Histone methyltransferase protein SETD2 interacts with p53 and selectively regulates its downstream genes

Cellular Signalling 20 (2008) 1671–1678

Contents lists available at ScienceDirect

Cellular Signalling j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / c e l l s i g

Histone methyltransferase protein SETD2 interacts with p53 and selectively regulates its downstream genes Ping Xie a,b, Chunyan Tian a,⁎, Liguo An b, Jing Nie a,c, Kefeng Lu a, Guichun Xing a, Lingqiang Zhang a,⁎, Fuchu He a,c,⁎ a b c

State Key Laboratory of Proteomics, Beijing Proteomics Research Center, Beijing Institute of Radiation Medicine, 27 Taiping Road, Beijing 100850, China Department of Life Science, Shandong Normal University, Ji Nan, Shandong Province 250014, China Department of Biological Sciences and Biotechnology, Tsinghua University, Beijing 100084, China

A R T I C L E

I N F O

Article history: Received 24 April 2008 Accepted 26 May 2008 Available online 27 June 2008 Keywords: Histone methyltransferase SETD2 p53 Transcriptional regulation

A B S T R A C T SETD2 (SET domain containing protein 2) is a histone H3K36 trimethyltransferase protein that associates with hyperphosphorylated RNA polymerase II and involves in transcriptional elongation. However, whether and how SETD2 is implicated in the specific regulation of gene transcription remains unknown. Here we show that SETD2 could interact with p53 and selectively regulate the transcription factor activity of p53. The interaction was dependent of C-terminal region of SETD2, which contains the SET and WW domains, and the N-terminal transactivation domain (residues 1–45) of p53. Overexpression of SETD2 upregulated the expression levels of a subset of p53 targets including puma, noxa, p53AIP1, fas, p21, tsp1, huntingtin, but downregulated that of hdm2. In contrast, it had no significant effect on those of 14-3-3σ, gadd45 and pig3. Consistently, knockdown of endogenous SETD2 expression by RNA interference resulted in converse effects as expected. In p53-deficient H1299 cells, SETD2 lost the ability to regulate these gene expression except hdm2, indicating the dependence of p53. Furthermore, we demonstrated that SETD2 downregulated hdm2 expression by targeting its P2 promoter and then enhanced p53 protein stability. Collectively, these findings suggest that the histone methyltransferase SETD2 could selectively regulate the transcription of subset genes via cooperation with the transcription factor p53. © 2008 Elsevier Inc. All rights reserved.

1. Introduction In eukaryotes, the regulation of chromatin structure modulates all DNA-templated processes such as DNA replication and transcription. One major mechanism that regulates the structure and function of chromatin is the covalent modification of histones [1,2]. These histone modifications, including acetylation, phosphorylation, ubiquitination, and methylation, create both synergistic and antagonistic signals that correlate with the transcriptional activity of a gene, through recruiting/dispelling some protein complexes or through changing the structure of chromatin to allow access for RNA polymerase to initiate transcription [3]. Moreover, these histone modifications and the consequent changes in chromatin structure may serve as an epigenetic marking system that is responsible for establishing and maintaining the heritable programs of gene expression during cellular differentiaAbbreviations: SETD2, SET domain containing protein 2; SmyD2, SET and MYND domain containing protein 2; HDM2, Human ortholog of murine double minutes 2 protein; HMTase, histone methyltransferase; HYPB, Huntingtin interaction protein B; RNAi, RNA interference. ⁎ Corresponding authors. Department of Genomics and Proteomics, Beijing Institute ^ of Radiation Medicine, 27 Taiping Road, Beijing 100850, China. Tel./fax: +86 10 68177417. E-mail addresses: [email protected] (C. Tian), [email protected] (L. Zhang), [email protected] (F. He). 0898-6568/$ – see front matter © 2008 Elsevier Inc. All rights reserved. doi:10.1016/j.cellsig.2008.05.012

tion and organism development. The function of many of these modifications is not well understood, although several of them have been linked to transcriptional activation and repression, DNA repair, and cell cycle regulation [4–6]. SET domain containing protein 2 (SETD2, also known as Huntingtin interacting protein B, HYPB) is a 230-kD protein that could trimethylate histone H3K36, which associates with hyperphosphorylated RNA polymerase II (RNAPII) and contains a transcriptional activation domain [7,8]. The SET domain is an evolutionarily conserved ~ 130aa sequence motif. It was originally identified in members of polycomb group (PcG), trithorax group (trxG), and Su(var) genes and was named after the genes Su(var)3–9, Enhancer of zeste (E(z)) and trithorax (trx) [9–11]. Most histone methyltransferases (HMTases) carry other functional domains such as transcriptional activation or repression domains, protein–protein interaction domains, and protein–DNA/RNA interaction domains [12]. These domains direct the HMTases to certain protein complexes and mediate some specific activities [13]. p53 is a transcription factor that plays a central role in tumor suppression by directing cellular responses to diverse stresses including DNA damage and oncogene activation resulting in diverse biological effects [14,15]. The wide range of p53's biological effects can in part be explained by its activation of expression of a number of target genes [16]. HDM2, a major ubiquitin E3 ligase for p53 and also

