Histone Acetylation by CBP and P300 at Double-Strand Break Sites Facilitates SWI/SNF Chromatin Remodeling and the Recruitment of Non-Homologous End Joining Factors

Oncogene (2011) 30, 2135–2146 & 2011 Macmillan Publishers Limited All rights reserved 0950-9232/11 www.nature.com/onc ORIGINAL ARTICLE Histone acetylation by CBP and p300 at double-strand break sites facilitates SWI/SNF chromatin remodeling and the recruitment of non-homologous end joining factors

H Ogiwara1,AUi1,2,3, A Otsuka1,2, H Satoh4, I Yokomi4,5, S Nakajima3, A Yasui3, J Yokota2 and T Kohno1

1Division of Genome Biology, National Cancer Center Research Institute, Chuo-ku, Tokyo, Japan; 2Division of Multistep Carcinogenesis, National Cancer Center Research Institute, Tokyo, Japan; 3Department of Molecular Genetics, Institute of Development, Division of Dynamic Proteome, Aging and Cancer, Tohoku University, Sendai, Japan; 4Laboratory of Tumor Cell Biology, Department of Medical Genome Sciences, Graduate School of Frontier Sciences, The University of Tokyo, Tokyo, Japan and 5Department of Pharmacology, School of Medicine, St Marianna University, Kanagawa, Japan

Non-homologous end joining (NHEJ) is a major repair Introduction pathway for DNA double-strand breaks (DSBs) generated by ionizing radiation (IR) and anti-cancer drugs. There- Non-homologous end joining (NHEJ) is a repair path- fore, inhibiting the activity of proteins involved in this way for DNA double-strand breaks (DSBs) that joins pathway is a promising way of sensitizing cancer cells two broken DNA ends without the need for sequence to both radiotherapy and chemotherapy. In this study, we homology. DSBs caused by ionizing radiation (IR) are developed an assay for evaluating NHEJ activity against preferentially repaired through NHEJ (Burma et al., DSBs in chromosomal DNA in human cells to identify the 2006; Lieber, 2008). Therefore, suppression of NHEJ chromatin modification/remodeling proteins involved in is a promising therapy that may sensitize cancer cells to NHEJ. We showed that ablating the activity of the homo- IR (Helleday et al., 2008). In addition, DSBs caused by logous histone acetyltransferases, CBP and p300, using anticancer drugs such as the topoisomerase II inhibitor, inhibitors or small interfering RNAs-suppressed NHEJ. etoposide, may also be repaired through NHEJ (Adachi Ablation of CBP or p300 impaired IR-induced DSB repair et al., 2003); therefore, suppression of NHEJ is also and sensitized lung cancer cells to IR and the anti-cancer a promising approach to sensitizing cancer cells to these drug, etoposide, which induces DSBs that are repaired by drugs (Zhao et al., 2006; Helleday et al., 2008). In vitro NHEJ. The CBP/p300 proteins were recruited to sites of studies have elucidated the molecular processes under- DSBs and their ablation suppressed acetylation of lysine 18 lying NHEJ of DNA ends, in which KU70/80 and within histone H3, and lysines 5, 8, 12, and 16 within DNA-dependent protein kinases (DNA-PKcs) function histone H4, at the DSB sites. This then suppressed the as core proteins (Burma et al., 2006; Lieber, 2008). recruitment of KU70 and KU80, both key proteins for Chemicals that inhibit DNA-PKcs activity and sensitize NHEJ, to the DSB sites. Ablation of CBP/p300 also cancer cells to IR and/or etoposide are awaiting eval- impaired the recruitment of BRM, a catalytic subunit of uation for clinical application (Helleday et al., 2008). the SWI/SNF complex involved in chromatin remodeling Thus, identification of novel proteins involved in NHEJ at DSB sites. These results indicate that CBP and p300 will help to develop methods for sensitizing cancer cells function as histone H3 and H4 acetyltransferases at DSB to both radiotherapy and chemotherapy. sites in NHEJ and facilitate chromatin relaxation. There- Chromosomal DNA and histones form a highly fore, inhibition CBP and p300 activity may sensitize cancer condensed structure known as chromatin. The accessi- cells to radiotherapy and chemotherapy. bility of proteins to chromosomal DNA in vivo is Oncogene (2011) 30, 2135–2146; doi:10.1038/onc.2010.592; required for a variety of intracellular processes, includ- published online 10 January 2011 ing DSB repair, and is regulated by two general chromatin remodeling mechanisms, histone modifica- Keywords: non-homologous end joining; DNA double- tion and alteration of the nucleosomal position (Osley strand break; DNA repair; chromatin; histone acetyl- and Shen, 2006). Acetylation of histones located at transferase DSB sites by histone acetyltransferases (HATs) is a critical chromatin modification required for DSB repair. In particular, the N-terminal lysine residues of histones H3 and H4 are acetylated during DSBs (Bird et al., Correspondence: Dr T Kohno, Division of Genome Biology, National 2002; Tamburini and Tyler, 2005). Consistent with this, Cancer Center Research Institute, 5-1-1, Tsukiji, Chuo-ku, Tokyo NuA4-Tip60 HAT acetylates histone H4 at DSB sites 104-0045, Japan. E-mail: [email protected] and facilitates DSB repair (Bird et al., 2002; Murr et al., Received 15 July 2010; revised 1 December 2010; accepted 3 December 2006). In addition to HATs, the SWI2/SNF2 super- 2010; published online 10 January 2011 family (INO80, SWR1, SWI/SNF and RSC) chromatin Role of CBP/p300 in non-homologous end joining H Ogiwara et al 2136 remodeling complexes are recruited to DSB sites and production of a transcript that enables translation of facilitate alterations in the nucleosomal position. enhanced green fluorescent protein (EGFP) instead Furthermore, the INO80, SWR1 and RSC complexes of the herpes simplex virus-thymidine kinase protein. are required for the recruitment of KU70/80 to DSB Therefore, the efficiency of NHEJ can be assessed sites (Shim et al., 2005; van Attikum et al., 2007). It is by monitoring EGFP production. In addition, the also suggested that the SWI/SNF complex is recruited DSBs produced by I-SceI and NHEJ of the two broken to DSB sites, where it interacts with acetylated DNA strands can be monitored by quantitative PCR histone proteins and g-H2AX, a phosphorylated (As shown in Figure 1a, ‘the proportion of uncut DNA’ H2AX protein generated upon DSBs (Lee et al., 2010). is used to assess the efficiency of DSB generation by Taken together, these studies suggest that histone I-SceI and ‘the proportion of joined DNA’ is used to acetylation facilitates chromatin remodeling and that assess the ligation efficiency of the DNA ends). the resulting ‘relaxed’ chromatin enables the recruitment We isolated two individual H1299 human lung cancer of DNA repair proteins to DSB sites. However, the cell clones, H1299dA3-1 #1 and #2, that stably carried HATs and chromatin remodeling complexes involved the IRES-TK-EGFP DNA within their genome in NHEJ have not been fully understood. In addition, (Supplementary Figure S1A and S1B). Fluorescence- the effect of inhibiting these proteins on the radio- activated cell sorting analysis showed that the propor- sensitivity and chemosensitivity of cancer cells remains tion of cells expressing EGFP increased 24–72 h after largely unknown. transfection of the I-SceI expression plasmid in both In this study, we developed an assay system to clones (Figure 1b and Supplementary Figure S1C). evaluate NHEJ activity after DSBs in chromosomal Quantitative PCR analysis revealed that the proportion DNA in human cells. Using this system, we found that of joined DNA increased 12–48 h post transfection, two HAT inhibitors, anacardic acid and curcumin, both whereas that of uncut DNA decreased during the 24 h known to radio-sensitize cancer cells (Khafif et al., 2005; post transfection (Figures 1c and d). Therefore, DSBs Sun et al., 2006; Javvadi et al., 2008), suppress NHEJ were introduced by I-SceI, and the subsequent joining of DSBs. The CBP and p300 proteins are homologous of the two broken DNA ends occurred in both the HATs, known to function as transcription co-activators, H1299dA3-1 #1 and #2 clones in vivo. Nucleotide and their enzymatic activity is inhibited by anacardic sequencing of the joined DNA revealed that the ligation acid and curcumin (Karamouzis et al., 2007). We required no (or very little) sequence homology between showed that CBP and p300 function as histone H3 the DNA ends (Supplementary Figure S1D), indicating and H4 acetyltransferases at DSB sites during NHEJ, that the DNA ends were joined via NHEJ. and facilitate the recruitment of KU70/80 proteins in We next examined the effect of inhibiting DNA-PKcs cooperation with the SWI/SNF chromatin remodeling activity on the joining of DNA ends using this assay complex. Ablation of CBP and p300 caused radio- system. In both the H1299dA3-1 #1 and #2 clones, the sensitization of lung cancer cells and sensitized the cells proportion of both EGFP-positive cells and joined to etoposide. These results indicate that the CBP and DNA decreased after treatment with the DNA-PKcs p300 proteins are involved in NHEJ in vivo and act as inhibitor, NU7026, at a concentration wherein phos- histone H3 and H4 acetyltransferases, facilitating phorylation of the serine 2056 residue of the DNA-PKcs chromatin relaxation. Inhibition of CBP and p300 protein (activated by autophosphorylation) by IR is activity may, therefore, be a way to sensitize cancer suppressed. This occurred without any increase in the cells to radiotherapy and chemotherapy. proportion of uncut DNA (Figures 1e–g and Supple- mentary Figure S1E). The proportion of EGFP-positive cells was also decreased by small interfering RNA Results (si)RNA-mediated ablation of DNA-PKcs (Figure 1h). These results confirmed that the two DNA ends Development of an NHEJ assay system for DSBs produced by I-SceI digestion were joined through in chromosomal DNA NHEJ, and indicated that a reduction in NHEJ activity The assay system for evaluating NHEJ activity against by treatment with inhibitors and siRNAs directed DSBs in chromosomal DNA in living human cells is against other proteins involved in NHEJ should also presented in Figure 1a. The IRES-TK-EGFP plasmid, be detectable by this assay. which contains two recognition sites for I-SceI endonuclease (Jasin, 1996) in the reverse direction, was integrated into the chromosomal DNA of human cells Suppression of NHEJ in vivo by ablation of CBP or p300 as a substrate for DSBs and subsequent NHEJ repair. To identify the HATs involved in NHEJ, we first exa- As there is no I-SceI site in the human genome, the mined whether the NHEJ in this assay could be suppressed I-SceI protein transiently expressed by transfection of by HAT inhibitors. The HAT inhibitors examined were the I-SceI expression plasmid specifically cleaves two anacardic acid (from cashew nuts), which inhibits PCAF, I-SceI sites in the substrate and produces DSBs with p300 and TIP60 (Balasubramanyam et al., 2003, 2004; incompatible ends. This results in DSBs in chromosomal Sun et al., 2006), and curcumin (from turmeric), which DNA in vivo. NHEJ of the two broken DNA strands inhibits CBP and p300 (Balasubramanyam et al.,2004). results in the deletion of the herpes simplex virus- These HAT inhibitors are known to radio-sensi- thymidine kinase open reading frame and leads to the tize cancer cells (Khafif et al.,2005;Sunet al.,2006;

