JQ1 affects BRD2-dependent and independent transcription regulation without disrupting H4-hyperacetylated chromatin states

Lusy Handoko1,6, Bogumil Kaczkowski1,6, Chung-Chau Hon1, Marina Lizio1, Masatoshi Wakamori2, Takayoshi Matsuda2, Takuhiro Ito2, Prashanti Jeyamohan1, Yuko Sato3, Kensaku Sakamoto2, Shigeyuki Yokoyama4, Hiroshi Kimura3, Aki Minoda1,*, Takashi Umehara2,5,*

1RIKEN Center for Life Science Technologies, Division of Genomic Technologies, 1- 7-22 Suehiro-cho, Tsurumi-ku, Yokohama, 230-0045 Japan. 2RIKEN Center for Life Science Technologies, Division of Structural and Synthetic Biology, 1-7-22 Suehiro-cho, Tsurumi-ku, Yokohama, 230-0045 Japan. 3Cell Biology Center, Institute of Innovative Research, Tokyo Institute of Technology, 4259 Nagatsuta-cho, Midori-ku, Yokohama, 226-8503, Japan. 4RIKEN Structural Biology Laboratory, 1-7-22 Suehiro-cho, Tsurumi-ku, Yokohama, 230-0045 Japan. 5PRESTO, Japan Science and Technology Agency (JST), 4-1-8 Honcho, Kawaguchi, Saitama 332-0012, Japan. 6These authors contributed equally to this work. *Corresponding authors.

Correspondence should be addressed to A.M. ([email protected]) or T.U. ([email protected]).

1 SUPPLEMENTARY INFORMATION

METHODS

Supplementary Figures 1. Supplementary Figure 1. Validation of H4K5acK8ac antibodies. 2. Supplementary Figure 2. Genome-wide characterization of H4K5acK8ac and BRD2 and association of BRD2 and histone H4 acetylation. 3. Supplementary Figure 3. Global chromatin states by ChromHMM. 4. Supplementary Figure 4. JQ1 effects on H3K27ac peaks. 5. Supplementary Figure 5. JQ1 effects on the expression in H23 cells. 6. Supplementary Figure 6. Enrichment of CTCF in the BRD2 peaks. 7. Supplementary Figure 7. Genome browser view of the whose promoter is not bound by BRD2.

Supplementary Tables 1. Supplementary Table 1. Peptide array analysis of anti-H4K5acK8ac antibodies and commercially available antibodies. 2. Supplementary Table 2. Crystallographic data collection and refinement statistics. 3. Supplementary Table 3. (Excel). Annotation of H4K5acK8ac, H3K27ac, and BRD2 peaks with the actively expressed CAGE signals based on FANTOMCAT. 4. Supplementary Table 4. (Excel). The list of H4K5acK8ac-based super-enhancers and the associated genes. 5. Supplementary Table 5. (Excel). (a) List of differentially expressed genes (CAGE analysis). (b) List of the genes validated by qPCR and qPCR results. 6. Supplementary Table 6. (Excel). Pathway analysis on the differentially expressed genes upon JQ1 treatment for 3h, 6h, 12h, and 24h (p-value < 0.01). 7. Supplementary Table 7. (Excel). Pathway analysis on the upregulated or downregulated BRD2-bound genes. 8. Supplementary Table 8. (Excel). The list of TF-binding motifs identified by MARA using CAGE data set. 9. Supplementary Table 9. (Excel). List of the primers for qPCR and ChIP-seq.

2 METHODS Generation and selection of mouse monoclonal antibodies The full-length human H4 containing acetyl residue at K5 and K8 (H4K5acK8ac) was synthesized in the cell-free system essentially as previously described21,47 to immunize mice. Generation, selection and purification of the antibodies were performed essentially as previously described48 with the following modification. After generating hybridomas, clones were screened by ELISA using plates immobilized with histidine-tagged H4 containing site-specific lysine acetylation. In this screening, one positive control of H4 protein, H4K5acK8ac, and four negative controls of H4 protein (unmodified H4, H4K5ac, H4K8ac, and H4K8acK12ac) were used. cDNAs were prepared from total RNAs isolated from the hit hybridoma clones 1A9D7 and 2A7D9 and sequenced essentially as previously described49.

Antibody validation by peptide array Peptide array was purchased from Active motif 13005. Blotting of the peptide arrays with the following antibodies were carried out as the manufacture’s protocol: anti- H4K5acK8ac antibody clone 1A9D7, dilution 1:2000; anti-H4K5acK8ac antibody clone 2A7D9, dilution 1:2000; anti-hyperacetylated histone H4 (Penta) antibody (Upstate 06-946, Lot no. 2631278), dilution factor 1:1500; anti-histone H4 (acetyl K5+K8+K12+K16) antibody (Abcam ab177790, Lot no. GR217018-4), dilution 1:15000.

Crystallization, data collection and refinement Anti-H4K5acK8ac IgGs were purified from hybridoma cell culture supernatants using serial chromatography with HiTrap Protein G and the HiTrap Q column. The Fab fragments were prepared by papain digestion of the IgGs, followed by protein-A affinity chromatography to remove Fc fragments and undigested IgGs. The Fab fragments were further purified with size exclusion chromatography using a HiLoad 16/60 Superdex 200 column equilibrated with a 20 mM Tris-HCl (pH 8.0) buffer containing 150 mM NaCl. The concentrations of the Fab fragments were determined by UV absorbance at 280 nm. Chemically synthesized H4K5acK8ac peptide, corresponding to residues 1–12 of human histone H4 containing acetyllysines at positions 5 and 8 [S-G-R-G-Kac-G-G-Kac-G-L-G-K] (Toray Research Center, Japan)

3 was dissolved in a 20 mM Tris-HCl buffer (pH 8.0) containing 150 mM NaCl. To prepare complexes of 1A9D7 Fab•H4K5acK8ac peptide and 2A7D9 Fab•H4K5acK8ac peptide, 340 µM Fab fragments were mixed with 680 µM of H4K5acK8ac peptide. Crystals of the 1A9D7 Fab•H4K5acK8ac peptide were obtained by the sitting drop vapor diffusion method from a drop containing equal volumes of the 1A9D7 Fab•H4K5acK8ac peptide solution and the reservoir solution containing 100 mM imidazole (pH 6.5), 150 mM zinc acetate, and 17.5% (w/v) PEG3000. The crystals of the 2A7D9 Fab•H4K5acK8ac peptide were similarly prepared with the reservoir solution containing 100 mM Bis-Tris buffer (pH 8.0), 200 mM ammonium acetate, and 25% (w/v) PEG3350. The crystals were immersed stepwisely in the reservoir solutions that contain increasing concentrations of glycerol [5%, 10%, 15%, and 20% (v/v)]. The crystals were flash-cooled with liquid nitrogen. X-ray diffraction datasets were collected using the beamline BL26B2 at SPring-8 (Harima, Japan). The datasets were integrated and scaled using the HKL-2000 software50. The initial phases were calculated by the molecular replacement method using the software Phaser51. The coordinate of the anti-OspA Fab fragment (PDB ID: 1FJ1) was used as a search model. The models were refined using the programs Coot52 and PHENIX53.

