Transcription Factor NF-Y Induces Apoptosis in Cells Expressing Wild-Type P53 Through E2F1 Upregulation and P53 Activation

Published OnlineFirst October 15, 2010; DOI: 10.1158/0008-5472.CAN-10-0721 Published OnlineFirst on November 30, 2010 as 10.1158/0008-5472.CAN-10-0721

Cancer Molecular and Cellular Pathobiology Research

Transcription Factor NF-Y Induces Apoptosis in Cells Expressing Wild-Type p53 through E2F1 Upregulation and p53 Activation

Aymone Gurtner1, Paola Fuschi1, Fabio Martelli2, Isabella Manni1, Simona Artuso1, Giacoma Simonte1, Valeria Ambrosino1, Annalisa Antonini2, Valentina Folgiero1, Rita Falcioni1, Ada Sacchi1, and Giulia Piaggio1

Abstract The CCAAT-binding transcription factor NF-Y plays a central role in regulating cellular proliferation by controlling the expression of genes required for cell-cycle progression such as cyclin A, cyclin B1, cyclin B2, cdc25A, cdc25C, and cdk1. Here we show that unrestricted NF-Y activity leads to apoptosis in an E2F1- and wild-type p53 (wtp53)-dependent manner. Unrestricted NF-Y activity induced an increase in E2F1 mRNA and protein levels. Furthermore, NF-Y directly bound the E2F1 promoter and this correlated with the appearance of open chromatin marks. The ability of NF-Y to induce apoptosis was impaired in cells lacking E2F1 and wtp53. Moreover, NF-Y overexpression elicited phosphorylation of wt p53Ser18 in an E2F1-dependent manner. Our findings establish that NF-Y acts upstream of E2F1 in p53-mediated apoptosis. Cancer Res; 70(23); 9711–20. 2010 AACR.

Introduction function approach, such as expression of dominant-negative NF-YA mutants and conditional deletion of the mouse NF-YA NF-Y is a ubiquitous protein, composed of 3 subunits, NF- gene. When a dominant-negative NF-YA mutant that interacts YA, -YB, and -YC, whose genes are highly conserved from yeast with -YB/YC but does not bind DNA is expressed in mouse to mammals. All 3 subunits are required for NF-Y binding to fibroblasts, retardation of cell growth is observed (12). More- the consensus sequence, the CCAAT-box. The NF-YB and -YC over, inactivation of the NF-YA gene in mouse embryonic subunits contain histone-like domains, and an activation fibroblasts results in inhibition of cell proliferation and growth domain is present in the NF-YA subunit (1). A bioinformatic arrest at various phases of the cell cycle (13). Taken together, analysis of promoters of cell-cycle regulatory genes shows an these studies demonstrated that the binding of NF-Y to abundance of CCAAT boxes in promoters regulated during the cellular promoters is essential for cell proliferation. G2/M transition (2). Consistent with this, the NF-Y complex Numerous findings indicate that NF-Y is involved in cancer. supports basal transcription of a class of regulatory genes We have demonstrated that NF-Y modulates the promoter responsible for cell-cycle progression, among which are mito- activity of several genes in response to DNA-damaging agents tic cyclin complexes (3–9). (14, 15). Next, we have shown that NF-Y interacts in vivo with NF-YA is the regulatory subunit of the trimer. Indeed, mutant p53 and increases DNA synthesis, which is impaired expression of the protein is modulated during the cell cycle upon abrogation of NF-YA expression (4). Clinical studies have (3) and its abrogation plays an important role in downregulat- indicated that patients with upregulated expression of NF-Y ing several cell-cycle control genes in differentiated muscle target genes have poor prognosis in multiple cancers (4, 16). cells (5–7, 10). Recently, it has been shown that NF-Y regulates Using global gene expression profiles, the involvement of NF-Y gene expression in embryonic stem cells, where it is required in cancer-associated pathways has been recently reported for their proliferation (11). Previous studies aimed at under- across human cancers (17). standing the biological role of NF-Y took advantage of a loss of Apoptosis and proliferation are intimately coupled. A tight linkage between cell-cycle regulation and apoptosis has been shown for c-myc, E2F1, and cyclins. These genes Authors' Affiliations: 1Experimental Oncology Department, Istituto may induce cell proliferation, cell-cycle arrest, or cell death, 2 Regina Elena and Laboratorio di Patologia Vascolare, Istituto Dermopa- with the different outcomes depending on variables such as tico dell’Immacolata, IRCCS, Rome, Italy cell type, cellular environment, and genetic background (18). Note: Supplementary data for this article are available at Cancer Research Online (http://cancerres.aacrjournals.org/). Intriguingly, many of these genes are NF-Y transcriptional targets (4, 5, 7, 15, 19–22). A. Gurtner and P. Fuschi contributed equally to this work. Multiple signaling pathways activated by cellular stress Corresponding Author: Giulia Piaggio, Experimental Oncology Depart- ment, Istituto Regina Elena, IRCCS, Via delle Messi D’Oro 156, 00158 converge on the wtp53 tumor suppressor whose activation Rome, Italy. Phone: 0039-06-52662531; Fax: 0039-06-4180526; E-mail: can lead to growth arrest or programmed cell death (23). [email protected]. Activation of p53 can result not only from physical and doi: 10.1158/0008-5472.CAN-10-0721 chemical stresses but also from oncogene imbalance. For 2010 American Association for Cancer Research. example, expression of c-myc leads to p53-dependent

www.aacrjournals.org 9711

Gurtner et al.