1672

P. Xie et al. / Cellular Signalling 20 (2008) 1671–1678

the target of p53, plays an important role in down-regulating p53 activities. Recently, increasing amounts of data suggest that p53 stability and degradation are more complex than once thought [17–19]. In the present study, we identified SETD2 as a novel interacting protein of p53 in vivo through co-immunoprecipitation assay. SETD2 enhances the transcriptional activity of p53 and regulates expression of select p53 target genes. We also demonstrate in detail SETD2 may affect the stability of p53 protein by regulating HDM2 mRNA and protein levels. The present work expanded the view that SETD2 may bind to p53 and coordinately regulate certain genes transcription. 2. Materials and methods 2.1. Plasmid constructs, antibodies and reagents Plasmids of SETD2 and its deletion mutants were constructed by PCR, followed by subcloning into various vectors. The restriction endonuclease enzymes used in the subcloning were listed below: (i) pCMV-Myc-SETD2, SalI/KpnI; (ii) pCMV2-Flag-SETD2, BglII/SalI; (iii) pCMV-Myc-SETD2(1–920aa), SalI/KpnI. pCMV2-Flag-SETD2(915–2061aa) was a kind gift from Dr. Zhu Chen and described previously [7]. pCMV-Flag-SmyD2 was a kind gift from Dr. Jing Huang and described [20]. pCMV-Myc-p53 (ND1,ND2,ND3,CD1) was a kind gift from Dr. Shengcai Lin [21]. The luciferase plasmids pG13-Luc, pMG15-Luc and p21-luc were gifts from Dr. Bert Vogelstein [22]. The luciferase plasmids of MDM2 was plasmids of MDM2 were a kind gift from Dr. Moshe Oren [23]. Anti-Myc antibody was purchased from Cell Signaling. Anti-Flag antibody, anti-Flag-HRP antibody, the proteasome inhibitor MG132, the DNA damage-mimicking reagents MMS, Etopside and Doxorubicin were from Sigma. Antibodies to p53 (DO-1) (Cat. No. OP43L) were from Oncogene and for immunoprecipitation, whereas the anti-p53 HRP (Cat. No. HAF1355) was from R&D Systems. HDM2 antibody (Cat. No. sc-965) and GAPDH antibody (Cat. No. FL-335) were from Santa Cruz. Antibody to SETD2 (Cat. No. AB31358) was from Abcam. 2.2. Cell culture, transfection HEK293T, MCF7 were cultured in DMEM (HyClone) supplemented with 10% FBS, penicillin, streptomycin and glutamine, and H1299 cells were maintained in RPMI medium 1640 (HyClone) with 10% FBS, penicillin, streptomycin and glutamine. Transfections were performed with Lipofectamine 2000 (Invitrogen) according to the manufacturer's instructions. 2.3. Immunoprecipitation and immunoblotting For general cell lysis and co-immunoprecipitation of SETD2 and p53, HEK293T cells were transfected with indicated expression vectors by Lipofectamine 2000. Cells were cultured for 48 h in DMEM medium, and were incubated in 400 μl ATM lysis buffer (50 mM Tris–HCl, pH 7.5, 150 mM NaCl, 1% [v/v] Tween 20, 0.2% NP-40, 10% glycerol) supplemented with protease inhibitor cocktail (Roche) and phosphatase inhibitors (10 mM NaF and 1 mM Na3VO4) [24]. The lysate was incubated with indicated antibody 3 h at 4 °C, then added Protein A/G-plus Agarose (Santa Cruz) and rotated gently more than 8 h at 4 °C. The immunoprecipitates were washed at least three times in lysis buffer, and proteins were recovered by boiling the beads in 2 × SDS sample buffer and analyzed by western blotting. 2.4. siRNA-mediated knockdown of SETD2 The knockdown of SETD2 was performed by transfection of U2OS cells for 48 h with two sets of on-target plus siRNA duplex targeting human SETD2 (5′-AGAGGAUCUUGAUCAAUUATT-3′/5′-UAAUUGAUCAAGAUCCUCUTT-3′) or (5′-GGAGUAUGCACGAAACAAATT-3′/ 5′-UUUGUUUCGUGCAUACUCCTT-3′), respectively, with on-target Negative Control siRNA Oligo Duplex (5′-UUCUCCGAACGUGUCACGUTT-3′/5′-ACGUGACACGUUCGGAGAATT-3′) as controls.