Cell culture and JQ1 treatment Non-small cell lung cancer, H23 cell line (NCI-H23) was purchased from ATCC (ATCC® CRL-5800™). Cells were cultured as recommended by ATCC in RPMI- 1640 Medium containing 10% FBS and Penicillin/Streptomycin and sub-cultured every two days at the ratio of 1:3. For JQ1 treatment, cells were prepared 48 hrs prior to the treatment, as followed: H23 cells were grown in a 15 cm plate at the density of 170,000 cells/mL (4 × 106 cells/plate). On the day of treatment, the medium was replaced with the fresh medium containing 500 nM JQ1 or an equivalent volume/percentage of DMSO. The cells were collected at 3, 6, 12, and 24 hrs after the addition of JQ1 and subjected to formaldehyde crosslinking for ChIP-seq experiment, RNA isolation for CAGE experiments or nuclear extraction for Western blot. JQ1 and DMSO treatment was done in three independent biological replicates.

Immunostaining Immunostaining was performed as previously described54. Anti-H4K5acK8ac

4 antibody was labeled with Cy3 (GE Healthcare) to yield 1:4 dye/protein ratio. H23 cells were fixed for 5 min at room temperature and permeabilized with 1% Triton X- 100 in PBS for 20 min at room temperature. After blocking with Blocking One P (Nacalai tesque) for 20 min at room temperature, cells were incubated with 2 µg/ml Cy3-conjugated anti-H4K5acK8ac antibody with 2-10 µg/ml Alexa Fluor 488- conjugated histone modification-specific antibody (H3K27ac/CMA309, H3K4me1/CMA302, or H3K4me3/CMA304)55 and 1 µg/ml Hoechst33342 for 2.5 hrs at room temperature. Fluorescence images were acquired using a confocal microscope (FV1000; Olympus) with a 60×UPlanApoN oil immersion lens (NA 1.40). Co- localization analysis was performed using NIS-elements analysis software ver. 4.30 (Nikon).

Chromatin extract preparation Cells were cross-linked with 1% formaldehyde for 10 min at room temperature. The crosslink reaction was stopped by addition of glycine at the final concentration of 0.2 M. H23 Chromatin extract was prepared as follows: the cell pellets were dissolved and incubated in the ice-cold 0.1% SDS lysis buffer (50 mM HEPES-KOH pH 7.5, 150 mM NaCl, 1 mM EDTA, 1% Triton X-100, 0.1% sodium deoxycholate, 0.1% SDS, and protease inhibitors) for 10 min on a rotation at 4°C. Following centrifugation, the cells were incubated in the ice-cold second lysis buffer (10 mM Tris-HCl pH 8.0, 200 mM NaCl, 1 mM EDTA, 500 µM EGTA, and protease inhibitors) as described above. The cells were then dissolved in 0.5% SDS sonication buffer (3×106 cells per 135 µg sonication buffer), directed to chromatin fragmentation by Covaris for 10 min and the fragmented DNAs were then examined by an Agilent Bioanalyzer. The chromatin was diluted using 5×dilution buffer (10 mM Tris-HCl pH 8.0, 210 mM NaCl, 1.4% Triton X-100, and protease inhibitors) and stored at -80°C.

ChIP-seq Chromatin immunoprecipitation was performed using the following antibodies: H3K27ac (mouse monoclonal, Diagenode, MAb-184-050, Lot no. 001-11), H4K5acK8ac (mouse monoclonal), H3K4me3 (mouse monoclonal, MAB307-34813, Lot no. 13005), H3K4me1 (rabbit polyclonal, Diagenode, C15410194, Lot no. A1862D), H3K27me3 (rabbit polyclonal, Diagenode, C15410195 (pAb-195-050), Lot

5 no. 00606C), H3K9me3 (rabbit polyclonal, Diagenode, C15410193 (pAb-193-050), Lot no. A1862D), BRD2 (rabbit monoclonal, Cell Signaling Technology, D89B4, Lot no. 002), and H3K36me3 (Abcam ab9050, Lot no. GR249065-1). The antibody (3 µg antibody/50 µl Dynabeads™ goat anti-mouse IgG (11033) or sheep anti-rabbit IgG (11203D) was incubated with the chromatin extract (4 × 106 cells/ChIP) overnight at 4°C. The beads were then washed using the following conditions: 2 times with wash buffer 1 (10 mM Tris-HCl pH 8.0, 0.1% SDS, 0.1% sodium deoxycholate, 1% Triton X-100, 150 mM NaCl, 1 mM EDTA, and 500 µM EGTA); 2 times with wash buffer 2 (10 mM Tris-HCl pH 8.0, 0.1% SDS, 0.1% sodium deoxycholate, 1% Triton X-100, 500 mM NaCl, and 1 mM EDTA); and 2 times with LiCl wash buffer (10 mM Tris- HCl pH 8.0, 250 mM LiCl, 0.5% sodium deoxycholate, 0.5% NP-40, 1 mM EDTA, and 500 µM EGTA). For BRD2 ChIP-seq, 10 µg antibody was used and the washing condition was: 4 times with wash buffer 1, 4 times with wash buffer 2, and 4 times with LiCl wash buffer. The chromatin was eluted from the bead using TE buffer containing 1% SDS and then de-crosslinked by incubating with 1 µl of Proteinase K (20 mg/ml, EC 3.4.21.14, Tritirachium album limber, CAS 39450-01-6) at 65°C for 3–4 hrs. The DNA fragments were subsequently isolated using Qiagen MinElute column purification. ChIP quality was tested by qPCR using specific primers and negative control region. One microgram of ChIP DNA was used for ChIP-seq library preparation using Mondrian and directed to 50 bp sequencing using HiSeq2000. ChIP-seq was done in two biological replicates for each factor.