apoptosis (24) through a mechanism largely dependent on Colony formation assay E2F1 (25), a member of the E2F family that controls For colony-forming assays cells were plated in 60-mm dish expression of genes important for cell-cycle progression and transfected with the indicated vectors. After 48 hours, and apoptosis (26). Ectopic expression of E2F1 induces puromycin (1 mg/mL) was added for selection for 2 weeks. p53-dependent apoptosis both in tissue culture (27–29) Surviving colonies were stained with crystal violet, photo- and in mouse models (30, 31), and a similar role for graphed, and colony number counted under a microscope. endogenous E2F1 is evidenced by the defect in apoptosis in E2F1-deficient mice (32). Different mechanisms have Chromatin immunoprecipitation been proposed for the ability of E2F1 to cause apoptosis Chromatin immunoprecipitation (ChIP) was performed as through both p53-dependent and -independent pathways. previously described (7). The following antibodies were used: In the first case, p53 is phosphorylated at residues that are anti-H4ac-Pan, (Upstate), anti-H3K9ac (Abcam), anti-H3tri-m- also phosphorylated in response to DNA damage (33–36). k9 (Abcam), anti-H4tri-m-k20 (Abcam), and anti-NF-YA The ability of NF-Y to control, like E2F1, the expression of (Rockland). PCR analysis was performed with HOT-MASTER many genes important for cell-cycle progression prompted us Taq (Eppendorf). Primer sequences specific for analysis were to investigate whether deregulation of NF-Y transcriptional as follows: mouse E2F1 promoter—for 50-gacagtgaacgctct- activity could trigger a tumor suppressor network capable of caggc, Rev 50-ccgaacgcctcctccattgg; human E2F1 promoter; inducing cell death. We overexpressed the regulatory subunit for 50-ggctacaggttgagggtcacg, Rev 50-gagcgccgccacaattggct. of the trimer, NF-YA, in mouse embryonic fibroblasts (MEF) and in human cells and showed that unrestrained NF-Y RNA extraction and real-time-PCR activity promotes apoptosis depending on E2F1 induction Total RNA was extracted using the Trizol reagent (Gibco and wtp53 activation. Our results demonstrate that NF-Y BRL) following the manufacturer's instructions. The first- activates a tumor suppressor network that functions at multi- strand cDNA was synthesized according to the instructions ple levels to efficiently induce cell death in wtp53-expressing for the M-MLV RT kit (Invitrogen). RT-PCR analysis was cells. performed with HOT-MASTER Taq (Eppendorf), using 2 mL of cDNA reaction. The primers were as follows: E2F1—for 50- Materials and Methods TCTGTACCACACAGCTGCAA, Rev 50-GCACAGGAAAACAT- CAATGG; aldolase, for 50-TGG ATG GGC TGT CTG AAC Cells lines GCT GT, Rev 50-AGT GAC AGC AGG GGG CAC TGT. Densito- Mouse embryonic fibroblasts were isolated from 13.5-day- metric analysis was performed using Scion-image software. old embryos as described previously (36). MEFs derived from p53 / mouse embryos (gift from Silvia Soddu), or from Quantitative PCR E2F1 / and p73 / mouse (gift from Massimo Levrero) Quantitative PCR (qPCR) was performed using SYBR Green and human breast carcinoma SKBR3 (mut p53-175H) were (Applied Biosystems) on an ABI Prism 7500 apparatus cultured in DMEM supplemented with 10% FBS (GIBCO BRL), (Applied Biosystems) in 2 independent experiments in tripli- 2 mmol/L of L-glutamine (Life Technologies Inc.), 100 U/mL of cate. The comparative threshold (DCt) method (7) was used. penicillin, and 100 mg/mL of streptomycin (Life Technologies Primer sequences were as follows: mouse E2F1 promoter—For Inc.). Primary breast epithelial cells MCF10A were cultured in 50-gggtggagaccacaggttga, Rev 50-ctggcgaagcgaacaaactt; mouse Ham's F-12 medium with 5% horse serum, 0.5 mg/mL of E2F1 MRNA—For 50-cacagctgcaactgctttcg, Rev 50-aaggtcctgg- hydrocortisone, 10 mg/mL of insulin, and 20 ng/mL of epi- caggtcacatag. dermal growth factor. Growth and lethality curves Viral vectors and infection Cells were harvested at different time points after infection The adenoviral vectors used for infections were as follows: or transfection and cell number was determined in duplicate, Ad-NF-YA, Ad-control (Ad-GFP; ref. 37), and Ad-E2F1. Cells with a Thomas hemocytometer. Lethality was determined by exponentially growing on plates were infected in DMEM counting the number of cells stained with trypan blue. without serum for 1 hour at 37C at a multiplicity of infection of 400 pfu per cell. Following infection, DMEM with 10% FBS FACS cell-cycle analysis was added and the cells were incubated at 37C and analyzed Cells were harvested, washed in PBS, and fixed in MetOH: at the indicated time points. acetic acid solution (4:1) for 60 minutes at þ4C. Cells were then incubated in 500 mL of staining solution (50 mg/mL of Plasmids and transfections propidium iodide, 50 mg/mL of RNAase, 0.1% Triton X-100 in Plasmids used in transfections were as follows: NF-YA, NF- PBS 1) for 1 hour at 4C and analyzed by flow cytometry. YB, dnNF-YA, empty vector (Poly II; ref. 1), E1A (E1A12S), Ha- Ras, and pBABE-PURO (vector bearing puromycin resistance Western blotting gene). Cells were transfected with Fugene6 (Roche) following Cells were harvested into 1 RIPA buffer (150 mmol/L of the manufacturer's instructions. Transfection efficiency was NaCl, 1% TritonX-100, 0.25% sodium deoxycholate, 0.1% SDS, normalized through b-galactosidase activity and protein 50 mmol/L of Tris/HCl, pH ¼ 8.0, 20 mmol/L of EDTA) concentrations. supplemented with 1 protease and phosphatase inhibitor

