P300 Provides a Corepressor Function by Cooperating with YY1 and HDAC3 to Repress C-Myc

Home , Promoter (genetics)

Oncogene (2008) 27, 5717–5728 & 2008 Macmillan Publishers Limited All rights reserved 0950-9232/08 $32.00 www.nature.com/onc ORIGINAL ARTICLE p300 provides a corepressor function by cooperating with YY1 and HDAC3 to repress c-Myc

N Sankar1, S Baluchamy1,3,4, R-K Kadeppagari1,3,5, G Singhal1,3, S Weitzman2 and B Thimmapaya1

1Department of Microbiology and Immunology, Feinberg School of Medicine, Northwestern University, Chicago, IL, USA and 2Department of Medicine, Feinberg School of Medicine, Northwestern University, Chicago, IL, USA

We showed earlier that p300/CBP plays an important role Myc expression in normal cells is critical and is likely to in G1 progression by negatively regulating c-Myc and be mediated at least in part by the chromatin remodeling thereby preventing premature G1 exit. Here, we have complexes. The chromatin remodeling proteins p300 studied the mechanism by which p300 represses c-Myc and and CBP are two highly conserved nuclear phospho- show that in quiescent cells p300 cooperates with histone proteins that are recruited by a large number of deacetylase 3 (HDAC3) to repress transcription. p300 and transcription factors to activate transcription. These HDAC3 are recruited to the upstream YY1-binding site of proteins also contain intrinsic histone acetyl transferase the c-Myc promoter resulting in chromatin deacetylation activity that acetylates the nucleosomal histones and and repression of c-Myc transcription. Consistent with modiﬁes the local chromatin structure to promote this, ablation of p300, YY1 or HDAC3 expression results transcription (Goodman and Smolik, 2000). Cell in chromatin acetylation and induction of c-Myc. These transformation by viral oncoproteins such as E1A is three proteins exist as a complex in vivo and form a dependent on its binding to and inactivating p300/CBP multiprotein complex with the YY1-binding site in vitro. (Frisch and Mymryk, 2002). p300 is mutated in several The C-terminal region of p300 is both necessary and forms of cancer suggesting a tumor suppressor role for sufﬁcient for the repression of c-Myc. These and other this protein (Iyer et al., 2004). results suggest that in quiescent cells the C-terminal We previously showed that both p300 and CBP play region of p300 provides corepressor function and facili- an important role in quiescent cells by repressing Myc tates the recruitment of p300 and HDAC3 to the YY1- and preventing premature cell cycle G1 exit and that the binding site and represses the c-Myc promoter. This repressive activity of p300 is not dependent on its HAT corepressor function of p300 prevents the inappropriate activity (Baluchamy et al., 2003, 2005). We also showed induction of c-Myc and S phase. that E1A induces Myc in quiescent cells in a p300- Oncogene (2008) 27, 5717–5728; doi:10.1038/onc.2008.181; dependent manner (Kolli et al., 2001; Baluchamy et al., published online 9 June 2008 2007). In this paper, we present evidence that in quiescent cells, p300 cooperates with transcription Keywords: c-Myc; p300; gene expression; cell cycle; factor YY1 and histone deacetylase 3 (HDAC3) to YY1; HDAC3 repress the Myc promoter. p300 and HDAC3 associate with the Myc promoter upstream YY1-binding site resulting in the deacetylation of the Myc chromatin. YY1, p300 and HDAC3 bind to the Myc promoter YY1 Introduction site in vitro as a complex. The three proteins also form complexes both in vivo and in vitro. These and other The proto-oncogene c-Myc (referred to as Myc in this results suggest that the C-terminal region of p300 report) plays a pivotal role in the control of cell provides an essential corepressor function in recruiting proliferation and its expression is deregulated in most HDAC3 to the Myc promoter and keeping Myc in a forms of cancer (Dang, 1999). Stringent regulation of repressed state in growth-arrested cells.

Correspondence: Dr B Thimmapaya, Department of Microbiology Results and Immunology, Northwestern University Medical School, 303 E. Chicago Avenue, olson 8452, Chicago, IL 60611, USA. E-mail: [email protected] Identification of p300 targets on human Myc promoter 3These authors contributed equally to this work. As p300 does not bind DNA directly, we reasoned that 4Current address: University of Illinois School of Medicine, Chicago, in quiescent cells a Myc promoter-specific transcription IL, USA. factor could recruit p300 and cooperate with deacety- 5Current address: LSU Health Sciences Center, New Orleans, LA, USA. lase(s) to repress the promoter. To identify the p300 Received 2 January 2008; revised 15 April 2008; accepted 1 May 2008; responsive elements, rat12 cells were transfected published online 9 June 2008 with Myc promoter–reporter constructs containing p300 is a corepressor of c-Myc N Sankar et al 5718 progressive deletions from the 50 end and maintained del-6 is repressed as efficiently as del-1 plasmid contain- under serum-free medium for 48 h, then serum stimu- ing 2130 bp promoter sequences (Figure 1b) and p300 lated for 2 h and the luciferase activity was quantified. fail to repress when the YY1-binding site in del-6 is Note that in this report, rat12 cells were used for mutated (Figure 1c). transient assays, whereas human breast epithelial cells (MCF10A) were used to study the endogenous Myc Recruitment of p300 and HDAC3 to the YY1 site, promoter. In our earlier studies, we used both MCF10A transcription initiation and coding regions and their and rat12 cells to show that the endogenous Myc effect on chromatin acetylation promoter is repressed by p300 (Baluchamy et al., 2003). Normally, the endogenous Myc promoter is minimally However, transfection efficiency of MCF10A cells is active transcriptionally and it is transiently induced after very poor and thus MCF10A cells could not be used for serum stimulation. We previously showed that the best transient transfection assays. Data shown in Figure 1b way to study the repression of the endogenous Myc indicate that a deletion mutant containing 647 bp promoter by p300 is to overexpress p300 in quiescent upstream from the cap site (del-6) was efficiently cells then induce Myc by serum stimulation (Baluchamy repressed by p300 (the numbering is relative to the et al., 2003). Thus, we overexpressed FLAG p300 in major P2 promoter start site; see Figure 1a; Zou et al., quiescent human MCF10A cells, serum stimulated for 1997). Although truncation of another 300 bp of the 2 h to induce Myc and then performed ChIP assays with promoter resulted in a significant loss of basal activity, primer pairs that encompass the upstream 647 bp of the this mutant also lost its response to p300 indicating that promoter sequences (Figure 2a shows the time course). transcription factors binding to sequences between À647 p300 is also reported to be associated with the and À347 bp contribute to the repression of Myc by transcriptional initiation and elongation complexes p300 (del-7; Figure 1b). Previous studies in rat and raising the possibility that in p300 overexpressing cells, mouse B cells have identified functional YY1 and increased levels of p300 may also be found in transcrip- Blimp1 sites abutting each other in the region between tional initiation and coding regions (Black et al., 2006; À350 and À450 bp of the Myc promoter (Figure 1a and Guermah et al., 2006). Therefore, we also scanned three d; Zou et al., 1997). Mutagenesis of the putative Blimp1 randomly chosen regions downstream of the RNA start site in the context of del-6 promoter did not affect site for p300 occupancy. A western blot shown in repression by p300, whereas mutation of the YY1 site Figure 2c indicated that only p300 levels increased in resulted in the complete loss of repression (plasmid p300 overexpressing cells. The occupancy of p300, YY1, YY1M, Figure 1c). Electrophoretic mobility shift assays HDAC1, -2 and -3 was quantified using real-time PCR (EMSA) using nuclear extracts prepared from human with the primer pairs shown in Figure 2b. Changes in MCF10A cells and an YY1-specific antibody showed the levels of occupancy of these proteins in different that human YY1 protein binds to this sequence (super- regions of the promoter in p300 overexpressing cells as shifted band shown by an open arrow in Figure 1e). compared to that of control samples are shown as a bar Note that in the absence of p300 overexpression the diagram in Figure 2d (see legend to Figure 2d for further reporter activity of YY1M construct is similar to that of details). These data show that the occupancy of YY1 the WTplasmid even though repression in YY1M and p300 was enriched by about 10- and 5-fold, construct is relieved (Figure 1c). We do not have a clear respectively, in the YY1-binding region (region II). A explanation for this at present. We assayed these 20-fold increase in the occupancy of HDAC3 but not plasmids 2 h after serum stimulation and it is possible HDAC1 or 2 was also noted in region II. HDAC1 and 2 that when stimulated with serum, in the absence of p300 occupancy levels did not change in any of the regions overexpression, the YY1-mediated repression of Myc is tested, and in fact, they were slightly reduced in several relieved in the WTplasmid and thus both plasmids show regions (Figure 2d). There was no increase in YY1 similar levels of reporter activity. Nonetheless, under occupancy levels in any other regions of the promoter. these assay conditions, when p300 is overexpressed, There was also no change in p300 or HDAC3 occupancy

