARF (p19ARF in the mouse) plays a central role in the Introduction regulation of cellular proliferation. Although the capacity of to induce a cell cycle arrest in depends Genotoxic stress induced by chemodrugs or ionizing on a functional /-signaling axis, the G2 arrest irradiation triggers an arrest in the of the cell triggered byp14 ARF is p53/p21-independent. Using iso- cycle to allow for the repair of damaged DNA. In most geneic HCT116 cells either wild-type or homozygously situations, this process depends on a functional p53/p21- deleted for p21, 14-3-3r or both, we further investigated signaling axis. In turn, cells deficient for p21 are the cooperative effect of p21 and 14-3-3r on cell cycle hampered in their ability to arrest in the G2 phase regulation and apoptosis induction byp14 ARF. In contrast following DNA damage and rapidly undergo mitotic to DNA damage, which induces mitotic catastrophe in catastrophe and subsequent cell death (Waldman et al., 14-3-3r-deficient cells, we show here that the expression 1996; Bunz et al., 1998; Bhonde et al., 2006b). In turn, of p14ARF triggers apoptotic cell death, as evidenced by forced expression of p21 strongly attenuates apoptosis nuclear DNA fragmentation and induction of pan-caspase sensitivities on DNA damage (Gorospe et al., 1996; activities, irrespective of the presence or absence of Martinez et al., 2002; Wendt et al., 2006). In addition to 14-3-3r. The activation of the intrinsic mitochondrial p21, 14-3-3s was shown to play a key role in regulating apoptosis pathwaybyp14 ARF was confirmed bycyto- G2/M checkpoint function (Hermeking et al., 1997). On chrome c release from mitochondria and induction of transcriptional upregulation by p53, 14-3-3s impairs caspase-9- (LEHDase) and caspase-3/7-like (DEVDase) mitotic entry by sequestering B1/cdc2 complexes activities. Moreover, 14-3-3r/p21 double-deficient cells (Hermeking et al., 1997; Chan et al., 1999; Hermeking, were exceedinglysensitive to apoptosis induction by 2003). While the expression of 14-3-3s blocks cells in the p14ARF as compared to wild-type cells or cells lacking G2 phase of the cell cycle, cells deficient for 14-3-3s are either gene alone. Notably, p14ARF-induced apoptosis was unable to maintain a stable G2 arrest, prematurely enter preceded byan arrest in the G2 phase of cell cycle,which and die by mitotic catastrophe (Chan et al., coincided with downregulation of cdc2 (cdk1) 1999). Notably, cells lacking both p21 and 14-3-3s are expression and lack of its nuclear localization. This highly sensitive to DNA damage-induced cell death, indicates that p14ARF impairs mitotic entrybytargeting indicating that p21 and 14-3-3s play complementary the distal DNA damage-signaling pathwayand induces roles and critically determine the response of the cells to apoptotic cell death, rather than mitotic catastrophe, out DNA damage (Chan et al., 2000; Hermeking, 2003). of a transient G2 arrest. Furthermore, our data delineate The p14ARF tumor suppressor (p19ARF in the mouse) is that the disruption of G2/M cell cycle checkpoint control encoded at the mammalian INK4a locus (Duro et al., criticallydetermines the sensitivityof the cell toward 1995; Mao et al., 1995; Quelle et al., 1995). Physiolo- p14ARF-induced mitochondrial apoptosis. gically, p14ARF (p19ARF) is rapidly upregulated on Oncogene (2008) 27, 6707–6719; doi:10.1038/onc.2008.193; oncogenic stress (de Stanchina et al., 1998; Radfar published online 22 September 2008 et al., 1998; Zindy et al., 1998). By stabilizing p53 through inhibition of hdm-2 (mdm-2), p14ARF triggers Keywords: p14ARF; 14-3-3s, p21; cell cycle; apoptosis; the expression of p53 target genes including p21 and mitotic catastrophe bax. Ectopic expression of p19ARF was shown to induce a p53-dependent arrest in the G1 and G2 phases of the cell cycle (Quelle et al., 1995). Similarly, the p19ARF- mediated induction of cycle arrest or apoptosis follow- Correspondence: Dr PT Daniel, Clinical and Molecular Oncology, ing oncogene expression depends on functional p53. Max-Delbru¨ ck-Center for Molecular Medicine, and University Med- Nonetheless, data from several in vitro models suggest ical Center Charite´ , Campus Berlin-Buch, Lindenberger Weg 80, 13125 that p14ARF (p19ARF) regulates cell cycle arrest and Berlin-Buch, Germany. apoptosis in a p53-independent manner as well (Eymin E-mail: [email protected] 3These authors contributed equally to this work. et al., 2001, 2003; Martelli et al., 2001; Hemmati et al., ARF Received 24 September 2007; revised 5 December 2007; accepted 11 2002; Normand et al., 2005). Furthermore, p14 was April 2008; published online 22 September 2008 shown to trigger the DNA damage-signaling pathway 14-3-3r and p21 in p14ARF-induced cell cycle arrest and apoptosis PG Hemmati et al 6708 through ATR/Chk-1 or ATM in a p53-independent of p21 protein in the 14-3-3s KO cells were not induced manner (Li et al., 2004; Rocha et al., 2005; Rocha and further on p14ARF expression. As expected, 14-3-3s Perkins, 2005; Eymin et al., 2006a, b), indicating that levels were induced on infection with Ad-p14ARF in the p14ARF may control cellular proliferation in a way much WT cells. In contrast, p21 KO cells exhibited high basal more complex than initially thought. levels of 14-3-3s protein, which were not further Although the G1 cell cycle arrest induced by p14ARF inducible by p14ARF. This constitutively high expression depends on a functional hdm-2/p53/p21-signaling axis, of 14-3-3s in the p21 KO cells may result from the expression of p14ARF in p53- and/or p21-deficient cells increased p53, which was reported to be hyperfunctional triggers an arrest in the G2 phase through downregulat- (Blagosklonny et al., 2002). Taken together, these results ing cdc2 (cdk1) expression and functional activity (Eymin indicate that the loss of 14-3-3s does not affect the et al., 2003, 2006a; Hemmati et al., 2005; Normand et al., ability of p14ARF to induce a G2/M cell cycle arrest in 2005). This is in line with the finding that p14ARF triggers p53 WT cells. the DNA damage-signaling pathway in a p53/p21- As compared to WT, p21 KO or 14-3-3s KO cells, independent manner (Rocha et al., 2003, 2005; Rocha there was no decrease in the relative number of cells in and Perkins, 2005; Eymin et al., 2006b). As 14-3-3s is the of the cell cycle on expression of p14ARF in involved in the sequestration of /cdc2 com- the DKO cells (Figure 1). This suggests that these cells plexes, the loss of 14-3-3s, either alone or in combination are impaired in the S phase control and still proliferate with loss of p21, may sensitize for p14ARF-induced following p14ARF expression. To confirm this hypothesis, cell death by deregulating G2/M checkpoint function HCT116 cells were infected with Ad-p14ARF (25 MOI) (Hermeking et al., 1997; Graves et al., 2001; Hermeking, for 48 h, pulse-labeled with 5-bromo-2-deoxyuridine 2003). We therefore investigated whether the absence of (BrdU) for 30 min and analysed by flow cytometry 14-3-3s, alone or in conjunction with loss of p21, alters (Figure 2a). As shown in Figure 2b, quantification of the cell cycle behavior and apoptosis sensitivities on BrdU-positive cells as percent of cells in S phase p14ARF expression. revealed that mock-treated or control vector-infected cells deficient for either p21, 14-3-3s or both displayed an overall higher degree of BrdU incorporation as compared to the WT cells. In particular, cells lacking Results p21, that is p21 KO and p21/14-3-3s DKO cells, exhibited the highest rates of BrdU incorporation. Induction of G2/M arrest by p14ARF is independent of p21 Although DKO cells were impaired to undergo growth and 14-3-3s arrest on p14ARF expression, a substantial decrease in the HCT116 cells, either wild-type (WT) or homozygously relative number of BrdU-positive cells was observed in deleted for either 14-3-3s (14-3-3s KO), p21 (p21 KO) the WT, p21 KO or 14-3-3s KO cells on p14ARF or both p21 and 14-3-3s (double knockout (DKO)), expression. This suggests that the combined loss of both were infected with an adenoviral vector expressing p21 and 14-3-3s impairs the S phase control following p14ARF (Ad-p14ARF) at 25 multiplicity of infection expression of p14ARF. (MOI) in parallel with mock-treated (control) or control Expression of p14ARF induced the accumulation of vector-infected cells (Ad-lacZ; 25 MOI) and assayed for cells in the G2/M phase irrespective of the presence or cell cycle distribution after 48 h (Figures 1a and b). absence of 14-3-3s and/or p21, as evidenced by an Although WT cells arrested in the G1 phase of the cell increase in the number of cells with a 4N DNA content. cycle showed a corresponding decrease in the relative To differentiate between cells arrested in the G2 phase number of cells in the S phase, p21 KO, 14-3-3s KO or versus M phase, we employed the bivariate analysis of DKO cells accumulated in the G2/M phase on expres- DNA content and expression of mitotic protein mono- sion of p14ARF. Western blot analysis revealed that clonal-2 (MPM-2) phosphoepitopes, which are selec- p14ARF induced both p53 and p21 protein levels in the tively exposed during mitosis (Taagepera et al., 1993). WT cells (Figure 1c). Notably, cells lacking functional As shown in Figure 3a, infection with Ad-p14ARF p21, that is, p21 KO and DKO cells, readily expressed (25 MOI) failed to induce MPM-2 expression in arrested high levels of p53 protein, as reported previously tetraploid HCT116 cells irrespective of the presence or (Giannakakou et al., 2001), which did not increase absence of p21 and/or 14-3-3s after 48 h. However, further on p14ARF expression. Similarly, the high levels treatment with 0.1 mg/ml nocodazole was paralleled by