9712 Cancer Res; 70(23) December 1, 2010 Cancer Research

Unrestricted NF-Y Activity Leads to Apoptosis

cocktail (Sigma-Aldrich), 1 mmol/L of PMSF (Sigma- Aldrich), 50 mmol/L of sodium fluoride (Sigma-Chemical A Co.), and 50 mmol/L of dithiothreitol (Bio-Rad Laboratories Inc.). Lysates were resolved onto SDS-PAGE. For p-Erk1/2 detection, lysates were boiled at 65C. After blotting, filters were immune reacted with anti-NF-YA (Rockland), anti-p53 (DOI, 5), anti-pS15-p53 and pS18 mouse (Cell Signaling), anti-pS139-H2A.X (Upstate), anti-pS1981-ATM (Rockland), anti-cdk1 (Santa Cruz), anti-cycA (Santa Cruz), anti-E2F1 (Santa Cruz), and (Santa Cruz), anti-Erk1/2 (Cell Signaling), anti-pErk1/2 (Cell Signaling), anti-HSP70 (StressGen), anti-a-actin (Calbiochem), anti-actin (Calbiochem), anti- caspase 3 (Cell Signaling), and anti-Parp (BD Biosciences). Signals were detected by ECL detection reagents (Amersham Biosciences). B Statistical analysis All experiments were performed in triplicate. Numerical data were reported as means SD. P values were determined using t test (GraphPad software). A 95% confidence interval (P < 0.05) was considered significant.

Results

Unrestricted NF-Y activity inhibits clonogenicity of mouse embryo fibroblasts To examine the effect of unrestricted NF-Y activity, MEFs were transfected with a plasmid encoding NF-YA or empty Figure 1. NF-YA overexpression inhibits clonogenicity. A, colony vector, along with a vector bearing puromycin resistance. formation assay performed on MEFs transfected with the indicated Puromycin selection was applied for 48 hours, and colonies vectors. Fold difference values are given versus the control sample. were stained and counted 2 weeks later. We found that NF-YA Means SD and P values of 8 independent experiments are reported. – overexpression led to more than 80% reduction of colony B, colony formation assay performed on Ha-Ras/E1A transformed MEFs transfected with NF-YA expressing vector or empty vector. Means SD formation. This reduction was NF-YA specific, as ectopic and P values of 3 independent experiments are reported. expression of NF-YB had no effect (Fig. 1A). This is in good agreement with the notion that the NF-YA is the regulatory subunit of the complex. As a control we used a dominant- negative NF-YA protein (dnNF-YA) carrying a mutant DNA- either alone or in combination with oncogenes, inhibits binding domain that interacts with the YB/YC dimer but colony formation. does not bind DNA. When overexpressed, this mutant acts as a competitive inhibitor of interaction between wild-type Unrestricted NF-Y activity inhibits cell NF-YA and the YB/YC dimer, forming an inactive complex in growth–inducing apoptosis terms of DNA binding activity. It has been previously To characterize the decrease in clonogenicity of NF-YA– described that the expression of this protein results in retar- overexpressing cells, MEFs were transduced with an adenoviral dation of cell growth (12). As expected, overexpression of the vector encoding NF-YA (Ad-NF-YA) or an empty vector as mutant protein in MEFs resulted in the inhibition of colony negative control (Ad-control). Figure 2A shows that NF-YA formation. Taken together, these results strongly indicate that expression reduced MEFs proliferation. Indeed, while control unbalanced NF-Y transcriptional activity leads to defects in cells more than doubled in number by 72 hours after infection, cell proliferation (Fig. 1A). NF-YA transduced cells did not. We used FACS analysis to To assess whether NF-Y could alter proliferation also of determine whether the reduced cell growth reflected an aber- transformed MEFs, we cotransfected MEFs with the Ha-Ras rant cell-cycle progression. As shown in Fig. 2B, NF-YA over- and E1A oncoproteins. As expected, the cotransfection expression induced a small but reproducible increase in the G2/ yielded numerous transformed foci as compared with M cell population and a corresponding decrease of G1 cells at 24 Ha-Ras or E1A alone (Supplementary Figure S1; ref. 37). and 48 hours after infection. Next, we asked whether the Ha-Ras/E1A-transformed foci were harvested, amplified, reduction of MEF proliferation reflected augmented cell death. and used in colony formation assays. Interestingly, over- Assessment of cell viability by trypan blue exclusion revealed expression of NF-YA in these cells still resulted in the that Ad-NF-YA infection induced cell death (Supplementary inhibition of colony formation (Fig. 1B). Taken together, Figure S2), and FACS analysis showed that the percentage of these results demonstrate that unrestricted NF-Y activity, sub-G1 cells increased compared with Ad-control–infected cells

www.aacrjournals.org Cancer Res; 70(23) December 1, 2010 9713

Gurtner et al.

AB ×

Figure 2. NF-YA overexpression inhibits cell growth and induces apoptosis. A, MEFs transduced with the indicated adenoviral vectors were plated in triplicate. Cell numbers were counted daily. Means SD of 3 independent experiments are reported. B, FACS analysis of Ad-NF-YA- C D and Ad-control–infected MEFs. C, subdiploid DNA content of Ad- NF-YA- and Ad-control–infected MEFs. D, PARP (Parp) and caspase 3 (Casp-3) cleavage were analyzed by Western blot on lysates of Ad-NF-YA- and Ad- control–infected MEFs. Tubulin was used as a loading control.