Oncogene Role of CBP/p300 in non-homologous end joining H Ogiwara et al 2137 12 I-SceI Uncut DNA #1 10 #2 CMV TK EGFP IRES Puror pA I-SceI 8 6 DSB induction by I-SceI & NHEJ 4

TK 2

r EGFP positive cells (%) CMV EGFP IRES Puro pA 0 Joined DNA 0 122436486072 Time after I-SceI transfection (hr)

1.2 #1 1.2 #1 #2 1 1 #2 0.8 0.8 0.6 0.6 0.4 0.4 0.2 0.2 Proportion of uncut DNA Proportion of joined DNA 0 0 0 122436486072 0 122436486072 Time after I-SceI transfection (hr) Time after I-SceI transfection (hr)

Proportion of EGFP-positive cells 0 0.2 0.4 0.6 0.8 1 1.2 #1 DMSO #2 NU7026

Proportion of joined DNA Proportion of uncut DNA 0 0.2 0.4 0.6 0.8 1 1.2 0 0.2 0.4 0.6 0.8 1 1.2

DMSO DMSO #1 #1 #2 NU7026 #2 NU7026

Proportion of EGFP-positive cells 0 0.2 0.4 0.6 0.8 1 1.2

siCTR #1 siCTR siDNA-PKcs DNA-PKcs siDNAPKcs #2 α Tubulin Figure 1 NHEJ assay system for detecting chromosomal DSBs in vivo.(a) Assay design. Two I-SceI sites in the reverse direction are indicated by yellow arrow heads. The locations of the PCR primers used for quantitative PCR to monitor DSB introduction by I-SceI (uncut DNA) and subsequent joining (joined DNA) are indicated by purple and red arrows, respectively. CMV, cytomegalovirus promoter/enhancer; IRES, internal ribosome entry site; pA, polyA signal. The structure of the pIRES-TK-GFP plasmid is shown in Supplementary Figure S1A. (b–d) Assessment of DSB generation and joining of DNA ends. Proportion of EGFP-positive cells assessed by ﬂuorescence-activated cell sorting analysis, 48 h after transfection of the I-SceI expression plasmid (b). Proportion of joined (c) and uncut (d) DNA as assessed by quantitative PCR. The proportion of joined DNA is expressed as a ratio versus the amount of joined DNA, 48 h after transfection of the I-SceI expression plasmid. The proportion of uncut DNA is expressed as a ratio versus the amount of uncut DNA present before I-SceI expression plasmid transfection. (e–g) Suppression of joining by inhibition of DNA-PKcs activity. The results 48 h after I-SceI expression plasmid transfection are shown. The proportion of EGFP-positive cells (e), joined (f) and uncut (g) DNA after DNA-PKcs inhibitor (NU7026, 50 mM) treatment is expressed as a ratio versus those seen after dimethyl sulfoxide treatment. In H1299dA3-1 #1 cells, 50 mM of NU7026 inhibited the phosphorylation of DNA-PKcs but not that of ATM or NBS1 (Supplementary Figure S1E). (h) Suppression of joining by inhibition of siRNA-mediated DNA-PKcs ablation. Results obtained 48 h after the transfection of the I-SceI expression plasmid are shown. The proportion of EGFP-positive cells treated with siRNA targeted to DNA-PKcs is expressed as a ratio versus those treated with non-targeting siRNA (siCTR). The results from immunoblot analysis are shown on the right.

Javvadi et al., 2008). As with NU7026, treatment with any signiﬁcant changes in the propor- anacardic acid and curcumin sensitized lung cancer tion of uncut DNA (Figures 2c and d). These results cells to IR (Figure 2a) and led to a reduction in the indicated that HAT inhibitors were able to suppress proportion of EGFP-positive cells in the NHEJ assay NHEJ, and that this activity contributes to the radio- (Figure 2b, Supplementary Figure S2). Quantitative PCR sensitization of cancer cells. analysis revealed that the proportion of joined DNA was We next examined the effect of siRNA-mediated reduced by both anacardic acid and curcumin, without ablation of PCAF, TIP60, CBP and p300 on NHEJ

Oncogene Role of CBP/p300 in non-homologous end joining H Ogiwara et al 2138 Survival Proportion of EGFP positive cells 0 20 40 60 80 100 120 140 0 0.2 0.4 0.6 0.8 1 1.2 DMSO DMSO NU7026 Anacardic acid Anacardic acid IR - Curcumin Curcumin IR +

Proportion of joined DNA Proportion of uncut DNA 0 0.2 0.4 0.6 0.8 1 1.2 0 0.2 0.4 0.6 0.8 1 1.2 DMSO DMSO Anacrdic acid Anacrdic acid Curcumin Curcumin

Proportion of EGFP-positive cells 0 0.2 0.4 0.6 0.8 1 1.2 1.4 siCTRsiPCAF siCTRsiTIP60 siCTR PCAF TIP60 siPCAF α Tubulin α Tubulin siTIP60 siCBP sip300 siCTRsiCBP siCTRsip300 CBP p300 α Tubulin α Tubulin

Proportion of EGFP-positive cells 0 0.2 0.4 0.6 0.8 1 1.2 siCTR

siCBP#1 siCTR siP300 #1 siP300 #2 siCTR siCBP #1 siCBP #2 siCBP#2 CBP p300 siP300#1 α Tubulin α Tubulin siP300#2