Spike in ChIP-seq and data processing To detect global changes of H4K5acK8ac and BRD2 upon BET inhibition by JQ1, we performed ChIP with anti-H4K5acK8ac or anti-BRD2 antibody by adding a minor fraction of D. melanogaster chromatin extract (Active Motif, 53083, Lot no. 11316004) and anti--specific histone variant, H2Av antibody (rabbit polyclonal, Active Motif, 61686, Lot no. 17316003) in the H4K5acK8ac or BRD2 ChIP reaction. This Drosophila ChIP serves as an internal control and is used for ChIP-seq read count normalization after sequencing. Spike in ChIP experiments with either H4K5acK8ac or BRD2 were performed in two biological replicate according to the above described ChIP-seq protocol with some modifications as follows: Anti- H4K5acK8ac or BRD2 antibody and anti-Drosophila H2Av antibody were mixed at

6 the following ratio: H4K5acK8ac:H2Av = 4:2; BRD2:H2Av = 10:4; and then incubated with the Dynabeads for overnight. H23 chromatin extract and Drosophila chromatin extract were mixed at the ratio of 400 to 1 (µg) for H4K5acK8ac ChIP and 600 to 1 (µg) for BRD2 ChIP. Following overnight incubation of the mixed chromatin extract and the mix antibodies, the beads were subsequently washed using the above- mentioned conditions. The ChIP quality control was done as the manufacture’s protocol and ChIP-seq library preparation was done as above described.

Analysis of Spike in ChIP-seq After mapping and alignment of the ChIP-seq reads to a human and Drosophila genome reference, differences in Drosophila read counts across samples were equalized or normalized. The human ChIP-seq read counts were subsequently normalized using the same ratio used to normalize the Drosophila reads. Using the normalized read counts we called ChIP-seq peaks using MACS2 peak caller with the following commands: H4K5ac8ac ChIP-seq peak call: macs2 callpeak -t -c --bw 400 - -broad --broad-cutoff 0.00005 -g hs BRD2 ChIP-seq peak call: macs2 callpeak -t -c --bw 350 -q 0.00005 -g hs --call- summits --slocal 2000 To identify sites with H4K5acK8ac signal significantly changed upon JQ1 treatment, we merged peaks from biological replicates 1 (BR1) and 2 (BR2), and count the reads within the merged peaks, identified peaks with significant differential signal (fdr < 0.05) using edgeR (version 3.6.8) with default settings56.

Transcriptome profiling using CAGE CAGE was carried out to analyze transcriptome changes in H23 cell line upon BET inhibition by JQ1. CAGE libraries were prepared as previously described20,57 using RNAs isolated from the nuclear fraction of the untreated/control (at 0 hr), 500 nM JQ1-and DMSO-treated H23 cells (at 3, 6, 12, and 24 hrs). All of these treatment conditions have three biological replicates. Since stability of RNA sample affects the promoter hit rate and transcript complexity of a CAGE library, the RNA integrity of each sample was measured using an Agilent Bioanalyzer and all samples used in this study had RIN above 9.0. At least 3 µg RNA was used for CAGE library preparation57. Each CAGE library was sequenced at least 15 millions reads. The

7 expression levels of genes were calculated based on the annotations from FANTOM CAT58 as described20. Briefly, the numbers of CAGE read 5′ ends (CAGE TSS) falling within the ± 50 nt region of the prominent TSS of all CAGE clusters in FANTOM CAT were counted in all samples. The expression levels of CAGE clusters were relative log expression (RLE) normalized across all libraries as counts per million (CPM) using edgeR (version 3.6.8)56 with default settings. Gene-based expression levels were calculated as the sum of CPM of their associated CAGE clusters. Genes differentially expressed between samples were identified, based on CAGE tag counts, using edgeR (version 3.6.8)56 with default settings. Sets of genes that are enriched in each differential expression analysis were investigated using GSEA pre-ranked of the Gene Set Enrichment Analysis59, with genes ranked by the log2 fold-changes from edgeR, in default settings. Selected gene sets in the 4 time points were clustered, based on the normalized enrichment score from GSEAPreranked, using the R package Pheatmap (clustering method = ward, distance = euclidean).

RNA isolation and cDNA synthesis Total RNA was isolated from the untreated (control: 0 h), DMSO-, and JQ1-treated cells at 0, 3, 6, 12, and 24 hrs. RNA isolation was performed using the TRIZOL reagent (Invitrogen, Thermo Fisher Scientific) according to the manufacture’s protocol. RNA was dissolved in 50 µl of 100 µM EDTA in RNase-free water. The purity of isolated RNA was determined by OD260/280 using a NanoDrop ND-1000 Spectrophotometer (Thermo Fisher Scientific). cDNA was synthesized using High- Capacity cDNA Reverse Transcription Kit with RNase inhibitor (Applied Biosystems, 4374966). First strand cDNA synthesis was performed using 2 µg of total RNA in recommended conditions per 20 µl reactions with the thermal cycling protocol of: 25 °C, 10 min; 37 °C, 120 min; 85°C, 5min; 4°C hold.

Validation of by quantitative PCR (qPCR) Gene expression was determined by qPCR (7300/7500 StepOnePlus Real-Time PCR Systems (Applied Biosystems) using SYBR Premix Ex Taq II (Tli RNase H plus, Takara Bio, RR820B). 1 µl of cDNA was used as a template in a 20 µl qPCR reaction with the following cycle parameter: 95°C for 30 sec, 40 cycles of 95°C for 5 sec and 60°C for 34 sec. qPCR of each validated gene was done in duplicates and

8 independently repeated two times for each biological replicate (primers listed in Supplementary Table 9). The relative expression level of the tested genes was calculated from the JQ1 and DMSO samples and normalized against GAPDH expression.

Assessing JQ1 effects on the apoptosis and proliferation of the cells To monitor cell proliferation during JQ1 treatment, red fluorescent protein (RFP)- expressing H23 cells were used. To quantify apoptotic cells, we used the IncuCyte Caspase-3/7 Green Apoptosis Assay Reagent couples the activated caspase-3/7 recognition motif (DEVD) to NucView™488, a DNA intercalating dye (IncuCyte, 4440). Apoptosis assay and proliferation assay were performed in a 96 well plate, as follows: RFP-H23 cells were seeded at a density of 8,000 cells/well one day before the experiments. On the day of experiments, the cells were treated with JQ1 at the concentrations of 5 nM, 50 nM, 500 nM, 1000 nM, and 5000 nM. For apoptosis assay, after adding the JQ1, Caspase-3/7 Green Apoptosis Assay Reagent was added as recommended according to the IncuCyte protocol. The cell apoptosis and proliferation was monitored every hour directly after the plate was put in the IncuCyte incubator for 24 hrs. The captured images were analyzed using the software provided by the IncuCyte by counting number of the objects (Caspase 3/7 object-green fluorescence signal count for apoptotic cells or red fluorescence protein for proliferating cells) per image. The fold change in proliferation and apoptosis from 0 hr was then calculated for every time point in every JQ1 treatment experiment.