(Fig. 2C), suggesting induction of apoptosis. To test this hypoth- hours after Ad-NF-YA infection (Fig. 3B). To further support esis, we investigated whether overexpression of NF-YA the role of NF-Y on E2F1 transcription, we investigated increases PARP and caspase 3 cleavage. Figure 2D shows a whether NF-Y modulated the chromatin structure of the reduction in the full-length form of PARP-1 and an increase of E2F1 promoter by ChIP experiments in Ad-NFYA and Ad- the cleaved caspase 3 in Ad-NF-YA-infected cells. control infected cells with anti- NF-YA, -H3K9ac, -H4ac, -H3K9me3, and -H4K20me3 antibodies. H3K9ac and H4ac E2F1 is induced by unrestricted NF-Y activity are well-known markers of open chromatin structure, and is necessary for NF-Y–mediated inhibition whereas H3K9me3 and H4K20me3 mark heterochromatin of colony formation (41–43). As expected, NF-Y was bound to the E2F1 promoter E2F1 has been shown to regulate both cell-cycle progression and overexpression of NF-YA led to a significant increase in and cell death when overexpressed. Moreover, both murine this binding (Fig. 3C). Of note, this increase correlated with and human E2F1 promoters contain 3 CCAAT boxes shown to a significant enrichment in the acetylation density of H3K9 be bound by NF-Y (38–40). These considerations prompted us and, to a lesser extent, of H4. The acetylation density of to assay whether the NF-Y effects on apoptosis were E2F1 histones associated with closed chromatin conformation mediated. was very low before and after NF-YA overexpression. NF-Y To determine whether NF-YA overexpression could directly binding to the E2F1 promoter occurs also in human cells, as induce E2F1 expression, MEFs were infected with either demonstrated by ChIP experiments on several human cell Ad-NF-YA or Ad-control and E2F1 expression was measured lines (Supplementary Figure S4). by Western blotting. Figure 3A shows that the E2F1 protein We further asked whether E2F1 mediates the apoptosis level was increased in NF-YA overexpressing cells. Next, we induced by NF-YA overexpression. MEFs derived from asked whether the downmodulation of NF-Y activity dimin- E2F1 / mice were transfected with a plasmid encoding ished E2F1 protein expression. Supplementary Figure S3 NF-YA or vector alone, and colony formation was assayed. shows that silencing of NF-YA by siRNA induced a time- Figure 3D shows that NF-YA failed to decrease the efficiency of dependent downmodulation of endogenous E2F1 expression. colony formation in the absence of E2F1 expression. Taken To investigate whether NF-Y modulates E2F1 expression together, these results demonstrate that (a) NF-Y increases at the transcriptional level, we analyzed the amount of E2F1 expression and (b) E2F1 is necessary for NF-Y growth E2F1 mRNA by quantitative real-time PCR (qRT-PCR) suppressive activity, strongly suggesting that NF-Y induces and RT-PCR. E2F1 mRNA was upregulated as early as 8 apoptosis through upregulation of E2F1 expression.

9714 Cancer Res; 70(23) December 1, 2010 Cancer Research

Unrestricted NF-Y Activity Leads to Apoptosis

Figure 3. E2F1 is transcriptionally induced by unrestricted NF-Y activity and is necessary for NF-Y– mediated growth inhibition. A and B, E2F1 protein and mRNA levels in Ad-NF-YA and Ad-control infected MEFs, respectively. A, tubulin was used as the loading control. B, qRT-PCR (top) and RT- PCR (bottom) analyses of E2F1 mRNA level after overexpression of NF-YA. Numbers on top represent the quantification of bands, normalized to the control (aldolase) and calculated as fold increase versus the Ad-control sample. For qRT-PCR, means SD and P values of 6 independent C D experiments are reported. C, ChIP (top) and Q-ChIP (bottom) analyses of the E2F1 promoter performed with the indicated antibodies in Ad-NF-YA and Ad- control infected MEFs. Cells were collected 24 hours postinfection. The enrichment of immunoprecipitated promoter fragments was calculated relative to the input. Sample without antibody (Ab) was used as the control. D, colony formation assay performed on MEFs from E2F1/ mice transfected with the indicated vectors. The fold C difference values are given versus the control sample. Means SD of 8 independent experiments are reported (bottom). ns, not significant.

NF-YA induces p53 activation via E2F1 cells were obtained (Fig. 4C). Similar levels of NF-YA protein E2F1 has been shown to induce apoptosis via multiple were expressed in p73 / and p53 / MEFs (Fig. 4D), pathways, some of which involve stabilization and activation excluding the possibility that the different biological effects of the tumor suppressor p53 (44). To assess the role of p53 in were due to different NF-YA amounts and demonstrating a NF-Y–dependent apoptosis, p53 / MEFs were used. Figure 4A specific p53 role in transducing NF-Y signaling. shows that the decrease in colony formation induced by NF-YA These results prompted us to analyze whether unrestricted is p53 dependent because no difference in colony formation NF-Y activity triggered p53 activation. Both p53 stabilization efficiency was found in p53 / MEFs overexpressing NF-YA. In and phosphorylation at serine-18 (Ser18; Ser15 in humans) agreement with this result, FACS analysis showed that the represent crucial steps in p53 activation mediated by onco- / percentage of sub-G1 cells did not change in p53 MEFs at 24 genic imbalance (45). We found that NF-YA overexpression and 48 hours post-Ad-NF-YA infection. As already shown in Fig. induced a small but reproducible increase in p53 protein level. 2C, wild-type MEFs infected with Ad-NF-A duplicate and Interestingly, it also increased p53Ser18 phosphorylation triplicate the number of sub-G1 cells at 24 and 48 hours (Fig. 5A). Because E2F1 is necessary for NF-Y–dependent postinfection, respectively (Fig. 4B). Of note, when the ability apoptosis, we assessed whether p53Ser18 phosphorylation of NF-YA to decrease colony formation was assayed in p73 / was mediated by E2F1. Figure 5B shows that NF-YA MEFs, results indistinguishable from those in wild-type MEF overexpression did not trigger p53Ser18 phosphorylation in

