Cdc25b Is Negatively Regulated by P53 Through Sp1 and NF-Y Transcription Factors

Oncogene (2011) 30, 2282–2288 & 2011 Macmillan Publishers Limited All rights reserved 0950-9232/11 www.nature.com/onc SHORT COMMUNICATION Cdc25B is negatively regulated by p53 through Sp1 and NF-Y transcription factors

M Dalvai1,2,5, O Mondesert1,2, J-C Bourdon3, B Ducommun1,2,4 and C Dozier1,2

1Universite´ de Toulouse, LBCMCP, Toulouse, France; 2CNRS, LBCMCP—UMR 5088 CNRS, University of Toulouse, Toulouse, France; 3Inserm-European Associated Laboratory U858, Department of Surgery and Molecular Oncology, College of Medicine, University of Dundee, Dundee, UK and 4CHU de Toulouse, Toulouse, France

Cdc25B phosphatases function as key players in G2/M CDK–cyclin complexes (Boutros et al., 2006). Cdc25B is cell cycle progression by activating the CDK1–cyclinB1 thought to act as the starter of mitosis, being responsible complexes. They also have an essential role in recovery for the initial dephosphorylation and activation of the from the G2/M checkpoint activated in response to DNA CDK1–cyclinB1 complex (Lindqvist et al., 2005). damage. Overexpression of Cdc25B results in bypass of Cdc25B is also essential for cell cycle recovery after the G2/M checkpoint and illegitimate entry into mitosis, DNA-damage induced checkpoint and its overexpres- and also causes replicative stress, leading to genomic sion results in bypass of the G2/M checkpoint and instability. Thus, fine-tuning of Cdc25B expression level illegitimate entry into mitosis, and also causes replica- is critical for correct cell cycle progression and G2 tive stress, leading to genomic instability (Miyata et al., checkpoint recovery. However, the transcriptional regula- 2001; Bugler et al., 2006). Thus, fine-tuning of Cdc25B tion of Cdc25B remains largely unknown. Earlier studies expression level is critical for correct cell cycle progres- have shown that the tumor suppressor p53 overexpression sion and for scheduled recovery from a G2 checkpoint transcriptionally represses Cdc25B; however, the mole- activated in response to DNA damage. cular mechanism of this repression has not yet been Upregulation of Cdc25B has been documented in a elucidated, although it was suggested to occur through the wide variety of human cancers, often associated with induction of p21. Here we show that Cdc25B is down- high-grade tumors and poor prognosis (Boutros et al., regulated by the basal level of p53 in multiple cell types. 2007). The mechanism leading to this upregulation is This downregulation also occurs in p21À/À cell lines, as yet unknown, but excludes gene amplification or indicating that p21 is not required for p53-mediated rearrangement. The components of the signaling path- regulation of Cdc25B. Deletion and mutation analyses of ways regulating Cdc25B levels are therefore likely to the Cdc25B promoter revealed that downregulation by be altered in cancer cells. This raises the major issue of p53 is dependent on the presence of functional Sp1/Sp3 understanding how and when Cdc25B levels are and NF-Y binding sites. Furthermore, chromatin immuno- regulated during cell cycle to allow entry into mitosis precipitation analyses show that p53 binds to the Cdc25B and recovery from a checkpoint arrest. promoter and mediates transcriptional attenuation The transcription factor p53 is activated in response through the Sp1 and NF-Y transcription factors. Our to a variety of cellular stresses, such as DNA damage results suggest that the inability to downregulate Cdc25B or oncogenic signal, and regulates expression of genes after loss of p53 might contribute to tumorigenesis. impinging on cell cycle arrest, senescence and apoptosis Oncogene (2011) 30, 2282–2288; doi:10.1038/onc.2010.588; (Riley et al., 2008). In particular, p53 represses impor- published online 17 January 2011 tant G2/M regulators such as cyclin B, Plk1, cdc2, Cdc25C, survivin and FoxM1 (Wang et al., 2010), Keywords: Cdc25B; p53; Sp1; NF-Y establishing a role for p53 at the G2/M transition. Recent studies show that p53 can regulate expression of genes in the absence of any cellular stresses, uncovering Introduction an additional role of the p53 protein that is expressed under basal physiological and cell-culture stress condi- The Cdc25 phosphatases (Cdc25A, B and C) regulate tions (Vousden and Prives, 2009). Using microarray eukaryotic cell-cycle progression through the depho- analysis, Cdc25B has been recently identified as a gene sphorylation and activation of their substrates, the transcriptionally repressed upon p53 overexpression (Scian et al., 2008). However, no molecular mechanism Correspondence: Dr C Dozier, LBCMCP—UMR 5088 CNRS, has been described concerning this repression. In part of University of Toulouse, Bat 4R3B1, 118 Route de Narbonne, our effort to decipher the regulation of the Cdc25B Toulouse 31062, France. phosphatase, we have analyzed the repression of E-mail: [email protected] Cdc25B by p53. Here we show that Cdc25B is down- 5Current address: CNRS, LBME-UMR5099, F-31062 Toulouse, France regulated by the basal level of the tumor suppressor p53 Received 8 July 2010; revised 1 November 2010; accepted 30 November in multiple cell types and elucidate the mechanism 2010; published online 17 January 2011 of this downregulation. Cdc25B is negatively regulated by p53 M Dalvai et al 2283 Results and discussion accumulation of p53 by nutlin-3 led to the downregulation of Cdc25B at both the mRNA and protein We observed that Cdc25B protein level was lower levels in p21À/À HCT116, indicating that p53 can in HCT116 human colon carcinoma cells expressing downregulate Cdc25B independently of p21. However, wild-type (WT) p53 than in their isogenic counterparts the downregulation was not as pronounced in p21À/À lacking p53 (HCT116 