Proportion of joined DNA Proportion of uncut DNA 0 0.2 0.4 0.6 0.8 1 1.2 0 0.2 0.4 0.6 0.8 1 1.2

siCTR siCTR siCBP#1 siCBP#1 siCBP#2 siCBP#2 siP300#1 siP300#1 siP300#2 siP300#2

Figure 2 Involvement of CBP and p300 in NHEJ. (a) Survival of HAT inhibitor-treated cells after IR. H1299 cells pretreated with dimethyl sulfoxide (DMSO), a DNA-PKcs inhibitor (NU7026: 50 mM), or HAT inhibitors (anacardic acid: 50 mM, or curcumin: 20 mM) for 1 h were irradiated at 5 Gy (control cells were not irradiated) and subjected to a clonogenic assay. The proportion of colonies derived from cells after IR and/or inhibitor treatment is expressed as a ratio versus cells not subjected to IR or inhibitor treatment. (b–d) Suppression of joining by inhibition of HAT activity. Results observed 48 h after transfection of the I-SceI expression plasmid are shown. (b) The proportion of EGFP-positive cells, (c) joined and (d) uncut DNA after HAT inhibitor (anacardic acid: 50 mM, or curcumin: 20 mM) treatment is expressed as a ratio versus that after DMSO treatment. (e) Suppression of NHEJ by ablation of HAT genes. The proportion of EGFP-positive cells 48 h after I-SceI expression plasmid transfection is shown. Mean values from three independent experiments are shown±s.d. The results from immunoblot analysis are shown on the right. The a-tubulin was used as a loading control. (f–h) Suppression of DNA joining by siRNA-mediated ablation of CBP and p300. Results observed 48 h after transfection of the I-SceI expression plasmid are shown. (f) The proportion of EGFP-positive cells, (g) joined and (h) uncut DNA. The proportion of cells observed after targeting siRNA (siCBP #1, #2 and sip300 #1, #2) treatment is expressed as a ratio versus that after non-targeting siRNA (siCTR) treatment. Mean values from three independent experiments are shown±s.d. The results from immunoblot analysis are shown on the right (f). The a-tubulin was used as a loading control.

(Figure 2e). Ablation of TIP60, a HAT thought to be joined DNA was reduced after ablation of CBP and involved in NHEJ (Robert et al., 2006; Sun et al., 2006), p300, without any significant change in the proportion led to a reduction in the proportion of EGFP-positive of uncut DNA (Figures 2g and h). This strongly suggests cells. Interestingly, ablation of CBP and p300 also the involvement of CBP, p300 and TIP60, but not caused a reduction in the proportion of EGFP-positive PCAF, in NHEJ. cells, whereas ablation of PCAF did not (Figure 2e). To further confirm these results, we examined the Similar reductions in the proportion of EGFP-positive effect of ablating CBP and p300 on the repair of cells after ablation of CBP and p300 were confirmed IR-induced DSBs using the neutral comet assay, which using other siRNAs (Figure 2f). The proportion of specifically detects DSBs. The repair of IR-induced

Oncogene Role of CBP/p300 in non-homologous end joining H Ogiwara et al 2139 Time after IR irradiation 0 min 15 min 60 min 120 min 180 min 35 siCTR 30 siCTR siCBP 25 sip300 20 15

siCBP Tail moment 10 5 0 0 15 60 120 180 Time after IR irradiation (min) sip300

100 100 120 siCTR 100 siCBP #1 10 10 80 siCBP #2 60 siP300 #1 siCTR 1 siCTR 1 40

siP300 #2 Survival (%) Survival (%) siCBP

Survival (%) siCBP 20 sip300 siP300 0.1 0.1 0 0 2.5 5 7.5 10 0 2.5 5 7.5 10 0 5 10 15 20 25 IR (Gy) IR (Gy) Etoposide (uM) Figure 3 Radiosensitization and chemosensitization of lung cancer cells by CBP and p300 ablation. (a) Defects in the repair of IR-mediated DSBs after CBP or p300 ablation. H1299 cells transfected with non-targeting siRNA (siCTR) or CBP/p300-targeting siRNA (siCBP, sip300) were exposed to 20 Gy IR and subjected to neutral comet analyses at the indicated time points. (Left) Representative comet images. (Right) Time course of DSB repair assessed by neutral comet assay. The mean values of the average Comet tail moment (percentage of DNA in tail Â tail length) for the three fields are shown±s.d. (b, c) Sensitivity to IR after CBP or p300 ablation. H1299 cells pre-transfected with siCTR, siCBP or sip300 were irradiated at 5 or 10 Gy (b). H1299 cells pre-transfected with non-targeting siRNA (siCTR) or CBP/p300-targeting siRNA (siCBP #1, #2, and sip300 #1, #2) were irradiated at 2.5, 5 or 10 Gy (c). (d) Sensitivity to etoposide after CBP or p300 ablation. H1299 cells pre-transfected with siCTR, siCBP or sip300 were treated with etoposide (10 or 25 mM) for 1 h. The number of colonies derived from treated cells is expressed as a ratio versus those derived from cells without IR (b, c) or etoposide (d) treatment. DSBs was suppressed in cells in which CBP and p300 enrichment of gH2AX, a phosphorylated form of were ablated (Figure 3a). This ablation also sensitized H2AX generated by DSBs. This was observed not only H1299 lung cancer cells to IR (Figures 3b and c). in the 0.3 kb region, but also in the 1.2 and 2.5 kb Consistently, ablation of CBP or p300 also sensitized the regions, consistent with a previous finding that gH2AX cells to etoposide, which causes DSBs that are preferen- lies B10 kb from the DNA ends (Bonner et al., 2008). tially repaired through NHEJ (Adachi et al., 2003) However, enrichment of CBP and p300 proteins was (Figure 3d). These results indicated that the CBP and only observed in the 0.3 kb region, and not in the 1.2 p300 proteins were involved in the NHEJ of DSBs and 2.5 kb regions, after I-SceI digestion. induced by IR and anti-cancer drugs, as well as those The CBP and p300 proteins in laser micro-irradiated induced by the I-SceI restriction enzyme. regions were co-localized with gH2AX and 53BP1, both markers for DSB sites (Figure 4b). This was confirmed using a different set of antibodies (Figure 4c). In Recruitment of CBP and p300 to DSB sites addition, exogenously expressed mouse Cbp protein Both CBP and p300 function as HATs during transcrip- tagged with yellow fluorescent protein (YFP-Cbp) tional regulation (Karamouzis et al., 2007). Ablation also accumulated in laser micro-irradiated regions of CBP/p300 did not significantly affect the expres- (Figure 4d). These results strongly indicated that CBP sion levels of NHEJ core proteins (Supplementary and p300 were recruited to the sites of DSBs in vivo. Figure S3), indicating that the suppression of NHEJ through ablation of CBP or p300 was caused by mechanisms other than the downregulation of proteins Impairment of DSB-induced histone H3/H4 acetylation engaged in NHEJ. Therefore, we next examined whether and KU recruitment by ablation of CBP or p300 the CBP and p300 proteins were recruited to sites of It is thought that CBP and p300 have HAT activity DSBs in chromosomal DNA in the same manner as against histones H3 and H4, as p300 acetylates lysines other proteins involved in NHEJ (Mari et al., 2006). 14 and 18 of histone H3, and lysines 5 and 8 of histone Recruitment of CBP and p300 to DSB sites generated H4, in vitro (Schiltz et al., 1999; Park et al., 2003; by I-SceI digestion was examined using chromatin Kouzarides, 2007). In addition, lysine 18 within histone immunoprecipitation (ChIP) analysis of DSBs in H3 is acetylated by CBP and p300 in vivo (Horwitz et al., H1299dA3-1 #1 cells (Figure 4a). Quantitative PCR 2008). Acetylation of histone H3 and H4 at DSB sites analysis of the 0.3, 1.2 and 2.5 kb regions downstream of has an important role in DSB repair (Bird et al., 2002; the second I-SceI site in the pIRES-TK-EGFP plasmid Tamburini and Tyler, 2005; Park et al., 2006; Robert sequence immunoprecipitated with control immuno- et al., 2006; Sun et al., 2006; Pandita and Richardson, globulin G or anti-gH2AX antibodies showed an 2009). In particular, acetylation of lysines 5, 8, 12 and