www.aacrjournals.org Cancer Res; 70(23) December 1, 2010 9715

Gurtner et al.

(MCF10A) interfered for E2F1 (Fig. 5C) and failed to induce A apoptosis in these cells unlike the case with wild-type MCF10A cells (Fig. 5D). One of the E2F1 targets is p14ARF (p19ARF in rodents). However, in several systems, ARF function seems to be independent from that of E2F1-induced p53 activation and apoptosis (44). Intriguingly, ARF expression was not modulated by NF-YA (data not shown), indicating that NF- Y–induced activation of p53 can be mediated by mechanisms that are ARF independent. It has been reported that the ataxia telangiectasia mutated (ATM) kinase is activated and directly phosphorylates p53 on Ser15 in response to DNA damage and/or oncogenic stress such as Myc overexpression (46).Weasked,therefore,whetherunrestrictedNF-Yactivity B triggered ATM activation. As shown in Supplementary Figure S6, NF-YA overexpression leads to the phosphorylation of ATMSer1981, a marker of ATM activation (47) as well of the ATM substrate histone H2A.X (on Ser139, g-H2A.X), suggesting that unrestricted NF-Y activity, activating ATM, can mimic DNA damage. Taken together, these results demonstrate that unrestricted NF-Y activity utilizes, in both murine and human cells, the E2F1 pathway to trigger p53 activation and suggest that it induces apoptosis via this mechanism.

Unrestricted NF-Y activity inhibits cell proliferation in human cells expressing wtp53 We tested the effect of unrestricted NF-Y activity on colony C formation in human MCF10A cells and lung embryonic fibroblast (IMR90) expressing endogenous wtp53 (Fig. 6A). The results show that NF-YA overexpression induced a 60% and 55.82% decrease in colony formation, respectively. Interest- ingly, colony formation also was inhibited in human tumor D cells (HCT116) expressing wtp53 but not in the same cells lacking p53 (HCT116 p53 / ), indicating that unrestricted NF-Y activity induces apoptosis through wtp53 regardless of the transformed status of the cells. Next, we measured cell proliferation and viability by trypan blue exclusion. As observed in MEFs, NF-YA overexpression reduced proliferation of MCF10A cells that express wtp53 (Fig. 6B). In addition, as already shown in Fig. 5D, the percentage of dead cells increased 24 hours posttransfection (Supplementary Figure S7). Conversely, NF- Figure 4. NF-YA induces p53 activation via E2F1. Colony formation assay YA overexpression increased the proliferation rate of SKBR3 / / performed on MEFs derived from p53 (A) and p73 (B) mice human breast cancer cells (SKBR3) harboring endogenous transfected with the indicated plasmids. Fold difference values are given versus the control sample. Means SD and P values of 8 independent mutant p53His175 (Fig. 6C) but had no effect on cell death in experiments are reported. ns, not significant. C, cell-cycle analysis of these cells (Supplementary Figure S4). As shown in Fig. 6D, p53/ and p73/ MEFs transfected with the indicated plasmids. Cells in MCF10A cells, NF-YA overexpression correlated with p53 were collected at the indicated time points. D, NF-YA protein level in empty accumulation and activation. Thus, in human cells too, vector and Ad-NF-YA–transfected p53/ and p73/ MEFs. unrestricted NF-Y activity increased E2F1 protein levels, and, as expected, cyclin Aandcdk1 expression and Erk E2F1 / MEFs. To confirm that the activation of p53 by NF-Y is phosphorylation, all markers of active proliferation, were E2F1 dependent, we reintroduced E2F1 in E2F1 / MEFs decreased. No effect on p53 and E2F1 levels was observed in and observed phosphorylation of the endogenous p53Ser18, SKBR3 cells, and cyclin Aandcdk1 expression and Erk demonstrating that aberrant levels of E2F1 induce p53 phosphorylation were induced in agreement with the activation in these cells (Supplementary Figure S5). Further- increased proliferation rate shown in Fig. 6C. more, NF-YA overexpression did not trigger p53Ser15 These results further support the hypothesis that unre- phosphorylation in human primary breast epithelial cells stricted NF-Y activity leads to apoptosis in a wtp53-dependent

9716 Cancer Res; 70(23) December 1, 2010 Cancer Research

Unrestricted NF-Y Activity Leads to Apoptosis

A B

Figure 5. NF-Y utilizes the E2F1 pathway to trigger p53 activation. Western blot analysis of protein extracts from Ad-NF-YA-, Ad- control-infected MEFs, and MEFs treated with 0.5 mg/mL of Adriamycin (A); Ad-NF-YA-, Ad- control-infected MEFs, and E2F1 / MEFs (B); and MCF10A C D following interference of E2F1 and/or NF-YA transfection (C). In all panels, the antibodies used are indicated. D, MCF10A cells following interference of E2F1 and/or NF-YA transfection were plated in triplicate, harvested, and quantified for the percentage of trypan blue–positive cells.

