- Email: [email protected]
Interaction of E2F–Rb family members with corepressors binding to the adjacent E2F site
Available online at www.sciencedirect.com
Biochemical and Biophysical Research Communications 364 (2007) 1050–1055 www.elsevier.com/locate/ybbrc
Interaction of E2F–Rb family members with corepressors binding to the adjacent E2F site Yusuke Nakajima a, Shumpei Yamada b, Nobuyuki Kamata c, Masa-Aki Ikeda b
b,*
a Section of Oral and Maxillofacial Surgery, Graduate School, Tokyo Medical and Dental University, Tokyo 113-8549, Japan Section of Molecular Embryology, Graduate School, Tokyo Medical and Dental University, 1-5-45 Yushima, Bunkyo-ku, Tokyo 113-8549, Japan c Department of Oral and Maxillofacial Surgery, Graduate School of Biomedical Sciences, Hiroshima University, Hiroshima 734-8553, Japan
Received 11 October 2007 Available online 30 October 2007
Abstract Cell cycle-dependent transcriptional repression of the E2F1 and B-myb promoters is mediated through E2F-binding sites and adjacent corepressor site (cell cycle gene homology region (CHR)/downstream repression site (DRS)). Here, we show that a factor binding to the B-myb CHR is co-purified with E2F DNA-binding activity, and coimmunoprecipitated with components of E2F/Rb-family repressor complexes, E2F4 and retinoblastoma (Rb) family proteins. In spite of structural and functional similarities, however, the E2F1 and B-myb CHRs exhibited distinct factor-binding specificities. Furthermore, substitution of E2F1 CHR with the B-myb CHR in the E2F1 promoter revealed that the B-myb CHR was unable to repress the E2F1 promoter completely in the G0 phase. These results suggest that transcriptional repression of the E2F1 and B-myb promoters is mediated by physical interaction of E2F/Rb-family repressor complexes with promoter-specific corepressors. 2007 Elsevier Inc. All rights reserved. Keywords: E2F1; B-myb; Cell cycle gene homology region (CHR); Downstream repression site (DRS); Rb; Corepressor; Transcription; Repression
The E2F family of transcription factors plays a critical role in the progression of the cell cycle by regulating transcription in response to mitogenic stimulation [1,2]. The mammalian E2F family consists of nine distinct gene products that are divided into several subgroups based on structural and functional properties. In particular, E2F1, E2F2, and E2F3a act as activators of transcription, whereas E2F4 and E2F5 function primarily as transcriptional repressors in concert with members of the Rb family of tumor suppressor proteins (Rb, p107, and p130). Binding of the Rb family proteins converts E2F4 and E2F5 to active repressors of transcription. In response to mitogenic stimulation, the Rb family proteins are phosphorylated by the cyclin dependent kinases, leading to the dissociation of the E2F/Rb family repressor complexes. Previous chromatin immunoprecipitation (ChIP) studies have revealed that an
*
Corresponding author. Fax: +81 3 5803 0213. E-mail address: [email protected] (M.-A. Ikeda).
0006-291X/$ - see front matter 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2007.10.130
E2F4/p130 complex predominantly occupies the E2Fresponsive promoters in the G0 phase. Following cell cycle re-entry, the repressor complex is replaced by E2F1, E2F2 and E2F3 activators, concomitant with the induction of E2F-target gene transcription [3,4]. The E2F1 and B-myb promoters are repressed in G0 and activated at the G1/S transition in an E2F-dependent manner. It has been shown that the E2F-binding sites on these promoters are required but not sufficient for the repression in G0 [5–9]. In addition to the E2F site, a CHR/DRS corepressor site, which is located adjacent to the E2F site, is required for repression [10]. The CHR was also found in the cyclin A, cdc2, cdc25C, cyclin B2, and Aurora B promoters, in which the CHR is located adjacent to a ‘cell-cycle-dependent element’ (CDE), and the CDE-CHR element contributes to the S/G2 phase-specific expression of these genes [11–14]. The E2F1 promoter has two E2F sites and a CHR which is located downstream of the distal E2F side (E2F site A). The relative spacing between the CHR and E2F site A is identical to that between the
Y. Nakajima et al. / Biochemical and Biophysical Research Communications 364 (2007) 1050–1055
CHR/DRS and E2F site in the B-myb promoter (Fig. 1). Recently, we have shown that both the CHR and E2F site A sites are required to suppress the promoter in G0, whereas the proximal E2F site (E2F site B) contributes to transcriptional activation at the G1/S boundary [9]. In agreement with this, ChIP assays have revealed that E2F4 and p130 proteins, components of the repressor complex, binds to the E2F site A, whereas E2F1 and E2F3 activators preferentially bind to the E2F site B. Furthermore, the recruitment of E2F4 and p130 to the distal E2F site A was impaired by the mutation of the CHR site [9], which is consistent with previous observations of the B-myb promoter [8]. These observations support the notion that the cooperation between the E2F and CHR sites is requited for recruiting the E2F4/p130 repressor complex to the E2F1 and B-myb promoters. Although a factor binding to the B-myb CHR has been detected in partially purified nuclear extracts [7], the nature of the protein(s) interacting with the CHRs remains largely unknown. Here, we show that the B-myb CHR binding activity is co-purified with E2F DNA-binding activity, and coimmunoprecipitated with E2F4 and Rb family proteins. Furthermore, we provide evidence that a factor binding to the E2F CHR is different from that of the B-myb CHR, suggesting that transcriptional repression of the E2F1 and B-myb promoters is mediated by interaction of E2F and Rb family members with promoter-specific corepressors. Materials and methods Cell culture. REF52 cells were cultured at 37 C in Dulbecco’s modified Eagle’s minimum essential medium (DMEM; Sigma–Aldrich) supplemented with 10% fetal calf serum (FCS). HL60 cells were maintained in RPMI 1640 medium (Nissui) supplemented with 20% FCS. Plasmids. The human E2F1 promoter reporter constructs, E2F1-Luc (242, wt), has been described in [5]. E2F1-Luc B-mybCHR were generated using the QuickChange II Site-Directed Mutagenesis Kit (Stratagene) with the oligonucleotides 5 0 -TCTTTCGCGGCAAAAAGGAAATGGC GCGTAAAAGTGGCCG-3 0 and 5 0 -CGGCCACTTTTACGCGCCATT TCCTTTTTGCCGCGAAAGA-3 0 .