2.5. Gene reporter assays MCF7 or H1299 cells at 60%–70% confluence were transfected using Lipofectamine 2000 (Invitrogen) according to manufacturer's protocol. After 48 h, cells were lysed in 100 μl of a passive lysis buffer (Promega). Luciferase activity was measured with the Dual Luciferase Assay System (Promega) according to the manufacturer's protocol [19]. 2.6. RT-PCR and real-time PCR Forty-eight hours after transfection, MCF7 or H1299 cells were harvested and total RNA was extracted using the TRIZOL reagent (Invitrogen) according to the manufacturer's instructions. Reverse transcription of 3 μg total RNA was performed by combining 1 μl ReverTraAce-α-™ (Toyobo.No.FSK-100), 4 μl 5×buffer, 1 μl 10 pmol/μl oligo(dT) primer, 2 μl 10 mM dNTP mix, 1 μl 10 U/μl RNase inhibitor, and water up to a volume of 20 μl. Reaction mixtures were incubated at 30 °C for 10 min, then 42 °C for 20 min, 99 °C for 5 min, 4 °C for 5 min and then diluted with 30 μl of water. Real-time RT-PCR was performed using the BioRad iQ5 PCR machine. Each PCR mixture contained 0.5 μl of cDNA template and primers at a concentration of 100 nM in a final volume of 25 μl of SYBR green reaction mix (Toyobo.No. QPK-201). Each PCR generated only the expected amplicon as shown by the melting temperature profiles of the final products and by gel electrophoresis. Standard curves were calculated using cDNA to determine the linear range and PCR efficiency of each primer pair. Reactions were done in triplicate, and relative amounts of cDNA were normalized to GAPDH. Primers used in these analyses were as follows: p53 F, 5′-GAGGGATGTTTGGGAGATGTAAGAAATG-3′ and p53 R, 5′-TTCACAGATATGGGCCTTGAAGTTAGAGAA-3′ [25]; puma F, 5′-GACCTCAACGCACAGTA-3′ and puma R, 5′-CTAATTGGGCTCCATCT-3′; HDM2 F, 5′ATCTTGGCCAGTATATTATG-3′ and HDM2 R, 5′-GTTCCTGTAGATCATGGTAT-3′ [26]; Noxa F, 5′AGAGCTGGAAGTCGAGTGT-3′ and Noxa R, 5′-GCACCTTCACATTCCTCTC-3′; p53AIP1 F, 5′TCTTCCTCTGAGGCGAGCT-3′ and p53AIP1 R, 5′-AGGTGTGTGTGTCTGAGCCC-3′ [27]; GADD45 F, 5′-TGCGAGAACGACATCAACAT-3′ and GADD45 R, 5′-TCCCGGCAAAAACAAATAAG-3′; Pig3 F, 5′-TCTCTGAAGCAACGCTGAAATTC-3′ and Pig3 R, 5′-ACGTTCTTCTCCCAGTAGGATCC-3′ [28]; 14-3-3σ F, 5′-GGCCATGGACATCAGCAAGAA-3′ and 14-3-3σ R, 5′-CGAAAGTGGTCTTGGCCAGAG-3′ [29]; p21 F, 5′-CACCGAGACACCACTGGAGG-3′ and p21 R, 5′GAGAAGATCAGCCGGCGTTT-3′; GAPDH F, 5′-GGGAAGGTGAAGGTCGGAGT-3′ and GAPDH R, 5′-TTGAGGTCAATGAAGGGGTCA-3′ [30]; Tsp1 F, 5′-CCCGTGGTCATCTTGTTCTGT-3′ and Tsp1 R, 5′-TTTCTTGCAGGCTTTGGTCTCC-3′ [31]; Fas F, 5′-GTGCTGGACCTCTTCCTGAA-3′ and Fas R,5′-CGGATGCCCAGGATGTGT-3′ [32]; Huntingtin F, 5′-AGTGATTGTTGCTATGGAGCGG-3′ and Huntingtin R, 5′-GCTGCTGGTTGGACAGAAACTC-3′ [33]; SETD2 F, 5′-TCATCGAGATATTAAGCGAATG-3′ and SETD2 R, 5′-TTTGGACACCGAGAAGAACA-3′.

3. Results 3.1. SETD2 interacts with p53 in vivo Most recently, evidences showed the SET domain-containing histone methyltransferase protein family, such as SET8, SET9 and SmyD2 could regulate the transcriptional activity of p53 [20,30,34]. Previous studies established that SETD2 protein is predominantly located within the nucleus to act as the histone methyltransferase. p53 is also mainly located in the nucleus as a transcriptional factor. In order to detect whether SETD2 could regulate the transcriptional activity of p53 like its family members, co-immunoprecipitation (Co-IP) assay was performed to investigate whether SETD2 protein could associate with p53 in vivo. Myc-tagged SETD2 and Flag-tagged p53 were transiently expressed in HEK293T cells, and the reciprocal Co-IP assays were performed with anti-Myc antibody or anti-Flag antibody. As shown in Fig. 1A, Flag-tagged p53 protein could be detected in SETD2 immunoprecipitate, and Myc-tagged SETD2 could also be detected in p53 immunoprecipitate, indicating their in vivo interactions.

Fig. 1. SETD2 interacts with p53 in vivo. (A) HEK293T cells were transfected with expression plasmids for Myc-SETD2 together with Flag-p53. Cell Lysates were immunoprecipitated with monoclonal anti-Flag or anti-Myc antibody. The cell lysates and immunoprecipitates were detected by western blot with anti-Myc HRP or anti-Flag HRP antibodies as indicated. (B) Endogenous SETD2 could interact with endogenous p53. HEK293 cells were immunoprecipitated with anti-p53 or normal IgG and immunoblotted with anti-SETD2 and anti-p53 HRP antibodies. Lys, lysate; IP, immunoprecipitate; IB, immunoblotting.

P. Xie et al. / Cellular Signalling 20 (2008) 1671–1678

To test if SETD2 is able to interact with p53 under physiological conditions, endogenous Co-IP assay was carried out. Cell lysates prepared from HEK293 cells were immunoprecipitated with an anti-p53 antibody or normal IgG as a control, followed by immunoblotting with anti-SETD2 antibody. As shown in Fig. 1B, endogenous SETD2 could be co-immunoprecipitated with endogenous p53 but not the control IgG. Taken together, the results showed that SETD2 could interact with p53 in vivo. 3.2. Mapping the regions required for the interaction between SETD2 and p53 To further characterize the p53-binding region in SETD2, two deletion truncates of SETD2 were used in the Co-IP assays with p53. HEK293T cells were co-transfected with Myc-SETD2, Myc-SETD2 (1– 920aa) or Flag-SETD2 (915–2061aa) and Flag-p53 or Myc-p53 as indicated. Cell lysates were immunoprecipitated with anti-Flag or anti-Myc antibody, and subjected to western blot analysis. Both SETD2 full-length and SETD2 (915–2061aa) could be detected in the immunoprecipitates with p53 (Fig. 2A). In contrast, the N-terminal 920 amino acids seem to be not required for this interaction (Fig. 2A), indicating C-terminal part of