manner. Moreover, they also suggest that unrestricted NF-Y some similarities to that induced by oncogenes. The role of activity increases cell proliferation in cells expressing mutant E2F1 as mediator of the NF-Y effect is clearly demonstrated by p53. the fact that unrestricted NF-Y activity does not induce apoptosis in E2F1 / cells. Furthermore, the increase in Discussion E2F1 mRNA by unrestricted NF-Y activity and the NF-Y– dependent enrichment on the E2F1 promoter of hyperacety- Over the past 10 years, much effort has been made toward lated histones H3 and H4 demonstrate that E2F1 is more establishing the role of NF-Y in controlling cell proliferation actively transcribed after NFY overexpression. Our results in vivo (4,6,7,12–14). Here, we examined the effect of support a model in which excessive E2F1, induced by NF-Y, unrestricted NF-Y activity on proliferation of primary cells. sensitizes cells to apoptosis. We show that sustained activity of NF-Y can regulate path- Unrestricted E2F1 activity induces p53-dependent apop- ways involved in apoptosis and we describe a novel linkage tosis in both cell cultures and mouse models (25–32), and between NF-Y, E2F1, and p53 that could help to understand our data reveal a specific role for p53 in transducing NF-Y the fine cross talk between cell cycle and apoptosis. The apoptotic signaling. It is, therefore, not surprising that p53 relevant new findings are that unrestricted NF-Y activity is also mutated or deleted in a large number of human (a) inhibits cell proliferation by inducing apoptosis and cancers (45). Our results further suggest that deregulated (b) leads to an E2F1- and wtp53-dependent apoptosis in NF-Y expression induces apoptosis by activating an E2F1- both mouse and human cells. dependentp53pathwayknownnottoinvolvep19ARF(44). The evidence that deregulated NF-Y expression induces This is in agreement with the ability of ectopically expressed apoptosis by activating an E2F1-dependent p53 pathway links ARF to target E2F1 for degradation (48). Furthermore, our loss of proliferation control to an apoptotic pathway, with evidence that unrestricted NF-Y activity induces ATM,

www.aacrjournals.org Cancer Res; 70(23) December 1, 2010 9717

Gurtner et al.

Figure 6. NF-Y activity leads to growth inhibition in human wtp53 expressing cells. A, colony formation assay was performed on the indicated cell lines transfected with the indicated vectors. Fold difference values are given versus the control sample. B and C, MCF10a and SKBR3 cells transfected with the indicated vectors were plated in triplicate CD and total cell numbers were counted daily. Means SD of 4 independent experiments are reported. D, 50mg of protein extract derived from MCF10A and SKBR3 cells transfected with the indicated vectors was subjected to Western blot analysis with the indicated antibodies. E2F1 protein was detected with both monoclonal (*) and polyclonal (**) antibodies.

which is known to phosphorylate and thus activate E2F1, NF-Y is apparently not overexpressed in tumor cells, strongly suggests that NF-Y induces p53-dependent apop- although its posttranslational regulation, occurring in normal tosis through this kinase. cells, seems to be lost in transformed cells expressing mutant On the basis of our results from NF-YA overexpression, we p53, leading to its constitutive expression (Manni et al., manu- speculate that the amount of this protein can dictate the cell script in preparation). In agreement with this, we have shown decision between division and death. We have recently shown here that unrestricted NF-Y activity does not lead to apoptosis that a posttranslational molecular mechanism tightly regu- but to increased proliferation in mutant p53 expressing cells. lates the transcriptional activity of NF-Y controlling the It is possible that the ability of NF-Y to drive apoptosis is stability of NF-YA, the regulatory subunit of the complex distinct from its ability to drive cell division. However, this (10). NF-YA is degraded by ubiquitin-mediated proteolysis, issue cannot be directly addressed until NF-Y target genes and this could be one mechanism evolved by cells to tightly essential for apoptosis have been clearly identified. The factors regulate NF-Y activity, failure of which would compromise cell that determine the decisions of inducing either cell division or survival and/or homeostasis. cell death need to be further investigated.

9718 Cancer Res; 70(23) December 1, 2010 Cancer Research

Unrestricted NF-Y Activity Leads to Apoptosis

Disclosure of Potential Conflicts of Interest Grant Support

No potential conflicts of interest were disclosed This work has been partially supported by grants from AIRC to G.P. and R.F., Ministero della Salute, and ISS-ACC (ICS-120.4/RA00-90; R.F.02/184) to G.P., ISS- Italia-USA to A.S., and RF07Onc-26/1 to F.M. V.F. is a recipient of a fellowship Acknowledgments from FIRC.

We thank Maria Pia Gentileschi for technical advice and Donatella Vitolo for secretary assistance. We thank Giulia Fontemaggi for a gift of siE2F1 oligos. Received 10/07/2010; revised 09/30/2010; accepted 03/05/2010; Special thanks go to Silvia Bacchetti for editing the manuscript. published OnlineFirst 10/15/2010.