1051
Transfection and luciferase assays. REF52 cells were transfected using FuGENE 6 Transfection Reagent (Roche Diagnostics) as described in [9]. A CMV b-gal vector was cotransfected as an internal control. The transfected cells were brought to quiescence in DMEM with 0.1% FCS for 48 h before harvest. Luciferase and b-galactosidase activities were measured using a luciferase reporter gene assay system (Toyo Ink) and the Aurora GAL-XE reporter gene assay system (ICN), respectively. DNA affinity purification. Nuclear extracts were prepared from growing HL60 cells as described in [15]. To generate E2F affinity beads, four tandem repeated fragments, each of which contains either wild-type (wt) or mutant versions of two E2F sites derived from adenovirus E2 promoter (gifts from K. Ohtani), were immobilized on magnetic beads (Dynabeads M-280; Dynal). Precleared HL60 nuclear extracts (100 lg) were diluted in shift buffer (20 mM Hepes (pH 7.9), 40 mM KCl, 6 mM MgCl2, 1 mM EGTA, 0.1% Nonidet P-40, 0.2 mM DTT, 10% glycerol) buffer, containing 1.2 mg/ml bovine serum albumin, 4% ficoll, 2 mg/ml sonicated salmon sperm DNA, 1% casein,, and 1 m PMSF, and then incubated with either wt or mutant versions of the E2F affinity beads for 1 h at room temperature with gentle rotation. The beads were washed with ice-cold shift buffer containing 100 mM KCl, and the bound complexes were eluted with shift buffer containing 500 mM NaCl and 1% casein. The supernatant was dialyzed with ice-cold shift buffer. Immunoprecipitation. Immunoprecipitation was performed as described in [15]. Briefly precleared HL60 nuclear extracts (100 lg) were incubated with antibody-conjugated Dynabeads for 4 h at 4 C with gentle rotation. After washing, the immunoprecipitates were dissociated by treating with 0.8% sodium deoxycholate (DOC) (Sigma–Aldrich) on ice, and then the released proteins were neutralized with Nonidet P-40 (final concentration of 1%). The following antibodies were used in this study: mouse monoclonal Rb (C-36, Santa Cruz Biotechnology) and p107 (a mixture of SD-2, SD-4, SD-6, SD-9, and SD-15; gifts from N. Dyson) antibodies, and rabbit polyclonal p130 (C-18) and E2F4 (C-108) antibodies (Santa Cruz Biotechnology). Gel mobility shift assays (EMSAs). EMSAs were performed as described in [15]. Sonicated salmon sperm DNA (500 ng) and Poly (dI–dC) (1 lg) were added as nonspecific competitors to analyze the E2F and CHR binding activities, respectively. The E2F (DHRF) probe was described previously [15]. The following double-stranded oligonucleotides were used as probes and/or unlabeled competitors: E2F wt and E2F mut; E2F1 wt and E2F1 mE2F (Mut E2F site A and B) [9]; E2F1 mCHR (5 0 -CGTGGCTCTTT CGCGGCAAAAAGGCCTTGGCGCGTAAAAGT-3 0 ); E2F1 E2F1 Bmyb-CHR (5 0 -CGTGGCTCTTTCGCGGCAAAAAGGAAATGGCGC GTAAAAGT-3 0 ); B-myb wt (5 0 -TTGCCGACGCACTTGGCGGGAGA TAGGAAAGTGGTTCTGAA-3 0 ); B-myb mE2F (5 0 -TTGCCGACGC ACYYGTATGGAGATAGGAAAGTGGTTCTGAA-3 0 ); B-myb mCHR (5 0 -TTGCCGACGCACTTGGCGGGAGATAGGCCTGTGGTTCTGA A-3 0 ); B-myb mE2F-mCHR (5 0 -TTGCCGA CGCACTTGTATGGAGATAGGCCTGTGGTTCTGAA-3 0 ).
Fig. 1. Alignment of human E2F1 and mouse B-myb promoters. Specific changes in these reporter constructs or the oligonucleotides used in this study are indicated below the wild-type sequences. E2F-binding sites are indicated by arrows and CHRs are boxed.