Figure 1 Identiﬁcation of the p300-responsive sequences on the human Myc promoter, and the YY1-binding site as a p300 target. (a) Map of the human Myc promoter and positions of YY1 and Blimp1 sites relative to the P2 promoter transcription start site. (b) Promoter–reporter assays using deletion constructs. Rat12 cells were transfected with the Myc promoter deletion constructs along with a p300 expression plasmid as described under Materials and methods. The vector DNA was included wherever appropriate to keep the DNA concentration constant. Luciferase units are shown for 5 mg of protein after normalizing with Rennila luciferase activity. Mean values from three independent experiments ±s.d. are shown. (c) Effect of YY1 and Blimp1 site mutations on repression by p300. Blimp1 and YY1 sites were mutated in the context of a deletion construct containing 647 bp from the P2 start site (del-6) and assayed as above. Note that these two sites are mutated in separate plasmids. YY1-M and Blimp-M refer to the plasmid constructs with mutations in YY1 and Blimp1 sites, respectively. (d) Comparison of the DNA sequences from À399 to À447 of the human Myc promoter with those of mouse and rat Myc promoters. The previously identiﬁed YY1 and Blimp1 sites of the mouse and rat Myc promoters and matching sequences in the human Myc promoter are underlined. Bases mutated in Blimp1 and YY1 sites are shown below the human sequence. (e) EMSA showing YY1 protein binding to the YY1 site. A 50 end-labeled double-stranded oligonucleotide containing the YY1 site was incubated in ice with 5 mg MCF10A nuclear extracts in an EMSA buffer and the DNA–protein complexes were resolved on 5% non- denaturing polyacrylamide gels at 4 1C (see Materials and methods). For super-shift assays, an a-YY1 antibody (H-414; Santa Cruz, Santa Cruz, CA, USA) was included during incubation. Open arrow shows the super-shifted complex. For competition assays, unlabeled probe was used at molar concentrations as indicated. YY1 refers to YY1–DNA complex and YY1m refers to mutant oligonucleotide as shown in (d). The sequence of the YY1-binding site probe is shown in (d).

Oncogene p300 is a corepressor of c-Myc N Sankar et al 5719 levels in regions III and IV between control and p300 However, these sites were not sensitive to p300 repres- overexpressing cells (Figure 2d). sion as transfection assays showed that promoter Region V was the only other promoter region that constructs with mutations on these sites continued to showed an increase in the p300 and HDAC3 occupancy be repressed by p300 (Figure 2e). As reported earlier, levels (3.3- and 9-fold, respectively). Regions IV and V deletion of the TATA box abolished the promoter contain binding sites for transcription factors MAZ activity (Albert et al., 2001). The increased occupancy of (Me1a2; Izzo et al., 1999), E2F (Albert et al., 2001), p300 in region V is likely to be due to its association with CTCF (Filippova et al., 1996) that function as protein complexes involved in transcription initiation, repressors and also TBP (TATA box; see Figure 2b). because p300 has been shown to be associated with the 444 (-190) 412 P1 P2 Blimp-1- YY1-

-1000-600 -500 -400 -300 -200 -100 +1

- p300 - p300 + p300 + p300 40

) ) 40 5 5

20 20 Luc activity (x10 Luc activity (x10

0 0

del-1 del-3 del-5 del-6 del-7 del-6 (-941) (-647) (-347) YY1-M (-2130) (-1544) Blimp1-M

Blimp1 YY1 Rat (-451) GTACAGAAAGGGAAAGGACCGACGCGCGAGCAGGAGAAAATGGTCGGGC (-403)

Mouse (-425) GTACAGAAAGGGAAAGGACTAGCGCGCGAGAAGAGAAAATGGTCGGGC (-378)

Human (-447) GCAGAGAAAGGGAGAGGGTTTGAGAGGGAGCAAAAGAAAATGGTAGGCG(-399) ATTCATACTCA TCGCGAATTCTT

α-YY1 - - + - - - - - Comp-YY1 - - - + + + - - Comp-YY1m ------+ + 5X 10X 25X 10X 25X

YY1

Free probe

Oncogene p300 is a corepressor of c-Myc N Sankar et al 5720 transcription initiation complex (Black et al., 2006) as suggesting that YY1 was not involved in the recruitment well as with the C-terminal region of RNA polymerase of p300 to these regions. Recent results indicate that II (Nakajima et al., 1997). p300 occupancy was also methylation of histone H3 also represses gene expression enriched in regions VI, VII and VIII consistent with (Schwartz and Pirrotta, 2007). Preliminary ChIP assays previous reports that p300 associates with proteins using an antibody speciﬁc for H3K9 indicated that involved in transcription elongation (factor SII, for H3K9 methylation was not involved in the p300- example, Guermah et al., 2006; see Discussion). mediated Myc repression (data not shown). Interestingly, HDAC3 was also enriched wherever The acetylation levels for each region in p300 over- p300 was enriched, suggesting that p300 cooperates expressing cells as compared to normal cells were with HDAC3 by associating with proteins involved determined using antiacetyl antibodies that recognize with transcription initiation and elongation. To rule out acetylated H3 and H4 and relevant primer pairs. Data that YY1 indirectly assisted p300 to associate with shown in Figure 2f indicate that acetylation was chromatin in regions VII and VIII, probes correspond- decreased in all the regions tested, although the degree ing to regions VII and VIII were tested in EMSA using to which the acetylation was reduced was variable and extracts containing epitope tagged p300 and YY1. No ranged from 30 to 90%. In summary, in p300 over- DNA–protein complexes were generated in these assays expressing cells, repression of Myc strongly correlates

Starve Infection 0hr/Stimul.