References 1. Mantovani R.The molecular biology of the CCAAT-binding factor NF- 17. Goodarzi H, Elemento O, Tavazoie S. Revealing global regulatory Y. Gene 1999;239:15–27. perturbations across human cancers. Mol Cell 2009;36:900–11. 2. Linhart C, Elkon R, Shiloh Y, Shamir R. Deciphering transcriptional 18. Vermeulen K, Berneman ZN, Van Bockstaele DR. Cell cycle and regulatory elements that encode specific cell cycle phasing by com- apoptosis. Cell Prolif 2003;36:165–75. parative genomics analysis. Cell Cycle 2005;4:1788–97. 19. Elkon R, Linhart C, Sharan R, Shamir R, Shiloh Y. Genome-wide in 3. Bolognese F, Wesner M, Lange-zu Dohna C, Gurtner A, Ronchi A, silico identification of transcriptional regulators controlling the cell Muller H, et al. The cyclin B2 promoter depends on NF-Y, a trimer cycle in human cells. Genome Res 2003;13:773–80. whose CCAAT-binding activity is cell cycle regulated. Oncogene 20. Salsi V, Caretti G, Wasner M, Reinhard W, Haugwitz U, Engeland K, 1999;18:1845–53. et al. Interactions between p300 and multiple NF-Y trimers govern 4. Di Agostino S, Strano S, Emiliozzi V, Zerbini V, Mottolese M, Sacchi A, cyclin B2 promoter function. J Biol Chem 2003;278:6642–50. et al. Gain of function of mutant p53: the mutant p53/NF-Y protein 21. Sciortino S, Gurtner A, Manni I, Fontemaggi G, Dey A, Sacchi A, et al. complex reveals an aberrant transcriptional mechanism of cell cycle The cyclin B1 gene is actively transcribed during mitosis in HeLa cells. regulation. Cancer Cell 2006;10:191–202. EMBO Rep 2001;2:1018–23. 5. Farina A, Manni I, Fontemaggi G, Tiainen M, Cenciarelli C, Bellorini 22. Yun J, Chae HD, Choi TS, Kim EH, Bang YJ, Chung J, et al. Cdk2- M, et al. Down-regulation of cyclin B1genetranscriptionintermin- dependent phosphorylation of the NF-Y transcription factor and its ally differentiated skeletal muscle cells is associated with loss of involvement in the p53-p21 signaling pathway. J Biol Chem functional CCAAT-binding NF-Y complex. Oncogene 1999; 2003;278:36966–72. 18:2818–27. 23. Jin S, Levine AJ. The p53 functional circuit. J Cell Sci 2001;114: 6. Gurtner A, Manni I, Fuschi P, Mantovani R, Guadagni F, Sacchi A, et al. 4139–40. Requirement for down-regulation of the CCAAT-binding activity of the 24. Hermeking H, Wolf DA, Kohlhuber F, Dickmanns A, Billaud M, Fanning NF-Y transcription factor during skeletal muscle differentiation. Mol E, et al. Role of c-myc in simian virus 40 large tumor antigen-induced Biol Cell 2003;14:2706–15. DNA synthesis in quiescent 3T3-L1 mouse fibroblasts. Proc Natl Acad 7. Gurtner A, Fuschi P, Magi F, Colussi C, Gaetano C, Dobbelstein M, Sci U S A 1994;91:10412–6. et al. NF-Y dependent epigenetic modifications discriminate between 25. Leone G, Sears R, Huang E, Rempel R, Nuckolls F, Park CH, et al. Myc proliferating and postmitotic tissue. PLoS One 2008;3:e2047. requires distinct E2F activities to induce S phase and apoptosis. Mol 8. Korner K, Jerome V, Schmidt T, Muller R. Cell cycle regulation of the Cell 2001;8:105–13. murine cdc25B promoter: essential role for NF-Y and a proximal 26. Johnson DG, Degregori J. Putting the oncogenic and tumor suppres- repressor element. J Biol Chem 2001;276:9662–9. sive activities of E2F into context. Curr Mol Med 2006;6:731–8. 9. Zwicher J, Lucibello FC, Wolfraim LA, Gross C, Truss M, Engeland K, 27. Kowalik TF, DeGregori J, Schwarz JK, Nevins JR. E2F1 overexpres- et al. Cell cycle regulation of the cyclin A, cdc25C and cdc2 genes is sion in quiescent fibroblasts leads to induction of cellular DNA synth- based on a common mechanism of transcriptional repression. EMBO esis and apoptosis. J Virol 1995;69:2491–500. J 1995;14:4514–22. 28. Qin XQ, Livingston DM, Kaelin WG Jr, Adams PD. Deregulated tran- 10. Manni I, Caretti G, Artuso S, Gurtner A, Emiliozzi V, Sacchi A, et al. scription factor E2F-1 expression leads to S-phase entry and p53- Posttranslational regulation of NF-YA modulates NF-Y transcriptional mediated apoptosis. Proc Natl Acad Sci U S A 1994;91:10918–22. activity. Mol Biol Cell 2008;19:5203–13. 29. Wu X, Levine AJ. p53 and E2F-1 cooperate to mediate apoptosis. 11. Grskovic M, Chaivorapol C, Gaspar-Maia A, Li H, Ramalho-Santos M. Proc Natl Acad Sci U S A 1994;91:3602–6. Systematic identification of cis-regulatory sequences active in mouse 30. Macleod K. pRb and E2f-1 in mouse development and tumorigenesis. and human embryonic stem cells. PLoS Genet 2007;3:e145. Curr Opin Genet Dev 1999;9:31–9. 12. Hu Q, Maity SN. Stable expression of a dominant negative mutant of 31. Matsumura I, Tanaka H, Kanakura Y. E2F1 and c-Myc in cell growth CCAAT binding factor/NF-Y in mouse fibroblast cells resulting in and death. Cell Cycle 2003;2:333–8. retardation of cell growth and inhibition of transcription of various 32. Iaquinta PJ, Lees JA. Life and death decisions by the E2F transcription cellular genes. J Biol Chem 2000;275:4435–44. factors. Curr Opin Cell Biol 2007;19:649–57. 13. Bhattacharya A, Deng JM, Zhang Z, Behringer R, de Crombrugghe B, 33. Hershko T, Chaussepied M, Oren M, Ginsberg D. Novel link between Maity SN. The B subunit of the CCAAT box binding transcription factor E2F and p53: proapoptotic cofactors of p53 are transcriptionally complex (CBF/NF-Y) is essential for early mouse development and upregulated by E2F. Cell Death Differ 2005;12:377–83. cell proliferation. Cancer Res 2003;63:8167–72. 34. Hiebert SW, Packham G, Strom DK, Haffner R, Oren M, Zambetti G, 14. Imbriano C, Gurtner A, Cocchiarella F, Di Agostino S, Basile V, Gostissa et al. E2F-1:DP-1 induces p53 and overrides survival factors to trigger M, et al. Direct p53 transcriptional repression: in vivo analysis of apoptosis. Mol Cell Biol 1995;15:6864–74. CCAAT-containing G2/M promoters. Mol Cell Biol 2005;25:3737–51. 35. Rogoff HA, Pickering MT, Debatis ME, Jones S, Kowalik TF. E2F1 15. Manni I, Mazzaro G, Gurtner A, Mantovani R, Haugwitz U, Krause K, induces phosphorylation of p53 that is coincident with p53 accumula- et al. NF-Y mediates the transcriptional inhibition of the cyclin B1, tion and apoptosis. Mol Cell Biol 2002;22:5308–18. cyclin B2, and cdc25C promoters upon induced G2 arrest. J Biol 36. Porrello A, Cerone MA, Coen S, Gurtner A, Fontemaggi G, Cimino L, Chem 2001;276:5570–6. et al. p53 regulates myogenesis by triggering the differentiation 16. Yamanaka K, Mizuarai S, Eguchi T, Itadani H, Hirai H, Kotani H. activity of pRb. J Cell Biol 2000;151:1295–304. Expression levels of NF-Y target genes changed by CDKN1B 37. Ruley HE. Adenovirus early region 1A enables viral and cellular correlate with clinical prognosis in multiple cancers. Genomics transforming genes to transform primary cells in culture. Nature 2009; 94:219–27. 1983;304:602–6.