1673

SETD2 containing the SET domain and WW domain was both necessary and sufficient for interaction with p53. Furthermore, we generated a series of p53 deletion mutants to define the SETD2-binding region in p53. Flag-SETD2 and Myc-tagged p53 truncates were co-expressed in HEK293T cells and cell lysates were immunoprecipitated with anti-Flag antibody. Western blot analysis of the immunoprecipitates showed that CD1 truncate which contains N-terminal TAD (transactivation domain, aa 1–45) and central DBD (DNA binding domain, aa 113–290) but lacks C-terminal oligomerization domain (OD) and regulatory domain (RD) (aa 290– 393) could interact with SETD2 (Fig. 2B). However, neither ND1 (deletion of TAD), ND2 (deletion of TAD and praline-rich region between TAD and DBD) nor ND3 (deletion of TAD and C-terminal 103 residues) could bind to p53, indicating the requirement of TAD domain of p53 in its interaction with SETD2. Thus, the C-terminal region of SETD2 and N-terminal TAD of p53 mediated their interaction. 3.3. SETD2 enhances the transcriptional activity of p53 We next investigated that whether SETD2 could influence the transcriptional activity of p53 by binding to p53. We transfected p53

Fig. 2. Mapping the region required between SETD2 and p53. (A) Flag-p53 was transfected into HEK293T cells together with or without Myc-SETD2 full-length or truncates 1–920aa (left). Myc-p53 was transfected with pCMV2-Flag vector or Flag-SETD2 (915–2061aa) (right). The immunoprecipitates were analyzed by immunoblotting with anti-Myc or anti-Flag antibody as indicated. (B) Co-immunoprecipitation assay of Flag-SETD2 full-length and Myc-p53 truncate, including p53-ND1 (45–393aa), ND2 (113–393aa), ND3 (45–290aa), CD1 (1–290aa). The cell lysates were prepared and immunoprecipitated with anti-Flag antibody, followed by western blotting analysis.

1674

P. Xie et al. / Cellular Signalling 20 (2008) 1671–1678

wild-type MCF7 and U2OS cells with a luciferase promoter reporter plasmid pG13L containing tandem 13 repeats of p53 responsive element or with a pMG15L mutant reporter plasmid as a control [19]. In this experiment, SmyD2 was used as a positive control, which had been shown to repress p53 transcriptional activity [20]. The results showed that the p53 transcriptional activity was significantly enhanced by SETD2 to about 3 folds in both cell lines, and inhibited by SmyD2 as expected, but neither SETD2 nor SmyD2 had any significant impact on pMG15L luciferase activity (Fig. 3A). To confirm the conclusion, exogenous p53 was expressed in p53-null H1299 cells together with pG13L, SETD2, SmyD2 or not as indicated. In the absence of exogenous p53, pG13L could not be activated, whereas in the presence of exogenous p53, pG13L could be activated to about 6 folds of the control (Fig. 3B). Coexpression of SETD2 could further enhance p53 activity to 1.5 folds and co-expression of SmyD2 could inhibit p53 activity to about 30% of the control (Fig. 3B). Importantly, knockdown of SETD2 in MCF7 cells resulted in a dramatic decrease of p53 transcriptional activity, indicating that endogenous SETD2 positively regulates p53 (Fig. 3C). Taken together, our results implicate that SETD2 could upregulate p53 activity. p53 is a transcription regulator that directs cellular responses to diverse stresses. We next examined whether SETD2 could impact on p53 transcriptional activity under DNA damage conditions [14,15].

MCF7 cells were treated with MMS, etoposide, and doxorubicin, which were chemicals that induce DNA damage resulting in the activation of a p53-responsive pathway. As shown in Fig. 3D, p53 transcriptional activity was obviously enhanced by SETD2 even under these conditions. In conclusion, SETD2 plays a positive role in regulating p53 both under normal condition or DNA damage conditions. 3.4. SETD2 regulates expression of select p53 target genes p53 acts as a transcriptional activator and induces the expression of a lot of downstream target genes. p53 regulation of these target genes contributes to the type and sensitivity of cellular response to different forms of stress. It is likely that unique sets of p53-regulated genes are responsible for different outcomes, such as cell cycle arrest, apoptosis or senescence, in response to distinct stimuli. To gain insight into the biological consequences for SETD2-mediated regulation of p53, the ability of SETD2 to regulate transcription of the p53 target genes was determined. We measured the mRNA level of typical target genes of p53 using quantitative real-time RT-PCR in the presence of overexpressed SETD2 or knockdown of endogenous SETD2. Overexpression of SETD2 in MCF7 cells significantly increased the levels of apoptosis-related genes, including puma, noxa, p53AIP1, and fas, as

Fig. 3. SETD2 enhances the transcriptional activity of p53. (A) Activity of the pG13L and pMG15L reporter gene in p53 wild-type MCF7 or U2OS cells transfected with SETD2, SmyD2 or mock vector. Reporter activity was assayed as described in Materials and methods and represented as the mean ± S.D. of three separate experiments. (B) Activity of the pG13L reporter gene in p53-deficient H1299 cells transfected with p53 in the absence or presence of SETD2, SmyD2 or mock vector. (C) siRNA ablation of SETD2 reduces the transcriptional activity of p53. MCF7 cells were transfected as indicated. (D) Activity of the pG13L reporter gene in MCF7 cells transfected with SETD2 or mock vector after DNA damage. The cells were treated with MMS (0.05%) for 2 h, Etoposide (20 μM), or Doxorubicin (25 μM) for 12 h before lysed in passive lysis buffer and the p53 activity was assayed.