p53À/À) (Figure 1a). Cdc25B HCT116 as that observed with WT cells. This results is an unstable protein degraded in a proteasome- probably from the upregulation of the p53 target gene dependent manner (Baldin et al., 1997a). Treatment of p21 in nutlin-treated WT HCT116 cells, as in earlier HCT116 WT or p53 null cells with the translational studies Cdc25B was also repressed by overexpression inhibitor, cycloheximide, or the proteasomal inhibitor, of p21 (Scian et al., 2008). Therefore, it is likely that MG132, at various times showed that the higher there exist p21-dependent and -independent pathways of Cdc25B protein level observed in HCT116 p53 null repression. cells was not the result of deregulation of protein It has been previously shown that a 941-bp fragment degradation or synthesis (Supplementary Figure S1). of the human Cdc25B promoter was able to mediate Analysis of the Cdc25B expression at the mRNA level the repression by overexpressed p53 (Scian et al., 2008). showed that HCT116 WT cells also express a lower In silico analyses of this sequence revealed no consensus Cdc25B mRNA level than p53À/À cells (Figure 1b). p53-binding sites. However, p53 can repress indirectly To examine further whether the correlation observed by interfering with transcriptional factors that usually between absence of p53 and increased Cdc25B levels in transactivate the genes (Riley et al., 2008; Wang et al., HCT116 p53À/À cells stems from p53-dependent effect, 2010). To identify promoter elements involved in p53- we invalidated p53 expression by RNA interference in mediated regulation of Cdc25B, progressively deleted HCT116 WT cells. Although p53 expression was fragments of the Cdc25B promoter were linked to the significantly downregulated in p53 siRNA-transfected luciferase reporter gene and the resulting constructs cells compared with mock- or scr siRNA-transfected were transfected in HCT116 cells previously transfected cells, increased levels of Cdc25B protein and mRNA with p53 siRNA or scr siRNA as control. Luciferase were observed (Figure 1c). No change in Cdc25B assays showed that invalidation of p53 increased the expression was observed in HCT116 p53À/À cells activities of the –911, –575, –227 and –122 luciferase transfected with p53 siRNA (Figure 1c). To determine constructs, but not that of the –51 luciferase construct, whether this observation could be extended to other suggesting the presence of p53 response elements within cell types, we invalidated p53 in IMR-90 and MCF-7 the –122 to –51 region of the Cdc25B promoter cells and found that Cdc25B was upregulated in p53 (Figure 3a). This region contains two GC-rich motifs siRNA-transfected cells (Figure 1c and Supplementary for binding Sp1/Sp3 and one CAAT box for binding Figure S2). As the expression of Cdc25B is cell cycle NF-Y, both conserved between human and mouse regulated, being initiated during S-phase, peaking in G2 genomes (data not shown), and important for transcrip- and declining in mitosis (Baldin et al., 1997b; Korner tional activity of the Cdc25B promoter (Korner et al., et al., 2001), we analysed the cell cycle distribution of the 2001), as well as two juxtaposed cell cycle-dependent siRNA-transfected cells and found that both scr and p53 element motifs (Figure 3b). There have been several siRNA-transfected cells showed similar cell-cycle dis- reports of p53 repressing genes via Sp1/Sp3, NF-Y or tribution profile (Figure 1d). Thus, increased Cdc25B cell cycle-dependent element sequences (Muller and expression in p53-depleted cells is a direct effect of p53 Engeland, 2009; Wang et al., 2010). To clarify the role and not due to alteration of cell-cycle progression. These of these elements in the p53-mediated regulation of data indicate that Cdc25B is downregulated by the basal Cdc25B, we mutated the binding sites for Sp1/Sp3 and level of p53. NF-Y as well as the cell cycle-dependent element motifs, It has been suggested that overexpression of p53 may singly or in combinations, within the –122 luciferase repress Cdc25B expression through the induction of p21 construct. The promoter mutants were then transfected (Scian et al., 2008). To investigate the role of p21 in the into HCT116 cells invalidated for p53 expression. p53-dependent regulation of Cdc25B, we invalidated Luciferase assays (Figure 3c and Supplementary Figure p53 in the p21 null derivative of HCT116 cells (HCT116 S3) showed that invalidation of p53 into HCT116 cells p21À/À). Our results showed increased levels of Cdc25B induced the cell cycle-dependent element mutant and the protein and mRNA in p53 siRNA-transfected cells WT promoters to comparable levels of activity, whereas compared with control scr siRNA-transfected cells both Sp1/Sp3 and NF-Y mutant promoters were (Figure 2a). This suggests that the mechanism of p53- induced to a lesser extent. However, when the Sp1/ mediated attenuation of Cdc25B does not require p21. Sp3-binding sites were mutated in conjunction with the To further validate this result, HCT116 p21À/À cells NF-Y site, no effect of p53 invalidation was detected on were treated with nutlin-3, a small molecule disrupting the Cdc25B mutant promoter. These results indicate the interaction between p53 and its major negative that Sp1/Sp3 and NF-Y binding sites are both involved regulator MDM2, which results in stabilization and in p53-mediated regulation of Cdc25B. accumulation of p53 in the absence of DNA damage, The protein p53 interacts with Sp1 and NF-Y but at a level comparable to that of the genotoxic drugs transcription factor (Imbriano et al., 2005; Koutsodontis etoposide and doxorubicin (Thompson et al., 2004; et al., 2005). This interaction either reduces the ability of Vassilev et al., 2004). We observed (Figure 2b) that these factors to bind to DNA, resulting in insufficient