Oncogene Role of CBP/p300 in non-homologous end joining H Ogiwara et al 2140 2.5kb 1.2kb 0.3kb

EGFP IRES puror pA I-SceI

IgG γH2AX 10 10 0hr 0hr 8 8 12hr 12hr 6 6 18hr 18hr 4 4 2 2 ChIP enrichment ChIP enrichment 0 0 0 0.5 1 1.5 2 2.5 0 0.5 1 1.5 2 2.5 Distance from I-SceI site (kb) Distance from I-SceI site (kb)

CBP p300 5 5 0hr 0hr 4 4 12hr 12hr 3 3 18hr 18hr 2 2 1 1 ChIP enrichment ChIP enrichment 0 0 0 0.5 1 1.5 2 2.5 0 0.5 1 1.5 2 2.5 Distance from I-SceI site (kb) Distance from I-SceI site (kb)

CBP γH2AX CBP 53BP1 0 min

p300 γH2AX p300 53BP1 YFP-Cbp 5 min

Figure 4 Recruitment of CBP/p300 proteins to DSB sites. (a) The locations of the primers used for quantitative PCR are shown at the top, along with information on the distances between the primers and the I-SceI sites. ChIP with a control immunoglobulin G or antibodies against gH2AX, CBP or p300 was performed 0, 12 and 18 h after transfection of the I-SceI expression plasmid into H1299dA3-1 #1 cells. The relative enrichment of proteins after 12 and 18 h (compared with 0 h) is shown. (b, c) Accumulation of endogenous CBP and p300 proteins at micro-irradiated regions. H1299dA3-1 #1 cells ﬁxed 5 min after laser micro-irradiation were immunostained with anti-CBP (b, sc-369; c, sc7300) or anti-p300 (b, sc584; c, sc48348), anti-gH2AX or -53BP1 antibodies, and 40,60-diamidino-2-phenylindole. Arrows indicate laser micro-irradiation regions. (d) Accumulation of yellow ﬂuorescent protein (YFP)- Cbp at micro-irradiated regions. H1299 cells transfected with the YFP-Cbp expression plasmid were examined 0 or 5 min after laser micro-irradiation. Arrows indicate the micro-irradiated regions.

16 within histone H4 (H4 K5/K8/K12/K16) is critical K16 and H3 K18 within the 0.3 kb region (designated as for NHEJ (Bird et al., 2002), whereas lysine 18 within the 300D locus) from the I-SceI site (Figures 5a and b) histone H3 (H3 K18) is also acetylated in DSBs without signiﬁcantly affecting their acetylation within (Tamburini and Tyler, 2005). However, whether CBP the GAPDH region located on a chromosome different and p300 function as histone H3 and H4 HATs during from that containing the I-SceI locus (Supplementary DSB repair is unknown. Therefore, we examined the Figure S4A and S4B). Acetylation at the 300D locus involvement of CBP and p300 in the acetylation of induced by I-SceI digestion was suppressed by ablation histones H3 and H4 at DSB sites by ChIP analysis using of CBP or p300 (Figures 5a and b). However, the anti-acetylated histone antibodies. DSBs induced by quantity of histone H3 at the 300D locus was not I-SceI digestion caused acetylation of H4 K5/K8/K12/ signiﬁcantly affected by I-SceI digestion and/or by CBP

Oncogene Role of CBP/p300 in non-homologous end joining H Ogiwara et al 2141 H4 AceK5K8K12K16 H3 AceK18 H3 8 8 8 7 No I-SceI 7 No I-SceI 7 No I-SceI 6 6 I-SceI 6 I-SceI I-SceI 5 5 5 4 4 4 3 3 3 2 2 2 at 300D locus at 300D locus at 300D locus ChIP enrichment ChIP enrichment 1 1 ChIP enrichment 1 0 0 0 siCTR siCBP sip300 siCTR siCBP sip300 siCTR siCBP sip300

GFP-KU70 0 sec 60 sec GFP-KU70 KU80 1.2 siCTR 6 siCTR 1 siCBP 5 No I-SceI 0.8 sip300 4 I-SceI 0.6 3 0.4 2 siCBP 0.2 at 300D locus 1 Relative intensity ChIP enrichment 0 0 20 40 60 80 100 120 140 siCTR siCBP sip300 Time after laser-irradiation (sec) siP300

Figure 5 Ablation of CBP or p300 caused a reduction in histone H3/H4 acetylation and KU70/KU80 recruitment at DSB sites. (a–d) H1299dA3-1 #1 cells pre-transfected with non-targeting siRNA (siCTR), or CBP- or p300-targeting siRNA (siCBP, sip300) were subjected to ChIP assay 0 h (No I-SceI) and/or 18 h (I-SceI) after transfection of the I-SceI expression plasmid. DNAs immunoprecipitated with anti-acetylated H4 (H4 AceK5/8/12/16) (a), anti-acetylated H3 K18 (H3 AceK18) (b), anti-H3 (c) and anti- KU80 (d) antibodies were subjected to quantitative PCR for the 0.3-kb region. The relative enrichment of proteins at DSBs after 18 h (I-SceI) against 0 h (No I-SceI) is shown. The mean values from three independent experiments are shown±s.d. (e, f) H1299 cells were transfected with siCTR, siCBP or sip300. After 48 h, the cells were transfected with the GFP-KU70 expression plasmid and examined at the indicated time points after laser micro-irradiation. (e) Representative fields for GFP-KU70 60 s after laser micro-irradiation are shown. (f) Real-time recruitment of GFP-KU70. The graph shows the quantification of the relative fluorescence intensities within the micro- irradiated regions subtracted from the background fluorescence in undamaged regions within the nucleus. The error bars indicate the s.d. or p300 ablation (Figure 5c). Acetylation of lysine 56 cooperation between histone modification and altera- within histone H3 at the 300D locus was not signi- tions in the nucleosomal position (Osley and Shen, ficantly suppressed by ablation of CBP or p300 ablation 2006). In fact, it was recently suggested that histone (Supplementary Figure S4C). We also examined acetyl- acetylation is required for the recruitment of the SWI/ ation of lysines 9 and 14 within histone H3 (H3 K9K14) SNF complex to DSB sites (Lee et al., 2010). SWI/SNF but could not detect any I-SceI-induced acetylation is a chromatin remodeling complex that enhances DSB (data not shown), although the DSB-induced acetyla- repair (Park et al., 2006). Thus, we examined the effects tion of H3 K9K14 was observed in another I-SceI- of BRM ablation (one of the two catalytic subunits of induced DSB system (Murr et al., 2006). These results the SWI/SNF complex) on NHEJ using H1299 dA3-1 indicated that the CBP/p300 proteins specifically acety- #1 cells. As H1299 cells lack a functional BRG1 gene lated H4 K5/K8/K12/K16 and H3 K18 at DSB sites. (Medina et al., 2008), which encodes the other catalytic Recruitment of KU70/KU80 proteins is a critical step subunit, it was predicted that the SWI/SNF complex in the ligation of DNA ends through NHEJ (Burma would be inactivated by ablation of BRM in our et al., 2006; Lieber, 2008). Therefore, we examined the assay system. BRM ablation caused a reduction in the effect of ablating CBP or p300 on the recruitment of proportion of both EGFP-positive cells and joined KU80 and KU70 proteins using ChIP analysis, and DNA, without any significant change in the propor- found that it was suppressed after ablation of CBP or tion of uncut DNA (Figures 6a and c). Furthermore, p300 (Figure 5d and Supplementary Figure S4D). BRM ablation also suppressed the accumulation of Furthermore, ablation of CBP or p300 also suppressed exogenously expressed GFP-KU70 protein at laser the accumulation of exogenously expressed human KU70 micro-irradiated sites (Figures 6d and e). These results tagged with green fluorescent protein (GFP-KU70) in suggested the involvement of BRM in NHEJ. Impor- laser micro-irradiated sites (Figures 5e and f). This tantly, ablation of the CBP or p300 proteins suppressed suggested that histone acetylation by CBP and p300 the accumulation of exogenously expressed GFP-tagged at DSB sites facilitates the recruitment of KU70/80. human BRM protein (GFP-BRM) at laser micro- irradiated sites (Figures 6f and g). ChIP analysis also indicated that the recruitment of BRM to DSB sites Impairment of SWI/SNF chromatin remodeling protein generated by I-SceI digestion was also suppressed by recruitment to DSB sites by CBP or p300 ablation CBP or p300 ablation (Supplementary Figure 5). These We next investigated how acetylation of histones H3 results suggest that histone acetylation at DSB sites and H4 by CBP/p300 contributes to NHEJ. The by CBP and p300 proteins facilitates the recruitment of accessibility of proteins to DNA in vivo is regulated by the SWI/SNF chromatin remodeling complex, and the