P. Xie et al. / Cellular Signalling 20 (2008) 1671–1678

well as angiogenesis-related gene Tsp1, and Huntington disease (HD) gene huntingtin (Fig. 4A). In contrast, the mRNA level of grow arrest-related genes gadd45, 14-3-3σ and cellular stress response gene pig3 had only a modest alteration when SETD2 was overexpressed, although p21 was also upregulated. Surprisingly, HDM2 mRNA level was significantly reduced by overexpressed SETD2 (Fig. 4A). Consistent with the results above, when SETD2 was knockdown, expression levels of puma, noxa, p53AIP1, fas, Tsp1, huntingtin and p21 were significantly reduced whereas those of gadd45, 14-3-3σ and pig3 were modestly altered, but HDM2 level was upregulated (Fig. 4B). These data strongly indicated that SETD2 selectively regulated certain subset of p53 target genes.

1675

Chromatin domains that are involved in establishment of heterochromatin boundaries and epigenetic gene regulation are defined by distinct sets of post-translationally modified histones [5,6]. Since SETD2 had been reported as a human H3K36-specific HMTase, this means SETD2 might play an important role in transcriptional regulation by itself. Then we examined in more detail SETD2 regulating p53 target genes in the absence of p53 in H1299 cells. In contrast to the case in p53 wild-type MCF7 cells, SETD2 overexpression had no significant effect on most of the p53 target genes examined except HDM2 (Fig. 4C), indicating that SETD2 regulated their expression via cooperation with p53. For HDM2, we proposed that SETD2

Fig. 4. SETD2 regulates expression of select p53 target genes. (A) MCF7 cells were transfected with Myc-SETD2 or mock vector. Total RNA was subjected to real-time RT-PCR analysis. Expression levels of puma, noxa, p53AIP1, fas, p21, 14-3-3σ, gadd45, pig3, Tsp1, huntingtin, and SETD2 RNAs were determined by the comparative threshold cycle method. Mean values and S.D. (error bar) are depicted. (B) MCF7 cells were transfected with SETD2 siRNA or control siRNA. Total RNA was subjected to real-time RT-PCR analysis. (C) H1299 cells were transfected with Myc-SETD2 or mock vector. Expression analysis of p53 target genes mRNA by quantitative real-time PCR. The results are given as fold increase of p53 target genes/GAPDH compared with the cells transfected with mock vector. Mean values and S.D. (error bar) are depicted. (D) SETD2 enhanced the transcriptional activity of p21 promoter. Activity of the p21-luc reporter gene in MCF7 cells transfected with SETD2 or mock vector, the right parallel transfected with SETD2 siRNA or control siRNA. Reporter activity was assayed as described in Materials and methods and represented as the mean ± S.D. of three separate experiments. (E) SETD2 repressed the transcriptional activity of HDM2 promoter. HDM2-luc reporter gene analysis in MCF7 cells transfected as in (D).

1676

P. Xie et al. / Cellular Signalling 20 (2008) 1671–1678

might regulate its transcription directly and in a p53-independent manner. To further confirm the regulatory role of SETD2 on p21 and HDM2 expression, we used the native promoter of p21 or HDM2 to drive the luciferase in reporter gene assay and tested the effect of SETD2. In line with the real-time PCR results, SETD2 overexpression enhanced (about 3–4 folds) and its knockdown by RNAi reduced the p21 promoter activity significantly (Fig. 4D). Converse effects were observed with the HDM2 promoter (Fig. 4E).

3.5. SETD2 downregulates expression of HDM2 and enhances the stability of p53 Our data above showed that SETD2 significantly suppressed the mRNA expression of HDM2 in a p53-independent manner. It has been well defined that HDM2 is a major negative regulator of p53 and mediates polyubiquitination and degradation of p53 in order to maintain p53 at low levels under normal conditions [35]. Therefore, we further explored the role and the possible mechanism of SETD2

Fig. 5. SETD2 downregulates expression of HDM2 and enhances the stability of p53. (A) SETD2 inhibits the protein level of endogenous HDM2. MCF7 cells were transiently transfected with plasmids expressing Myc-SETD2 (0, 400, 800, 1600 ng) or mock vector. Cell lysates were analyzed by western blotting. (B) SETD2 could not affect the protein level of exogenous HDM2. HEK293T cells were transiently transfected with plasmids expressing Myc-HDM2 and Myc-SETD2 or mock vector. (C) SETD2 overexpression downregulates expression of HDM2. H1299 cells were transfected with the plasmids as indicated. Cells were grown in medium containing MG132 (10 μM) for 12 h (lanes 5 to 8). Cell lysates were immunoblotted with anti-p53, HDM2, SETD2, GAPDH antibodies. (D) SETD2 negatively regulated HDM2 promoter. MCF7 cells were co-transfected with truncated HDM2 reporter constructs including P1, P2, P1 + P2 promoter, and Myc-SETD2 or mock vector. Luciferase activity was assayed as described. (E) SETD2 may affect the stability of p53. MCF7 cells were transfected with SETD2 siRNA or control siRNA. Cell lysates were immunoblotted with anti-p53, HDM2, SETD2, GAPDH antibodies. Total RNA was subjected to real-time RT-PCR analysis with p53-specific primers.