ChIP-seq peak calling We used MACS260 peak-caller with the significance threshold: p-value = 0.01 on each ChIP-seq replicate separately and on pooled bam files from ChIP-seq BR1 and BR2. We performed the irreproducibility discovery rate (IDR) analysis by running the batch-consistency-analysis.r script on the peaks from ChIP-seq BR1 and BR2 (https://sites.google.com/site/anshulkundaje/projects/idr). For the downstream analysis, we used the peaks from pooled bam that overlap with the peaks obtained after IDR analysis: IDR > 0.01 for robust peaks, and IDR > 0.25 for permissive peaks.

Analysis of ChIP-seq peak profile Genomic distribution/localization of H4K5acK8ac and BRD2 (Fig. 2a and

9 Supplementary Fig. 2a), was determined using peak annotation function of HOMER31. To examine and visualize enrichment of histone marks or BRD2 in defined regions, such as promoter, enhancer, eRNA, we performed meta-region analysis using genomation61. Average normalized peak intensities (i.e. read per million; RPM) of histone modification marks or BRD2 were plotted around the center of defined genomic regions (± 5 kb).

Super-enhancer (SE) and CAGE expression Super-enhancers were called on H4K5acK8ac reproducible peaks (IDR > 0.01) using the publicly available tool ROSE15,62 with default parameters. The stitched regions that are not classified as super enhancers by ROSE are defined as typical enhancer (TE) To associate the CAGE peaks to these regions, we used the bedtools closestBed function to find the peak closest to the SE or TE. Three biological replicates of CAGE for each time point were used. A set (3 BRs) of CAGE from the untreated cells was used as control. The figures were generated using the R function boxplot.

Unified set of active regions To obtain a common set of active sites with acetylation marks, we merged the robust peaks (IDR < 0.01) of H3K27ac (20,305 peaks) and H4K5acK8ac (22,882 peaks) and obtained 25,448 merged regions. We merged the peaks separated by less than 100 nucleotides, which reduced the number to peaks to 24,045. To perform integrative analysis of histone modification and transcription initiation events we aimed to identify the center and if possible the strand of transcription initiation. First, we overlapped the active acetylated regions with CAGE based TSS annotation from FANTOM520 and found that 71,188 out of 201,802 known TTS (DPIs) are overlapping or within 200 nt window from the active acetyl regions in H23 cell line. We clustered TSS separated by less that 500 nt and selected the center and strand of strongest TSS/DPI as a representative for a given group of TSS cluster, which led to the selection of 21,251 centers with strand information (representing 14,685 out of 24,045 merged acetylation peaks). Second, we overlapped the remaining 9,360 acetyl peaks with bidirectionally transcribed enhancer annotated from FANTOM563,64 and found 3,161 enhancer-overlapping 2,448 acetyl sites. We clustered enhancer separated by less that 500 nt and selected the strongest enhancer within each group, 2,815 in total. As the enhancers are bi-directionally transcribed, there is no

10 strand/direction information. Third, we clustered the remaining 6,912 peaks separated by less that 500 nt and selected the strongest peak within each cluster, 6,177 in total. As we used peak midpoints as the centers, no strand information was available. In total, we obtained 30,243 active centers: 21,251 directional TTS, 2,815 bidirectional enhancers and 6,177 acetylation peaks with little or no CAGE signal.

ChIP-seq and CAGE signal at the active sites We calculated read per million within (RPM) ± 5 kb from the peak center, using strand/directionality information where available. The 10 kb window was divided in to 100 bins, and mean RPM within each bin were calculated using ngs.plot package65 and using bam files of pooled replicates as an input. The RPM profile within 10 kb window was calculated for all 30,243 active centers for each histone marks and CAGE, including time JQ1 time-course and input. 156 regions with mean RPM > 0.15 in the input controls were blacklisted and removed from further analyses. To calculate the difference in ChIP-seq and CAGE signal between histone marks or between time points, we calculated the mean RPM in 1 kb window of the center (±

500 bp) and calculated the log2 fold changes, adding 0.25 offset to mean RPM values.

ChromHMM We used ChromHMM software24 to identify chromatin states and to perform genome wide state annotation based on combination of H3K4me1, H3K4me3, H3K27ac, H4KacK8ac, H3K36me3, H3K9me3, H3K27me3, and BRD2. We used BinarizeBam to binarize each the genome into 200 nt bins. We used pooled bam files as an input, a default significance threshold, and fold change threshold > 2. The segmentation was performed using LearnModel function, setting the number of states to 20 with default settings. As a reproducibility check we also performed binarization for set of BR1 and BR2 separately. When ChromHMM was performed using the ChIP-seq data set from biological replicate 1 and 2, same findings were obtained, suggesting high reproducibility of our analysis (Supplementary Fig. 3b).

De novo transcription factor binding motif search To identify TFs, which control gene expression in BRD2-dependent or independent manners, we performed de novo motif search using HOMER31. To identify DNA- binding motifs of TFs that might control the downregulation or upregulation of BRD2

11 bound genes, BRD2 peaks that overlap with the promoter of the downregulated genes or with the promoter of the upregulated genes were used to run HOMER motif search using the following parameter: findMotifsGenome.pl input hg19 -size -150,150 -h -mask -S 10 -bg -len 8,10,12,15 input sequences: summits of the above indicated peaks generated by MACS2 -bg/background sequences: BRD2 peaks associated with the downregulated genes was used for the motif search using BRD2 peaks associated with the upregulated genes and vice versa. To search motifs of non BET proteins which are recruited to acetylated chromatin, we performed motif search using H4K5acK8ac peaks that do not overlap with BRD2 as input sequences using the following parameter: findMotifsGenome.pl input hg19 -size -150,150 -h -mask -S 10 -bg -len 6,8,12 background sequences: summits of all H4K5acK8ac peaks. To visualize enrichment and occurrence of the identified motifs in the peak regions, HOMER’s “the annotate peak” function was performed.

Motif Activity Response Analysis (MARA) To identify key TFs, which might be involved in transcriptional regulation affected by JQ1 and the transcriptional activities of their target genes, we carried out MARA30. We submitted the bam files with mapped CAGE sequencing tags (hg19) to ISMARA online tool (The Integrated System for Motif Activity Response Analysis, http://ismara.unibas.ch/fcgi/mara). ISMARA integrates the gene expression data and the predicted transcription factor binding sites to provide estimates of TF motif activities.

REFERENCES 47. Mukai, T. et al. Genetic-code evolution for protein synthesis with non-natural

amino acids. Biochem. Biophys. Res. Commun. 411, 757–761 (2011).