32hr 16hr ChIP protein Luc assay

β-gal Ad Adp300 (-190) α P1 P2 - p300 CTCF +2 Blimp-1-444 YY1-412 PuF-310 TATA -28 R Me1a2-108 E2F-68 α- HDAC1 -400 -300 -200 -100 +1 α- HDAC2 α- HDAC3 IIIIII IVV VI VII VIII α- YY1 F -985 F -494 F -385 F -192 F -89 F +371 F +2067 F +4109 α- Actin R -783 R -366 R -244 R -77 R +54 R +599 R +2313 R +4315

25 I II III IV V VI VII VIII 20 15 10

Fold Enriched 5 1 0 α-p300 α-YY1 α-HDAC3 α-HDAC1 α-HDAC2

5 120 - p300 α -Ac-H3 ) + p300 4 4 α -Ac-H4

3 80

1 % Change 40 Luc activity (x 10

del6 0 TATA

E2F-M IIVVI VII VIII ∆ CTCF-M Me1a2-M Control p300

Oncogene p300 is a corepressor of c-Myc N Sankar et al 5721 with increased occupancy of p300 and HDAC3 on the it was not uniform. The H4 acetylation in all regions was promoter and the coding regions and reduced acetyla- increased by about twofold, whereas increased H3 tion of Myc chromatin. acetylation was evident for only regions II and V. Nonetheless, the increase in H4 acetylation in p300- depleted cells correlates with the increase in Myc Effect of YY1, HDAC3 and p300 knockdown on their expression. recruitment to the YY1 site, and chromatin acetylation Next, we examined whether the occupancy of HDAC3 and p300 on the YY1 site would be affected when the Formation of a multiprotein complex containing YY1, expression of p300 or YY1 is knocked down by small p300 and HDAC3 with the Myc promoter YY1-binding hairpin RNAs (shRNAs). Expression of the shRNAs site specific for YY1, p300 or HDAC3 in quiescent cells led EMSA was carried out using nuclear extracts prepared to a significant reduction in the accumulation of the from SaOS2 cells (we avoided the usual 293Tor COS relevant protein (Figure 3a), Also, p300 shRNA cells for extract preparation, as these cells contain E1A expression did not affect the expression of CBP or large Tthat bind to p300) transfected with HA-p300, (Figure 3a), YY1, HDAC1, 2 or 3 (Figure 3b). As FLAG-YY1 and His-HDAC3 simultaneously and with shown in Figure 3c, p300 knockdown led to a 12-fold an oligonucleotide containing the Myc promoter YY1 induction of Myc reporter activity (luciferase shRNA site. The gel was run until the YY1 DNA-protein was used as control). Knockdown of YY1 and HDAC3 complex migrated to the bottom of the gel (shown by an individually led to a sixfold induction of the Myc open arrow in Figure 4a). When epitope-specific activity. Ablation of HDAC1 expression had no effect antibodies were used in the assays, several distinct on the Myc activity. The increase in transcriptional super-shifted bands were visible whose identities could activity of the Myc protein also correlated with the be established based on the migration of the complexes increase in Myc protein levels (Figure 3d). We pre- and the type of the antibodies used. For example, when viously showed that the changes in Myc protein and a-FLAG antibody (YY1 specific) was included in the activity levels correlate with Myc RNA levels (Baluchamy assay, a distinct super-shifted band appeared (C2), et al., 2003, 2005). Under p300 knockdown conditions, confirming the presence of YY1 in the complex (lane neither p300 nor HDAC3 were detected at the YY1- 2). Similarly, with HDAC3-specific a-His antibody a binding site and YY1 was detected at much reduced different super-shifted band appeared (C1), confirming levels. Likewise, when YY1 expression was blocked, the association of HDAC3 with the YY1–DNA complex there was no detectable occupancy of p300 and HDAC3 (lane 4). It is interesting that the DNA-HisHDAC3 at the YY1-binding site (Figure 3e). super-shifted band migrates faster than the DNA-FLA- ChIP assays as described above showed that there was GYY1. This may be due to unique conformation of an overall increase in the acetylation of chromatin H3 these complexes or due to different stoichiometry of the and H4 both in p300 and HDAC3 downregulated antibodies binding DNA–protein complexes that alter samples (Figure 3f ). The acetylation of H3 and H4 in the mobility of the complexes. When a-HA (p300 HDAC3 shRNA samples increased by about three- to specific) antibody was included in the incubation mix fivefold in the regions tested. There was also an increase along with a-FLAG antibody, a slowly migrating band in the acetylation of the p300 shRNA samples although near the origin appeared (C4, lane 5). Similarly, in the

Figure 2 Increased occupancy of p300 and HDAC3 on the Myc promoter and coding regions in p300 overexpressing cells. (a) Schematic representation of the time course of serum starvation, vector infection and harvesting of cells. (b) Schematic representation of the Myc promoter showing the location of key transcription factor-binding sites within the upstream 500 bp of the P2 promoter and the location of the primer pairs on the promoter and coding regions used in ChIP assays. (c). Levels of p300, HDAC1, 2 and 3 and YY1 proteins in MCF10A cells infected with Adp300 vector. Serum starved cells were infected with Adp300 for 16 h, serum stimulated for 2 h then lysed and the proteins were analysed by western immunoblotting using antibodies as shown. The sources of the antibodies are: p300 (N-15), YY1 (H-414) and HDAC3 (B-10) were from Santa Cruz; HDAC1 (2E10) and HDAC2 (3F3) were from Upstate (Charlottesville, VA, USA). (d) ChIP assays showing increased occupancy of p300, HDAC3 and YY1 on the upstream YY1 site and increased occupancy of p300 and HDAC3 on the transcription initiation and coding regions of the human Myc promoter. The ChIP assay was carried out essentially as described in Materials and methods using primer pairs shown in the figure to generate about 150–200 bp products. The chromatin was immunoprecipitated with antibodies specific for p300 (N-15), YY1 (H-414), HDAC1 (2E10), 2 (3F3) and 3 (B-10) as shown and the amplified products were quantified by real-time PCR. Percent input for each region was calculated for control and experimental group and the fold change for p300 samples over control samples for each region were calculated. As additional controls, occupancy of all three proteins in each region were quantified using Immunoglobulin G antibodies and in p300 unresponsive region I. In both cases using real-time PCR, negligible quantities of DNA were recovered. The ChIP assay was repeated twice and the data shown are mean±s.d. (e) Luciferase assays for plasmids with mutations in transcription factor-binding sites. The Me1a2 site that maps at À108 (GAGAAGGGCCAGGGC) was mutated in which two underlined bases were changed to TT using a site directed mutagenesis procedure. The E2F site located at À68 (TTGGCGGGA) was mutated such that the five underlined bases were changed to AATTC. The three underlined bases in the CTCF site located at þ 2 (CTAACTCGCTGTAGTAATTCCAGCGAG) were mutated to ACA. The TATA box mutation included a 6 bp deletion in the TATA box identical to that of the APA2 mutant described earlier (Albert et al., 2001). (f) Bar diagram showing changes in H3 and H4 acetylations in promoter and coding regions in p300 overexpressing cells. The ChIP assays were carried out using primer pairs as shown in (b). The percent input for each region in control and the experimental samples were calculated as above. The values shown are the percent change observed for p300 overexpression samples for each region as compared to the corresponding regions of the control samples. The value for each control region was taken as 100%. Regions III and IV were not assayed. The ChIP assay was repeated twice and the data shown are mean±s.d.