Figure 1 Induction of G2/M cell cycle arrest by p14ARF is independent of p21 and 14-3-3s.(a) HCT116 colorectal cancer cells, wild-type (WT) or homozygously deleted for either p21 (p21 KO), 14-3-3s (s KO) or both (DKO), were infected with an adenoviral vector expressing p14ARF (Ad-p14ARF; 25 MOI) for 48 h. In parallel, cells were either treated mock (control) or transduced with a control vector expressing b-galactosidase (Ad-lacZ; 25 MOI). Cell cycle distribution was analysed by flow cytometry. The relative number of cells in each phase of the cell cycle (G1, S, G2/M) is given in percent. Representative histograms are shown. (b) Cells were treated as described in (a). Data represent the mean of the relative number of cells in the G2/M phase of the cell cycle±s.d. from three experiments. (c) Cells were infected with the indicated adenoviruses for 48 h. Total cellular protein (20 mg/lane) was subjected to SDS–polyacrylamide gel electrophoresis and western blot analysis with antibodies against p14ARF, p53, p21 and 14-3-3s. Equal loading was confirmed by re- probing with an antibody against actin. MOI, multiplicity of infection.

Oncogene 14-3-3r and p21 in p14ARF-induced cell cycle arrest and apoptosis PG Hemmati et al 6709

Oncogene 14-3-3r and p21 in p14ARF-induced cell cycle arrest and apoptosis PG Hemmati et al 6710

Figure 2 DNA synthesis is not impaired in p21/14-3-3s double-deficient cells on p14ARF expression. (a) HCT116 colorectal cancer cells were infected with Ad-p14ARF (ARF; 25 MOI) for 48 h. In parallel, cells were either treated mock (control) or infected with a control vector expressing b-galactosidase (lacZ; 25 MOI). Before harvest, cells were pulse-labeled with 5-bromo-2-deoxyuridine (BrdU). Incorporation of BrdU across the cell cycle was analysed by flow cytometry using a FITC-labeled monoclonal antibody. BrdU-positive cells were gated and counted as the percent of cells in the S phase. (b) Cells were treated as described in (a). Data represent the mean of the relative number of BrdU-positive cells in the S phase of the cell cycle±s.d. from three experiments. FITC, fluorescein-isothiocyanate; MOI, multiplicity of infection.

an increase of MPM-2-positive tetraploid cells. In fluorescence after staining with Hoechst 33258 contrast, the relative number of tetraploid M-phase (Figure 3c). In contrast, a strong increase in the number cells was markedly reduced upon p14ARF expression of mitotic cells was observed upon treatment with (Figure 3b). Consistently, HCT116 cells of all genotypes 0.1 mg/ml nocodazole for 18 h. Western blot analysis lacked condensed chromatin or mitotic figures 48 h on revealed that p14ARF triggered the downregulation of infection with Ad-p14ARF when studied by immuno- cdc2 protein expression in G2-arrested p21 KO, 14-3-3s

Oncogene 14-3-3r and p21 in p14ARF-induced cell cycle arrest and apoptosis PG Hemmati et al 6711