48. Hayashi-Takanaka, Y. et al. Distribution of histone H4 modifications as

revealed by a panel of specific monoclonal antibodies. Res. 23,

753–766 (2015).

12 49. Zhu, Y. Y., Machleder, E. M., Chenchik, A., Li, R. & Siebert, P. D. Reverse

transcriptase template switching: A SMARTtm approach for full-length cDNA

library construction. BioTechniques 30, 892–897 (2001).

50. Otwinowski, Z. & Minor, W. Processing of X-ray diffraction data collected in

oscillation mode. Methods in Enzymology 276, 307–326 (1997).

51. McCoy, A. J. et al. Phaser crystallographic software. J. Appl. Crystallogr. 40,

658–674 (2007).

52. Emsley, P., Lohkamp, B., Scott, W. G. & Cowtan, K. Features and development

of Coot. Acta Crystallogr. D Biol. Crystallogr. 66, 486–501 (2010).

53. Adams, P. D. et al. PHENIX: A comprehensive Python-based system for

macromolecular structure solution. Acta Crystallogr. D Biol. Crystallogr. 66,

213–221 (2010).

54. Sato, Y. et al. A Genetically Encoded Probe for Live-Cell Imaging of H4K20

Monomethylation. J. Mol. Biol. 428, 3885–3902 (2016).

55. Kimura, H., Hayashi-Takanaka, Y., Goto, Y., Takizawa, N. & Nozaki, N. The

Organization of Histone H3 Modifications as Revealed by a Panel of Specific

Monoclonal Antibodies. Cell Struct. Funct. 33, 61–73 (2008).

56. Robinson, M. D., McCarthy, D. J. & Smyth, G. K. edgeR: a Bioconductor package

for differential expression analysis of digital gene expression data.

Bioinforma. Oxf. Engl. 26, 139–40 (2010).

57. Kodzius, R. et al. CAGE: cap analysis of gene expression. Nat. Methods 3, 211–

222 (2006).

58. Hon, C.-C. et al. An atlas of human long non-coding RNAs with accurate 5′

ends. Nature 543, 199–204 (2017).

13 59. Subramanian, A. et al. Gene set enrichment analysis: A knowledge-based

approach for interpreting genome-wide expression profiles. Proc. Natl. Acad.

Sci. 102, 15545–15550 (2005).

60. Feng, J., Liu, T., Qin, B., Zhang, Y. & Liu, X. S. Identifying ChIP-seq enrichment

using MACS. Nat. Protoc. 7, 1728–1740 (2012).

61. Akalin, A., Franke, V., Vlahoviček, K., Mason, C. E. & Schübeler, D. Genomation:

A toolkit to summarize, annotate and visualize genomic intervals.

Bioinformatics 31, 1127–1129 (2015).

62. Whyte, W. A. et al. Master transcription factors and mediator establish super-

enhancers at key cell identity genes. Cell 153, 307–319 (2013).

63. Arner, E. et al. Transcribed enhancers lead waves of coordinated

transcription in transitioning mammalian cells. Science 347, 1010–4 (2015).

64. Andersson, R. et al. An atlas of active enhancers across human cell types and

tissues. Nature 507, 455–61 (2014).

65. Shen, L., Shao, N., Liu, X. & Nestler, E. ngs.plot: Quick mining and visualization

of next-generation sequencing data by integrating genomic databases. BMC

Genomics 15, 284 (2014).

14 Supplementary Figure 1 a b unmodified 1 2 3 4 5 6 7 K5ac+K8ac 1. unmodified 0.4 K8ac+K12ac 2A7D9 2. K5ac K5ac ) 3. K8ac 0.3 K8ac 4. K5ac+K8ac 1A9D7 5. K8ac+K12ac+K16ac D(495 0.2 O H4ac 6. K5ac+K12ac+K16ac 7. K5ac+K8ac+K12ac+K16ac 0.1

0 3-1 3-3 3-5 3-7 (%) Serum dilution c VH CDR H1 2A7D9: 1 QIQLVQSGPELKKPGETVKISCKASGYTFTDYSMHWVKQAPGKGLKWMG 1A9D7: 1 ...... CDR H2 CDR H3 WINTATGEPTYADDFKG RFAFSLETSASTAYLQINNLKNEDTATYFCGRDYWGQGTTLTVSS 111 ....E...... T...... 111

VL CDR L1 CDR L2 2A7D9: 1 DIQMIQSPSSLSASLGGKVTITCKASQDINKYIA WFQHKPGKGPTLLIYYTSTLQP 1A9D7: 1 ....T...... D...... R...... N...... R...... CDR L3 GIPSRFSGSGSGRHYSFSISNLEPEDIATYYCLQYDNLRTFGGGTKLEIK 106 ...... D.....R...... 106

Supplementary Figure 1. Validation of H4K5acK8ac antibodies. (a) ELISA assay for 1A9D7 antibody. Histone H4 protein containing indicated site- specific acetyllysine(s) are used as antigens. (b) Western blot analysis for 2A7D9 and 1A9D7 antibodies to acetylated H4 proteins. Lanes are as follows: Lane 1, Non- acetylated H4; lane 2, H4 with K5ac; lane 3, H4 with K8ac; lane 4, H4 with K5ac and K8ac; lane 5, H4 with K8ac, K12ac and K16ac; lane 6, H4 with K5ac, K12ac and K16ac; and lane 7, H4 with K5ac, K8ac, K12ac and K16ac. The bottom panel shows Coomassie Brilliant Blue (CBB)-stained H4 proteins used for western blotting. (c) Sequence alignment of 2A7D9 (top) and 1A9D7 (bottom). In the alignment, the conserved (identical) amino acid is denoted by a dot. Complementarity-determining regions (CDRs) are highlighted in red. The 54th amino acids of VH are circled.