Oncogene p300 is a corepressor of c-Myc N Sankar et al 5722 presence of both a-HA and a-His antibodies, a band in faint super-shifted band was visible on longer exposures the position of C4 was also observed (lane 7). The (shown next to the Figure 4a). Nonetheless, the presence complex C4 appeared only when a-HA antibody was of p300 in the complex was conﬁrmed because the used indicating the presence of p300 in the YY1 complex complex C4 (lanes 5, 7, and 8) appears only when a-HA along with HDAC3. When all three antibodies were antibody is used in the assay. Furthermore, C4 complex included, the C4 band appears with much higher shown in Figure 4a is not the result of an artifact due to intensity (lane 8) when compared to those observed in HA antibody added to the assay, as addition of anti-HA lanes 5 and 7. This is most likely due to higher quantities antibody to the EMSA using extracts containing only of radioactivity super-shifted due to three antibodies in His-HDAC and FLAG-YY1 did not generate C4 the assay. We could not detect a strong super-shifted complex (Figure 4b, compare lane 4 of 4b with lane 7 band using only a-HA antibody perhaps because the of 4a, lane 6 of 4b with lane 4 of 4a and lane 8 of 4b with epitope is burried in the protein (lane 3). However, a lane 8 of 4a). Complexes marked C3 in lanes 6 and 8

a c Lucsh shRNA b α-p300 12 p300 α -CBP Lucsh p300sh α-Actin p300 8

α-YY1 HDAC1 YY1 4 α-Actin HDAC2 Luc act. (fold change) (fold Luc act. HDAC3 α -HDAC3 0 HDAC3 YY1 α-Actin Luc YY1

Actin p300 α-HDAC1 HDAC1 HDAC3 HDAC1 shRNA α-Actin

d e shRNA -p300 -HDAC3 -YY1 Input α α α α IgG

Serum p300 YY1 HDAC3 HDAC1 Luc Lucsh c-Myc p300sh Actin YY1sh

f 6 AcH3 AcH4

Fold change 2

0 II VVII VIII II VVII VIII p300-sh HDAC3sh

Oncogene p300 is a corepressor of c-Myc N Sankar et al 5723 most likely are due to the presence of a portion of the bound to GST-YY1, consistent with a previous report complexes that contain only YY1 and HDAC3. The (Figure 5d; Yao et al., 2001). Similarly, as reported specificity of these bands was further confirmed in EMSA previously, only the C-terminal region of p300 (amino using nuclear extracts prepared from cells transfected with acids (aa) 1572–2370) bound to GST-YY1 (Figure 5d; the expression plasmids described above in various Lee et al., 1995). Direct interactions between HDAC3 combinations as well as incubating the probe with and the C-terminal region of p300 could be demonstrated antibodies alone (data not shown). Together, the above by incubating the GST-p300 with radio-labeled HDAC3 data indicate that much of the DNA–YY1 complex also or GST-HDAC3 with radio-labeled p300C (Figure 5e). contains HDAC3 and p300 and strongly argue that When GST-YY1 beads were incubated with HDAC3 HDAC3 and p300 are associated with YY1 protein that and the C-terminal region of p300 simultaneously, YY1 is bound to the Myc promoter YY1 site. bound to both p300 and HDAC3 indicating a direct interaction of YY1 with p300 and HDAC3 (Figure 5f). To confirm that p300 can bind to YY1 and HDAC3 p300 interactions with YY1 and HDAC3 in vivo and simultaneously, we used a mutant of YY1 in which aa in vitro 170–200 deleted. Published results show that in vivo two Co-immunoprecipitation experiments using cell extracts regions of YY1 (aa 170–200 and 261–333) are involved prepared from quiescent MCF10A cells indicated that in YY1–HDAC3 interactions, whereas in vitro YY1 uses the three proteins form a complex in vivo. For example, only aa 170–200 region for binding to HDAC3 (Yao et al., anti-p300 antibody can co-immunoprecipitate both YY1 2001). Therefore, GST-YY1D170-200 immobilized on and HDAC3 from cell extracts prepared from quiescent GSTbeads should not bind to HDAC3. If both HDAC3 cells (Figure 5a). Similarly, both p300 and HDAC3 were and p300 can be recovered from GST-YY1D170-200 co-immunoprecipitated by an anti-YY1 antibody in- beads, it would indicate that p300 can simultaneously bind dicating that these three proteins are likely to be present to both YY1 and HDAC3. Accordingly, GST-YY1D170- as a complex in vivo. 200 was immobilized on GSTbeads and incubated Further, any two of these three proteins can be co- with radio-labeled HDAC3 and p300C in various immunoprecipitated using extracts prepared from cells combinations. As shown in Figure 5g, beads containing in which expression of the third protein is knocked GST-YY1D170-200 bound to radio-labeled p300C but down using appropriate shRNAs (Figure 5b shows the not HDAC3 (lanes 2 and 4). In contrast, when levels of the three proteins in extracts prepared cells GST-YY1D170-200 was incubated with labeled HDAC3 expressing relevant shRNAs). For example, ablation of and p300C simultaneously, significant amounts of both p300 expression did not affect the formation of YY1– HDAC3 and p300C could be recovered from the beads HDAC3 complex (Figure 5c, Luc-sh refers to control). (lane 6). These results show that at least in vitro, p300 can Similarly, ablation of YY1 expression did not affect bind to both HDAC3 and YY1 simultaneously. formation of the p300–HDAC3 complex and ablation of HDAC3 expression did not affect p300–YY1 complex formation. Taken together, these results suggest that the p300 cooperation with HDAC3 and YY1 in Myc three proteins can form a ternary complex in vivo. repression To further study the protein–protein interactions, we In transient assays p300 and HDAC3 together was more employed glutathione S-transferase (GST) pull-down effective in inhibiting the Myc promoter than either assays. When GST-YY1 was incubated with radio- plasmid alone indicating that p300 and HDAC3 labeled HDAC3, significant quantities of HDAC3 cooperate with YY1 in repressing the Myc promoter