Oncogene Cdc25B is negatively regulated by p53 M Dalvai et al 2284

HCT116 ** 3 WT p53-/- p53 2 cdc25B 1

* mRNA level

actin Relative Cdc25B 0 WT p53-/-

HCT116 WT IMR-90 HCT116 p53-/- siRNA siRNA siRNA - scr p53 -scr p53 -scr p53 p53 cdc25B cdc25B * * p21

actin actin Real-time RT-PCR ** 2.5 2.5 2.5 * 2.0 2.0 2.0 1.5 1.5 1.5 1.0 1.0 1.0

mRNA level 0.5 mRNA level 0.5 mRNA level 0.5 Relative Cdc25B Relative Cdc25B Relative Cdc25B 0.0 0.0 0.0 - scr p53 - scr p53 - scr p53 siRNA siRNA siRNA

HCT116 WT IMR-90 100 100 G1 80 80 S 60 60 G2/M 40 40 % of total % of total 20 20 0 0 - scr p53 - scr p53 siRNA siRNA Figure 1 Inactivation of endogenous p53 leads to upregulation of Cdc25B (a) Increased levels of Cdc25B protein are observed in HCT116 cells lacking p53. Total cell extracts isolated from wild-type (WT) and p53À/À HCT116 (human colon carcinoma) cells were analysed by immunoblotting with anti-p53 (mouse DO-1, Santa Cruz Biotechnology, Santa Cruz, CA, USA), anti-Cdc25B (rabbit C-20, Santa Cruz Biotechnology), and anti-actin (mouse, Chemicon, Temecula, CA, USA) as loading control. The asterisk indicates an alternatively spliced variant of Cdc25B lacking the domain B and detected in some cell lines. (b) Cdc25B mRNA expression was upregulated in HCT116 lacking p53, determined by quantitative real-time RT–PCR. Total RNA was extracted from WT and p53À/À HCT116 cells using RNeasy Mini Kit (Invitrogen, Carlsbad, CA, USA) and the first strand was synthesized using Thermoscript RT–PCR system (Invitrogen). The real-time RT–PCR was performed using Platinum SYBR Green qPCR SuperMix- UDG (Invitrogen) with primers Cdc25B: Fw, 50-TCATTTTCCACTGTGAATTCTCATC-30 and Rv, 50-AGTCGTTGACAG CACGGTCTC-30. The relative Cdc25B mRNA level was normalized to human GAPDH levels: Fw, 50-GAAGGTGAAGGTCG GAGTCA-30 and Rv, 50-GAAGATGGTGATGGGATTTC-30, and set to 1 for WT cells. (c) Silencing of endogenous p53 increases Cdc25B expression at the mRNA and protein levels. IMR90 (human normal fibroblasts), WT and p53À/À HCT116 cells were mock- transfected or transfected with p53 siRNA (50-GACUCCAGUGGUAAUCUAC-dTdT-30) (Brummelkamp et al., 2002) or scr siRNA (Dharmacon, Inc, Chicago, IL, USA) using the Nucleofector device (Lonza Ltd, Basel, Switzerland). After 48 h, transfected cells were subjected to western blot analyses (top panels) as in (a), using also anti-p21 antibody (rabbit, Santa Cruz), and quantitative real-time RT–PCR (bottom panels) as in (b). The relative Cdc25B mRNA level was set to 1 for each mock-transfected conditions (À). (b, c, real- time RT–PCR) Results are means from three independent experiments with s.e.m. indicated. Statistical analyses were performed using a parametric paired t-test. *Po0.05 and **Po0.01. (d) FACS analysis of the cell cycle distribution profile of WT HCT116 and IMR90 mock transfected or transfected with p53 or scr siRNAs.