Oncogene Role of CBP/p300 in non-homologous end joining H Ogiwara et al 2142 Proportion of EGFP-positive cells 0 0.2 0.4 0.6 0.8 1 1.2 siCTR siBRM siCTR BRM siBRM β actin

Proportion of joined DNA Proportion of uncut DNA 0 0.2 0.4 0.6 0.8 1 1.2 0 0.2 0.4 0.6 0.8 1 1.2

siCTR siCTR siBRM siBRM

GFP-KU70 GFP-KU70 1.2 siCTR 0 sec 60 sec 1 siBRM 0.8 siCTR 0.6 0.4

Relative intensity 0.2 0 siBRM 0 20 40 60 80 100 120 140 Time after laser-irradiation (sec)

GFP-BRM 0 sec 60 sec GFP-BRM 1.2 siCTR siCTR 1 siCBP 0.8 sip300 0.6

siCBP 0.4

Relative intensity 0.2 0 20 40 60 80 100 120 140 siP300 Time after laser-irradiation (sec)

Figure 6 Functional interaction between BRM and CBP/p300 during NHEJ. (a–c) Assay for NHEJ of chromosomal DSBs in vivo.Results observed 48 h after transfection of the I-SceI expression plasmid are shown. (a) The proportion of EGFP-positive cells, and (b) joined and (c) uncut DNA. The proportion of cells observed after targeting siRNA (siBRM) treatment is expressed as a ratio versus that observed after non-targeting siRNA (siCTR) treatment. The mean values from three independent experiments are shown±s.d. The results from the immunoblot analysis are shown on the right (a). The b-actin was used as a loading control. (d, e) Impaired recruitment of KU70 protein to DSB sites after BRM ablation. H1299 cells were transfected with siCTR or siBRM. After 48 h, the cells were transfected with the GFP-KU70 expression plasmid and examined at the indicated time points after laser micro-irradiation. (d) Representative fields for GFP-KU70, 60 s after micro-irradiation are shown. (e) Real-time accumulation of GFP-KU70 at the micro-irradiation sites. (f, g)Impaired recruitment of BRM protein to DSB sites after CBP or p300 ablation. H1299 cells were transfected with siCTR,siCBPorsip300.After48h, the cells were transfected with the GFP-BRM expression plasmid and examined at the indicated time points after laser micro-irradiation. (e) Representative fields for GFP-BRM 60 s after micro-irradiation are shown. (f) Real-time accumulation of GFP-BRM at micro- irradiation sites. The graphs in e and g show quantification of the relative fluorescence intensities within the micro-irradiated regions subtracted from the background fluorescence within undamaged regions in the nucleus. The error bars indicate the s.d.

resulting ‘relaxed’ chromatin enables the accumulation of the breakpoint junctions within DSBs generated by of the core NHEJ proteins. I-SceI endonuclease, indicate that ligation is mediated by NHEJ in this system. However, the treatment of cells with the HAT inhibitors, anacardic acid and curcumin, Discussion also suppressed NHEJ in this system. Acetylation of the N-terminal lysines 5, 8, 12 and 16 within histone H4 We have developed an assay system that allows the (H4 K5/K8/K12/K16) is critical for NHEJ (Bird et al., evaluation of ligation activity in chromosomal DNA 2002). TIP60 functions as an acetyltransferase for the in vivo. Suppression of DNA joining by inhibitors or N-terminal residues of histone H4 at DSB sites in siRNAs targeting DNA-PKcs, along with the structure both yeast and mammalian cells (Downs et al., 2004;

Oncogene Role of CBP/p300 in non-homologous end joining H Ogiwara et al 2143 Murr et al., 2006). However, it was not known whether forms a barrier to DNA–protein interactions, it is likely TIP60 is required for NHEJ in vivo. In this study, we that NHEJ requires chromatin remodeling. This consists showed that siRNA-mediated ablation of TIP60, a of histone modification by HATs and alteration of the target of anacardic acid (Sun et al., 2006), suppressed nucleosome positions within the chromatin by chroma- NHEJ. The results also indicated that NHEJ is accomp- tin remodeling complexes. Importantly, ablation of CBP lished through chromatin modification by HATs or p300 HATs impaired the recruitment of the KU70/80 (Valerie and Povirk, 2003; Downs, 2007), and that this proteins to DSB sites. This result complements a recent assay system can be used to examine the molecular result showing that inhibition of histone deacetylases mechanisms involved in NHEJ in vivo, including that of enhances the association of NHEJ factors with DSB HATs and other chromatin remodeling proteins. sites (Miller et al., 2010). Taken together, these results ChIP analysis showed that DSB-induced acetylation strengthen the importance of histone acetylation at DSB of lysines 5, 8, 12 and 16 within histone H4 occurs in sites for the recruitment of NHEJ factors. Members of both yeast and mammalian cells (Downs et al., 2004; the SWI2/SNF2 superfamily (INO80, SWR1, SWI/SNF Tamburini and Tyler, 2005; Murr et al., 2006). This was and RSC) of chromatin remodeling complexes were confirmed by our system. DNA damage-induced acetyl- found at single DSBs in budding yeast (van Attikum ation of N-terminal lysines 9, 14, 18, 23 and 56 within et al., 2004; Chai et al., 2005; Shim et al., 2005). There- histone H3 in mammalian cells was observed by fore, these complexes were postulated to have a role in immunoblot analysis (Das et al., 2009; Tjeertes et al., DSB repair by altering the nucleosome positions at the 2009; Yuan et al., 2009; Lee et al., 2010; Miller et al., DSB sites. In this study, ablation of both the catalytic 2010). However, the results were controversial. Further- subunits of the SWI/SNF complex (BRM (by RNAi) more, these studies only examined global acetylation, and BRG1 (inactivated by mutation in H1299 cells used not specific acetylation at DSB sites. The significance of for analysis)) impaired the recruitment of KU70 to the H3 acetylation at DSB sites was demonstrated by DSB sites in human cells. Significantly, accumulation showing acetylation of H3 K9/K14 and deacetylation of BRM at DSB sites was also impaired by the ablation of H3 K56 at these sites by ChIP analysis (Murr et al., of CBP or p300. This indicated that histone acetylation 2006; Miller et al., 2010). In our study, acetylation of by CBP or p300 at DSB sites facilitates the recruitment histone H3 upon DSBs was also observed using ChIP of the SWI/SNF complex. This result is consistent with analysis, validating the significance of histone H3 a report that nucleosomes containing gH2AX, and acetylation at DSB sites. Interestingly, H3 K18, rather histone H3 with mutations in the acetylation sites, lack than H3 K9/K14 or H3 K56, was acetylated at the DSB the ability to interact with the SWI/SNF complex (Lee sites in our study. The inconsistent result regarding H3 et al., 2010). In summary, the present results strongly K9/K14 might be due to differences in the assay systems indicate that the CBP and p300 proteins function as used; the present assay reproduced repair by NHEJ, HATs for histone H3 and H4 at DSB sites and facilitate whereas that of Murr et al. reproduced repair by the recruitment of the SWI/SNF chromatin remodeling homologous recombination (Murr et al., 2006). complex. The resultant ‘relaxed’ chromatin allows Ablation of CBP, p300 and TIP60 suppressed NHEJ, NHEJ by facilitating the recruitment of core NHEJ whereas ablation of PCAF did not. These results proteins. These results are supported by the fact that indicate that multiple, but not all, HATs are involved the INO80, SWR1 and RSC complexes are required for in the facilitation of NHEJ via acetylation of histones at the recruitment of KU70/80 proteins to DSB sites in DSB sites. Acetylation of H3 K18 and H4 K5/K8/K12/ budding yeast (Shim et al., 2005; van Attikum et al., K16, but not H3 K56, at DSB sites was suppressed by 2007). However, both CBP and p300 have a role during the ablation of CBP or p300. Importantly, CBP and transcription, either as a physical bridge for diverse p300 accumulated at the DSB sites. Therefore, CBP and proteins, or as a scaffold for the formation of multi- p300 acetylate these N-terminal histone residues at the component complexes. In addition, acetylation of non- DSB sites to facilitate NHEJ. Another HAT, GCN5, is histone proteins, such as p53, by CBP and p300 also thought to be involved in the acetylation of histone enhances their DNA-binding activity and regulates H3 at DSB sites, although its involvement in NHEJ has protein-DNA recognition (Karamouzis et al., 2007). not been investigated (Park et al., 2006). Thus, at least Thus, the functions of CBP and p300, other than that four HATs (TIP60, CBP, p300 and GCN5) are likely to of HATs, might also help the recruitment of KU70/80 have redundant roles in histone acetylation at DSB proteins. It was reported that CBP associates with and sites in human cells. HATs, including CBP and p300, acetylates KU70 in vitro (Cohen et al., 2004). However, also acetylate non-histone proteins (Karamouzis et al., we could not detect any interaction between CBP or 2007). Further study is needed to elucidate the NHEJ p300 and KU70/80 in the cell nucleus (data not shown). mechanisms facilitated by these HATs. Thus, CBP/p300 might not have a role in NHEJ as an NHEJ proceeds in a stepwise manner employing acetyltransferase for KU70 itself, but might function as multiple proteins. First, the broken DNA ends are a scaffold and/or acetyltransferase for non-histone recognized by KU70/80. DNA-PKcs then forms a proteins to recruit KU70/80. complex with the KU70/80 proteins that brings the Enhanced DSB repair activity within cancer cells, two DNA ends together. Finally, the two DNA ends which allows them to evade IR and anti-cancer drug- are joined by the LIGIV/XRCC4 complex (Burma et al., induced cell death, is a problematic issue in cancer 2006; Lieber, 2008). As the structure of chromatin therapy. Thus, the suppression of NHEJ activity is a