Figure 3 Induction of the G2 but not the M phase arrest by p14ARF in cells lacking p21 and/or 14-3-3s.(a) HCT116 cells were transduced with the indicated adenoviral vectors at 25 MOI or treated mock (control) for 48 h. Incubation with nocodazole (0.1 mg/ml) for 18 h was used to arrest cells in mitosis. Bivariate analysis of cells stained with a FITC-conjugated antibody against MPM-2 and DNA content (propidium iodide; PI) was performed, and mitotic cells were identified by a 4N DNA content and MPM-2 positivity. Representative histograms are shown. (b) Quantification of three independent experiments as described in (a). Bars denote the mean±s.d. (c) HCT116 cells were infected with Ad-p14ARF (ARF) or Ad-lacZ (lacZ) control vector for 48 h. To arrest cells in mitosis, cells were treated with 0.1 mg/ml nocodazole (Noc) for 16 h. Mitotic cells displaying condensed chromatin were identified by fluorescence microscopy after staining with Hoechst 33258. Representative high-power fields (magnification  400) are shown. (d) HCT116 cells were infected with the indicated adenoviral vectors at 25 MOI and harvested 48 h after infection. Total cellular protein (20 mg/lane) was subjected to SDS–polyacrylamide gel electrophoresis and western blot analysis with the indicated antibodies. Re-probing with an antibody against actin served as a loading control. FITC, fluorescein-isothiocyanate; MOI, multiplicity of infection.

KO and DKO cells, whereas cdc2 protein levels Progression from the G2 phase to mitosis depends remained stable in G1-arrested WT cells 48 h upon on the timely activation of the cdc2 kinase. This involves expression of p14ARF (Figure 3d). Notably, expression of the upregulation of cyclin B1 expression and the cyclin B1 decreased in WT, 14-3-3s KO and DKO cells, removal of inhibitory phosphorylation of cdc2, but not in p21 KO cells. In contrast, p14ARF induced a that is at Tyr15, prior to the nuclear translocation of decrease in cdc25C protein expression in all HCT116 cyclin B1/cdc2 complexes (Jin et al., 1996, 1998; sublines. These data indicate that p14ARF induces an Takizawa and Morgan, 2000). As shown in Figure 4, arrest in the G2 phase but not in the M phase of the cell expression of p14ARF (Ad-p14ARF, 25 MOI) abrogated cycle, which is independent of both p21 and 14-3-3s and nuclear localization of cdc2 as compared to Ad-lacZ dissociates from cyclin B1 and cdc25C expressions. The (25 MOI)-infected cells 36 h after transduction (green; G2 arrest correlated, however, with the downregulation left panel). Similarly, p14ARF-expressing cells lacked of cdc2 protein expression. phosphohistone H3 (Ser10) expression (red; middle

Oncogene 14-3-3r and p21 in p14ARF-induced cell cycle arrest and apoptosis PG Hemmati et al 6712

Figure 4 Expression of p14ARF abrogates nuclear localization of cdc2. HCT116 cells were grown on cover slides and infected with Ad- p14ARF (25 MOI) or Ad-lacZ control vectors (25 MOI) for 36 h. Cellular localization of cdc2 (green; left panel) and phospho-histone H3 (Ser10) (red; middle panel) was studied by immunofluorescence. Nuclei were counterstained with Hoechst 33258 (right panel). Representative high-power fields (magnification  400) are shown. MOI, multiplicity of infection.

Oncogene 14-3-3r and p21 in p14ARF-induced cell cycle arrest and apoptosis PG Hemmati et al 6713 panel) and chromatin condensation or mitotic figures (blue; right panel). This indicates that p14ARF abrogates G2/M progression by downmodulating cdc2 expression and by interfering with its nuclear localization regardless of the presence or absence of 14-3-3s or p21. Cells void of 14-3-3s succumb to mitotic catastrophe on prolonged exposure to doxorubicin (Chan et al., 1999, 2000). We therefore investigated whether p14ARF induces mitotic catastrophe in cells lacking 14-3-3s, that is sKO and DKO cells, as compared to WT cells or cells solely deficient for p21. HCT116 WT, p21 KO, 14-3-3s KO and DKO cells were either infected with Ad-p14ARF (25 MOI) or treated mock (Co: medium; lacZ: Ad-lacZ, 25 MOI) for 72 h. As a positive control, cells were exposed to epirubicin (0.1 mg/ml) for 72 h. Staining with Hoechst 33258 revealed that p14ARF-expressing cells lacked micronucleation or enlarged nuclei indicative of mitotic catastrophe (Bhonde et al., 2006a) (Figure 5). As expected, expression of p14ARF was paralleled by a reduction in the number of mitotic cells as compared to mock-treated or control vector-infected cells. In con- trast, prolonged exposure to epirubicin resulted in a substantial increase in the number of 14-3-3s KO or DKO cells displaying micronucleation and enlarged nuclei. This was not the case in the p21 KO or the WT cells. This indicates that in contrast to treatment with the DNA-damaging agents, expression of p14ARF does not induce mitotic catastrophe in cells void of 14-3-3s. This may be due to the downregulation of cdc2 expression and interference with mitotic entry by p14ARF. Figure 5 Expression of p14ARF does not induce mitotic cata- Prolonged expression of p14ARF resulted in a progres- strophe in 14-3-3s-deficient cells. HCT116 cells were infected with sive loss of viable cells. To clarify if this occurs through the indicated adenoviral vectors (25 MOI) or treated mock (Co). As a positive control, cells were exposed to epirubicin (0.1 mg/ml). induction of apoptosis, HCT116 cells were infected with Cells displaying mitotic chromatin or micronucleation and swollen ARF Ad-p14 (25 MOI) and assayed for apoptotic DNA hypertrophic nuclei, the latter two indicative of mitotic cata- fragmentation. As shown in Figures 6a and b, infection strophe, were quantified by staining nuclei with Hoechst 33258 with Ad-p14ARF induced apoptotic DNA fragmentation at 72 h after infection. Bars represent the mean±s.d. from at least after 72 h. Notably, apoptosis was substantially aug- 3 Â 100 cells counted in high-power fields. MOI, multiplicity of infection. mented in p21-deficient cells as compared to WT cells. Likewise, the single loss of 14-3-3s resulted in a moderate increase in cells displaying fragmented geno- were most strongly induced in cells lacking both p21 and mic DNA, whereas cells lacking both p21 and 14-3-3s ARF 14-3-3s. Taken together, the combined loss of p21 were most sensitive to p14 -induced apoptosis. This and 14-3-3s substantially augments p14ARF-induced effect occurred in a dose-dependent manner (Figure 6c) mitochondrial apoptosis, which involves the release of starting as early as 48 h after expression and further cytochrome c subsequently followed by the induction of increased at 72 h (Figure 6d). caspase-9- and caspase-3/7-like activities. As compared to WT cells, loss of p21 or 14-3-3s enhanced the induction of pan-caspase activities on expression of p14ARF, as determined by the use of a fluorescein-isothiocyanate (FITC)-labeled derivative of Discussion the pan-caspase substrate zVADfmk (Figures 7a and b). The combined loss of both p21 and 14-3-3s markedly Members of the 14-3-3 family of play essential enhanced pan-caspase activities in a dose-dependent roles in the regulation of cell cycle control and apoptosis manner (Figure 7c). As shown in Figure 7d, this was sensitivities (Hermeking, 2003). Upon DNA damage, preceded by the release of cytochrome c into the cytosol ataxia telangiectasia mutated (ATM) triggers the activa- 48 h on expression of p14ARF. In the same vein, p14ARF- tion of the chk2 kinase, which leads to the phosphoryla- induced apoptosis was paralleled by the induction tion of cdc25C on Ser216. This allows for the of caspase-9-like (LEHDase) (data not shown) and sequestration of the cdc25C phosphatase through binding caspase-3/-7-like (DEVDase) activities (Figure 7e). to 14-3-3s. In turn, the depression of cdc25C inhibits G2/ In analogy, both LEHDase and DEVDase activities M progression by impeding the dephosphorylation of