15 Supplementary Figure 2 a

Hoechst33342 H4K5acK8ac H3K27ac H4K5acK8ac H4K5acK8ac

H3K27ac 10 µm H3K27ac Hoechst33342 H4K5acK8ac H3K4me1 H4K5acK8ac H4K5acK8ac H3K4me1 H3K4me1 Hoechst33342 H4K5acK8ac H3K4me3 H4K5acK8ac H4K5acK8ac H3K4me3 H3K4me3 c b Promoter-associated H4K5acK8ac Enhancer-associated H4K5acK8ac

1.3% 0.2% 4.5% 5'UTR P300 P300 Promoter BRD4 BRD4 15.8% BRD2 BRD2 Exon 1.6% 0.3% Intron 3'UTR

21.7% 47.5% average score Stop site average score 0.1 0.2 0.3 Intergenic 0.1 0.3 0.6 ncRNA −4−2 0 2 4 −4−2 0 2 4 7.2% Kb to the peak center Kb to the peak center

Supplementary Figure 2. Genome-wide characterization of H4K5acK8ac and BRD2 and association of BRD2 and histone H4 acetylation. (a) Co-localization of H4K5acK8ac with H3K27ac (upper), H3K4me1 (middle), and H3K4me3 (lower) as validated by immunofluorescence (IF). Hoechst33342 (left) represents nuclei staining. Scale bar, 10 µm. The most right panel indicates correlation between IF signals derived from the immunostaining of the indicated antibodies. (b) Genomic localization of BRD2 binding sites. Promoter is defined as a region within ± 1 kb from transcription start sites/TSS (RefSeq). Intergenic region is a region outside 1 kb upstream of TSS or 1 kb downstream from transcription stop site. (c) BRD2, BRD4, and EP300 are enriched within ± 5 kb of the center of promoter- and enhancer-associated H4K5acK8ac peaks (left and right). H4K5acK8ac peaks, which are located within or outside ± 5 kb from TSS, are defined as promoter- or enhancer-associated H4K5acK8ac, respectively. Publicly available BRD4 ChIP-seq from SCLC cell line H2171 and EP300 ChIP-seq A549 were used.

16

Supplementary Figure 3 a b Chromatin mark frequency ChromHMM from ChIP-seq, biological replicate 1, 20 state 2.2 17.3 0.3 71.1 89.9 4.4 20.3 0.5 1 Genomic 0.5 0.0 42.7 96.6 98.6 77.9 18.9 0.2 2 Emission Position to TSS (bp) Distribution 0.7 1.1 1.2 98.4 99.3 95.2 99.4 31.0 3 0.4 0.7 3.0 7.7 99.8 94.2 65.9 0.3 4 80 0.1 0.1 0.4 0.0 100.0 99.9 100.0 63.1 5 9.4 0.2 2.5 8.7 0.0 61.2 16.4 0.3 6 60 1.9 0.1 4.6 96.6 0.2 99.7 96.3 8.2 7 0.5 0.3 0.0 61.6 0.9 19.2 71.7 1.5 8 0.1 0.0 0.2 0.8 0.0 0.4 0.4 0.2 9 40 0.0 0.0 3.6 97.7 0.0 87.1 10.7 0.2 10 0.1 0.1 1.3 77.0 0.0 3.5 0.5 0.2 11 20 0.3 0.0 79.6 77.6 0.1 12.1 0.0 0.2 12 2.2 0.0 94.0 78.5 0.1 93.3 16.0 0.4 13 0 0.2 0.0 13.8 0.9 0.0 0.3 0.0 0.1 14 0.2 0.0 90.8 1.0 0.0 3.4 0.0 0.1 15 75.3 1.4 83.1 0.9 0.0 1.5 0.1 0.2 16 53.8 3.4 0.2 0.1 0.0 0.1 0.1 0.2 17 0.5 0.3 0.0 0.1 0.0 0.0 0.0 0.1 18 71.2 77.1 0.0 0.5 0.1 0.0 0.2 0.6 19 1.7 25.8 0.0 0.3 0.0 0.0 0.0 0.1 20 H3K9me3 H3K27me3 H3K36me3 H3K4me1 H3K4me3 H3K27ac H4K5K8diac BRD2

ChromHMM from ChIP-seq, biological replicate 2, 20 state Genomic Emission Position to TSS (bp) Distribution

c H3K27ac enhancer (E2, E10) H4K5acK8ac enhancer (E8)

0.20 BRD2 BRD4 P300 average score 0.10 0.08 0.12 0.16 0.20 −4−2 se 2 4 −4−2 se 2 4 Kb to start (s) and end(e) Kb to start (s) and end(e) of the state of the state

Supplementary Figure 3. Global chromatin states by ChromHMM. (a) Frequency of each histone mark or transcription factor BRD2 in each state. The frequency ratio is used to predict the chromatin states shown in Fig. 3c. The color density indicates enrichment. (b) ChromHMM using ChIP-seq data set derived from the biological replicate 1 (upper) and biological replicate 2 (lower). The same types of chromatin states were identified from both biological replicates and from the pooled ChIP-seq data set shown in Fig. 3b. (c) Normalized average plots showing enrichment of BRD2, BRD4 (H2171) and EP300 (A450) ChIP-seq intensities in the H3K27ac enhancer state (left, E2 and E10) and H4K5acK8ac enhancer state (right, E8).

17

Supplementary Figure 4 a b H3K27ac within H3K27ac within decreased H4K5acK8ac peaks increased H4K5acK8ac peaks

JQ1 DMSO JQ1 1.2 DMSO 0.8 0.4 0.6 01.01.5 average score 0.4 average score 00.81.0

−4−2 0 2 4 −4−2 0 2 4 −4−2 0 2 4 Kb to the peak center Kb to the peak center Kb to the peak center

Supplementary Figure 4. JQ1 effect on H3K27ac peaks. (a) 500 nM JQ1 treatment for 24 hrs did not affect the global H3K27ac sites, as indicated by normalized average plot. H3K27ac ChIP-seq signals from JQ1 (red) or DMSO (blue) treated cells were plotted into the robust H3K27ac ChIP-seq peaks from the untreated cells (5 kb ± from the peak summit). (b) Decrease of H3K27ac intensities was detected in the JQ1-decreased H4K5acK8ac peaks upon JQ1 treatment (red), compared to the DMSO (blue). (c) Increase of H3K27ac intensities was observed in the JQ1-increased H4K5acK8ac peaks upon JQ1 treatment (red), compared to the DMSO (blue).

18

Supplementary Figure 5 a Downregulated promoters Upregulated promoters Downregulated 0 n=1,538 n=3,066 10 6 n=1,406 -2 8 n=1,257 n=2,336 n=3,505 4 Upregulated -4 6 -6 4 2 n=2,048 Number of -8 2 genes (x1,000) n=6,500 Log2FC cage signal 0 Log2FC cage signal -10 0 361224 361224 361224 Hours after treatment Hours after treatment Hours after treatment b 8

7 3h 6 6h 5 12h 24h 4 DMSO 3 (compared to DMSO) relative expression level 2

1

0

MYC BRD4 KRAS RAF1 BCL2 E2F4 JUN NRF1 CFLAR WNT2 FOSL1 PYGO-2 WNT2B WNT5A-2 CTNNB1 TNFRSF11 BRD2 BRD21st TSS 2nd TSS

CATG00000039382.1 c d expression by CAGE expression by CAGE 0 0 BCL2 HOXC10 FLIP/CFLAR -0.5 HOXC4 HOXC6 -1 -1 -1.5 -2 -1.8 -2.5 Log2FC JQ1 vs. DMSO vs. JQ1 Log2FC Log2FC JQ1 vs. DMSO vs. JQ1 Log2FC 3 6 12 24 3 6 12 24 Hours after JQ1 treatment Hours after treatment