Oncogene Role of CBP/p300 in non-homologous end joining H Ogiwara et al 2144 promising therapeutic method for sensitizing cancer cells Lipofectamine 2000 reagent (Invitrogen), and two clones, to radiotherapy and chemotherapy (Helleday et al., H1299dA3-1 #1 and #2, resistant to puromycin (2.5 mg/ml) 2008). In this study, siRNA-mediated ablation of CBP/ were obtained. Retention of pIRES-TK-EGFP DNA was p300 increased the radio-sensitivity of lung cancer confirmed by genomic PCR. Integration of the pIRES-TK- cells. Therefore, inhibition of CBP/p300 activity using EGFP DNA into chromosomal DNA was confirmed by specific inhibitors or siRNAs would be a novel way fluorescence in situ hybridization analysis (Yoshida et al., 2001). A 2.7-kb fragment covering herpes simplex virus- of sensitizing cancer cells to radiotherapy. Ablation of thymidine kinase and GFP from the pIRES-TK-EGFP CBP/p300 also increased the sensitivity of lung cancer plasmid was used as a probe, together with a whole cells to the anti-cancer drug, etoposide, which induces chromosome painting probe for chromosome 15 (CAMBIO, DSBs that are repaired by NHEJ (Adachi et al., 2003). Cambridge, UK). IRES-TK-EGFP DNA was detected in Therefore, inhibition of CBP/p300 would also be an a marker chromosome in #1 cells in 41 (91%) out of 45 effective way to sensitize cancer cells to chemotherapy. metaphases examined (Supplementary Figure S1B). IRES-TK- EGFP DNA was integrated in one to three chromosomes in #2 cells in 47 (94%) out of 50 metaphases examined. The pCBASce plasmid DNA (0.8 mg) was introduced into Materials and methods 7.5 Â 104 of H1299dA3-1 #1 and #2 cells per well in a 24-well plate by transfection with Lipofectamine 2000 reagent. For Construction of plasmids fluorescence-activated cell sorting analysis, cells were harvested The pIRES-TK-EGFP plasmid used for the NHEJ assay by trypsinization, washed with phosphate-buffered saline and (Supplementary Figure S1A) was constructed by inserting a applied to a FACSCalibur (Becton Dickinson). For quantita- DNA fragment containing herpes simplex virus-thymidine tive PCR analysis, genomic DNA was purified from cells using kinase open reading frame from the pAD-TK plasmid and a Illustra Tissue and Cells GenomicPrep Mini Spin kits (GE DNA fragment containing the EGFP open reading frame from Healthcare, Tokyo, Japan), and 20 ng of DNA was subjected the pmaxGFP plasmid (Amaxa, Cologne, Germany) into the to quantitative PCR to quantify the uncut and joined DNAs pIRESpuro3 plasmid (BD Bioscience Clontech, Mountain using an ABI 7900HT fast real-time PCR system analyzer View, CA, USA). The pCBASce, an I-SceI expression plasmid, (Applied Biosystems, Foster City, CA, USA). PCR was was obtained from Dr Maria Jasin (Johnson et al., 1999). performed for 45 cycles consisting of denaturation at 95 1C Plasmids expressing human KU70 and BRM complementary for 30 s, and annealing and extension at 60 1C for 60 s using a DNA were constructed by subcloning complementary DNA QuantiTect Probe PCR kit (Qiagen, Valencia, CA, USA). derived from HeLa cell RNA into a modified EGFP-C1 GAPDH DNA was co-amplified and quantified as an internal plasmid (Hong et al., 2008), which contains multiple cloning control. The primers and probes used for quantitative PCR are sites for the introduction of complementary DNAs with in- listed in Supplementary Table S1. The proportion of uncut frame XhoI NotI sites at the start and stop codons, DNA (reflecting the efficiency of I-SceI digestion) was respectively. The integrity of the constructs was verified by reproducibly observed to be 0.5–0.6, regardless of the siRNA sequencing. Plasmid DNAs were propagated in, and purified or inhibitor treatments described below. To examine the effects from, DH10B E.coli cells (Invitrogen, Carlsbad, CA, USA). of inhibitors on DNA joining, an inhibitor or dimethyl sulfoxide (Sigma) alone (used as a solvent for the inhibitors) was added at the time of pCBASce plasmid transfection. To Materials examine the effects of siRNAs on DNA joining, cells were H1299 and H1299dA3-1 human lung cancer cells were subjected to the NHEJ assay and immunoblot analysis 48 h cultured in RPMI 1640 supplemented with 10% fetal bovine after siRNA transfection. The siRNAs were transfected at a serum at 37 1C with 5% CO2. NU7026 and curcumin were concentration of 50 nM using Lipofectamine RNA MAX purchased from Sigma (St Louis, MO, USA), and anacardic (Invitrogen). The inhibitor concentrations were as follows: acid was purchased from Merck (Hamburg, Germany). The NU7026: 50 mM; anacardic acid: 50 mM; curcumin: 20 mM. following siRNA duplexes were purchased from Dharmacon (Yokohama, Japan): siCTR (L-001810-10), siDNAPKcs (L-005030-00), siKU70 (L-005084-00), siKU80 (L-010491-00), Clonogenic assay siTIP60 (L-006301-00), siCBP (L-003477), siCBP#1 (J-003477- H1299 cells were pre-treated with dimethyl sulfoxide or each 06), siCBP#2 (J-003477-09), sip300 (L-003486-00), sip300#1 HAT inhibitor (anacardic acid: 25 or 50 mM; curcumin: 10 or (J-03489-11), sip300#2 (J-03489-12), siPCAF (L-005055-00) and 20 mM) for 1 h, or pre-transfected with siRNA for 48 h, and siBRM (L-017253-00). Antibodies were purchased from Sigma subjected to (or not) irradiation at 2.5, 5 or 10 Gy. The cells (a-tubulin, T6199; b-actin, A5441), Abcam (Cambridge, MA, were plated onto 60-mm dishes and surviving colonies were USA) (DNA-PKcs, pS2056 and ab18192; PCAF, ab12188; counted after 10 days. The number of colonies derived from XRCC4, ab12069; 53BP1, ab36823), Calbiochem (San Diego, treated cells is expressed as a ratio against that of untreated CA, USA) (DNA-PKcs, NA57), Santa Cruz Biotechnology cells (control). (San Diego, CA, USA) (CBP, sc369 and sc7300; p300, sc585 and sc48343; KU70, sc9033; KU80, sc9034; LIGIV, sc11750), Becton Immunoblot analysis Dickinson (Franklin Lakes, NJ, USA) (BRM, 610390), Active Cells were harvested, washed with phosphate-buffered saline Motif (Carlsbad, CA, USA) (Histone H3, 39164), Epitomics and lysed in 100 ml NETN300 buffer (20 mM Tris-HCl, pH 7.5, (Burlingame, CA, USA) (acetylated H3 (K56), 2134-1) and 300 mM NaCl, 0.5% NP-40, 1 mM EDTA) containing 1 mM Millipore (Billerica, MA, USA) (gH2AX, 05-636; TIP60, 07-03; phenylmethylsulfonyl fluoride, 10 mM sodium butyrate, a acetylated H4 (K5/8/12/16), 06-866; acetylated H3 (K18), 07-354). proteinase inhibitor cocktail and a phosphatase inhibitor cocktail (Roche Diagnostics, Indianapolis, IN, USA). After NHEJ assay 30 min of incubation on ice, the samples were centrifuged for The pIRES-TK-EGFP DNA, linearized by FspI restriction 10 min at 14 000 r.p.m. at 4 1C. Supernatants were mixed with enzyme digestion, was transfected into H1299 cells using the SDS sample buffer, boiled and subjected to SDS–polyacryl-