Oncogene Cdc25B is negatively regulated by p53 M Dalvai et al 2285 activation of their target promoter (Jung et al., 2001; general transcription factors (Imbriano et al., 2005; Jin Zhou et al., 2003; Bocangel et al., 2009), or interferes et al., 2008). To elucidate the mechanism by which p53 with the downstream recruitment of coactivators and/or downregulates Cdc25B, we performed chromatin immunoprecipitation (ChIP) analyses. For this purpose, HCT116 p21-/- we used the EJp53 cells that have lost functional siRNA scr - + - endogenous p53 but express WT p53 under the control of a tetracycline-regulated promoter (Sugrue et al., siRNA p53 - - + 1997). Upon tetracycline removal, p53 levels increased with an upregulation of p21 and a decrease of Cdc25B p53 expression at the mRNA and protein level (Supplementary Figure S4). This downregulation of Cdc25B after p53 cdc25B induction was not due to p53-mediated alteration of cell * cycle, because after 10 h of induction there was only a actin very minor effect on the cell cycle profile. Chromatin fragments were immunoprecipitated from either EJp53 Real-time cells left uninduced or induced to express p53 for RT-PCR 24 h, using anti-p53, anti-Sp1, anti-Sp3 and anti-NF-Y antibodies, or an irrelevant antibody as control. The 3 precipitated DNA was amplified by quantitative RT– PCR using primers designed to specifically amplify the proximal region of the Cdc25B promoter. Our results 2 showed (Figure 4a) that Sp3 was bound to the Cdc25B promoter in vivo in uninduced cells, and no change in * 1 Sp3 recruitment was observed following p53 induction. mRNA level Although in uninduced cells p53 and Sp1 were not or Relative Cdc25B were poorly detected on the Cdc25B promoter, a 0 recruitment of these factors was observed in p53-induced cells. NF-Y was bound on the Cdc25B promoter in - scr p53 siRNA uninduced cells but was absent in p53-induced cells. This result suggests that p53 downregulates Cdc25B by reducing the binding of NF-Y on the Cdc25B promoter, WT p21-/- thus preventing the full activation of this promoter, as Nutlin-3 (µM) 020020 has already been shown for cyclinB and cdk1 (Jung et al., 2001). To demonstrate that NF-Y is involved in p53 p53 regulation of Cdc25B, we inhibited NF-YB expression, one of the three subunits of the NF-Y factor, by cdc25B transfecting HCT116 cells with NF-YB siRNA. Redu- * cing expression of NF-YB (by 25% only, because of poor efficiency of the siRNA), reduced Cdc25B mRNA p21 and protein levels, and silencing p53 in NF-Y siRNA- transfected cells increased Cdc25B expression only very slightly, compared with scr-transfected cells (Figure 4b). MDM2 These results, in agreement with the luciferase experiments (Figure 3c and Supplementary Figure S3), actin demonstrate that NF-Y contributes both to the expression and to the p53 regulation of Cdc25B. To test WT p21-/- whether the reduced binding of NF-Y on the Cdc25B promoter observed after induction of p53 interferes with the recruitment of coactivators, we performed ChIP 1.0

Figure 2 Downregulation of Cdc25B by p53 can be p21 independent. (a) p21À/À HCT116 cells were mock-transfected or transfected with p53 or control scr siRNAs for 48 h and subjected 0.5 to western blot (top panel) and quantitative real-time RT–PCR (bottom panel). The relative Cdc25B mRNA level was set to 1 for

mRNA level the mock-transfected condition (-). Results are the means from Relative Cdc25B three independent experiments with s.e.m. *Po0.05. (b) WT and p21À/À HCT116 cells were treated with 20 mM Nutlin-3 (Sigma, St 0.0 Louis, MO, USA) for 24 h before extraction for protein (top panel) and mRNA (bottom panel) analyses. MDM2 was detected with 0 20 0 20 mouse anti-MDM2 from Santa Cruz. Relative Cdc25B mRNA Nutlin-3 (µM) level was set to 1 for the untreated conditions.