19 e REACTOME_SIGNALING REACTOME_APOPTOSIS _BY_WNT 0.5 0.5 ES

NES 2.2 ES NES 2.33 0.0 FDR 6.214575E-5 0.0 FDR 0.0 Hits Hits

Upregulated Downregulated Upregulated Downregulated f KEGG_CYTOKINE_CYTOKINE HALLMARK_KRAS REACTOME_GPCR_ RECEPTOR INTERACTION _SIGNALING_DN DOWNSTREAM_SIGNALING 0.5 0.0 0.0 0.0 ES ES NES -1.39 NES -1.78 ES NES -1.97 FDR 0.5 -0.35 -0.45 FDR 0.09 -0.5 FDR 0.01 Hits Hits Hits Upregulated Downregulated Upregulated Downregulated Upregulated Downregulated g SCHLOSSER_MYC_TARGETS_ SCHUHMACHER_MYC_UP DANG_MYC_TARGETS_UP 0.6 REPRESSED_BY_SERUM 0.6 0.5 ES ES ES NES 2.58 NES 2.74 NES 2.77 FDR 1.0E-10 FDR 1.0E-10 FDR 1.0E-10 0 0 0 Hits Hits Hits Upregulated Downregulated Upregulated Downregulated Upregulated Downregulated

E2F_03 0.3

ES NES 1.72 FDR 0.01 0 Hits Upregulated Downregulated

Supplementary Figure 5. JQ1 effects on the gene expression in H23 cells. (a) Number of the genes that were downregulated (blue) or upregulated (red) increased with the prolonged JQ1 treatment (left panel); the middle panel shows the number of downregulated gene time (log2FC of CAGE signal ≤ -1) increased with the incubation; the right panel: the number of upregulated genes (log2FC of CAGE signal ≥ 1). (b) Validation of the expression of 20 genes upon JQ1 treatment at 3, 6, 12, and 24 hrs. The gene expression level was determined from the JQ1- and DMSO-treated cells and normalized against the housekeeping gene GAPDH. (c) Bar plot showing the decrease of BCL2 and CFLAR expression as upon JQ1 treatment. Log2FC of CAGE signals calculated from the JQ1- and DMSO-treated cells. (d) Bar plot showing the decrease of HOXC10, HOXC6, and HOXC4 expression as upon JQ1 treatment. Log2FC of CAGE signals calculated from the JQ1- and DMSO-treated cells. (e-g) GSEA of differentially expression genes showing the enrichment of: (e) Apoptosis pathway and signaling by Wnt in the upregulated genes; (f) Cytokine-Cytokine receptor interaction, target genes of KRAS signaling, and G-coupled protein receptor (GPCR) in the downregulated genes; (g) MYC target genes from multiple datasets: Schumacher, Schlosser, Dang (upper), and E2F (lower) target genes in the upregulated genes. ES: normalized score, NES: normalized enrichment score, Hits: upregulated genes (red) and downregulated genes (blue).

20 Supplementary Figure 6

a 4.0e-05 CTCF motif (MA0139.1) downregulated 2 upregulated

bits 1

1 2 3 4 5 6 7 8 9 10 11 1213141516171819 -5 0 5 Kb to the peak center

b Upregulated genes Downregulated genes

CTCF BRD2 average score 0.2 0.4 0.6 0.8 0.2 0.4 0.6 0.8

−4−2 0 2 4 −4−2 0 2 4 Kb to the peak center Kb to the peak center

Supplementary Figure 6. Enrichment of CTCF in the BRD2 peaks. (a) CTCF motif enrichment in the BRD2 peaks. Motif enrichment analysis was done using HOMER with CTCF motif from Jaspar (left panel). The right panel: y-axis: motif per base-pair per peak within 6 kb ± from the peak summit (x-axis) in two different sets of BRD2 peaks at the downregulated (orange) and upregulated (green) genes. (b) Enrichment of CTCF ChIP-seq signals (ENCODE ChIP-seq in A549) within the BRD2 peaks associated with the upregulated genes (left) and downregulated genes (right). Red: CTCF, green: BRD2 signals.

21 Supplementary Figure 7 chr14:75734229-75760184 FOS 3.8 H4K5acK8ac 5.7 H3K27ac 0.5 BRD2 H3K4me3 8.9 H3K4me1 0.5

chr7:22765552-22772835 IL6 0.7 H4K5acK8ac H3K27ac 0.8 BRD2 0.5 H3K4me3 0.07 H3K4me1 1

Supplementary Figure 7. Genome browser view of the genes whose promoter is not bound by BRD2: FOS (upper) and IL6 (lower).

22 Supplementary Table 1

Anti-Histone H4 (acetyl K5 + K8 + K12 + K16) Anti-hyperacetylated Histone H4 (Penta) Anti H4K5acK8ac Anti H4K5acK8ac antibody (Abcam ab177790) antibody (Upstate 06-946) antibody - 1A9D7 antibody- 2A7D9