Oncogene Role of CBP/p300 in non-homologous end joining H Ogiwara et al 2145 amide gel electrophoresis, followed by blotting onto a stained with 90% glycerol/phosphate-buffered saline contain- polyvinylidene difluoride membrane. The membranes were ing 1 mg/ml of paraphenylenediamine and 0.5 mg/ml of 40,60- blocked overnight with Tris-buffered saline containing 0.1% diamidino-2-phenylindole and observed using the FV-500 Tween 20 and 0.5% bovine serum albumin, and probed with confocal scanning laser microscopy system. To examine yellow primary antibodies. After washing with Tris-buffered saline fluorescent protein-Cbp, GFP-KU70 or GFP-BRM localiza- containing 0.1% Tween 20, the membranes were incubated tion, H1299 cells were transfected with the yellow fluorescent with horseradish peroxidase-conjugated anti-mouse or anti- protein-Cbp (gifted from Dr Marc Tini), GFP-KU70 or GFP- rabbit secondary antibodies and treated with an enhanced BRM expression plasmids either with or without siRNA, 24 h chemiluminescence reagent (GE Healthcare). before micro-irradiation. Fluorescence images were obtained after micro-irradiation. To quantitate protein accumulation at ChIP assay the laser-generated DSBs, the mean fluorescence intensity of H1299dA3-1 cells (2 Â 106) were plated in 100-mm dishes and the damaged regions within each cell was subtracted from that transfected with or without 16 mg of the pCBASce plasmid of the non-damaged nuclear regions. The values represent the using Lipofectamine RNAMAX (Invitrogen). At 0, 12 or 18 h mean±s.d. after transfection, cells were treated with 1% formaldehyde for 10 min at room temperature to cross-link proteins to DNA, Comet assay followed by addition of glycine (0.125 M) to stop the cross- The neutral comet assay was performed using a Trevigen’s linking. Cells were collected with a cell scraper and washed in Comet Assay kit (4250-050-K, Trevigen, Gaithersburg, MD, ice-cold phosphate-buffered saline. The ChIP assay was USA) according to the manufacturer’s instructions. H1299 performed using a ChIP IT Express Enzymatic kit (Active cells were exposed to 20 Gy of g-irradiation and subjected to Motif) according to the manufacturer’s protocol. Briefly, re- comet analysis at the indicated time points. After staining with suspended cells were subjected to DNA shearing for 10 min at SYBR green, comet images were captured by fluorescence 37 1C, followed by the addition of EDTA to stop the reaction. microscopy. Average Comet tail moment (percentage of DNA A portion of the samples was set aside for input control, in tail Â tail length) was scored for three fields using the whereas the remaining portion was subjected to immunopre- CometScore software (TriTek, Sumerduck,VA, USA). The cipitation performed at 4 1C overnight. The cross-links were values represent the mean±s.d. reversed in a reverse cross-linking buffer for 15 min at 95 1C, and the samples were de-proteinized for 1 h at 37 1C using proteinase K. Aliquots were subjected to quantitative PCR using the primer/probe set shown in Supplementary Table S1 Abbreviations and an ABI 7900HT fast real-time PCR system analyzer (Applied Biosystems). PCR was performed for 45 cycles DSB, DNA double-strand breaks; NHEJ, non-homologous consisting of denaturation at 95 1C for 30 s, and annealing end joining; HAT, histone acetyltransferase; ChIP, chromatin and extension at 60 1C for 60 s using a QuantiTect Probe PCR immunoprecipitation; IR, ionizing radiation. kit (Qiagen). The proportion of DNAs from the test loci cross- linked to proteins was calculated based on the quantity of DNA in the input control samples at each time point. The relative enrichment of a protein at the test sites was expressed Conflict of interest as the amount of immunoprecipitated DNA divided by the amount of input DNA, and plotted as ratios against the values The authors declare no conflict of interest. for 0 h. The antibodies used in the ChIP assay were: CBP (sc369), p300 (sc585), KU70 (sc9033), KU80 (sc9034), BRM (610390), histone H3 (39164), gH2AX (05-636), acetylated H4 Acknowledgements (K5/8/12/16) (06-866), acetylated H3 (K18) (07-354) and acetylated H3 (K56) (2134-1). We thank Dr Maria Jasin of the Memorial Sloan-Kettering Cancer Center, USA, for the pCBASce plasmid, Dr Marc Tini Laser micro-irradiation of the University of Western Ontario for the yellow Laser micro-irradiation was performed using the FV-500 fluorescent protein-Cbp plasmid and Dr Kazunori Aoki of confocal scanning laser microscopy system (Olympus, Tokyo, the National Cancer Center Research Institute for the Japan) as reported previously (Lan et al., 2005; Nakajima pAD-TK plasmid. This work was supported by Grants- et al., 2006). Briefly, cells in glass-bottomed dishes were micro- in-Aid from the Ministry of Education, Culture, Sports, irradiated with a 405 nm pulse laser (Olympus). The laser was Science and Technology of Japan for Scientific Research on focused through a Â 40 objective lens. To examine endogenous Innovative Areas; from the Ministry of Health, Labor and CBP, p300, p53BP1 and gH2AX proteins, micro-irradiated Welfare of Japan for the 3rd-term Comprehensive 10-year H1299 cells were fixed and permeabilized with MeOH and Strategy for Cancer Control and for Cancer Research (19-9); acetone for 10 min at À20 1C, 10 min after micro-irradiation. and from the Program for Promotion of Fundamental Studies The cells were subjected to an immunofluorescence assay using in Health Sciences of the National Institute of Biomedical primary antibodies and secondary antibodies. Samples were Innovation (NiBio).

References

Adachi N, Suzuki H, Iiizumi S, Koyama H. (2003). Hypersensitivity of Balasubramanyam K, Swaminathan V, Ranganathan A, nonhomologous DNA end-joining mutants to VP-16 and ICRF- Kundu TK. (2003). Small molecule modulators of 193: implications for the repair of topoisomerase II-mediated DNA histone acetyltransferase p300. J Biol Chem 278: damage. J Biol Chem 278: 35897–35902. 19134–19140.