Figure 3 Myc induction and occupancy of p300, HDAC3 and YY1 on the YY1 site of the Myc promoter in cells expressing short hairpin RNAs (shRNAs) targeting p300, YY1 and HDAC3. An Ad vector expressing shRNAs specific for luciferase gene was used as a control in all the experiments shown in this figure. (a) Western blot showing reduced expression of p300, YY1, HDAC1 and HDAC3 in quiescent MCF10A cells expressing appropriate shRNAs. MCF10A cells were serum starved for 36 h, infected with AdshRNA vectors, and 18 h after infection, harvested and the protein samples were subjected to western immunoblotting using antibodies as shown. Actin levels were assayed by reprobing the membranes with an a-b-actin antibody. All left lanes are control samples in which cells were infected with AdLucsh. AdLucsh expresses shRNAs targeting luciferase. Samples prepared from cells infected with Ad vectors expressing relevant shRNAs are shown on the right lanes. Sources of antibodies were described in legend to Figure 2c. (b) Levels of p300, HDACs and YY1 in cells expressing p300 shRNA. Cell extracts prepared from quiescent MCF10A cells infected Ad vector expressing shRNAs targeting luciferase for p300 were analysed in western immunoblots using antibodies described in legend to Figure 2c. (c) Assay of Myc transcriptional activation activity in YY1-, p300-, HDAC1-, HDAC3- and luciferase shRNA-expressing cells. Serum starved MCF10A cells were co-infected with Ad vectors expressing appropriate shRNAs along with the Myc responsive luciferase reporter virus (AdM4). The luciferase activity was measured after 18 h of viral infection. Values (fold change) from three independent sets as compared to that of AdLucsh samples ±s.d. are shown. (d) Western blot showing Myc protein levels in p300, HDAC1, HDAC3 and YY1 depleted MCF10A cells infected with Ad vectors expressing the respective shRNAs as above. Lane showing serum is a positive control where cells were treated with serum for 2 h. Sources of antibodies are as described in legend to Figure 2c. (e) Occupancy of p300 and HDAC3 in quiescent cells expressing p300, and YY1 shRNAs. Serum-starved cells were infected with AdYY1sh or Adp300sh for 16 h and the ChIP assays were performed using antibodies as described in legend to Figure 2d. (f) H3 and H4 acetylation of the Myc promoter and the coding regions in cells expressing shRNAs specific for p300 and HDAC3. Bar diagram showing the real-time PCR quantification of the DNA products generated in ChIP assays. The percent input values for each region in control and shRNA-expressing cells were calculated as above and the fold increase when compared to that of AdLucsh infected cells are shown. Mean values ±s.d. obtained from two independent sets are shown. The primer pairs used here are as those used in Figure 2d.

Oncogene p300 is a corepressor of c-Myc N Sankar et al 5724 α-FLAG -+- -++-+ -+- - (see Supplementary Figure 1 for further details). Similar α-HA --+-+-++ --+- cooperative effect was observed between YY1 and p300 α-His ---+-+++ ---+ in inhibiting the Myc promoter. In contrast to HDAC3, HDAC1 had only a modest effect in this assay, while C4 HDAC2 had no effect at all. Western immunoblotting showed that the three proteins were expressed at similar C3 levels. Thus, consistent with ChIP assay results shown in Figure 2d, only HDAC3 cooperated with p300 at C2 signiﬁcant levels in repressing Myc promoter.

C1 Role of SUMO modification of p300 in Myc repression Others have shown that p300 contains a transcriptional repression domain also known as cell cycle regulatory domain (CRD domain) that is dependent on sumoylation of the two sumoylation sites between aa 1004 and 1045 (Girdwood et al., 2003). Transient transfection assays using a mutant p300 in which the two SUMO YY1 modification sites were deleted (aa 1004–1045) showed complex that this mutant repressed the Myc promoter as 12345678 efficiently as the WTp300 (Figure 6a and b shows FLAGYY1 + HAp300 + HisHDAC3 different p300 domains and the structure of the p300 mutants). Data shown in Figure 5 suggest that both α-FLAG ---- ++++ YY1 and HDACs bind to the C-terminal region of p300 α-His --++ --++ downstream of the sumoylation sites (also see Lee et al., α-HA ---+ -+-+ 1995). These observations prompted us to examine whether the C-terminal half of p300 would be sufficient to repress Myc. Data shown in Figure 6c indicate that the C-terminal half of the protein (aa 1145–2414) C3 repressed the Myc promoter as efficiently as the full- C2 length p300. The N-terminal half of p300 (aa 1-1301) did C1 not show any repressive activity. Western blots showed that the truncated forms of p300 were expressed efficiently in transiently transfected cells (Figure 6d). These results suggest that the Myc repressive activity of p300 resides in the C-terminal half of p300.

Discussion

YY1 In this paper, we showed that p300 cooperates with complex HDAC3 at the Myc promoter YY1-binding site to 12345678 repress the promoter. HDAC3 recruitment to Myc FLAGYY1 + HisHDAC3 chromatin correlated with chromatin deacetylation and reduced Myc expression. These results provide a Figure 4 YY1, p300 and HDAC3 form a complex with Myc YY1 mechanistic explanation for the changes in Myc expres- site in vitro.(a) EMSA showing multiprotein complexes containing YY1, p300 and HDAC3 on the Myc promoter YY1 site in vitro. sion when p300 levels are altered in the cell. The Nuclear extracts were prepared from SaOS2 cells transfected with repressive activity of p300 is not dependent on plasmids expressing FLAG-YY1, HA-p300, and His-HDAC3, and its intrinsic HATactivity (Baluchamy et al., 2003) or 2 mg protein was incubated with a 50end labeled double-stranded the SUMO modification and it is associated with the oligonucleotide containing the YY1 site. For super-shift assays, antibodies specific for the epitopes were added to the assay (see C-terminal half of p300 to which the transcriptional Materials and methods). The gel was run until the YY1–DNA repressors YY1 and HDAC3 bind (Figures 5 and 6). protein complex reached to the bottom of the gel (4 h) to resolve Thus, it is likely that the C-terminal half of p300 serves different high molecular weight complexes, dried and autoradio- as a corepressor and facilitates the recruitment of graphed. A darker exposure of the first four lanes of the gel shown HDAC3 to the YY1 site. in (a) is shown next to (a). Sources of the antibodies were: a-HA (12CA5), a-FLAG (M2; Sigma, St Louis, MO, USA) and The endogenous Myc in quiescent cells is minimally a-His (sc-803; Santa Cruz). (b) EMSA showing that complex C4 active transcriptionally. Thus, to study the mechanism does not appear in the absence of HA-p300 in the extracts. EMSA of Myc repression in its natural context, we over- was carried using nuclear extracts prepared from cells transfected expressed p300 in quiescent cells and then serum with FLAG-YY1 and His-HDAC3 expression plasmids and incubated with YY1-binding site and antibodies as shown. stimulated to transcriptionally induce Myc. Under these conditions Myc is partially repressed (Baluchamy et al., 2003). Therefore, increased occupancy of p300 on the