L L L L M M M M N N N N O O O O P P P P 1 24 1 24 1 24 1 24 Antibody Modification number Location Sequence Name Mod1 Mod2 Mod 3 Mod 4 N-terminus ab177790 Upstate 06-946 1A9D7 2A7D9 275 L11 R K S A P A T G G V Kac K P H R Y R P G H3 26-45 K36ac acetylated 276 L12 S G R G K G G K G L G K G G A K R H R H4 1-19 unmod free 277 L13 pS G R G K G G K G L G K G G A K R H R H4 1-19 S1P free 278 L14 S G Rme2s G K G G K G L G K G G A K R H R H4 1-19 R3me2s free 279 L15 S G Rme2a G K G G K G L G K G G A K R H R H4 1-19 R3me2a free 280 L16 S G R G Kac G G K G L G K G G A K R H R H4 1-19 K5ac free 281 L17 S G R G K G G Kac G L G K G G A K R H R H4 1-19 K8ac free 282 L18 S G R G K G G K G L G Kac G G A K R H R H4 1-19 K12ac free 286 L22 pS G R G Kac G G K G L G K G G A K R H R H4 1-19 S1P K5ac free 287 L23 S G Rme2s G Kac G G K G L G K G G A K R H R H4 1-19 R3me2s K5ac free 288 L24 S G Rme2s G K G G Kac G L G K G G A K R H R H4 1-19 R3me2s K8ac free 289 M 1 S G Rme2a G Kac G G K G L G K G G A K R H R H4 1-19 R3me2a K5ac free 290 M 2 S G Rme2a G K G G Kac G L G K G G A K R H R H4 1-19 R3me2a K8ac free 291 M 3 S G R G Kac G G Kac G L G K G G A K R H R H4 1-19 K5ac K8ac free 292 M 4 S G R G K G G Kac G L G Kac G G A K R H R H4 1-19 K8ac K12ac free 293 M 5 S G R G K G G Kac G L G K G G A Kac R H R H4 1-19 K8ac K16ac free 294 M 6 S G R G K G G K G L G Kac G G A Kac R H R H4 1-19 K12ac K16ac free 295 M 7 pS G Rme2s G Kac G G K G L G K G G A K R H R H4 1-19 S1P R3me2s K5ac free 296 M 8 pS G Rme2a G Kac G G K G L G K G G A K R H R H4 1-19 S1P R3me2a K5ac free 297 M 9 S G Rme2s G Kac G G Kac G L G K G G A K R H R H4 1-19 R3me2s K5ac K8ac free 298 M10 S G Rme2a G Kac G G Kac G L G K G G A K R H R H4 1-19 R3me2a K5ac K8ac free 299 M11 S G R G Kac G G Kac G L G Kac G G A K R H R H4 1-19 K5ac K8ac K12ac free 300 M12 S G R G K G G Kac G L G Kac G G A Kac R H R H4 1-19 K8ac K12ac K16ac free weak 301 M13 pS G Rme2s G Kac G G Kac G L G K G G A K R H R H4 1-19 S1P R3me2s K5ac K8ac free 302 M14 pS G Rme2a G Kac G G Kac G L G K G G A K R H R H4 1-19 S1P R3me2a K5ac K8ac free 303 M15 S G Rme2s G Kac G G Kac G L G Kac G G A K R H R H4 1-19 R3me2s K5ac K8ac K12ac free 304 M16 S G Rme2a G Kac G G Kac G L G Kac G G A K R H R H4 1-19 R3me2a K5ac K8ac K12ac free 305 M17 S G R G Kac G G Kac G L G Kac G G A Kac R H R H4 1-19 K5ac K8ac K12ac K16ac free 306 M18 G K G G A K R H R K V L R D N I Q G I T H4 11-30 unmod acetylated 307 M19 G Kac G G A K R H R K V L R D N I Q G I T H4 11-30 K12ac acetylated 308 M20 G K G G A Kac R H R K V L R D N I Q G I T H4 11-30 K16ac acetylated 309 M21 G K G G A K Rme2s H R K V L R D N I Q G I T H4 11-30 R17me2s acetylated 310 M22 G K G G A K Rme2a H R K V L R D N I Q G I T H4 11-30 R17me2a acetylated 311 M23 G K G G A K R H Rme2s K V L R D N I Q G I T H4 11-30 R19me2s acetylated 312 M24 G K G G A K R H Rme2a K V L R D N I Q G I T H4 11-30 R19me2a acetylated 319 N 7 G Kac G G A Kac R H R K V L R D N I Q G I T H4 11-30 K12ac K16ac acetylated 328 N16 G Kac G G A Kac R H R Kme1 V L R D N I Q G I T H4 11-30 K12ac K16ac K20me1 acetylated 329 N17 G Kac G G A Kac R H R Kme2 V L R D N I Q G I T H4 11-30 K12ac K16ac K20me2 acetylated 330 N18 G Kac G G A Kac R H R Kme3 V L R D N I Q G I T H4 11-30 K12ac K16ac K20me3 acetylated 331 N19 G Kac G G A Kac R H R Kac V L R D N I Q G I T H4 11-30 K12ac K16ac K20ac acetylated 366 P 6 P D P A K S A P A P K Kac G S K K A V T H2B 1-19 K12ac free weak 369 P 9 P D P A Kac S A P A P K Kac G S K K A V T H2B 1-19 K5ac K12ac free weak 373 P13 P D P A K S A P A P K Kac G S Kac K A V T H2B 1-19 K12ac K15ac free 376 P16 P D P A Kac S A P A P K Kac G S Kac K A V T H2B 1-19 K5ac K12ac K15ac free 377 P17 P D P A Kac S A P A P K K G pS Kac K A V T H2B 1-19 K5ac S14P K15ac free weak 378 P18 P D P A K S A P A P K Kac G pS Kac K A V T H2B 1-19 K12ac S14P K15ac free weak weak 379 P19 P D P A Kac S A P A P K Kac G pS Kac K A V T H2B 1-19 K5ac K12ac S14P K15ac free weak weak

Supplementary Table 1. Peptide array analysis using anti-H4K5acK8ac antibodies and two commercially available antibodies. The upper panel shows images of ECL detection of the MODified histone peptide array (Active Motif, 13001) containing 384 different patterns of histone tail modification from the indicated antibodies (from the left to right): anti-histone H4 (acetyl K5+K8+K12+K16) antibody (Abcam, ab177790); anti-hyperacetylated histone H4 (Penta) antibody (Upstate, 06-946); 1A9D7 antibody; and 2A7D9 antibody. The lower table is the list of the peptides, which were recognized by the tested antibodies.

23 Supplementary Table 2

2A7D9 Fab 1A9D7 Fab

+ H4(1–12) K5acK8ac + H4(1–12) K5acK8ac Data collection

Space group P212121 P21 Cell dimensions

a, b, c (Å) 39.95, 85.98, 122.85 39.35, 164.09, 73.70 α, β, γ (°) 90.00, 90.00, 90.00 90.00, 104.50, 90.00 50.00–1.70 50.00–1.80 Resolution (Å) (1.73–1.70)* (1.83–1.80)*

Rsym 0.070 (0.623) 0.053 (0.334) I / σI 26.8 (2.3) 28.3 (2.5) Completeness (%) 98.8 (93.6) 99.0 (96.8) Redundancy 6.7 (6.2) 3.5 (3.1)

Refinement

Resolution (Å) 36.97–1.70 37.77–1.80 No. reflections 46,824 82,524

Rwork / Rfree 0.166 / 0.195 0.194 / 0.224 No. atoms

Protein 3,315 6,599 Ligand / ion - 26 Water 522 789 B-factors

Protein 16.57 22.31 Ligand / ion - 54.54 Water 30.61 34.32 R.m.s. deviations

Bond lengths (Å) 0.006 0.004 Bond angles (°) 0.87 0.75

PDB ID 5YE3 5YE4

Supplementary Table 2. Crystallographic data collection and refinement statistics *Values in parentheses are for highest-resolution shell.

24