www.aacrjournals.org Cancer Res; 70(23) December 1, 2010 9719

Gurtner et al.

38. Kabe Y, Yamada J, Uga H, Yamaguchi Y, Wada T, Handa H. NF-Y is 43. Li B, Carey M, Workman JL. The role of chromatin during transcription. essential for the recruitment of RNA polymerase II and inducible Cell 2007;128:707–19. transcription of several CCAAT box-containing genes. Mol Cell Biol 44. Polager S, Ginsberg D. p53 and E2F: partners in life and death. Nat 2005;25:512–22. Rev Cancer 2009;9:738–48. 39. Caretti G, Salsi V, Vecchi C, Imbriano C, Mantovani R. Dynamic 45. Brosh R, Rotter V. When mutants gain new powers: news from the recruitment of NF-Y and histone acetyltransferases on cell-cycle mutant p53 field. Nat Rev Cancer 2009;9:701–13. promoters. J Biol Chem 2003;278:30435–40. 46. Hong S, Paulson QX, Johnson DG. E2F1 and E2F3 activate ATM 40. Van Ginkel PR, Hsiao KM, Schjerven H, Farnham PJ. E2F-mediated through distinct mechanisms to promote E1A-induced apoptosis. Cell growth regulation requires transcription factor cooperation. J Biol Cycle. 2008;7:391–400. Chem 1997;272:18367–74. 47. Bakkenist CJ, Kastan MB. DNA damage activates ATM through 41. Barski A, Cuddapah S, Cui K, Roh TY, Schones DE, Wang Z, et al. intermolecular autophosphorylation and dimer dissociation. Nature High-resolution profiling of histone methylations in the human gen- 2003;421:499–506. ome. Cell 2007;129:823–37. 48. Martelli F, Hamilton T, Silver DP, Sharpless NE, Bardeesy N, Rokas M, 42. Berger SL. The complex language of chromatin regulation during et al. p19ARF targets certain E2F species for degradation. Proc Natl transcription. Nature 2007;447:407–12. Acad Sci USA 2001;98:4455–60.

9720 Cancer Res; 70(23) December 1, 2010 Cancer Research

Transcription Factor NF-Y Induces Apoptosis in Cells Expressing Wild-Type p53 through E2F1 Upregulation and p53 Activation

Aymone Gurtner, Paola Fuschi, Fabio Martelli, et al.

Cancer Res Published OnlineFirst October 15, 2010.

Updated version Access the most recent version of this article at: doi:10.1158/0008-5472.CAN-10-0721

Supplementary Access the most recent supplemental material at: Material http://cancerres.aacrjournals.org/content/suppl/2010/10/14/0008-5472.CAN-10-0721.DC1

E-mail alerts Sign up to receive free email-alerts related to this article or journal.

Reprints and To order reprints of this article or to subscribe to the journal, contact the AACR Publications Subscriptions Department at [email protected].

Permissions To request permission to re-use all or part of this article, use this link http://cancerres.aacrjournals.org/content/early/2010/11/25/0008-5472.CAN-10-0721. Click on "Request Permissions" which will take you to the Copyright Clearance Center's (CCC) Rightslink site.