Oncogene 14-3-3r and p21 in p14ARF-induced cell cycle arrest and apoptosis PG Hemmati et al 6714

Figure 6 Apoptotic DNA fragmentation induced by p14ARF is enhanced in the absence of p21 and 14-3-3s.(a) HCT116 cells were infected with 25 MOI of the indicated adenoviral vector or treated mock (control). Cells were harvested 72 h after infection and assayed for apoptotic DNA fragmentation by flow cytometry. Representative DNA histograms are shown. (b) Cells were treated as described in (a). Bars represent the mean±s.d. from three experiments. (c) HCT116 cells were infected with an increasing MOI of Ad-p14ARF, harvested 72 h after infection, and assayed for apoptotic DNA fragmentation by flow cytometry. (d) HCT116 cells were infected at an MOI of 25 and assayed for apoptotic DNA fragmentation at the indicated time points. The data given represent the mean±s.d. from three experiments. MOI, multiplicity of infection.

cdc2 (Peng et al., 1997; Sanchez et al., 1997; Zeng and G2 phase of the cell cycle, irrespective of the presence or Piwnica-Worms, 1999; Graves et al., 2001). Conse- absence of 14-3-3s. Notably, cells deficient for p21 or quently, the loss of 14-3-3s, that is by somatic gene for both p21 and 14-3-3s also undergo a G2 arrest knockout, impairs the induction of a stable G2 cell cycle underscoring that both p53 and p21 are dispensable for arrest on treatment with DNA-damaging agents. Instead, executing a p14ARF-induced G2 arrest (Hemmati et al., these cells prematurely enter mitosis and subsequently die 2005; Normand et al., 2005). The failure of these G2- following mitotic catastrophe (Chan et al., 1999, 2000). arrested cells to enter mitosis is underlined by the In contrast, we show here that the forced expression decreased number of cells expressing the MPM-2 of p14ARF induces the accumulation of cells in the phosphoepitope, that is mitotic cells displaying a

Oncogene 14-3-3r and p21 in p14ARF-induced cell cycle arrest and apoptosis PG Hemmati et al 6715

Figure 7 Induction of pan-caspase activity during p14ARF-induced apoptosis. (a) HCT116 sublines were infected with the indicated adenoviral vectors at a MOI of 50 or treated mock (control) for 72 h. The relative number of cells displaying pan-caspase activation was determined by flow cytometry using an FITC-labeled derivative of the pan-caspase substrate VADfmk. Representative histograms are shown. (b) HCT116 cells were infected as described in (a). Bars denote the means of cells displaying pan-caspase activation±s.d. from three experiments. (c) HCT116 cells were infected with an increasing MOI of Ad-p14ARF, harvested 72 h after infection, and assayed for pan-caspase activation by flow cytometry. Data represent the mean of cells displaying pan-caspase activation±s.d. from three experiments. (d) HCT116 cells were infected with the indicated adenoviral vectors at a MOI of 50 or treated mock (control) for 48 h. Cytochrome c release from the mitochondria was determined by western blot analysis of cytosolic protein extracts. (e) Caspase-3/7-like (DEVDase) activities were determined by the use of a fluorescein (FAM)-coupled derivative of zDEVDfmk 72 h after infection with Ad-p14ARF (ARF). Bars represent the mean±s.d. from three experiments. FITC, fluorescein-isothiocyanate; MOI, multiplicity of infection.