1052
Y. Nakajima et al. / Biochemical and Biophysical Research Communications 364 (2007) 1050–1055
Results and discussion Co-purification of B-myb CHR binding activities with E2F DNA-binding activities To examine the physical interaction of the CHR binding factor with E2F4 or the Rb family member proteins, we purified E2F activities from nuclear extracts prepared from growing HL60 cells by using E2F DNA affinity beads that were conjugated with either wt or mutant E2F-binding sequences derived from the adenovirus E2 promoter, which has no recognizable CHR. EMSAs with a probe containing a E2F site confirmed that samples eluted from the wt, but not the mutant E2F affinity beads contained free E2F, which was abolished by unlabeled wt, but not mutant, E2F competitors (Fig. 2A, lanes 2 and 3). The wt E2Faffinity purified samples were then analyzed with three different probes containing wt and mutant B-myb promoter sequences that were mutated in either the E2F site (mE2F) or CHR (mCHR) as described previously [7] (Fig. 1). Fig. 2B shows that E2F specific activity was detected with wt and mCHR probes (lanes 1 and 3). Moreover, we detected bands, although faint, with the mE2F probe containing the intact CHR (lane 2, arrows). Fig. 2C shows that the bands are successfully abolished by competitors containing the CHR (wt and mE2F) (lanes 2 and 3), but not by competitors mutated in CHR (mCHR) (lane 4). Furthermore, that the bands detected with the Bmyb mE2F probe were competed by wt, but not mCHR,
both of which have no mutated E2F site (Fig. 2C, 3B, and 4B. see below), together with the fact that the double mutant (B-myb mE2F-mCHR), although it has the mutated E2F site, failed to compete the bands (Fig. 4B and see below), excludes the possibility that the mutation of the E2F site might create a binding site irrelevant to the CHR-binding factor. These results indicate that the CHR-specific DNA binding activity is co-purified with the E2F DNA-binding activity. Interaction of the B-myb CHR binding activities with components of the E2F/Rb family member complexes To examine the interaction of the CHR binding factor with E2F4 and the Rb family proteins, HL60 nuclear extracts were immunoprecipitated with either E2F4 or Rb antibody, and then the precipitated complexes were dissociated by treating with DOC as described previously [15]. EMSAs with the E2F probe confirmed that free E2F was released from the E2F4 and Rb immunoprecipitates (Fig. 3A, and data not shown). Competition experiments with the B-myb mE2F probe revealed that DNA binding activities dependent on the CHR, but not E2F site, were detected in Rb immunoprecipitates (Fig. 3B, lane 2–4, and see below). Furthermore, the CHR-specific bands, which were confirmed by competition assays (data not shown), were also found in association with E2F4, p107, and p130 immunoprecipitates (Fig. 3C). These results indicate that the binding activities specific to the B-myb CHR
Fig. 2. CHR specific DNA-binding activities were co-purified with E2F-specific DNA affinity beads. (A) Detection of E2F DNA-binding activity. The samples purified with the wild-type (wt, lanes 1–3) and mutated version (mut, lane 4) of E2F affinity beads were subjected to EMSA with the 32P-endlabled E2F (DHRF) probe in the presence or absence of non-labeled competitor DNA (E2F wt or E2F mut) at 100-fold molar excess. (B,C) Detection of CHR specific DNA-binding activities. Samples purified by wt E2F affinity beads were subjected to EMSA with the indicated B-myb probes (Fig. 1). Arrows and an asterisk indicate CHR (B, lane 2; C, lane 1 and 4) and non-specific bands, respectively. EMSA was performed in the presence or absence of the indicated cold competitors (Fig. 1) at 100-fold molar excess (C).
Y. Nakajima et al. / Biochemical and Biophysical Research Communications 364 (2007) 1050–1055
1053
Fig. 3. Detection of CHR specific DNA-binding activities in immunoprecipitates with E2F4 and the Rb family member antibodies. (A) The supernatant of DOC-treated immunoprecipitates with antibodies specific to E2F4 or Rb was assayed for E2F DNA binding as in Fig. 2A. NE: HL60 nuclear extracts. (B,C) DOC-treated immunoprecipitates with antibodies specific to E2F4 or the indicated Rb family members were assayed for CHR DNA binding with the B-myb mE2F probe as in Fig. 2C in the presence or absence of the indicated cold competitors (Fig. 1) at 100-fold molar excess. The arrows and asterisk indicate CHR (B, lane 4 and C lanes 2–5) and non-specific bands, respectively. IgG refers to immunoprecipitates with a control rabbit IgG, as a negative control.
are associated with components of the E2F/Rb family member complexes. The E2F1 and B-myb CHRs exhibits distinct factor-binding specificities Structural and functional similarities of the E2F1 and Bmyb CHRs suggest that the same or similar factors bind to these CHRs. To examine the DNA-binding specificity of the CHR-binding factors, we performed competition assays using a probe that contains the E2F1 CHR (E2F1 mE2F). Fig. 4A shows that a shifted band was specifically abolished by competitors containing the E2F1 CHR (E2F1 wt and mutE2F) (lanes 3–5), but not by CHR-mutated competitors (E2F1 mCHR) (lane 4). Unexpectedly, however, the band was not affected by competitors containing the B-myb CHR (B-myb mE2F), even at a higher concentration (500-fold molar excess) (lanes 7 and 8). Conversely, two bands which specifically bound to the B-myb CHR probe (B-myb mE2F) were not abolished by E2F1 CHR competitors (E2F1 mE2F) (Fig. 4B, lane 8). These results suggest that the factor binding to the E2F CHR is different from that binding to the B-myb CHR. The E2F1 and B-myb CHRs are not equivalent in repressing the E2F1 promoter To examine whether the B-myb CHR can act as a corepressor site in the E2F1 promoter, we made use of an E2F1 reporter construct in which its CHR was substituted with the B-myb CHR sequences (E2F1 B-myb-CHR). EMSAs using a B-myb mE2F probe confirmed that the shifted
bands were abolished by the E2F1 B-myb-CHR competitors (Fig. 4C). Luciferase assays using REF52 cells revealed that the E2F1 CHR substitution with the B-myb CHR led to a sevenfold reduction of the E2F1 promoter activity relative to the CHR-mutated E2F1 promoter under serumstarved conditions (Fig. 4D). However, we reproducibly detected twofold higher activity in the chimeric promoter relative to the wt E2F1 promoter, indicating that the Bmyb CHR is functionally similar but not equivalent to the E2F1 CHR in terms of the repression of the E2F1 promoter. In this study, biochemical evidence is presented for the physical interactions of the putative CHR binding factors with the components of the E2F/Rb-family repressor complexes. This provides a basis for the cooperation between the CHR and E2F sites in repression of the E2F1 and Bmyb promoters. It is likely that the interaction of the E2F/Rb-family members with the CHR binding factor stabilizes binding of the repressor complexes to these promoters, thereby playing a critical role for recruiting the repressor complexes in vivo. Furthermore, we show that, in spite of structural and functional similarities, the factor binding to the E2F CHR is different from that binding to the B-myb CHR, and that the B-myb CHR is functionally similar but not equivalent to the E2F1 CHR in the repression of the E2F1 promoter. In B-myb promoter, substitution of A to T at the 5th position of the CHR, which is identical to that of the E2F1 CHR, results in the derepression of the B-myb promoter in G0 [8]. Therefore, these observations indicate that promoter-specific CHRs are required to repress the E2F1 and B-myb promoters. The promoter-specific function of the CHRs is also supported
1054
Y. Nakajima et al. / Biochemical and Biophysical Research Communications 364 (2007) 1050–1055
Fig. 4. The E2F1 and B-myb CHRs were functionally similar but not equivalent. (A, B, and C) Distinct DNA binding specificities of factors binding to the E2F1 and B-myb CHRs. The supernatant of DOC-treated immunoprecipitates with the Rb antibody were assayed for CHR DNA binding as in Fig. 2C with the E2F1 mE2F (A) and B-myb mE2F (B, C) probes in the presence or absence of cold competitors (Fig. 1) at the indicated molar excess. The arrows indicate CHR-specific bands. The probe only refers to the absence of the supernatant (C). (D) Corepressor activity of the B-myb CHR in the E2F1 promoter. REF52 cells were transfected with either E2F1-Luc wt [5], mCHR [9], or B-myb-CHR, along with pCMV b-gal as an internal control. Cells were induced into quiescence by serum starvation, and then harvested 48 h after transfection. Luciferase and b-galactosidase activities were measured as described previously [9]. Luciferase activity was normalized to b-galactosidase activity. Data represent average of three independent experiments, each performed in duplicate.