Oncogene GST ohpoen r eetd(ae6.( 6). (lane detected are proteins both Sbaswr eovda ecie bv.Nt hti ae hr DC a nuae ihGST-YY1 with incubated was HDAC3 where lanes GST-YY1 where in lanes that in Note whereas missing above. is described band as HDAC3 resolved were GSTbeads ( above. incubated as was identified were p300 proteins of bound region the and together HDAC3 ( above. as identified blots Western of p300– Details 2), 4). (panel (panel cells of cells depleted presence depleted p300 HDAC3 the ( in in 2c. showing complex Figure complex assays p300–YY1 to YY1–HDAC3 Co-immunoprecipitation and legend 1), 3), protein. in (panel (panel third as cells cells are the depleted control of YY1 in absence in HDAC3 complex the and HDAC3 in YY1 proteins p300, ( three containing the (shRNAs). complex of RNAs two hairpin any short containing relevant expressing vectors a ofimdb D–AEfloe ihCoasebu tiig(aantson.( shown). not (data case staining each blue in were Coomassie GSTproteins proteins p300 with of bound of Expression the followed region autoradiography. and SDS–PAGE reactions by by incubation followed separate confirmed (PAGE) in was electrophoresis GST-YY1 gel with for SDS–polyacrylamide incubated were by 1572–2370) used identified aa (p300C, protein p300 of fragments the of vitro In 5% contained and lanes buffer input lysis the NP40 and in lysed immunoprecipitation were for cells used MCF10A starved was Serum protein ( p300. of immunoprecipition. and milligrams HDAC3 YY1, Three containing complex using immunoprecipitated a of presence sdi h idn sassoni ( in shown assays binding the in used iue5 Figure HDAC3 GST-p300CT IP:

rnltdHA3o h -emnl(30,aioais(a –9) ide(30M a7417) n h C-terminal the and 744–1571), aa M, (p300 middle 1–596), (aa) acids amino (p300N, N-terminal the or HDAC3 translated α-p300

nvivo In GST

p300C Luc-sh nvitro in nvitro in

nvitro in α GST-HDAC3 -YY1 Input and

HDAC3 p300C α

f -HDAC3 idn fbt 30 n DC oYY1 to HDAC3 and p300C both of Binding ) b S-30 a nuae with incubated was GST-p300C . Sfso rti otiigadlto uato Y nwiha 7–0 eee (GST-YY1 deleted 170–200 aa which in YY1 of mutant deletion a containing protein GSTfusion A . with etr ltsoigdpeino 30 Y n DC nqisetMF0 el netdwt Ad with infected cells MCF10A quiescent in HDAC3 and YY1 p300, of depletion showing blot Western ) nvitro in IgG α-p300 a IP: nvitro in p0 nioy(et or (left) antibody -p300 d neatosbtenY1 30adHA3 ( HDAC3. and p300 YY1, between interactions Spldw sassoigbnigo h -emnlrgo fp0 n DC oYY1 to HDAC3 and p300 of region C-terminal the of binding showing assays GSTpulldown ) α-YY1 p300-sh IgG GST aee DC rp0Cidvdal ri obnto ssonadtepoen on othe to bound proteins the and shown as combination in or individually p300C or HDAC3 labeled c α α α h nu aiatvt.1%o the of 10% radioactivity. Input ) and α-HDAC3 -p300 -HDAC3 -YY1 GST-YY1 d eelae nt h e.Atreetohrssteglwsdidadautoradiographed. and dried was gel the electrophoresis After gel. the to on loaded were )

HDAC3 p300-C IgG

a α-p300 Y1atbd rgt olwdb muolte ihatbde sindicated. as antibodies with immunoblotted by followed (right) antibody -YY1 YY1-sh

nvitro in Input α-HDAC3 D 7-0 a nuae ihlbldp0CadHA3simultaneously, HDAC3 and p300C labeled with incubated was 170-200 123456 c aee DC lf)o ievra(ih)adtebudpoen were proteins bound the and (right) versa vice or (left) HDAC3 labeled oimnpeiiainasy hwn h rsneo complexes of presence the showing assays Co-immunoprecipitation ) GST α HDAC3 -YY1

IgG IP: GST-YY1 nvitro in g

∆ Sankar of N corepressor a is p300

iutnosbnigo Y n DC oteC-terminal the to HDAC3 and YY1 of binding Simultaneous ) ( 170-200) IgG 35 GST HDAC3-sh α

ehoielbldHA3adtetrep0 fragments p300 three the and HDAC3 Methionine-labeled -p300 S-Y a nuae with incubated was GST-YY1 .

p300C α α α

GST-YY1 -YY1 -HDAC3 -p300 tal et

a ∆ α

) ( 170-200) -YY1 nvivo In HDAC3 GST + GST-YY1 IgG p300C

∆ α α α oimnpeiiainasy hwn the showing assays co-immunoprecipitation ( 170-200) HDAC3 -HDAC3 -YY1 -p300 e IB: p300C c-Myc idn fHA3adteC-terminal the and HDAC3 of Binding ) Luc-sh

p300-sh

YY1-sh

nvitro in HDAC3 GST HDAC3-sh Tubulin HDAC3 YY1 p300 D aee 30 and p300C labeled GST-YY1 N 7-0 ln 2), (lane 170-200 HDAC3 C M N p300 Input MC D

170-200) p300 nvitro in . Oncogene 5725 p300 is a corepressor of c-Myc N Sankar et al 5726 c-Myc repression 1 CH1 CH2 CH3 2414 CRD1 Bromo p300-FL + 1 2414 p300- dl1004-1045 + dl1005-1045 1145 2414 p300-CT + 1 CH1 1301 p300-NT - 1011 1029 SUMO

p300 4 4 ) ) 5 4 FL FL CT NT NT CT

250-

2 2 150- Luc activity (x10 Luc activity (x10 -N15 -RW128

0 0 p300FL - + - p300FL- + - - p300 - - + p300NT - - + - p300CT - - - + (dl1005-1045) Figure 6 SUMO modification of p300 is not required for Myc repression and the repressive activity is associated with the C-terminal region of p300. (a) Diagram of full-length p300 showing different domains, a p300 mutant lacking SUMO modification sites and the truncated p300 fragments used in the assay. CH1, 2 and 3 refer to the three cysteine-and histidine-rich domains. Bromo, bromodomain; CRD1, cell cycle regulatory domain 1; p300dl1004-1045, a 40 aa deletion mutant in which the two SUMO modification sites deleted. (b) Transient assay showing Myc repression by p300 is independent of SUMO modification. Rat12 cells were transfected with plasmids-expressing WTp300 or p300dl1004-1045 along with the Myc promoter–reporter construct (del-6) as described in legend to Figure 1, and the luciferase activities in the harvested cells are shown. Mean values from three independent experiments ±s.d. are shown. (c) Transient assay showing the association of the repression activity with the C-terminal half of p300. Rat12 cells were transfected with the N- and the C-terminal halves of p300 along with Myc promoter-luciferase reporter construct as described in legend to Figure 1 and the luciferase activity is presented. Mean values from three independent experiments ±s.d. are shown. (d) Western immunoblot showing the expression of the WT p300 and p300NT and p300CT. FL, full-length p300; NT, N-terminal p300; CT, C-terminal p300; a-N-15 recognizes N-terminal epitope; a-RW128 recognizes C-terminal epitope.