Oncogene 14-3-3r and p21 in p14ARF-induced cell cycle arrest and apoptosis PG Hemmati et al 6716 or other upstream signaling components of the DNA damage-signaling pathway controlling the cyclin B1/ cdc2/cdc25C rheostat (Hermeking, 2003). This further enforces the hypothesis that p14ARF regulates G2/M progression by targeting downstream components of the DNA damage-signaling cascade, that is cdc2 (Eymin et al., 2003; Normand et al., 2005). However, this does not exclude that p14ARF also acts upstream of cyclin B1/ cdc2, that is by regulation of cdc25C (Eymin et al., 2006a). Consequently, p14ARF was shown to link oncogene-triggered signaling events to the DNA- damage pathway by inducing ATR/Chk1 interaction and nucleolar sequestration of ATR/BRCA1 complexes (Rocha et al., 2005). Likewise, ATM may contribute to the antiproliferative effects of p14ARF (Li et al., 2004), suggesting that, based on their functional redundancy, both ATM/chk2 and ATR/chk1 may be involved in the p53-independent DNA-damage response triggered by p14ARF (Eymin et al., 2006b). Notably, we observed the p21/14-3-3s-independent downregulation of cdc25C following p14ARF expression. Previous data establish that p14ARF-induced apoptosis is independent of p53 (Hemmati et al., 2002). Instead, loss of functional p21 strongly sensitized for p14ARF- induced cell death indicating that intact control at the G1/S and G2/M boundaries is critical (Hemmati et al., 2005). Therefore, we investigated whether the loss of 14-3-3s, an important regulator at the G2/M restriction point, influences p14ARF-induced cell death. Indeed, loss of 14-3-3s substantially en- hanced sensitivity toward p14ARF-induced apoptosis, Figure 8 Model for p14ARF-induced cell cycle arrest and apopto- which was further augmented by the additional loss sis. An arrest in the G1 phase of the cell cycle by p14ARF primarily depends on p53 and p21 (Normand et al., 2005). In contrast, the of p21. This is in line with a study investigating induction of a G2 cell cycle arrest is mediated by the functional the apoptosis sensitivity of cells deficient for p21 and/ inactivation of cdc2 (cdk-1) by p14ARF, particularly in cells lacking or 14-3-3s as compared to WT cells (Chan et al., 2000). functional p53 and/or p21. This p14ARF-induced G2 arrest There, p21/14-3-3s double-deficient cells were most is followed by apoptotic cell death (a). Inactivation of cdc2 impairs sensitive toward apoptosis induction by exogenous p53 mitotic entry. Therefore, and in contrast to DNA damage (b), 14-3-3s is dispensable for blocking mitotic catastrophe during or 5-FU as compared to cells lacking either gene alone. p14ARF-induced cell death. Nevertheless, cells lacking p21 and/or The same group showed that 14-3-3s-deficient cells 14-3-3s are more sensitive toward p14ARF-induced apoptosis due to succumb to mitotic catastrophe on exposure to DNA- the lack of a stable G1 arrest on forced expression of p14ARF. damaging agents (Chan et al., 1999, 2000). Here, expression of p14ARF failed to induce morphological features of mitotic catastrophe (Bhonde et al., 2006a). 4N DNA content on expression of p14ARF. This was Instead, these cells showed condensed nuclei typical for paralleled by the absence of cells displaying mitotic apoptotic cell death. In contrast, loss of 14-3-3s allowed figures. These findings are well in line with the for the induction of cell death by mitotic catastrophe observation that cdc2 is downregulated in p21 and following exposure to doxorubicin. This is remarkable 14-3-3s single- or double-deficient cells. Consequently, considering that both anticancer drugs and p14ARF were cells arrested in the G2 phase on expression of p14ARF shown to trigger a DNA-damage response by activation failed to accumulate nuclear cdc2, which in turn impairs of checkpoint kinases (Rocha et al., 2005; Rocha and mitotic entry. In the same vein, G2-arrested cells lack Perkins, 2005). Moreover, p14ARF-induced cell death is phosphohistone H3. Given that the nuclear accumula- subject to inhibition by Bcl-xL (Hemmati et al., 2002) tion of cyclin B1/cdc2 is a prerequisite for G2/M and relies largely on Bax and/or Bak (Hemmati et al., progression (Jin et al., 1996, 1998; Morgan et al., 2006). This, together with the massive induction of 1998), this explains why p14ARF-expressing cells arrest in caspase activities, indicates that p14ARF induces apopto- G2 phase and cannot progress to mitosis. tic cell death and not mitotic catastrophe. This and the Furthermore, this explains why 14-3-3s is dispensable decrease of cells entering mitosis delineate that apopto- for a G2 arrest in this experimental setting, as the sis occurs out of G2 phase of the cell cycle without prior downregulation of total cdc2 protein expression and its mitotic entry. cytoplasmic localization are sufficient to execute a G2 Nonetheless, the question arises how the loss of arrest irrespective of the presence or absence of 14-3-3s 14-3-3s sensitizes for apoptosis induction following