Oncogene Cdc25B is negatively regulated by p53 M Dalvai et al 2286 luciferase activity (% of scr) NF-Y +104 50 100 150 200 Sp1/Sp3 CDE -911 // luc -911 -575 // luc -575 -227 luc -227 -122 luc -122 -51 luc -51 *

luciferase activity (% of scr) -122 GCCCTGCGGGCCCCGCCCT 50 100 150 200 Sp1/Sp3 -103 CAGTCCCGCCCTCATCTAA wt Sp1/Sp3 Sp* - 84 CCCGCTACCCCATTGGTGG NF-Y NF-Y* - 65 CGTCCGGCGGCGCGGCTG Sp*NF-Y* CDE CDE CDE *

Figure 3 Downregulation of Cdc25B by p53 occurs via GC-rich and CAAT box motifs. (a) Silencing of endogenous p53 increases the Cdc25B promoter activity. WT HCT116 cells were transfected with p53 or control scr siRNAs as described above, and, 24 h later, transiently co-transfected (Jet PEI, Poly plus transfection, Poly plus, Illkirch, France) with progressively deleted Cdc25B promoter luciferase constructs (pGL4.10 [luc2] vector, Promega) (represented schematically in the left panel, with sites for Sp1/Sp3, NF-Y and the CDE motifs indicated) and SV40 renilla, used as internal control. Cells were harvested 48 h post-transfection for dual-luciferase assays (Promega, Madison, WI, USA) (right panel). The normalized luciferase activity of each construct in the p53 siRNA transfected cells was compared with that of the scr siRNA transfected cells, which is expressed as 100%. Results are the means from three independent experiments with s.e.m indicated. *Po0.05. (b) Cdc25B promoter sequence encompassing the p53-responsive region, highlighting Sp1/Sp3 and NF-Y binding sites as well as the CDE motifs. (c) Both Sp1/Sp3 and NF-Y binding sites are involved in p53-mediated regulation of Cdc25B. The À122 Cdc25B promoter luciferase constructs WT or mutated (*) in binding sites for Sp1/Sp3 (Sp, with ACAT instead of CGCC) or NF-Y (TGAA instead of CAAT), or the CDE motifs (mutated into CTATG) were analysed for response to silencing of p53 as in (a).

experiments as above, using anti-histone acetyltransfer- on the Cdc25B promoter after p53 induction (Figure 4c). ase p300, a coactivator known to interact with NF-Y Altogether, our results show that the binding of p53 (Mantovani, 1999). Results show that, while p300 was on the Cdc25B promoter resulted in a switch between bound on the Cdc25B promoter in EJp53 uninduced the recruitment of the coactivator p300 and the core- cells, a release of this factor was observed after pressor DNMT1, thereby causing the downregulation of induction of p53 (Figure 4c). Cdc25B. Our ChIP experiment (Figure 4a) revealed that p53 Our ChIP experiments (Figure 4a) suggest that Sp3 binds to the Cdc25B promoter in vivo. However, p53 activates the Cdc25B promoter in vivo. In addition, could bind farther away than the proximal promoter. To transfection of Sp3-encoding vectors in HCT116 cells test this possibility, we performed ChIP experiments upregulated Cdc25B (Supplementary Figure S7). using primers designed to amplify sequences located Although Sp3 is a basal transcription factor, it is also upstream or downstream of the proximal Cdc25B involved in tumorigenesis, being overexpressed in cancer promoter. We observed no recruitment of p53 in p53- cells, often associated with poor prognosis and tumor induced cells, indicating that p53 binds to the proximal aggressiveness (Essaﬁ-Benkhadir et al., 2009; Lambertini promoter region (Supplementary Figure S5). Co- et al., 2010). Thus, elevated Cdc25B expression observed immunoprecipitation experiments performed in EJp53 cells in cancers could be due to Sp3 overexpression. uninduced or induced to express p53 for 24 h showed In this study, we report that Cdc25B is downregulated that p53 interacts with Sp1 (Supplementary Figure S6). by basal levels of p53. Elevated levels of Cdc25B cause As there is no p53 binding site in the proximal Cdc25B cells with an S-phase DNA content to enter mitosis promoter, our results suggest that p53 interacts with Sp1 prematurely, and also cause replicative stress, leading to and the formed complexes are recruited on the Cdc25B chromosomal abnormalities and genomic instability, a promoter, as this has already been shown in the source of tumorigenesis (Karlsson et al., 1999; Varmeh promoters of different genes (Torgeman et al., 2001; and Manfredi, 2009; Bugler et al., 2010). Thus, a tight Innocente and Lee, 2005; Lin et al., 2010). The p53–Sp1 regulation of Cdc25B by p53 ensures proper cell cycle complexes could then act as a loading platform for progression. Data from many studies point out the fact repressor complexes such as the histone deacetylase that the vast majority of genes displaying upregulation HDAC1 or the DNA methyltransferase DNMT1 (Wierstra, in p53-null or mutant cells, or p53 RNAi cell lines, are 2008). The ChIP experiments performed with anti- proliferation-related genes, some of them being impli- DNMT1 showed that DNMT1 was indeed recruited cated in the G2/M transition (Brosh and Rotter, 2010).