P. Xie et al. / Cellular Signalling 20 (2008) 1671–1678

on HDM2 regulation. Western blot analysis showed that protein levels of endogenous HDM2 in MCF7 cells were gradually decreased accompanied by the expression of increasing amount of SETD2 (Fig. 5A). Strikingly, the level of exogenous HDM2 was hardly affected by SETD2 (Fig. 5B), suggesting that SETD2 might target the regulatory element such as promoter of HDM2 gene and control its transcription rather than translation or post-translational stability. HDM2 is a well-known RING finger-type E3 ubiquitin ligase, which could reduce its stability by self-ubiquitination followed by proteasome-mediated degradation. To rule out the possibility SETD2 might impact HDM2 degradation, we tested the effect of SETD2 on HDM2 proteins in H1299 cells in which p53 is deficient and then endogenous HDM2 level is very low. We indeed observed low levels of HDM2 in normal H1299 cells (Fig. 5C, lanes 1–2) even after MG132 (a proteasome inhibitor) treatment due to lack of p53 (lanes 6–7), the major transcription factor of HDM2 gene. Ectopic p53 expression resulted in HDM2 transcription and then high levels of HDM2 proteins (lane 3). Further co-expression of SETD2 significantly reduced HDM2 protein level (lane 4), which effect could not be blocked by MG132 treatment (lane 8). This result suggested that SETD2 had significant effect on HDM2 synthesis but little effect on HDM2 degradation. HDM2 expression is controlled by two different promoters, leading to alternatively spliced transcripts that differ in their 5′-untranslated regions. Transcription from the first promoter P1 is independent of p53 and yields mRNA with exon 2 spliced out. Conversely, transcription from the second promoter P2 is p53-dependent and gives rise to a transcript lacking exon 1 but containing exon 2 [23,36]. To further study the cis-acting elements required for SETD2-regulated HDM2 transcription, a genomic DNA fragment consisting of both basal (P1) and p53-responsible (P2) promoter regions of the HDM2 gene was isolated and ligated to luciferase reporter gene (Fig. 5D). The reporter construct was then co-transfected with a fixed amount of SETD2 in MCF7 cells. As shown in Fig. 5D, SETD2 primarily inhibited the activity of HDM2 promoter P2 and then secondly P1. The ability of SETD2 to downregulate HDM2 expression promotes us to speculate the effect of SETD2 on p53 stability as HDM2 is the major E3 ligase of p53. Indeed, we detected an observable decline of p53 protein level together with an increase of HDM2 level by SETD2 knockdown (Fig. 5E, left). Interestingly, the level of p53 mRNA was hardly affected by SETD2 knockdown (Fig. 5E, right), which is consistent with the fact that SETD2 may affect the stability of p53 protein by regulating HDM2 mRNA and then protein levels. 4. Discussion SETD2/HYPB is one of the H3K36 methyltransferases in humans. Previous findings suggest that SETD2 may serve as a linker between histone H3K36 methylation and transcriptional regulation in mammals [7,8]. As far as we know, three SET domain-containing proteins, SET9, SmyD2 and SET8, have been sequentially shown to regulate p53 activity via direct methylation of p53 at its C-terminus [20,30,34]. In the present study, we for the first time established the interaction between SETD2 and p53. SETD2 could be co-immunoprecipitated with p53 under both exogenous and endogenous conditions (Fig. 1). Interestingly, deletion analysis showed that C-terminal region comprising aa 915–2061 of SETD2 and N-terminal TAD domain (1–45aa) of p53 were required for this interaction (Fig. 2). Different from those three HMTases, SETD2 is a huge protein of more than 230 kDa. In addition to the SET domain, low charged region and WW domain, whether it possesses other functional domains remains unclear and its physiological functions are still mysterious. TAD region of p53 is known to bind HDM2, the major ubiquitin ligase of p53. Phosphorylation of Ser15 within the TAD domain results in p53 dissociation with HDM2 and enhanced p53 stability. In our study, we also showed that SETD2 could inhibit HDM2 transcription and downregulate the expression level of HDM2 proteins, resulting in enhanced p53 stability (Fig. 5). Consistent with this notion, SETD2