TATA box and the coding regions is consistent with the required for the repression of Myc in quiescent cells recently published results showing that p300 associates (Rajabi et al., 2005). In addition, published reports show with transcriptional initiation and elongation factors that in vivo, these two proteins are present in limiting during gene transcription (Black et al., 2006; Guermah amounts and that they show dose-dependent effects on et al., 2006). Thus, it appears that the corepressor normal cell proliferation and development (Yao et al., function of p300 is extended to the transcription 1998; Kung et al., 2000; Iyer et al., 2007). initiation and coding regions where p300 appears to It is at present unclear what factors contribute to the facilitate the recruitment of HDAC3 to the chromatin. increased Myc chromatin acetylation in p300 depleted We believe that the primary role played by p300 in cells (Figure 3). One possibility is that when the repressing the Myc promoter is the recruitment of repressor complex is removed from the Myc promoter, HDAC3 to the YY1-binding site. This is based on the the chromatin undergoes changes such that it is poised observation that p300 fails to repress a promoter for acetylation by other acetylases such as CBP and construct in which YY1 site is mutated (YY1-M in GCN5/PCAF. Using a recent literature survey, we did Figure 1). not ﬁnd a situation exactly similar to that reported here. Although CBP was present in p300-depleted cells at However, others have shown that the interferon-b normal levels, it did not compensate for p300 in promoter acetylation by GCN5/PCAF when p300 or repressing Myc in p300 knockdown cells (Figure 3). CBP is depleted (Agalioti et al., 2002). This is consistent with our previously published data It is noteworthy that HDAC1 and 2 are not involved showing that normal levels of both p300 and CBP are in p300 mediated repression of Myc. Studies have

Oncogene p300 is a corepressor of c-Myc N Sankar et al 5727 shown that HDACs exist as distinct multi-subunit b-galactosidase (Adb-gal), and Myc responsive reporter complexes with HDAC3 present in one complex cassette (AdM4) were described earlier (Baluchamy et al., while HDAC1 and HDAC 2 are found together in a 2003). Adp300sh is an Ad vector expressing sh RNAs using 0 0 separate complex (Hassig et al., 1998; Zhang et al., DNA sequence 5 -ACCAGATGATGCCTCGAATAA-3 1999). At present is unclear whether p300 is a part of the ( þ 2501 to þ 2521 from the AUG codon of p300) from human U6 promoter (DNA sequence was chosen using HDAC3 containing complex in quiescent cells. www.genescript.com site). The double stranded oligonucleo- In p300 depleted cells, we found reduced occupancy tide containing sh RNA sequence was first cloned into of YY1 on the YY1-binding site as compared to normal pSiren-RetroQ vector (BD biosciences) and then transferred cells (Figure 3e). This may be due to reduced affinity of to Ad vector. Ad vectors expressing shRNAs directed against YY1 to the YY1-binding site in the absence of p300 as it YY1, HDAC1 and HDAC3 were constructed based on has been suggested that YY1 may be present in cells as a shRNA sequences published earlier (Ishizuka and Lazar, 2003; complex with p300 (Lee et al., 1995). It may not be Sui et al., 2004). Del-1 is a human Myc promoter-reporter related to changes in the DNA binding activity of YY1 plasmid containing 2.3 kb upstream promoter sequences due to p300 mediated acetylation (Yao et al., 2001) as (He et al., 1998). Deletion and substitution mutants of this the HATactivity of p300 is not involved in the plasmid were constructed using the Stratagene site directed mutagenesis kit. Other expression plasmids used in this report repression mechanism. were HA-p300, FLAG-YY1, FLAG-HDAC3, His-HDAC3, The repressive properties of p300 that we describe pVR1012p300 and pVR1012DCRD1 (a CMV enhancer driven here are consistent with a corepressor role for p300 in p300 and a mutant in which aa 1004–1045 were deleted, transcriptional repression (Santoso and Kadonaga, Girdwood et al., 2003). 2006). In several other cases the mechanism of repression of gene expression by p300 appears to be indirect Promoter reporter assays, in vitro and in vivo protein-binding (Munshi et al., 1998; Waltzer and Bienz, 1998; Poizat assays, EMSA and ChIP assays et al., 2000; Hong and Chakravarti, 2003; Guidez et al., Procedures for GST-pull down assays and EMSA were as 2005). However, our studies show that p300 plays a described (Jayaraman et al., 1999). For detecting the multi- more direct role at the chromatin level in repressing protein complexes, nuclear extracts were prepared from SaOS2 Myc. Currently we do not know whether p300 also plays cells after transfecting with expression plasmids expressing a role in the YY1 mediated repression of other HA-p300, FLAG-YY1 and His-HDAC3 were incubated with promoters. The c-Fos promoter has been shown to be an YY1 probe and in the presence of appropriate antibodies repressed by YY1 (Murphy et al., 1999). However, our for 20 min. Promoter reporter assays were carried out using rat12 cells. previously published results show that in the context of 5 G1, the c-Fos promoter is not repressed by p300 Briefly, cells (4–6 Â 10 per well) were transfected with Myc promoter–reporter plasmids along with expression plasmids overexpression (Baluchamy et al., 2003). using lipofectamine-2000 (Invitrogen, Carlsbad, CA, USA). p300 may play an essential role in cell cycle Cells were maintained in serum-free media for 48 h, then serum progression by co-activating a number of cell cycle stimulated for 2 h, harvested, and the luciferase activity was related transcription factors (Goodman and Smolik, determined. ChIP assays were carried out essentially as 2000). Thus, it is tempting to speculate a repressor role described (Christova and Oelgeschlager, 2002). Typically, 5% for p300 that is similar to that of Rb. That is, in of the immunoprecipitated sample was used for PCR analysis. quiescent cells, an underphosphorylated form of p300 in The primer pairs used in this study will be available upon association with chromatin repressor complexes may request. The PCR generated products were about 150–300 bp repress gene expression. During G1-S transition, growth in size. The PCR generated products were quantified by factors stimulation may phosphorylate p300 (Yaciuk quantitative real-time PCR using Applied Biosystems 1200HT system. and Moran, 1991) leading to the dissociation of YY1 and HDAC3 from the complex and gaining its acetylase function to activate transcription. Acknowledgements

We are very grateful to Drs E Seto (Mofﬁts Cancer Materials and methods Center), Dr Wu (University of Oregon), Yang Shi (Harvard University), N Perkins and R Hayward (University of Dundee, Cells, Ad vectors, and plasmids UK), and G Chinnadurai (St Louis University) for providing Growth conditions for the human MCF10A (Soule et al., various plasmids and viruses. We also thank Dr Kathy 1990) and rat12 cells were described earlier (Baluchamy et al., Rundell for helpful discussion and editing this article. This 2003) The Ad vectors expressing FLAG-p300 (Adp300), work was supported by the NIH Grant CA73303.

References

Agalioti T, Chen G, Thanos D. (2002). Deciphering the transcriptional conditionally overexpressing p300 that is not dependent on its histone acetylation code for a human gene. Cell 111: 381–392. histone acetyltransferase activity. Proc Natl Acad Sci USA 100: Albert T, Wells J, Funk JO, Pullner A, Raschke EE, Stelzer G et al. 9524–9529. (2001). The chromatin structure of the dual c-myc promoter P1/P2 is Baluchamy S, Sankar N, Navaraj A, Moran E, Thimmapaya B. regulated by separate elements. J Biol Chem 276: 20482–20490. (2007). Relationship between E1A binding to cellular Baluchamy S, Rajabi HN, Thimmapaya R, Navaraj A, Thimmapaya B. proteins, c-myc activation and S-phase induction. Oncogene 26: (2003). Repression of c-Myc and inhibition of G1 exit in cells 781–787.