Oncogene 14-3-3r and p21 in p14ARF-induced cell cycle arrest and apoptosis PG Hemmati et al 6717 expression of p14ARF. Loss of 14-3-3s allows for an p14ARF induces a G2 arrest by downregulating cdc2 untimed exit from the G2 phase following DNA expression, which may explain why mitotic entry is damage. This results in massive cell death by mitotic impaired. This downregulation of cdc2 expression by catastrophe (Chan et al., 1999, 2000). In contrast, the p14ARF occurs on the post-translational level by desta- expression of p14ARF triggers the downregulation of bilizing the cdc2 protein (Eymin et al., 2003; Normand cdc2, which primarily arrests cells in the G2 phase. In et al., 2005). In turn, downregulation of cdc2 expression the 14-3-3s-deficient cells, this is paralleled by a by p14ARF may abrogate the untimed entry of cdc2 into decrease in cyclin B1 expression. However, cells do not the nucleus even in checkpoint-deficient cells that lack progress into the M phase, but die. This suggests that p21 and/or 14-3-3s. Cdc2-mediated mitotic entry has p14ARF generates conflicting G2/M checkpoint signals, been proposed to play a central role in mitotic which then trigger the mitochondrial apoptosis pathway catastrophe (Chan et al., 1999, 2000). Therefore, our out of the G2 phase. Furthermore, 14-3-3s was shown data are in agreement with the notion that 14-3-3s is to sequester the pro-apoptotic BH3-only protein Bad, required to prevent mitotic catastrophe following DNA which may attenuate apoptosis sensitivities through a damage. However, the p14ARF-induced G2 arrest is relative increase of anti-apoptotic Bcl-2 or Bcl-xL (Zha independent of 14-3-3s (and p21). Instead of promoting et al., 1996; Datta et al., 2000). In the same vein, 14-3-3s mitotic catastrophe, loss of 14-3-3s and/or p21 facili- was proposed to negatively regulate the functional tated a G2 arrest followed by apoptotic cell death out of activity of Bax (Samuel et al., 2001; Nomura et al., the G2 phase but not out of the M phase of the cell cycle 2003). (Figure 8). This appears to be related with the As reported previously, HCT116 cells lacking p21 sequestration of cdc2 by p14ARF in checkpoint-deficient express high levels of p53 (Figure 1c) (Giannakakou cells. et al., 2001). Furthermore, p53 was shown to be hyperfunctional in these cells (Blagosklonny et al., 2002). Consequently, loss of p21 may sensitize for Materials and methods p14ARF-induced apoptosis through activation of a p53- triggered mitochondrial death pathway (Hemmati et al., Cell culture 2006). This would explain the increased sensitivities of HCT116 WT cells and their isogeneic sublines, HCT116 the DKO cells toward p14ARF-induced apoptosis as p21–/– (p21 KO) (Waldman et al., 1995), HCT116 compared to the 14-3-3s KO cells. In the same vein, loss 14-3-3s–/– (14-3-3s KO) (Chan et al., 1999) and HCT116 of p21 disables the induction of a p53-mediated cell 14-3-3s–/–/p21–/– (DKO) (Chan et al., 2000), were generously provided by Dr Bert Vogelstein, Johns Hopkins Cancer cycle arrest. This, in turn, was shown to sensitize for Center, Baltimore, MD, USA. Cells were cultured in McCoy’s apoptosis induction following diverse pro-apoptotic 5A medium supplemented with 10% fetal calf serum, 100 U/ml stimuli (Hemmati et al., 2005; Sohn et al., 2006; Wendt penicillin and 0.1 mg/ml streptomycin (all from Invitrogen, et al., 2006; Ja¨ nicke et al., 2007). Recently, p21 was Karlsruhe, Germany). reported to block apoptosis downstream of mitochon- dria by inhibiting caspase-9 activation (Sohn et al., Adenoviral vectors, antibodies and reagents 2006). This would explain why there is no major Ad5-CMVp14ARF (Ad-p14ARF) and Ad-lacZ expressing difference in the degree of cytochrome c release, b-galactosidase were constructed and employed as described although caspase-9- and caspase-3/-7-like activities as (Hemmati et al., 2002). Mouse monoclonal anti-p14ARF anti- well as apoptosis sensitivities markedly differ between body (clone 14P02) was purchased from NeoMarkers (Fre- the HCT116 sublines on expression of p14ARF. mont, CA, USA). Antibodies against p21 (clone 6B6) and p53 In p53-deficient cells, p14ARF-induced apoptosis is (clone DO-1) were murine monoclonal antibodies from BD fully independent of p53 and Bax, but depends on Bak. Pharmingen (Heidelberg, Germany) and were used at a ARF dilution of 1:1000. Antibodies against 14-3-3s (Oncogene In contrast, p14 -induced apoptosis is mediated by a Research Products, Boston, MA, USA), Cdc25C (Santa Cruz Bax/Bak-dependent signaling cascade in p53-proficient Biotechnology, Santa Cruz, CA, USA) and actin (Sigma- cells. This differential requirement for the activation of Aldrich, St Louis, MI, USA) were rabbit polyclonal antibodies Bax and/or Bak suggests that the regulation and used at dilutions of 1:200 and 1:500, respectively. Anti-p34cdc2 activation of BH3-only proteins, the main activators of and anti-phospho-cdc2 (Tyr15) were both rabbit polyclonal Bax and Bak, may play a central role in p14ARF-induced antibodies from Cell Signaling Technologies, Beverly, MA, apoptosis. Consequently, expression of p14ARF may USA, and were used at a dilution of 1:1000. Secondary goat regulate the expression and functional activity of BH3- anti-mouse IgG and goat anti-rabbit IgG antisera coupled to only proteins other than Bad to sensitize for mitochon- horseradish peroxidase were obtained from Southern Bio- drial apoptosis. technologies (Birmingham, AL, USA) and were used at a dilution of 1:10 000. Caspase activation is functionally relevant in p14ARF- induced cell death, and apoptosis can be inhibited in the presence of broad spectrum or DEVDfmk peptide Immunofluorescence and microscopy Cells were seeded on sterile coverslips, infected with adenoviral inhibitors (Hemmati et al., 2002, 2005). Our data on vectors at 25 MOI for 36 h, fixed and stained for anti-cdc2 mitotic arrest and mitotic catastrophe indicate that (dilution 1:200), an anti-p14ARF (dilution 1:200) or an anti- p14ARF utilizes a signaling pathway that is, at least in phosphohistone H3 (Ser10) antibody (dilution 1:200) at 4 1C part, distinct from that triggered by DNA-damaging overnight as described (Normand et al., 2005). Thereafter, agents. In line with this, we show that expression of slides were incubated with an Alexa-fluor-488-labeled goat

Oncogene 14-3-3r and p21 in p14ARF-induced cell cycle arrest and apoptosis PG Hemmati et al 6718 anti-rabbit or an Alexa-fluor-594-labeled goat anti-mouse quantified using ModFit. Apoptosis was determined on the antibody (Molecular Probes, Leiden, Netherlands) for 2 h. single-cell level by measuring the DNA content of individual Slides were washed with phosphate-buffered saline containing cells with a logarithmic amplification in the FL-3 channel as Hoechst 33258 at 0.5 mg/ml (Sigma-Aldrich) and mounted with described (Daniel et al., 1999). Data are given in percent of Mowiol solution (Merck, Bad Soden, Germany). hypoploid cells, that is cells with a sub-G1 DNA content, reflecting apoptotic cells with fragmented genomic DNA. SDS–polyacrylamide gel electrophoresis and western blot analysis Cells were harvested, washed and lysed as described (Hemmati Detection of caspase activation on the single-cell level et al., 2002). Samples were mixed with sample buffer (125 mM Measurement of pan-caspase activation using an FITC-labeled Tris/HCl, 288 mM b-mercaptoethanol, 20% glycerol, 2% SDS, conjugate of the cell permeable pan-caspase inhibitor VAD-fmk 10 mg/ml bromophenol blue) and boiled for 5 min. Equal (valyl-alanyl-aspartyl-(O-methyl)-fluoro-methylketone) from amounts of protein (20 mg per lane) were separated by SDS– Promega (Mannheim, Germany) was performed as described polyacrylamide gel electrophoresis (SDS–PAGE). Immuno- (Hemmati et al., 2005). Green fluorescent cells displaying blotting and visualization of the proteins were performed using activated caspases were quantified by flow cytometry. enhanced chemiluminescence (Hemmati et al., 2002). Equal protein loading was confirmed by re-probing membranes with MPM-2 labeling antibodies against actin. Cells were infected with adenoviral vectors at 25 MOI for 48 h and stained with MPM-2 as described (Normand et al., 2005). Detection of cytochrome c release Bivariate analysis of DNA content (propidium iodide) and Cytosolic extracts were prepared as described (von Haefen MPM-2 antibody (FITC) was performed by flow cytometry. et al., 2003). Briefly, cells were harvested, washed in M-phase cells were identified as MPM-2 positive with a 4N phosphate-buffered saline and lysed in 40 ml of hypotonic DNA content. Incubation with 0.1 mg/ml nocodazole (Sigma- buffer (20 mM 4-(2-hydroxyethyl)-1-piperazineethanesulfonic Aldrich) for 18 h served as a positive control for cells arrested acid (pH 7.4), 10 mM KCl, 2 mM MgCl2,1mM EDTA) in mitosis. supplemented with 1 mM phenylmethylsulfonyl fluoride and 0.75 mg/ml digitonin (both from Sigma-Aldrich). After in- BrdU incorporation cubation on ice for 3 min, debris was pelleted by centrifuga- Cells were pulse-labeled with BrdU (Sigma-Aldrich; final tion. Supernatants were mixed with 10 mlof5Â SDS–PAGE concentration 10 mM sample buffer and boiled for 5 min. Equal amounts of protein ) 48 h after infection as described (Nor- et al (20 mg) were separated by SDS–PAGE and subjected to mand ., 2005). Analysis of DNA synthesis, that is western blot analysis. incorporation of BrdU across the cell cycle, was performed using an FITC-labeled monoclonal antibody against BrdU (BD Pharmingen). BrdU-positive cells were gated and counted and measurement of apoptotic cell death as percent of cells in S phase. Cells were infected with adenoviral vectors and harvested at the indicated time points. Cell cycle analysis was performed on the single-cell level by measuring the DNA content of Acknowledgements individual cells by flow cytometry as described (Normand et al., 2005). Analysis was carried out with a linear amplifica- This work was supported by the Deutsche Krebshilfe Grant tion in the FL-2 channel of a FACScan flow cytometer (BD 10-2088-Da3 to PTD and PGH. We would like to thank Ms Pharmingen) equipped with the CellQuestPro software and Antje and Anja Richter for expert technical assistance.