Oncogene Cdc25B is negatively regulated by p53 M Dalvai et al 2287 Promoter ChIP Assay Figure 4 Importance of Sp1 and NF-Y in downregulation of Cdc25B Uninduced by p53. (a) Induction of p53 in EJp53 cells leads to recruitment of p53 p53 induced 0.4 0.025 and Sp1 on the Cdc25B promoter in vivo and decrease of NF-Y binding. ChIP analyses were performed as described previously 0.020 0.3 (Iacovoni et al., 2010) from EJp53 cells left uninduced or induced to 0.015 express p53 for 24 h. Chromatin fragments were immunoprecipitated 0.2 using antibodies against p53 (mouse DO-1, Santa-Cruz), Sp1 (rabbit 0.010 sc-59X, Santa Cruz), Sp3 (rabbit sc-644X, Santa Cruz), NF-YB (goat

0.1 ChIP efficiency C-20X, Santa Cruz) or an irrelevant antibody (irr) as control. The

ChIP efficiency 0.005 precipitated DNA was amplified by real-time PCR, with primer sets 0.0 0.000 designed to amplify the proximal region of the Cdc25B promoter irr p53 Sp1 Sp3 irr NF-YB (primers Fw: 50-CCTCATCTAACCCGCTACCC-30 and Rv: 50- antibodies antibodies AGCCAGGGAAGCCTCTAGTT-30). The values of ChIP efficiencies are given as % of input with s.e.m. indicated. (b) HCT116 cells were 0 Uninduced transfected with NF-YB siRNA (5 -GGCAUUUACUAACCA- 0 siRNA p53 induced GUUA-dTdT-3 , Benatti et al., 2008), p53 siRNA, both together, or 0.03 scr siRNA for 48 h, and subjected to western blot and quantitative real- NF-YB time RT–PCR analyses. NF-YB was detected with rabbit anti-NF-YB (sc-13045) from Santa Cruz. Cdc25B, NF-YB and p53 protein levels 1 0.66 0.97 0.71 0.02 are indicated below, relative to protein levels measured in control scr siRNA-transfected cells set to 1. The relative Cdc25B mRNA level was p53 * 0.01 set to 1 for the scr-transfected cells. (c) Induction of p53 in EJp53 cells 1 0.9 0.21 0.30 resulted in a switch between the recruitment of the histone acetyl ChIP efficiency cdc25B transferase p300 and that of DNMT1 on the Cdc25B promoter in vivo. * 0.00 ChIP analyses were performed as above using antibodies against p300 1 0.56 1.68 0.93 irr p300 DNMT1 (rabbit sc-585, Santa Cruz), DNMT1 (mouse IMG-261A, IMGENEX) actin antibodies or an irrelevant antibody (irr) as control. The values of ChIP efficiencies are given as % of input with s.e.m. indicated. Real-time 2.00 RT-PCR 1.75 activity of NF-Y (Di Agostino et al., 2006), the 1.50 increased expression of Cdc25B in mutant p53 tumors 1.25 (Troester et al., 2006; Langerod et al., 2007) may stem 1.00 not only from loss of WT p53 repressive activity but also 0.75 *

mRNA level 0.50 from the gain of functional properties of the mutant Relative Cdc25B 0.25 protein. 0.00 Collectively, our results suggest that the inability to scrNF-YB p53 p53 downregulate Cdc25B after loss of p53 might contribute NF-YB siRNA to tumorigenesis.