1677

overexpression augmented expression of the majority of p53 target genes we examined, whereas SETD2 knockdown led to their downregulation (Fig. 4). Whether SETD2 binding to p53 TAD domain interferes p53-HDM2 interaction should be investigated in the future. Among the SET domain-containing proteins, SET9 enhanced transcriptional activity of p53, even in the presence of DNA damage [30], which is similar to the manner of SETD2. In contrast, SET8 and SmyD2 repressed transcriptional activity of p53, and SmyD2 antagonized with SET9 to control p53 activation [20,34]. Although SET9, SmyD2 and SET8 could directly methylate p53 at its C-terminus, other SET domain-containing proteins could not methylate p53 at least in vitro, including Suv39h1, Suv4-20h1 and hDOT1L [34]. These results suggest that not all of histone methyltransferases could methylate transcription factor such as p53. In this respect, whether SETD2 could methylate p53 or not needs further investigation. Nonetheless, we showed strong evidence that SETD2 could regulate p53 activity and selectively regulate its targets. We analyzed a total of eleven typical target genes of p53, whose products are critical for cell apoptosis, cell cycle arrest, senescence or angiogenesis. Among them, all of the apoptosis-related genes were shown upregulated by SETD2 and in a p53-dependent manner, suggesting that SETD2 might affect p53-mediated apoptosis specifically. Two of three examined cell cycle arrest-related genes, Gadd45 and 14-3-3σ were not affected significantly by SETD2, implying that SETD2 might have a weaker effect on cell cycle regulation compared with apoptosis. Only p21, the well-known CDK inhibitor in G1 phase, was also upregulated by SETD2. Interestingly, p21 was also shown to be involved in apoptosis regulation. For instance, SET9 could upregulate p21 expression and increase p53-mediated apoptosis [30]. We note that compared to SETD2, whether and how SET9, SmyD2 and SET8 regulate different sets of p53 target genes still remain unclear. p53 is a critical tumor suppressor that lies at the center of DNA damage response network. How p53 gains the ability to control more than 150 downstream genes precisely under diverse physiological conditions and in a variety of cell types is attractive. Recent evidence showed that histone methylation might play a role in the determination. Chromatin modifications are often termed ‘epigenetic’ marks. These modifications have been broadly classified into repressing and activating—in other words, they correlate with, and perhaps directly regulate, gene repression and induction [37,38]. The human cellular apoptosis susceptibility protein (hCAS/CSE1L) associates with chromatin within a subset of p53 target promoters, including PIG3, in a p53-autonomous manner. hCAS/CSE1L silencing leads to increased methylation of histone H3 lysine 27 (H3K27) within the PIG3 gene, decreases PIG3 transcription and reduces apoptosis [39]. Both SETD2 and SmyD2 can methylate H3K36 and regulate p53 activity. SETD2 is responsible for virtually all global and transcription dependent H3K36 trimethylation. In this regard, SETD2 could be a critical co-operator of p53 to determine specificity of target genes, since SETD2 silencing led to dramatic decrease the expression of subset genes which are important regulators of cell fate (Fig. 4B). Thus, we speculate that SETD2 together with SmyD2 might also link the histone methylation within certain p53 target genes to p53-mediated cellular responses, which need to be elucidated. Human SETD2 was also identified as a huntingtin interacting protein, implicated in the pathogenesis of Huntington disease (HD). HD is a devastating neurologic disorder that is characterized by abnormal expression of a CAG repeat in the first exon of the huntingtin (Htt) gene. The molecular pathogenesis of HD has not been fully elucidated, but recent studies indicated that SETD2 interaction with Htt protein might be implicated HD pathogenesis [40]. Most recently, p53 has been found to transactivate the human and murine Htt promoters, and there is a complex interaction exists between Htt and p53, in which p53 modifies Htt level, and downstream effectors of mutant Htt may feedback on p53. These results suggest that p53 also plays a role in the pathology of Huntington's disease, and that the normal function of p53 may actually promote the disease process [41,42]. We postulate that

1678

P. Xie et al. / Cellular Signalling 20 (2008) 1671–1678

the Huntington's disease pathogenesis might be correlated to perplexing interaction and regulation among p53, Htt protein and SETD2. Our work might open a perspective for the further study of the pathogenesis of HD. Interestingly, we also found that SETD2 could regulate HDM2 expression in the p53-independent manner, as in both p53 wild-type MCF7 and p53-deficient H1299 cells, SETD2 overexpression and silencing had significant effect on HDM2 gene expression (Fig. 4). Further analysis revealed that SETD2 regulated HDM2 promoter primarily the P2 and then secondly the P1 promoter (Fig. 5), although P2 has been shown to contain p53 binding element and HDM2 can be activated by p53 dependent of P2 promoter [36]. Our results suggest that HDM2 regulation might be more complex than once thought. In addition to p53 binding to P2 directly, SETD2 might coordinate histone methylation (H3K36) with HDM2-P2 regulation. Furthermore, it is not surprising that SETD2 possesses p53-independent functions due to the fact that p53 originates from worm but SETD2 from yeast. Its yeast ortholog SET2 also displays the HMTase activity [43]. Considering this high conservation, it is not strange that SETD2 might be a functionally essential protein. Preparation of SETD2 gene knockout mice might contribute to full understandings of its physiological role. Acknowledgements We thank Drs. Zhu Chen, Qiuhua Huang and Xiao-Jian Sun (Shanghai First People's Hospital Affiliated to Shanghai Jiao Tong University) for SETD2 plasmids; Dr. Shengcai Lin (Department of Biology in Xiamen University) for p53 truncates; Drs. Moshe Oren (The Weizmann Institute of Science) and Bert Vogelstein (Johns Hopkins Kimmel Comprehensive Cancer Center) for reporter gene plasmids; Drs. Shelley L. Berger and Jing Huang (The Wistar Institute) for SmyD2 expression constructs; Dr. Juntao Yang for help in designing the SETD2 siRNA sequence. This work was partially supported by Chinese National Natural Science Foundation Projects (30770518 and 30621063), the National Basic Research Programs of China (2007CB914601 and 2006CB910802) and Beijing Science and Technology NOVA Program (2007A063). References [1] [2] [3] [4] [5] [6]

V.A. Spencer, J.R. Davie, Gene 240 (1999) 1. B.D. Strahl, C.D. Allis, Nature 403 (2000) 41. T. Jenuwein, C.D. Allis, Science 293 (2001) 1074. B.M. Turner, BioEssays 22 (2000) 836. A.P. Wolffe, BioEssays 37 (2001) 45. Y. Zhang, D. Reinberg, Genes Dev. 15 (2001) 2343.