by the fact that substitution of the CHR in the Cdc25C promoter with the B-myb CHR causes derepression of the promoter activity in G0 [7]. Based on the study presented here, and in combination with other findings, we propose that transcriptional repression of the E2F1 and B-myb promoters would result from physical interaction of E2F/Rb-family repressor complexes with promoter-specific corepressors, and that such promoter-specific interaction is required for proper promoter functions of E2F1 and B-myb genes. Whether the CHR binding factors interact with the repressor complexes through E2F4 or Rb family members, and whether such interactions change during
the cell cycle remain unclear at present. Identification of the CHR binding factors will be required to address these issues.
Acknowledgments We thank K. Ohtani, and N. Dyson for materials, and K. Ohtani for critical reading of the manuscript. This study was supported in part by Grants-in-Aid for Scientific Research from the Japan Society for the Promotion of Science.
Y. Nakajima et al. / Biochemical and Biophysical Research Communications 364 (2007) 1050–1055
References [1] C. Attwooll, E. Lazzerini Denchi, K. Helin, The E2F family: specific functions and overlapping interests, EMBO J. 23 (2004) 4709–4716. [2] D.K. Dimova, N.J. Dyson, The E2F transcriptional network: old acquaintances with new faces, Oncogene 24 (2005) 2810–2826. [3] Y. Takahashi, J.B. Rayman, B.D. Dynlacht, Analysis of promoter binding by the E2F and pRB families in vivo: distinct E2F proteins mediate activation and repression, Genes Dev. 14 (2000) 804–816. [4] J. Wells, K.E. Boyd, C.J. Fry, S.M. Bartley, P.J. Farnham, Target gene specificity of E2F and pocket protein family members in living cells, Mol. Cell Biol. 20 (2000) 5797–5807. [5] D.G. Johnson, K. Ohtani, J.R. Nevins, Autoregulatory control of E2F1 expression in response to positive and negative regulators of cell cycle progression, Genes Dev. 8 (1994) 1514–1525. [6] J.D. Bennett, P.G. Farlie, R.J. Watson, E2F binding is required but not sufficient for repression of B-myb transcription in quiescent fibroblasts, Oncogene 13 (1996) 1073–1082. [7] N. Liu, F.C. Lucibello, J. Zwicker, K. Engeland, R. Muller, Cell cycle-regulated repression of B-myb transcription: cooperation of an E2F site with a contiguous corepressor element, Nucleic Acids Res. 24 (1996) 2905–2910. [8] S. Catchpole, F. Tavner, L. Le Cam, C. Sardet, R.J. Watson, A Bmyb promoter corepressor site facilitates in vivo occupation of the adjacent E2F site by p107 · E2F and p130 · E2F complexes, J. Biol. Chem. 277 (2002) 39015–39024.
1055
[9] K. Araki, Y. Nakajima, K. Eto, M.A. Ikeda, Distinct recruitment of E2F family members to specific E2F-binding sites mediates activation and repression of the E2F1 promoter, Oncogene 22 (2003) 7632–7641. [10] J. Zwicker, R. Muller, Cell-cycle regulation of gene expression by transcriptional repression, Trends Genet. 13 (1997) 3–6. [11] J. Zwicker, F.C. Lucibello, L.A. Wolfraim, C. Gross, M. Truss, K. Engeland, R. Muller, Cell cycle regulation of the cyclin A, cdc25C and cdc2 genes is based on a common mechanism of transcriptional repression, EMBO J. 14 (1995) 4514–4522. [12] N. Liu, F.C. Lucibello, K. Korner, L.A. Wolfraim, J. Zwicker, R. Muller, CDF-1, a novel E2F-unrelated factor, interacts with cell cycle-regulated repressor elements in multiple promoters, Nucleic Acids Res. 25 (1997) 4915–4920. [13] M. Kimura, C. Uchida, Y. Takano, M. Kitagawa, Y. Okano, Cell cycle-dependent regulation of the human aurora B promoter, Biochem. Biophys. Res. Commun. 316 (2004) 930–936. [14] F. Bolognese, M. Wasner, C.L. Dohna, A. Gurtner, A. Ronchi, H. Muller, I. Manni, J. Mossner, G. Piaggio, R. Mantovani, K. Engeland, The cyclin B2 promoter depends on NF-Y, a trimer whose CCAAT-binding activity is cell-cycle regulated, Oncogene 18 (1999) 1845–1853. [15] M.A. Ikeda, L. Jakoi, J.R. Nevins, A unique role for the Rb protein in controlling E2F accumulation during cell growth and differentiation, Proc. Natl. Acad. Sci. USA 93 (1996) 3215– 3220.