Oncogene Role of CBP/p300 in non-homologous end joining H Ogiwara et al 2146 Balasubramanyam K, Varier RA, Altaf M, Swaminathan V, Siddappa complexes requires interaction between Ku70/80 and XRCC4. NB, Ranga U et al. (2004). Curcumin, a novel p300/CREB-binding Proc Natl Acad Sci USA 103: 18597–18602. protein-specific inhibitor of acetyltransferase, represses the acetyla- Medina PP, Romero OA, Kohno T, Montuenga LM, Pio R, Yokota J tion of histone/nonhistone proteins and histone acetyltransferase- et al. (2008). Frequent BRG1/SMARCA4-inactivating mutations in dependent chromatin transcription. J Biol Chem 279: 51163–51171. human lung cancer cell lines. Hum Mutat 29: 617–622. Bird AW, Yu DY, Pray-Grant MG, Qiu Q, Harmon KE, Megee PC Miller KM, Tjeertes JV, Coates J, Legube G, Polo SE, Britton S et al. et al. (2002). Acetylation of histone H4 by Esa1 is required for DNA (2010). Human HDAC1 and HDAC2 function in the DNA-damage double-strand break repair. Nature 419: 411–415. response to promote DNA nonhomologous end-joining. Nat Struct Bonner WM, Redon CE, Dickey JS, Nakamura AJ, Sedelnikova OA, Mol Biol 17: 1144–1151. Solier S et al. (2008). GammaH2AX and cancer. Nat Rev Cancer 8: Murr R, Loizou JI, Yang YG, Cuenin C, Li H, Wang ZQ et al. (2006). 957–967. Histone acetylation by Trrap-Tip60 modulates loading of repair Burma S, Chen BP, Chen DJ. (2006). Role of non-homologous end proteins and repair of DNA double-strand breaks. Nat Cell Biol 8: joining (NHEJ) in maintaining genomic integrity. DNA Repair 91–99. (Amst) 5: 1042–1048. Nakajima S, Lan L, Kanno S, Usami N, Kobayashi K, Mori M et al. Chai B, Huang J, Cairns BR, Laurent BC. (2005). Distinct roles for the (2006). Replication-dependent and -independent responses of RAD18 RSC and Swi/Snf ATP-dependent chromatin remodelers in DNA to DNA damage in human cells. JBiolChem281: 34687–34695. double-strand break repair. Genes Dev 19: 1656–1661. Osley MA, Shen X. (2006). Altering nucleosomes during DNA double- Cohen HY, Lavu S, Bitterman KJ, Hekking B, Imahiyerobo TA, strand break repair in yeast. Trends Genet 22: 671–677. Miller C et al. (2004). Acetylation of the C terminus of Ku70 by Pandita TK, Richardson C. (2009). Chromatin remodeling finds its CBP and PCAF controls Bax-mediated apoptosis. Mol Cell 13: place in the DNA double-strand break response. Nucleic Acids Res 627–638. 37: 1363–1377. Das C, Lucia MS, Hansen KC, Tyler JK. (2009). CBP/p300-mediated Park EJ, Chan DW, Park JH, Oettinger MA, Kwon J. (2003). DNA- acetylation of histone H3 on lysine 56. Nature 459: 113–117. PK is activated by nucleosomes and phosphorylates H2AX within Downs JA. (2007). Chromatin structure and DNA double-strand the nucleosomes in an acetylation-dependent manner. Nucleic Acids break responses in cancer progression and therapy. Oncogene 26: Res 31: 6819–6827. 7765–7772. Park JH, Park EJ, Lee HS, Kim SJ, Hur SK, Imbalzano AN et al. Downs JA, Allard S, Jobin-Robitaille O, Javaheri A, Auger A, (2006). Mammalian SWI/SNF complexes facilitate DNA double- Bouchard N et al. (2004). Binding of chromatin-modifying activities strand break repair by promoting gamma-H2AX induction. to phosphorylated histone H2A at DNA damage sites. Mol Cell 16: EMBO J 25: 3986–3997. 979–990. Robert F, Hardy S, Nagy Z, Baldeyron C, Murr R, Dery U et al. Helleday T, Petermann E, Lundin C, Hodgson B, Sharma RA. (2008). (2006). The transcriptional histone acetyltransferase cofactor DNA repair pathways as targets for cancer therapy. Nat Rev Cancer TRRAP associates with the MRN repair complex and plays a role 8: 193–204. in DNA double-strand break repair. Mol Cell Biol 26: 402–412. Hong Z, Jiang J, Lan L, Nakajima S, Kanno S, Koseki H et al. (2008). Schiltz RL, Mizzen CA, Vassilev A, Cook RG, Allis CD, Nakatani Y. A polycomb group protein, PHF1, is involved in the response to (1999). Overlapping but distinct patterns of histone acetylation by DNA double-strand breaks in human cell. Nucleic Acids Res 36: the human coactivators p300 and PCAF within nucleosomal 2939–2947. substrates. J Biol Chem 274: 1189–1192. Horwitz GA, Zhang K, McBrian MA, Grunstein M, Kurdistani SK, Shim EY, Ma JL, Oum JH, Yanez Y, Lee SE. (2005). The yeast Berk AJ. (2008). Adenovirus small e1a alters global patterns of chromatin remodeler RSC complex facilitates end joining repair of histone modification. Science 321: 1084–1085. DNA double-strand breaks. Mol Cell Biol 25: 3934–3944. Jasin M. (1996). Genetic manipulation of genomes with rare-cutting Sun Y, Jiang X, Chen S, Price BD. (2006). Inhibition of histone endonucleases. Trends Genet 12: 224–228. acetyltransferase activity by anacardic acid sensitizes tumor cells to Javvadi P, Segan AT, Tuttle SW, Koumenis C. (2008). The ionizing radiation. FEBS Lett 580: 4353–4356. chemopreventive agent curcumin is a potent radiosensitizer of Tamburini BA, Tyler JK. (2005). Localized histone acetylation and human cervical tumor cells via increased reactive oxygen species deacetylation triggered by the homologous recombination pathway production and overactivation of the mitogen-activated protein of double-strand DNA repair. Mol Cell Biol 25: 4903–4913. kinase pathway. Mol Pharmacol 73: 1491–1501. Tjeertes JV, Miller KM, Jackson SP. (2009). Screen for DNA-damage- Johnson RD, Liu N, Jasin M. (1999). Mammalian XRCC2 promotes responsive histone modifications identifies H3K9Ac and H3K56Ac the repair of DNA double-strand breaks by homologous recombi- in human cells. Embo J 28: 1878–1889. nation. Nature 401: 397–399. Valerie K, Povirk LF. (2003). Regulation and mechanisms of Karamouzis MV, Konstantinopoulos PA, Papavassiliou AG. (2007). mammalian double-strand break repair. Oncogene 22: 5792–5812. Roles of CREB-binding protein (CBP)/p300 in respiratory epithe- van Attikum H, Fritsch O, Gasser SM. (2007). Distinct roles for lium tumorigenesis. Cell Res 17: 324–332. SWR1 and INO80 chromatin remodeling complexes at chromoso- Khafif A, Hurst R, Kyker K, Fliss DM, Gil Z, Medina JE. (2005). mal double-strand breaks. EMBO J 26: 4113–4125. Curcumin: a new radio-sensitizer of squamous cell carcinoma cells. van Attikum H, Fritsch O, Hohn B, Gasser SM. (2004). Recruitment Otolaryngol Head Neck Surg 132: 317–321. of the INO80 complex by H2A phosphorylation links ATP- Kouzarides T. (2007). Chromatin modifications and their function. dependent chromatin remodeling with DNA double-strand break Cell 128: 693–705. repair. Cell 119: 777–788. Lan L, Nakajima S, Komatsu K, Nussenzweig A, Shimamoto A, Yoshida S, Satoh H, Mitsuya T, Tate G. (2001). Assignment of the Oshima J et al. (2005). Accumulation of Werner protein at DNA human dickkopf (Xenopus) homolog 4 (DKK4) to chromosome double-strand breaks in human cells. J Cell Sci 118: 4153–4162. 8p11.2–4p11.1 by fluorescence in situ hybridization. Cytogenet Cell Lee HS, Park JH, Kim SJ, Kwon SJ, Kwon J. (2010). A cooperative Genet 94: 88–89. activation loop among SWI/SNF, gamma-H2AX and H3 acetyla- Yuan J, Pu M, Zhang Z, Lou Z. (2009). Histone H3-K56 acetylation is tion for DNA double-strand break repair. Embo J 29: 1434–1445. important for genomic stability in mammals. Cell Cycle 8: 1747–1753. Lieber MR. (2008). The mechanism of human nonhomologous DNA Zhao Y, Thomas HD, Batey MA, Cowell IG, Richardson CJ, end joining. J Biol Chem 283: 1–5. Griffin RJ et al. (2006). Preclinical evaluation of a potent novel Mari PO, Florea BI, Persengiev SP, Verkaik NS, Bruggenwirth HT, DNA-dependent protein kinase inhibitor NU7441. Cancer Res 66: Modesti M et al. (2006). Dynamic assembly of end-joining 5354–5362.

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

Oncogene