References

Bhonde MR, Hanski ML, Budczies J, Cao M, Gillissen B, Moorthy D Daniel PT, Sturm I, Ritschel S, Friedrich K, Do¨ rken B, Bendzko P et al. (2006a). DNA damage-induced expression of p53 suppresses et al. (1999). Detection of genomic DNA fragmentation during mitotic checkpoint kinase hMps1: the lack of this suppression apoptosis (DNA ladder) and the simultaneous isolation of RNA in p53MUT cells contributes to apoptosis. J Biol Chem 281: from low cell numbers. Anal Biochem 266: 110–115. 8675–8685. Datta SR, Katsov A, Hu L, Petros A, Fesik SW, Yaffe MB Bhonde MR, Hanski ML, Notter M, Gillissen BF, Daniel PT, Zeitz M et al. (2000). 14-3-3 proteins and survival kinases cooperate to et al. (2006b). Equivalent effect of DNA damage-induced apoptotic inactivate BAD by BH3 domain phosphorylation. Mol Cell 6: cell death or long-term cell cycle arrest on colon carcinoma cell 41–51. proliferation and tumour growth. Oncogene 25: 165–175. de Stanchina E, McCurrach ME, Zindy F, Shieh SY, Ferbeyre G, Blagosklonny MV, Demidenko ZN, Fojo T. (2002). Inhibition of Samuelson AV et al. (1998). E1A signaling to p53 involves the transcription results in accumulation of Wt p53 followed by delayed p19(ARF) tumor suppressor. Genes Dev 12: 2434–2442. outburst of p53-inducible proteins: p53 as a sensor of transcriptional Duro D, Bernard O, Della Valle V, Berger R, Larsen CJ. (1995). A new integrity. Cell Cycle 1: 67–74. type of p16INK4/MTS1 gene transcript expressed in B-cell Bunz F, Dutriaux A, Lengauer C, Waldman T, Zhou S, Brown JP malignancies. Oncogene 11: 21–29. et al. (1998). Requirement for p53 and p21 to sustain G2 arrest after Eymin B, Claverie P, Salon C, Brambilla C, Brambilla E, Gazzeri S. DNA damage. Science 282: 1497–1501. (2006a). p14ARF triggers G2 arrest through ERK-mediated Chan TA, Hermeking H, Lengauer C, Kinzler KW, Vogelstein B. Cdc25C phosphorylation, ubiquitination and proteasomal degrada- (1999). 14-3-3Sigma is required to prevent mitotic catastrophe after tion. Cell Cycle 5: 759–765. DNA damage. Nature 401: 616–620. Eymin B, Claverie P, Salon C, Leduc C, Col E, Brambilla E et al. Chan TA, Hwang PM, Hermeking H, Kinzler KW, Vogelstein B. (2006b). p14ARF activates a Tip60-dependent and p53-independent (2000). Cooperative effects of genes controlling the G(2)/M ATM/ATR/CHK pathway in response to genotoxic stress. Mol Cell checkpoint. Genes Dev 14: 1584–1588. Biol 26: 4339–4350.