Biochemical and Biophysical Research Communications 364 (2007) 1050–1055 www.elsevier.com/locate/ybbrc
Interaction of E2F–Rb family members with corepressors binding to the adjacent E2F site Yusuke Nakajima a, Shumpei Yamada b, Nobuyuki Kamata c, Masa-Aki Ikeda b
b,*
a Section of Oral and Maxillofacial Surgery, Graduate School, Tokyo Medical and Dental University, Tokyo 113-8549, Japan Section of Molecular Embryology, Graduate School, Tokyo Medical and Dental University, 1-5-45 Yushima, Bunkyo-ku, Tokyo 113-8549, Japan c Department of Oral and Maxillofacial Surgery, Graduate School of Biomedical Sciences, Hiroshima University, Hiroshima 734-8553, Japan
Received 11 October 2007 Available online 30 October 2007
Abstract Cell cycle-dependent transcriptional repression of the E2F1 and B-myb promoters is mediated through E2F-binding sites and adjacent corepressor site (cell cycle gene homology region (CHR)/downstream repression site (DRS)). Here, we show that a factor binding to the B-myb CHR is co-purified with E2F DNA-binding activity, and coimmunoprecipitated with components of E2F/Rb-family repressor complexes, E2F4 and retinoblastoma (Rb) family proteins. In spite of structural and functional similarities, however, the E2F1 and B-myb CHRs exhibited distinct factor-binding specificities. Furthermore, substitution of E2F1 CHR with the B-myb CHR in the E2F1 promoter revealed that the B-myb CHR was unable to repress the E2F1 promoter completely in the G0 phase. These results suggest that transcriptional repression of the E2F1 and B-myb promoters is mediated by physical interaction of E2F/Rb-family repressor complexes with promoter-specific corepressors. 2007 Elsevier Inc. All rights reserved. Keywords: E2F1; B-myb; Cell cycle gene homology region (CHR); Downstream repression site (DRS); Rb; Corepressor; Transcription; Repression
The E2F family of transcription factors plays a critical role in the progression of the cell cycle by regulating transcription in response to mitogenic stimulation [1,2]. The mammalian E2F family consists of nine distinct gene products that are divided into several subgroups based on structural and functional properties. In particular, E2F1, E2F2, and E2F3a act as activators of transcription, whereas E2F4 and E2F5 function primarily as transcriptional repressors in concert with members of the Rb family of tumor suppressor proteins (Rb, p107, and p130). Binding of the Rb family proteins converts E2F4 and E2F5 to active repressors of transcription. In response to mitogenic stimulation, the Rb family proteins are phosphorylated by the cyclin dependent kinases, leading to the dissociation of the E2F/Rb family repressor complexes. Previous chromatin immunoprecipitation (ChIP) studies have revealed that an
*
Corresponding author. Fax: +81 3 5803 0213. E-mail address: [email protected] (M.-A. Ikeda).
0006-291X/$ - see front matter 2007 Elsevier Inc. All rights reserved. doi:10.1016/j.bbrc.2007.10.130
E2F4/p130 complex predominantly occupies the E2Fresponsive promoters in the G0 phase. Following cell cycle re-entry, the repressor complex is replaced by E2F1, E2F2 and E2F3 activators, concomitant with the induction of E2F-target gene transcription [3,4]. The E2F1 and B-myb promoters are repressed in G0 and activated at the G1/S transition in an E2F-dependent manner. It has been shown that the E2F-binding sites on these promoters are required but not sufficient for the repression in G0 [5–9]. In addition to the E2F site, a CHR/DRS corepressor site, which is located adjacent to the E2F site, is required for repression [10]. The CHR was also found in the cyclin A, cdc2, cdc25C, cyclin B2, and Aurora B promoters, in which the CHR is located adjacent to a ‘cell-cycle-dependent element’ (CDE), and the CDE-CHR element contributes to the S/G2 phase-specific expression of these genes [11–14]. The E2F1 promoter has two E2F sites and a CHR which is located downstream of the distal E2F side (E2F site A). The relative spacing between the CHR and E2F site A is identical to that between the
Y. Nakajima et al. / Biochemical and Biophysical Research Communications 364 (2007) 1050–1055
CHR/DRS and E2F site in the B-myb promoter (Fig. 1). Recently, we have shown that both the CHR and E2F site A sites are required to suppress the promoter in G0, whereas the proximal E2F site (E2F site B) contributes to transcriptional activation at the G1/S boundary [9]. In agreement with this, ChIP assays have revealed that E2F4 and p130 proteins, components of the repressor complex, binds to the E2F site A, whereas E2F1 and E2F3 activators preferentially bind to the E2F site B. Furthermore, the recruitment of E2F4 and p130 to the distal E2F site A was impaired by the mutation of the CHR site [9], which is consistent with previous observations of the B-myb promoter [8]. These observations support the notion that the cooperation between the E2F and CHR sites is requited for recruiting the E2F4/p130 repressor complex to the E2F1 and B-myb promoters. Although a factor binding to the B-myb CHR has been detected in partially purified nuclear extracts [7], the nature of the protein(s) interacting with the CHRs remains largely unknown. Here, we show that the B-myb CHR binding activity is co-purified with E2F DNA-binding activity, and coimmunoprecipitated with E2F4 and Rb family proteins. Furthermore, we provide evidence that a factor binding to the E2F CHR is different from that of the B-myb CHR, suggesting that transcriptional repression of the E2F1 and B-myb promoters is mediated by interaction of E2F and Rb family members with promoter-specific corepressors. Materials and methods Cell culture. REF52 cells were cultured at 37 C in Dulbecco’s modified Eagle’s minimum essential medium (DMEM; Sigma–Aldrich) supplemented with 10% fetal calf serum (FCS). HL60 cells were maintained in RPMI 1640 medium (Nissui) supplemented with 20% FCS. Plasmids. The human E2F1 promoter reporter constructs, E2F1-Luc (242, wt), has been described in [5]. E2F1-Luc B-mybCHR were generated using the QuickChange II Site-Directed Mutagenesis Kit (Stratagene) with the oligonucleotides 5 0 -TCTTTCGCGGCAAAAAGGAAATGGC GCGTAAAAGTGGCCG-3 0 and 5 0 -CGGCCACTTTTACGCGCCATT TCCTTTTTGCCGCGAAAGA-3 0 .