[7] X. Sun, J. Wei, X. Wu, M. Hu, L. Wang, H. Wang, Q. Zhang, S. Chen, Q. Huang, Z. Chen, J. Biol. Chem. 280 (2005) 35261. [8] J.W. Edmunds, L.C. Mahadevan, A.L. Clayton, EMBO J. 27 (2008) 406. [9] D.C. Schultz, K. Ayyanathan, D. Negorev, G.G. Maul, F.J. Rauscher, Genes Dev. 16 (2002) 919. [10] L. Zeleznik, A.M. Harden, J.D. Rowley, Proc. Natl. Acad. Sci. U. S. A. 91 (1994) 10610. [11] N. Huang, B.E. Vom, J.M. Garnier, T. Lerouge, J.L. Vonesch, Y. Lutz, P. Chambon, R. Losson, EMBO J. 17 (1998) 3398. [12] P.M. Ayton, E.H. Chen, M.L. Cleary, Mol. Cell. Biol. 24 (2004) 10470. [13] W.A. Krajewski, T. Nakamura, A. Mazo, E. Canaani. Mol. Cell. Biol. 25 (2005) 1891. [14] A.J. Levine, Cell 88 (1997) 323. [15] C. Prives, P.A. Hall, J. Pathol. 187 (1999) 112. [16] S. Bates, K.H. Vousden, Cell. Mol. Life Sci. 55 (1999) 28. [17] B. Vogelstein, D. Lane, A.J. Levine, Nature 408 (2000) 307. [18] O. Laptenko, C. Prives, Cell Death Differ. 13 (2006) 951. [19] L. Li, B. Deng, G. Xing, Y. Teng, C. Tian, X. Cheng, X. Yin, J. Yang, X. Gao, Y. Zhu, Q. Sun, L. Zhang, X. Yang, F. He, Proc. Natl. Acad. Sci. U. S. A. 104 (2007) 7951. [20] J. Huang, L.P. Burgos, B.J. Placek, R. Sengupta, M. Richter, J.A. Dorsey, S. Kubicek, S. Opravil, T. Jenuwein, S.L. Berger, Nature 444 (2006) 629. [21] Y. Rui, Z. Xu, S. Lin, Q. Li, H. Rui, W. Luo, H. Zhou, P. Cheung, Z. Wu, Z. Ye, P. Li, J. Han, S. Lin, EMBO J. 23 (2004) 4583. [22] J. Yu, Z. Wang, K.W. Kinzler, B. Vogelstein, L. Zhang, Proc. Natl. Acad. Sci. U. S. A. 100 (2003) 1931. [23] A. Zauberman, D. Flusberg, Y. Haupt, Y. Barak, M. Oren, Nucleic Acids Res. 23 (1995) 14. [24] L. Zhang, Y. Tie, C. Tian, G. Xing, Y. Song, Y. Zhu, Z. Sun, F. He, Cell. Signal. 18 (2006) 1386. [25] J. Yan, J. Jiang, C.A. Lim, Q. Wu, H.H. Ng, K.C. Chin, Proc. Natl. Acad. Sci. U. S. A. 104 (2007) 1841. [26] N.G. Iyer, S.F. Chin, H. Ozdag, Y. Daigo, D. Hu, M. Cariati, K. Brindle, S. Aparicio, C. Caldas, Proc. Natl. Acad. Sci. U. S. A. 101 (2004) 7386. [27] S. Ohtani, S. Kagawa, Y. Tango, T. Umeoka, N. Tokunaga, Y. Tsunemitsu, J.A. Roth, Y. Taya, N. Tanaka, T. Fujiwara, Mol. Cancer Ther. 14 (2000) 981. [28] J.G. Jackson, O.M. Pereira-Smith, Cancer Res. 66 (2006) 8356. [29] Y. Liu, H. Liu, B. Han, J. Zhang, Cancer Res. 66 (2006) 3248. [30] S. Chuikov, J.K. Kurash, J.R. Wilson, B. Xiao, N. Justin, G.S. Ivanov, K. McKinney, P. Tempst, C. Prives, S.J. Gamblin, N.A. Barlev, D. Reinberg, Nature 432 (2004) 353. [31] K. Simmons, P. Nde, Y. Kleshchenko, M. Lima, F. Villalta, FEBS Lett. 580 (2006) 2365. [32] J. Chai, Y. He, J. Chongqing Med. 34 (2005) 866. [33] Z. Chen, B.K. Kren, P. Wong, W.C. Low, C.J. Steer, Biochem. Biophy. Res. Commun. 329 (2005) 646. [34] X. Shi, L. Kachirskaia, H. Yamaguchi, L.E. West, H. Wen, E.W. Wang, S. Dutta, E. Appella, O. Gozan, Mol. Cell 27 (2007) 636. [35] C.L. Brooks, W. Gu, Mol. Cell 21 (2006) 307. [36] C. Chang, J.F. Danie, H. Wu, J. Biol. Chem. 279 (2004) 29841. [37] C.R. Vakoc, M.M. Sachdeva, H. Wang, G.A. Blobel, Mol. Cell. Biol. 26 (2006) 9185. [38] W.G. Müller, D. Rieder, T.S. Karpova, S. John, Z. Trajanoski, J.G. McNally, J. Cell. Biol. 177 (2007) 957. [39] T. Tanaka, S. Ohkubo, I. Tatsuno, C. Prives, Cell 130 (2007) 638. [40] R. Susanne, S. Thorsten, I.C. Dae, P. Barbara, E. Helmut, B. Walter, M. Guido, M.P. Brigitte, Mol. Cell. Neuroscience. 18 (2001) 68. [41] B.I. Bae, H. Xu, S. Igarashi, M. Fujimuro, N. Agrawal, Y. Taya, S.D. Hayward, T.H. Moran, C. Montell, C.A. Ross, S.H. Snyder, A. Sawa, Neuron 47 (2005) 29. [42] A.B. Ryan, S.O. Zeitlin, H. Scrable, Neurobiol. Dis. 24 (2006) 419. [43] B.D. Strahl, P.A. Grant, S.D. Briggs, Z. Sun, J.R. Bone, J.A. Caldwell, S. Mollah, R.G. Cook, J. Shabanowitz, D.F. Hunt, C.D. Allis, Mol. Cell. Biol. 22 (2002) 1298.