Thus, in the absence of any cellular stresses, p53 may Conflict of interest serve as a regulator for the fine-tuning of genes implicated in the G2/M transition, such as Cdc25B. The authors declare no conflict of interest. Overexpression of Cdc25B is observed in many cancers, but, until now, the mechanisms leading to this upregulation were unknown. Our data suggest that loss of p53 (by inactivation or mutation), one of the most common Acknowledgements events occurring in cancer, might contribute to the upregulation of Cdc25B observed in many types of We thank Dr Bert Vogelstein (Howard Hughes Medical tumors. In agreement with this, Cdc25B was identified Institute, Baltimore USA) for providing the WT, p53À/À and p21À/À HCT116 cells. MD was a recipient of a post- in a set of genes more highly expressed in primary breast doctoral fellowship from the CNRS (Centre National de la tumors with p53 mutations (Troester et al., 2006; Recherche Scientifique). This work was supported by CNRS, Langerod et al., 2007). As many p53 mutants not only Universite´Paul Sabatier, la re´gion Midi-Pyreńeés, l’Institut act through dominant-negative function but also National du Cancer, the Cance´ropoˆle Grand Sud-Ouest and la acquire a gain-of-function, notably by enhancing the Ligue Nationale Contre le Cancer (Equipe labelliseé 2008).

References

Baldin V, Cans C, Knibiehler M, Ducommun B. (1997a). Phospho- Benatti P, Basile V, Merico D, Fantoni LI, Tagliaﬁco E, Imbriano C. rylation of human CDC25B phosphatase by CDK1-cyclin A (2008). A balance between NF-Y and p53 governs the pro- and anti- triggers its proteasome-dependent degradation. J Biol Chem 272: apoptotic transcriptional response. Nucleic Acids Res 36: 1415–1428. 32731–32734. Bocangel D, Sengupta S, Mitra S, Bhakat KK. (2009). p53-Mediated Baldin V, Cans C, Superti-Furga G, Ducommun B. (1997b). down-regulation of the human DNA repair gene O6-methylgua- Alternative splicing of the human CDC25B tyrosine phosphatase. nine-DNA methyltransferase (MGMT) via interaction with Sp1 Possible implications for growth control? Oncogene 14: 2485–2495. transcription factor. Anticancer Res 29: 3741–3750.