1051
Transfection and luciferase assays. REF52 cells were transfected using FuGENE 6 Transfection Reagent (Roche Diagnostics) as described in [9]. A CMV b-gal vector was cotransfected as an internal control. The transfected cells were brought to quiescence in DMEM with 0.1% FCS for 48 h before harvest. Luciferase and b-galactosidase activities were measured using a luciferase reporter gene assay system (Toyo Ink) and the Aurora GAL-XE reporter gene assay system (ICN), respectively. DNA affinity purification. Nuclear extracts were prepared from growing HL60 cells as described in [15]. To generate E2F affinity beads, four tandem repeated fragments, each of which contains either wild-type (wt) or mutant versions of two E2F sites derived from adenovirus E2 promoter (gifts from K. Ohtani), were immobilized on magnetic beads (Dynabeads M-280; Dynal). Precleared HL60 nuclear extracts (100 lg) were diluted in shift buffer (20 mM Hepes (pH 7.9), 40 mM KCl, 6 mM MgCl2, 1 mM EGTA, 0.1% Nonidet P-40, 0.2 mM DTT, 10% glycerol) buffer, containing 1.2 mg/ml bovine serum albumin, 4% ficoll, 2 mg/ml sonicated salmon sperm DNA, 1% casein,, and 1 m PMSF, and then incubated with either wt or mutant versions of the E2F affinity beads for 1 h at room temperature with gentle rotation. The beads were washed with ice-cold shift buffer containing 100 mM KCl, and the bound complexes were eluted with shift buffer containing 500 mM NaCl and 1% casein. The supernatant was dialyzed with ice-cold shift buffer. Immunoprecipitation. Immunoprecipitation was performed as described in [15]. Briefly precleared HL60 nuclear extracts (100 lg) were incubated with antibody-conjugated Dynabeads for 4 h at 4 C with gentle rotation. After washing, the immunoprecipitates were dissociated by treating with 0.8% sodium deoxycholate (DOC) (Sigma–Aldrich) on ice, and then the released proteins were neutralized with Nonidet P-40 (final concentration of 1%). The following antibodies were used in this study: mouse monoclonal Rb (C-36, Santa Cruz Biotechnology) and p107 (a mixture of SD-2, SD-4, SD-6, SD-9, and SD-15; gifts from N. Dyson) antibodies, and rabbit polyclonal p130 (C-18) and E2F4 (C-108) antibodies (Santa Cruz Biotechnology). Gel mobility shift assays (EMSAs). EMSAs were performed as described in [15]. Sonicated salmon sperm DNA (500 ng) and Poly (dI–dC) (1 lg) were added as nonspecific competitors to analyze the E2F and CHR binding activities, respectively. The E2F (DHRF) probe was described previously [15]. The following double-stranded oligonucleotides were used as probes and/or unlabeled competitors: E2F wt and E2F mut; E2F1 wt and E2F1 mE2F (Mut E2F site A and B) [9]; E2F1 mCHR (5 0 -CGTGGCTCTTT CGCGGCAAAAAGGCCTTGGCGCGTAAAAGT-3 0 ); E2F1 E2F1 Bmyb-CHR (5 0 -CGTGGCTCTTTCGCGGCAAAAAGGAAATGGCGC GTAAAAGT-3 0 ); B-myb wt (5 0 -TTGCCGACGCACTTGGCGGGAGA TAGGAAAGTGGTTCTGAA-3 0 ); B-myb mE2F (5 0 -TTGCCGACGC ACYYGTATGGAGATAGGAAAGTGGTTCTGAA-3 0 ); B-myb mCHR (5 0 -TTGCCGACGCACTTGGCGGGAGATAGGCCTGTGGTTCTGA A-3 0 ); B-myb mE2F-mCHR (5 0 -TTGCCGA CGCACTTGTATGGAGATAGGCCTGTGGTTCTGAA-3 0 ).
Fig. 1. Alignment of human E2F1 and mouse B-myb promoters. Specific changes in these reporter constructs or the oligonucleotides used in this study are indicated below the wild-type sequences. E2F-binding sites are indicated by arrows and CHRs are boxed.