Oncogene p300 is a corepressor of c-Myc N Sankar et al 5728 Black JC, Choi JE, Lombardo SR, Carey M. (2006). A mechanism for premature G1 exit and up-regulation of c-MYC. Proc Natl Acad Sci coordinating chromatin modification and preinitiation complex USA 98: 4646–4651. assembly. Mol Cell 23: 809–818. Kung KL, Bronson RT, Chang L, Sieff CA, Livingston DM, Yao T. Christova R, Oelgeschlager T. (2002). Association of human TFIID- (2000). Gene dose-dependent control of hematopoiesis and hema- promoter complexes with silenced mitotic chromatin in vivo. Nat tologic tumor suppression by CBP. Genes Dev 14: 272–277. Cell Biol 4: 79–82. Lee JS, Galvin KM, See RH, Eckner R, Livingston D, Moran E et al. Dang CV. (1999). c-Myc target genes involved in cell growth, (1995). Relief of YY1 transcriptional repression by adenovirus E1A apoptosis, and metabolism. Mol Cell Biol 19: 1–11. is mediated by E1A-associated protein p300. Genes Dev 9: Filippova GN, Fagerlie S, Klenova EM, Myers C, Dehner Y, 1188–1198. Goodwin G et al. (1996). An exceptionally conserved transcriptional Munshi N, Merika M, Yie J, Senger K, Chen G, Thanos D. (1998). repressor, CTCF, employs different combinations of zinc fingers to Acetylation of HMG I(Y) by CBP turns of IFNbeta expression by bind diverged promoter sequences of avian and mammalian c-Myc disrupting the enhanceosome. Cell 2: 457–467. oncogenes. Mol Cell Biol 16: 2802–2813. Murphy DJ, Hardy S, Engel DA. (1999). Human SWI-SNF Frisch SM, Mymryk JS. (2002). Adenovirus-5 E1A: paradox and component BRG1 represses transcription of the c-fos gene. Mol paradigm. Nat Rev Mol Cell Biol 6: 441–4452. Cell Biol 19: 2724–2733. Girdwood D, Bumpass D, Vaughan OA, Thain A, Anderson LA, Nakajima T, Uchida C, Anderson SF, Lee CG, Hurwitz J, Parvin JD Snowden AW et al. (2003). p300 transcriptional repression is et al. (1997). RNA helicase A mediates association of CBP with mediated by SUMO modification. Molecular Cell 11: 1043–1054. RNA polymerase II. Cell 90: 1107–1112. Goodman RH, Smolik S. (2000). CBP/p300 in cell growth, transfor- Poizat C, Sartorelli V, Chung G, Kloner RA, Kedes L. (2000). mation, and development. Genes Dev 14: 1553–1577. Proteasome-mediated degradation of the coactivator p300 impairs Guermah M, Palhan VB, Tackett AJ, Chait BT, Roeder RG. (2006). cardiac transcription. Mol Cell Biol 20: 8643–8654. Synergistic functions of SII and p300 in productive activator- Rajabi HN, Baluchamy S, Kolli S, Nag A, Srinivas R, Raychaudhuri P dependent transcription of chromatin templates. Cell 125: 275–286. et al. (2005). Effects of depletion of CREB-binding protein on c-Myc Guidez F, Howell L, Isalan M, Cebrat M, Alani RM, Ivins S et al. regulation and cell cycle G1-S transition. J Biol Chem 280: 361–374. (2005). Histone acetyltransferase activity of p300 is required for Santoso B, Kadonaga JT. (2006). Reconstitution of chromatin transcriptional repression by the promyelocytic leukemia zinc finger transcription with purified components reveals a chromatin-specific protein. Mol Cell Biol 25: 5552–5566. repressive activity of p300. Nat Struc Mol Biol 2: 131–139. Hassig CA, Tong JK, Fleischer TC, Owa T, Grable PG, Ayer DE et al. Schwartz YB, Pirrotta V. (2007). Polycomb silencing mechanisms and (1998). A role for histone deacetylase activity in HDAC1-mediated the management of genomic programmes. Nat Rev Genet 8: 9–22. transcritpional repression. Proc Natl Acad Sci USA 95: Soule HD, Maloney TN, Wolman SR, Peterson WDJ, Brenz R, 3519–3524. McGrath CM et al. (1990). Isolation and characterization of a He TC, Sparks AB, Rago C, Hermeking H, Zavel L, DaCosta LT et al. spontaneously immortalized human breast epithelial cell line (1998). Identification of c-Myc as a target of the APC pathway. MCF-10. Cancer Res 50: 6075–6086. Science 281: 1509–1512. Sui G, Affar el B, Shi Y, Brignone C, Wall NR, Yin P et al. (2004). Hong R, Chakravarti D. (2003). The human proliferating cell nuclear Yin Yang 1 is a negative regulator of p53. Cell 117: 859–872. antigen regulates transcriptional coactivator p300 activity and Waltzer L, Bienz M. (1998). Drosophila CBP represses the transcrip- promotes transcriptional repression. J Biol Chem 278: 44505–44513. tion factor TCF to antagonize wingless signalling. Nature 395: Ishizuka T, Lazar MA. (2003). The N-CoR/histone deacetylase 3 521–525. complex is required for repression by thyroid hormone receptor. Yaciuk P, Moran E. (1991). Analysis with specific polyclonal Mol Cell Biol 15: 5122–5131. antiserum indicates that the E1A-associated 300-kDa product is a Iyer NG, Ozdag H, Caldas C. (2004). p300/CBP and cancer. Oncogene stable nuclear phosphoprotein that undergoes cell cycle phase- 23: 4225–4231. specific modification. Mol Cell Biol 11: 5389–5397. Iyer NG, Xian J, Chin SF, Bannister AJ, Daigo Y, Aparicio S et al. Yao TP, Oh SP, Fuchs M, Zhou ND, Ch’ng LE, Newsome D et al. (2007). p300 is required for orderly G1/S transition in human cancer (1998). Gene dosage-dependent embryonic development and pro- cells. Oncogene 26: 21–29. liferation defects in mice lacking the transcriptional integrator p300. Izzo WW, Strachan GD, Stubbs MC, Hall DJ. (1999). Transcriptional Cell 93: 361–372. repression from the c-Myc P2 promoter by the zinc finger protein Yao YL, Yang WM, Seto E. (2001). Regulation of transcription factor ZF87/MAZ. J Biol Chem 274: 19498–19506. YY1 by acetylation and deacetylation. Mol Cell Biol 21: 5979–5991. Jayaraman G, Srinivas R, Duggan C, Ferreira E, Swaminathan S, Zhang Y, Ng HH, Erdjument-Bromage H, Tempst P, Bird A, Somasundaram K et al. (1999). p300/cAMP-responsive element- Reinberg D. (1999). Analysis of the NuRD subunits reveals a binding protein interactions with ets-1 and ets-2 in the transcrip- histone deacetylase core complex and a connection with DNA tional activation of the human stromelysin promoter. J Biol Chem methylation. Genes Dev 13: 1924–1935. 274: 17342–17352. Zou X, Lin Y, Rudchenko S, Calame K. (1997). Positive and negative Kolli S, Buchmann AM, Williams J, Weitzman S, Thimmapaya B. regulation of c-Myc transcription. Curr Top Microbiol Immunol 224: (2001). Antisense-mediated depletion of p300 in human cells leads to 57–66.

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

Oncogene