Oncogene Cdc25B is negatively regulated by p53 M Dalvai et al 2288 Boutros R, Dozier C, Ducommun B. (2006). The when and wheres of Langerod A, Zhao H, Borgan O, Nesland JM, Bukholm IR, Ikdahl T et al. Cdc25 phosphatases. Curr Opin Cell Biol 18: 185–191. (2007). TP53 mutation status and gene expression profiles are powerful Boutros R, Lobjois V, Ducommun B. (2007). CDC25 phosphatases in prognostic markers of breast cancer. Breast Cancer Res 9: R30. cancer cells: key players? Good targets? Nat Rev Cancer 7: 495–507. Lin RK, Wu CY, Chang JW, Juan LJ, Hsu HS, Chen CY et al. (2010). Brosh R, Rotter V. (2010). Transcriptional control of the proliferation Dysregulation of p53/Sp1 control leads to DNA methyltransferase- cluster by the tumor suppressor p53. Mol Biosyst 6: 17–29. 1 overexpression in lung cancer. Cancer Res 70: 5807–5817. Brummelkamp TR, Bernards R, Agami R. (2002). A system for stable Lindqvist A, Kallstrom H, Lundgren A, Barsoum E, Rosenthal CK. expression of short interfering RNAs in mammalian cells. Science (2005). Cdc25B cooperates with Cdc25A to induce mitosis but has a 296: 550–553. unique role in activating cyclin B1-Cdk1 at the centrosome. J Cell Bugler B, Quaranta M, Aressy B, Brezak MC, Prevost G, Biol 171: 35–45. Ducommun B. (2006). Genotoxic-activated G2-M checkpoint exit Mantovani R. (1999). The molecular biology of the CCAAT-binding is dependent on CDC25B phosphatase expression. Mol Cancer Ther factor NF-Y. Gene 239: 15–27. 5: 1446–1451. Miyata H, Doki Y, Yamamoto H, Kishi K, Takemoto H, Fujiwara Y Bugler B, Schmitt E, Aressy B, Ducommun B. (2010). Unscheduled et al. (2001). Overexpression of CDC25B overrides radiation- expression of CDC25B in S-phase leads to replicative stress and induced G2-M arrest and results in increased apoptosis in DNA damage. Mol Cancer 9: 29. esophageal cancer cells. Cancer Res 61: 3188–3193. Di Agostino S, Strano S, Emiliozzi V, Zerbini V, Mottolese M, Sacchi A Muller GA, Engeland K. (2009). The central role of CDE/CHR et al. (2006). Gain of function of mutant p53: the mutant p53/NF-Y promoter elements in the regulation of cell cycle-dependent gene protein complex reveals an aberrant transcriptional mechanism of cell transcription. FEBS J 277: 877–893. cycle regulation. Cancer Cell 10: 191–202. Riley T, Sontag E, Chen P, Levine A. (2008). Transcriptional control Essafi-Benkhadir K, Grosso S, Puissant A, Robert G, Essafi M, of human p53-regulated genes. Nat Rev Mol Cell Biol 9: 402–412. Deckert M et al. (2009). Dual role of Sp3 transcription factor as an Scian MJ, Carchman EH, Mohanraj L, Stagliano KE, Anderson MA, inducer of apoptosis and a marker of tumour aggressiveness. PLoS Deb D et al. (2008). Wild-type p53 and p73 negatively regulate One 4: e4478. expression of proliferation related genes. Oncogene 27: Iacovoni JS, Caron P, Lassadi I, Nicolas E, Massip L, Trouche D et al. 2583–2593. (2010). High-resolution profiling of gamma H2AX around DNA double Sugrue MM, Shin DY, Lee SW, Aaronson SA. (1997). Wild-type p53 strand breaks in the mammalian genome. EMBO J 29: 1446–1457. triggers a rapid senescence program in human tumor cells lacking Imbriano C, Gurtner A, Cocchiarella F, Di Agostino S, Basile V, functional p53. Proc Natl Acad Sci USA 94: 9648–9653. Gostissa M et al. (2005). Direct p53 transcriptional repression: Thompson T, Tovar C, Yang H, Carvajal D, Vu BT, Xu Q et al. (2004). in vivo analysis of CCAAT-containing G2/M promoters. Phosphorylation of p53 on key serines is dispensable for transcriptional Mol Cell Biol 25: 3737–3751. activation and apoptosis. JBiolChem279: 53015–53022. Innocente SA, Lee JM. (2005). p53 is a NF-Y- and p21-independent, Torgeman A, Mor-Vaknin N, Zelin E, Ben-Aroya Z, Lochelt M, Sp1-dependent repressor of cyclin B1 transcription. FEBS Lett 579: Flugel RM et al. (2001). Sp1-p53 heterocomplex mediates activation 1001–1007. of HTLV-I long terminal repeat by 12-O-tetradecanoylphorbol-13- Jin W, Chen Y, Di GH, Miron P, Hou YF, Gao H et al. (2008). acetate that is antagonized by protein kinase C. Virology 281: Estrogen receptor (ER) beta or p53 attenuates ER alpha-mediated 10–20. transcriptional activation on the BRCA2 promoter. J Biol Chem Troester MA, Herschkowitz JI, Oh DS, He X, Hoadley KA, Barbier 283: 29671–29680. CS et al. (2006). Gene expression patterns associated with p53 status Jung MS, Yun J, Chae HD, Kim JM, Kim SC, Choi TS et al. (2001). in breast cancer. BMC Cancer 6: 276. p53 and its homologues, p63 and p73, induce a replicative Varmeh S, Manfredi JJ. (2009). Inappropriate activation of cyclin- senescence through inactivation of NF-Y transcription factor. dependent kinases by the phosphatase Cdc25b results in premature Oncogene 20: 5818–5825. mitotic entry and triggers a p53-dependent checkpoint. J Biol Chem Karlsson C, Katich S, Hagting A, Hoffmann I, Pines J. (1999). Cdc25B 284: 9475–9488. and Cdc25C differ markedly in their properties as initiators of Vassilev LT, Vu BT, Graves B, Carvajal D, Podlaski F, Filipovic Z mitosis. J Cell Biol 146: 573–584. et al. (2004). in vivo activation of the p53 pathway by small- Korner K, Jerome V, Schmidt T, Muller R. (2001). Cell cycle regulation of molecule antagonists of MDM2. Science 303: 844–848. the murine cdc25B promoter: essential role for nuclear factor-Y and a Vousden KH, Prives C. (2009). Blinded by the light: the growing proximal repressor element. JBiolChem276: 9662–9669. complexity of p53. Cell 137: 413–431. Koutsodontis G, Vasilaki E, Chou WC, Papakosta P, Kardassis D. Wang B, Xiao Z, Ko HL, Ren EC. (2010). The p53 response element (2005). Physical and functional interactions between members of the and transcriptional repression. Cell Cycle 9: 870–879. tumour suppressor p53 and the Sp families of transcription factors: Wierstra I. (2008). Sp1: emerging roles–beyond constitutive activation importance for the regulation of genes involved in cell-cycle arrest of TATA-less housekeeping genes. Biochem Biophys Res Commun and apoptosis. Biochem J 389: 443–455. 372: 1–13. Lambertini C, Pantano S, Dotto GP. (2010). Differential control of Zhou Y, Mehta KR, Choi AP, Scolavino S, Zhang X. (2003). DNA Notch1 gene transcription by Klf4 and Sp3 transcription factors in damage-induced inhibition of securin expression is mediated by p53. normal versus cancer-derived keratinocytes. PLoS One 5: e10369. J Biol Chem 278: 462–470.

Supplementary Information accompanies the paper on the Oncogene website (http://www.nature.com/onc)

Oncogene