1052
Y. Nakajima et al. / Biochemical and Biophysical Research Communications 364 (2007) 1050–1055
Results and discussion Co-purification of B-myb CHR binding activities with E2F DNA-binding activities To examine the physical interaction of the CHR binding factor with E2F4 or the Rb family member proteins, we purified E2F activities from nuclear extracts prepared from growing HL60 cells by using E2F DNA affinity beads that were conjugated with either wt or mutant E2F-binding sequences derived from the adenovirus E2 promoter, which has no recognizable CHR. EMSAs with a probe containing a E2F site confirmed that samples eluted from the wt, but not the mutant E2F affinity beads contained free E2F, which was abolished by unlabeled wt, but not mutant, E2F competitors (Fig. 2A, lanes 2 and 3). The wt E2Faffinity purified samples were then analyzed with three different probes containing wt and mutant B-myb promoter sequences that were mutated in either the E2F site (mE2F) or CHR (mCHR) as described previously [7] (Fig. 1). Fig. 2B shows that E2F specific activity was detected with wt and mCHR probes (lanes 1 and 3). Moreover, we detected bands, although faint, with the mE2F probe containing the intact CHR (lane 2, arrows). Fig. 2C shows that the bands are successfully abolished by competitors containing the CHR (wt and mE2F) (lanes 2 and 3), but not by competitors mutated in CHR (mCHR) (lane 4). Furthermore, that the bands detected with the Bmyb mE2F probe were competed by wt, but not mCHR,
both of which have no mutated E2F site (Fig. 2C, 3B, and 4B. see below), together with the fact that the double mutant (B-myb mE2F-mCHR), although it has the mutated E2F site, failed to compete the bands (Fig. 4B and see below), excludes the possibility that the mutation of the E2F site might create a binding site irrelevant to the CHR-binding factor. These results indicate that the CHR-specific DNA binding activity is co-purified with the E2F DNA-binding activity. Interaction of the B-myb CHR binding activities with components of the E2F/Rb family member complexes To examine the interaction of the CHR binding factor with E2F4 and the Rb family proteins, HL60 nuclear extracts were immunoprecipitated with either E2F4 or Rb antibody, and then the precipitated complexes were dissociated by treating with DOC as described previously [15]. EMSAs with the E2F probe confirmed that free E2F was released from the E2F4 and Rb immunoprecipitates (Fig. 3A, and data not shown). Competition experiments with the B-myb mE2F probe revealed that DNA binding activities dependent on the CHR, but not E2F site, were detected in Rb immunoprecipitates (Fig. 3B, lane 2–4, and see below). Furthermore, the CHR-specific bands, which were confirmed by competition assays (data not shown), were also found in association with E2F4, p107, and p130 immunoprecipitates (Fig. 3C). These results indicate that the binding activities specific to the B-myb CHR
Fig. 2. CHR specific DNA-binding activities were co-purified with E2F-specific DNA affinity beads. (A) Detection of E2F DNA-binding activity. The samples purified with the wild-type (wt, lanes 1–3) and mutated version (mut, lane 4) of E2F affinity beads were subjected to EMSA with the 32P-endlabled E2F (DHRF) probe in the presence or absence of non-labeled competitor DNA (E2F wt or E2F mut) at 100-fold molar excess. (B,C) Detection of CHR specific DNA-binding activities. Samples purified by wt E2F affinity beads were subjected to EMSA with the indicated B-myb probes (Fig. 1). Arrows and an asterisk indicate CHR (B, lane 2; C, lane 1 and 4) and non-specific bands, respectively. EMSA was performed in the presence or absence of the indicated cold competitors (Fig. 1) at 100-fold molar excess (C).
Y. Nakajima et al. / Biochemical and Biophysical Research Communications 364 (2007) 1050–1055
1053
Fig. 3. Detection of CHR specific DNA-binding activities in immunoprecipitates with E2F4 and the Rb family member antibodies. (A) The supernatant of DOC-treated immunoprecipitates with antibodies specific to E2F4 or Rb was assayed for E2F DNA binding as in Fig. 2A. NE: HL60 nuclear extracts. (B,C) DOC-treated immunoprecipitates with antibodies specific to E2F4 or the indicated Rb family members were assayed for CHR DNA binding with the B-myb mE2F probe as in Fig. 2C in the presence or absence of the indicated cold competitors (Fig. 1) at 100-fold molar excess. The arrows and asterisk indicate CHR (B, lane 4 and C lanes 2–5) and non-specific bands, respectively. IgG refers to immunoprecipitates with a control rabbit IgG, as a negative control.
are associated with components of the E2F/Rb family member complexes. The E2F1 and B-myb CHRs exhibits distinct factor-binding specificities Structural and functional similarities of the E2F1 and Bmyb CHRs suggest that the same or similar factors bind to these CHRs. To examine the DNA-binding specificity of the CHR-binding factors, we performed competition assays using a probe that contains the E2F1 CHR (E2F1 mE2F). Fig. 4A shows that a shifted band was specifically abolished by competitors containing the E2F1 CHR (E2F1 wt and mutE2F) (lanes 3–5), but not by CHR-mutated competitors (E2F1 mCHR) (lane 4). Unexpectedly, however, the band was not affected by competitors containing the B-myb CHR (B-myb mE2F), even at a higher concentration (500-fold molar excess) (lanes 7 and 8). Conversely, two bands which specifically bound to the B-myb CHR probe (B-myb mE2F) were not abolished by E2F1 CHR competitors (E2F1 mE2F) (Fig. 4B, lane 8). These results suggest that the factor binding to the E2F CHR is different from that binding to the B-myb CHR. The E2F1 and B-myb CHRs are not equivalent in repressing the E2F1 promoter To examine whether the B-myb CHR can act as a corepressor site in the E2F1 promoter, we made use of an E2F1 reporter construct in which its CHR was substituted with the B-myb CHR sequences (E2F1 B-myb-CHR). EMSAs using a B-myb mE2F probe confirmed that the shifted
bands were abolished by the E2F1 B-myb-CHR competitors (Fig. 4C). Luciferase assays using REF52 cells revealed that the E2F1 CHR substitution with the B-myb CHR led to a sevenfold reduction of the E2F1 promoter activity relative to the CHR-mutated E2F1 promoter under serumstarved conditions (Fig. 4D). However, we reproducibly detected twofold higher activity in the chimeric promoter relative to the wt E2F1 promoter, indicating that the Bmyb CHR is functionally similar but not equivalent to the E2F1 CHR in terms of the repression of the E2F1 promoter. In this study, biochemical evidence is presented for the physical interactions of the putative CHR binding factors with the components of the E2F/Rb-family repressor complexes. This provides a basis for the cooperation between the CHR and E2F sites in repression of the E2F1 and Bmyb promoters. It is likely that the interaction of the E2F/Rb-family members with the CHR binding factor stabilizes binding of the repressor complexes to these promoters, thereby playing a critical role for recruiting the repressor complexes in vivo. Furthermore, we show that, in spite of structural and functional similarities, the factor binding to the E2F CHR is different from that binding to the B-myb CHR, and that the B-myb CHR is functionally similar but not equivalent to the E2F1 CHR in the repression of the E2F1 promoter. In B-myb promoter, substitution of A to T at the 5th position of the CHR, which is identical to that of the E2F1 CHR, results in the derepression of the B-myb promoter in G0 [8]. Therefore, these observations indicate that promoter-specific CHRs are required to repress the E2F1 and B-myb promoters. The promoter-specific function of the CHRs is also supported
1054
Y. Nakajima et al. / Biochemical and Biophysical Research Communications 364 (2007) 1050–1055
Fig. 4. The E2F1 and B-myb CHRs were functionally similar but not equivalent. (A, B, and C) Distinct DNA binding specificities of factors binding to the E2F1 and B-myb CHRs. The supernatant of DOC-treated immunoprecipitates with the Rb antibody were assayed for CHR DNA binding as in Fig. 2C with the E2F1 mE2F (A) and B-myb mE2F (B, C) probes in the presence or absence of cold competitors (Fig. 1) at the indicated molar excess. The arrows indicate CHR-specific bands. The probe only refers to the absence of the supernatant (C). (D) Corepressor activity of the B-myb CHR in the E2F1 promoter. REF52 cells were transfected with either E2F1-Luc wt [5], mCHR [9], or B-myb-CHR, along with pCMV b-gal as an internal control. Cells were induced into quiescence by serum starvation, and then harvested 48 h after transfection. Luciferase and b-galactosidase activities were measured as described previously [9]. Luciferase activity was normalized to b-galactosidase activity. Data represent average of three independent experiments, each performed in duplicate.