Oncogene 14-3-3r and p21 in p14ARF-induced cell cycle arrest and apoptosis PG Hemmati et al 6719 Eymin B, Karayan L, Seite P, Brambilla C, Brambilla E, Larsen CJ and p21-deficient cells by down-regulating p34cdc2 kinase activity. et al. (2001). Human ARF binds E2F1 and inhibits its transcrip- J Biol Chem 280: 7118–7130. tional activity. Oncogene 20: 1033–1041. Peng CY, Graves PR, Thoma RS, Wu Z, Shaw AS, Piwnica-Worms Eymin B, Leduc C, Coll JL, Brambilla E, Gazzeri S. (2003). p14ARF H. (1997). Mitotic and G2 checkpoint control: regulation of 14-3-3 induces G2 arrest and apoptosis independently of p53 leading protein binding by phosphorylation of Cdc25C on serine-216. to regression of tumours established in nude mice. Oncogene 22: Science 277: 1501–1505. 1822–1835. Quelle DE, Zindy F, Ashmun RA, Sherr CJ. (1995). Alternative Giannakakou P, Robey R, Fojo T, Blagosklonny MV. (2001). Low reading frames of the INK4a tumor suppressor gene encode two concentrations of induce cell type-dependent p53, p21 and unrelated proteins capable of inducing cell cycle arrest. Cell 83: G1/G2 arrest instead of mitotic arrest: molecular determinants of 993–1000. paclitaxel-induced cytotoxicity. Oncogene 20: 3806–3813. Radfar A, Unnikrishnan I, Lee HW, DePinho RA, Rosenberg N. Gorospe M, Wang X, Guyton KZ, Holbrook NJ. (1996). Protective (1998). p19(Arf) induces p53-dependent apoptosis during abelson role of p21(Waf1/Cip1) against prostaglandin A2-mediated virus-mediated pre-B cell transformation. Proc Natl Acad Sci USA apoptosis of human colorectal carcinoma cells. Mol Cell Biol 16: 95: 13194–13199. 6654–6660. Rocha S, Campbell KJ, Perkins ND. (2003). p53- and Mdm2- Graves PR, Lovly CM, Uy GL, Piwnica-Worms H. (2001). Localiza- independent repression of NF-kappa B transactivation by the tion of human Cdc25C is regulated both by nuclear export and ARF tumor suppressor. Mol Cell 12: 15–25. 14-3-3 protein binding. Oncogene 20: 1839–1851. Rocha S, Garrett MD, Campbell KJ, Schumm K, Perkins ND. (2005). Hemmati PG, Gillissen B, von Haefen C, Wendt J, Starck L, Gu¨ ner D Regulation of NF-kappaB and p53 through activation of ATR and et al. (2002). Adenovirus-mediated overexpression of p14(ARF) Chk1 by the ARF tumour suppressor. EMBO J 24: 1157–1169. induces p53 and Bax-independent apoptosis. Oncogene 21: Rocha S, Perkins ND. (2005). ARF the integrator: linking NF- 3149–3161. kappaB, p53 and checkpoint kinases. Cell Cycle 4: 756–759. Hemmati PG, Gu¨ ner D, Gillissen B, Wendt J, von Haefen C, Samuel T, Weber HO, Rauch P, Verdoodt B, Eppel JT, McShea A Chinnadurai G et al. (2006). Bak functionally complements for loss et al. (2001). The G2/M regulator 14-3-3 sigma prevents apoptosis of Bax during p14ARF-induced mitochondrial apoptosis in human through sequestration of Bax. J Biol Chem 276: 45201–45206. cancer cells. Oncogene 25: 6582–6594. Sanchez Y, Wong C, Thoma RS, Richman R, Wu Z, Piwnica-Worms Hemmati PG, Normand G, Verdoodt B, von Haefen C, Hasenjager A, H et al. (1997). Conservation of the Chk1 checkpoint pathway in Gu¨ ner D et al. (2005). Loss of p21 disrupts p14 ARF-induced G1 mammals: linkage of DNA damage to Cdk regulation through cell cycle arrest but augments p14 ARF-induced apoptosis in human . Science 277: 1497–1501. carcinoma cells. Oncogene 24: 4114–4128. Sohn D, Essmann F, Schulze-Osthoff K, Ja¨ nicke RU. (2006). p21 Hermeking H. (2003). The 14-3-3 cancer connection. Nat Rev Cancer blocks irradiation-induced apoptosis downstream of mitochondria 3: 931–943. by inhibition of cyclin-dependent kinase-mediated caspase-9 activa- Hermeking H, Lengauer C, Polyak K, He TC, Zhang L, Thiagalingam tion. Cancer Res 66: 11254–11262. S et al. (1997). 14-3-3 sigma is a p53-regulated inhibitor of G2/M Taagepera S, Rao PN, Drake FH, Gorbsky GJ. (1993). DNA progression. Mol Cell 1: 3–11. topoisomerase II alpha is the major chromosome protein recognized Ja¨ nicke RU, Sohn D, Essmann F, Schulze-Osthoff K. (2007). The by the mitotic phosphoprotein antibody MPM-2. Proc Natl Acad multiple battles fought by anti-apoptotic p21. Cell Cycle 6: 407–413. Sci USA 90: 8407–8411. Jin P, Gu Y, Morgan DO. (1996). Role of inhibitory CDC2 Takizawa CG, Morgan DO. (2000). Control of mitosis by changes in phosphorylation in radiation-induced G2 arrest in human cells. the subcellular location of cyclin-B1-Cdk1 and Cdc25C. Curr Opin J Cell Biol 134: 963–970. Cell Biol 12: 658–665. Jin P, Hardy S, Morgan DO. (1998). Nuclear localization of cyclin B1 von Haefen C, Wieder T, Essmann F, Schulze-Osthoff K, Do¨ rken B, controls mitotic entry after DNA damage. J Cell Biol 141: 875–885. Daniel PT. (2003). Paclitaxel-induced apoptosis in BJAB cells Li Y, Wu D, Chen B, Ingram A, He L, Liu L et al. (2004). ATM proceeds via a death receptor-independent, caspases-3/-8-driven activity contributes to the tumor-suppressing functions of p14ARF. mitochondrial amplification loop. Oncogene 22: 2236–2247. Oncogene 23: 7355–7365. Waldman T, Kinzler KW, Vogelstein B. (1995). p21 is necessary for Mao L, Merlo A, Bedi G, Shapiro GI, Edwards CD, Rollins BJ et al. the p53-mediated G1 arrest in human cancer cells. Cancer Res 55: (1995). A novel p16INK4A transcript. Cancer Res 55: 2995–2997. 5187–5190. Martelli F, Hamilton T, Silver DP, Sharpless NE, Bardeesy N, Rokas Waldman T, Lengauer C, Kinzler KW, Vogelstein B. (1996). M et al. (2001). p19ARF targets certain species for degradation. Uncoupling of S phase and mitosis induced by anticancer agents Proc Natl Acad Sci USA 98: 4455–4460. in cells lacking p21. Nature 381: 713–716. Martinez LA, Yang J, Vazquez ES, Rodriguez-Vargas Mdel C, Olive Wendt J, Radetzki S, von Haefen C, Hemmati PG, Gu¨ ner D, Schulze- M, Hsieh JT et al. (2002). p21 modulates threshold of apoptosis Osthoff K et al. (2006). Induction of p21CIP/WAF-1 and G2 arrest induced by DNA-damage and growth factor withdrawal in prostate by ionizing irradiation impedes caspase-3-mediated apoptosis in cancer cells. Carcinogenesis 23: 1289–1296. human carcinoma cells. Oncogene 25: 972–980. Morgan DO, Fisher RP, Espinoza FH, Farrell A, Nourse J, Zeng Y, Piwnica-Worms H. (1999). DNA damage and replication Chamberlin H et al. (1998). Control of eukaryotic cell cycle checkpoints in fission yeast require nuclear exclusion of the Cdc25 progression by phosphorylation of cyclin-dependent kinases. Cancer phosphatase via 14-3-3 binding. Mol Cell Biol 19: 7410–7419. J Sci Am 4(Suppl 1): S77–S83. Zha J, Harada H, Yang E, Jockel J, Korsmeyer SJ. (1996). Serine Nomura M, Shimizu S, Sugiyama T, Narita M, Ito T, Matsuda H et al. phosphorylation of death agonist BAD in response to survival (2003). 14-3-3 Interacts directly with and negatively regulates factor results in binding to 14-3-3 not BCL-X(L). Cell 87: 619–628. pro-apoptotic Bax. J Biol Chem 278: 2058–2065. Zindy F, Eischen CM, Randle DH, Kamijo T, Cleveland JL, Sherr CJ Normand G, Hemmati PG, Verdoodt B, von Haefen C, Wendt J, et al. (1998). Myc signaling via the ARF tumor suppressor regulates p53- Gu¨ ner D et al. (2005). p14ARF induces G2 cell cycle arrest in p53- dependent apoptosis and immortalization. Genes Dev 12: 2424–2433.

Oncogene