by the fact that substitution of the CHR in the Cdc25C promoter with the B-myb CHR causes derepression of the promoter activity in G0 [7]. Based on the study presented here, and in combination with other findings, we propose that transcriptional repression of the E2F1 and B-myb promoters would result from physical interaction of E2F/Rb-family repressor complexes with promoter-specific corepressors, and that such promoter-specific interaction is required for proper promoter functions of E2F1 and B-myb genes. Whether the CHR binding factors interact with the repressor complexes through E2F4 or Rb family members, and whether such interactions change during
the cell cycle remain unclear at present. Identification of the CHR binding factors will be required to address these issues.
Acknowledgments We thank K. Ohtani, and N. Dyson for materials, and K. Ohtani for critical reading of the manuscript. This study was supported in part by Grants-in-Aid for Scientific Research from the Japan Society for the Promotion of Science.
Y. Nakajima et al. / Biochemical and Biophysical Research Communications 364 (2007) 1050–1055
References [1] C. Attwooll, E. Lazzerini Denchi, K. Helin, The E2F family: specific functions and overlapping interests, EMBO J. 23 (2004) 4709–4716. [2] D.K. Dimova, N.J. Dyson, The E2F transcriptional network: old acquaintances with new faces, Oncogene 24 (2005) 2810–2826. [3] Y. Takahashi, J.B. Rayman, B.D. Dynlacht, Analysis of promoter binding by the E2F and pRB families in vivo: distinct E2F proteins mediate activation and repression, Genes Dev. 14 (2000) 804–816. [4] J. Wells, K.E. Boyd, C.J. Fry, S.M. Bartley, P.J. Farnham, Target gene specificity of E2F and pocket protein family members in living cells, Mol. Cell Biol. 20 (2000) 5797–5807. [5] D.G. Johnson, K. Ohtani, J.R. Nevins, Autoregulatory control of E2F1 expression in response to positive and negative regulators of cell cycle progression, Genes Dev. 8 (1994) 1514–1525. [6] J.D. Bennett, P.G. Farlie, R.J. Watson, E2F binding is required but not sufficient for repression of B-myb transcription in quiescent fibroblasts, Oncogene 13 (1996) 1073–1082. [7] N. Liu, F.C. Lucibello, J. Zwicker, K. Engeland, R. Muller, Cell cycle-regulated repression of B-myb transcription: cooperation of an E2F site with a contiguous corepressor element, Nucleic Acids Res. 24 (1996) 2905–2910. [8] S. Catchpole, F. Tavner, L. Le Cam, C. Sardet, R.J. Watson, A Bmyb promoter corepressor site facilitates in vivo occupation of the adjacent E2F site by p107 · E2F and p130 · E2F complexes, J. Biol. Chem. 277 (2002) 39015–39024.
1055
[9] K. Araki, Y. Nakajima, K. Eto, M.A. Ikeda, Distinct recruitment of E2F family members to specific E2F-binding sites mediates activation and repression of the E2F1 promoter, Oncogene 22 (2003) 7632–7641. [10] J. Zwicker, R. Muller, Cell-cycle regulation of gene expression by transcriptional repression, Trends Genet. 13 (1997) 3–6. [11] J. Zwicker, F.C. Lucibello, L.A. Wolfraim, C. Gross, M. Truss, K. Engeland, R. Muller, Cell cycle regulation of the cyclin A, cdc25C and cdc2 genes is based on a common mechanism of transcriptional repression, EMBO J. 14 (1995) 4514–4522. [12] N. Liu, F.C. Lucibello, K. Korner, L.A. Wolfraim, J. Zwicker, R. Muller, CDF-1, a novel E2F-unrelated factor, interacts with cell cycle-regulated repressor elements in multiple promoters, Nucleic Acids Res. 25 (1997) 4915–4920. [13] M. Kimura, C. Uchida, Y. Takano, M. Kitagawa, Y. Okano, Cell cycle-dependent regulation of the human aurora B promoter, Biochem. Biophys. Res. Commun. 316 (2004) 930–936. [14] F. Bolognese, M. Wasner, C.L. Dohna, A. Gurtner, A. Ronchi, H. Muller, I. Manni, J. Mossner, G. Piaggio, R. Mantovani, K. Engeland, The cyclin B2 promoter depends on NF-Y, a trimer whose CCAAT-binding activity is cell-cycle regulated, Oncogene 18 (1999) 1845–1853. [15] M.A. Ikeda, L. Jakoi, J.R. Nevins, A unique role for the Rb protein in controlling E2F accumulation during cell growth and differentiation, Proc. Natl. Acad. Sci. USA 93 (1996) 3215– 3220.