Reagents and cell lines

Antibodies recognizing H3K4me3 (Millipore, 04–745, RRID: AB_1163444; CST, 9751, RRID:AB_2616028), H3K27ac (Abcam, ab4729, RRID: AB_2118291), MLL1 (CST, 14,197, RRID:AB_2688010), JUN (CST, 9165, RRID:AB_2130165), CD3 (Abcam, ab16669, RRID:AB_443425), CD4 (Abcam, ab183685, RRID:AB_2686917), CD8 (Abcam, ab217344, RRID:AB_2890649), F4/80(CST, 70,076, RRID:AB_2799771), CD206 (Proteintech, 60143–1-Ig, RRID:AB_2144924), NKp46 (Abcam, ab233558, RRID:AB_2904203), Ki67(Abcam, ab15580, RRID:AB_443209) were purchased from indicated commercial sources. Protein G-Sepharose beads (GE Healthcare), AOM (MP Biomedicals, 183971), DSS (MP Biomedicals, 160110), CBX8 (ABclonal Cat# A6222, RRID:AB_2766831), CBX2 (ABclonal Cat# A3294, RRID:AB_2863036), CBX4 (ABclonal Cat# A5109, RRID:AB_2863448), RPS6KA5 (ABclonal Cat# A5699, RRID:AB_2766458), GPR68 (ABclonal Cat# A7348, RRID:AB_2767885), CCND1 (ABclonal Cat# A19038, RRID:AB_2862530), JAK3 (ABclonal Cat# A0748, RRID:AB_2757376), KRT18 (ABclonal Cat# A19778), PTP4A3(ABclonal Cat# A2004, RRID:AB_2764028), SNAI1 (ABclonal Cat# A11794, RRID:AB_2758760), TNFSF10 (ABclonal Cat# A2138, RRID:AB_2764157), ACTB ( ABclonal Cat# AC038, RRID:AB_2863784), GAPDH (ABclonal Cat# AC033, RRID:AB_2769570) were purchased from the indicated companies. PCR primers were custom synthesized by Tsingke Biotechnology and siRNAs by GenePharma. HEK293T, HCT116 and RKO Cell lines were purchased from Cell Bank of Chinese Academy and cultured under recommended conditions according to the manufacturer’s instruction with 10% FBS. Cell lines were verified by the merchant before shipping, and no mycoplasma was detected.

Animal housing and ethics approval

The BALB/C nude mice and C57BL/6 J mice were purchased from GemPharmatech Co., Ltd. All the mice were born and maintained under pathogen-free condition at 20 ~ 24℃ with a humidity of 40 ~ 70% and a 12/12-h dark/light cycle (lights on at 7:00 AM, lights off at 7:00 PM), with free access of water and food (Animal Center of College of Life Sciences, Wuhan University).

Generation of colitis-associated colorectal cancer mice model

The 8-week-old C57BL/6 J mice were randomly divided into two experimental groups. Mice in the experimental groups were given a single intraperitoneal injection of AOM (10 mg/kg body weight). Seven days after the AOM injection, the mice were given 2.5% DSS (w/v) in drinking water for 7 days, and then fed with distilled water for 14 days, which was repeated at 4-week and 7-week. The mice were intraperitoneally injected with DMSO or OICR-9429 (5 mg/kg) at 3-week, 6-week, and 9-week (three time per week). All the mice were sacrificed at 10-week. The colorectum tissues were divided to distal, middle, proximal fragments to anus, and only the proximal fragments were collected for experiments.

ChIP assay

ChIP assay was performed as previously described [55, 56]. Briefly, cells were cross-linked with 1% formaldehyde for 10 min at room temperature and quenched with 0.125 M glycine for 5 min. Cross-linked cells were washed twice with PBS, then collected by centrifugation. Then cells were re-suspended with the ChIP digestion buffer (50 mM of Tris–HCl, pH 8.0, 1 mM of CaCl2, 0.2% Triton X-100). Chromatin was digested to 150- 300 bp by MNase (M0247S, NEB) at 37 °C for 20 min and quenched with EDTA. Supernatant was collected and equally divided after diluted with five times of dilution buffer (20 mM of Tris–HCl, pH 8.0, 150 mM of NaCl, 2 mM of EDTA, 1% Triton X-100, 0.1% SDS). Samples were then incubated with protein G beads and antibodies at 4 °C overnight. Next day, the beads were washed once with wash buffer I (20 mM Tris–HCl, pH 8.0, 150 mM NaCl, 2 mM EDTA, 1% Triton X-100, 0.1% SDS), once with wash buffer II (20 mM Tris–HCl pH 8.0, 500 mM NaCl, 2 mM EDTA, 1% Triton X-100, 0.1% SDS), once with wash buffer III (10 mM Tris–HCl pH 8.0, 250 mM LiCl, 1 mM EDTA, 1% Na-deoxycholate, 1% NP-40) and twice with TE (10 mM Tris–HCl pH 8.0, 1 mM EDTA). The beads were eluted twice with 100 μL elution buffer (1% SDS, 0.1 M NaHCO3, 0.2 mg/ml Proteinase K (Biosharp)) at room temperature. The elution was incubated at 65 °C for 6 h and then purified with DNA purification kit (TIANGEN DP214-03). The purified DNA was assayed by quantitative PCR with quantitative PCR. Primers for ChIP-qPCR were in Additional file 3: Table S2.

Native ChIP

Briefly, 2 × 107 cells were centrifuged at 3,000 rpm for 5 min at 4℃. After washing them once with cold PBS, re-suspend the cells in nuclear extraction buffer (10 mM of Tris, pH 8.0, 10 mM NaCl, 0.2% NP-40, proteinase inhibitor cocktail (1x), 1 mM DTT and fresh PMSF) for 30 min at 4℃. Nuclei was collected by centrifuge and resuspended in MNase digestion buffer (50 mM Tris–HCl, pH7.6, 1 mM CaCl2, 0.2% Triton X-100 or NP-40, proteinase inhibitor cocktail (1x), 1 mM DTT and fresh PMSF). Chromatin was digested by MNase (M0247S, NEB) at 37℃ for 20 min and quenched with EDTA to a final concentration of 5 mM. Supernatant was collected and equally divided after diluted with five times of dilution buffer (20 mM of Tris–HCl, pH 8.0, 150 mM of NaCl, 2 mM of EDTA, 1% Triton X-100, 0.1% SDS). Samples were then incubated with protein G beads and antibodies at 4℃ overnight. Next day, the beads were washed once with wash buffer I, II, III, and the following procedures were performed as described in ChIP assay. Primer information was in Additional file 3: Table S2.

Capture-ChIP

2 × 107 FB-EGFP or FB-JUN/BirA-expressing HCT116 or RKO stable cells were harvested, cross-linked with 2 mM SMCC (MCE) for 45 min and 0.2% formaldehyde for 10 min at room temperature, and then quenched with 0.125 M glycine for 5 min. Cross-linked cells were washed twice with PBS, then collected by centrifugation. Cells were lysed in 1 mL RIPA buffer (10 mM Tris–HCl pH 8.0, 1 mM EDTA, 0.1% sodium deoxycholate, 0.1% SDS, 1% Triton X-100), and rotated for 15 min at 4℃. Cell lysates were centrifuged at 2,300 g for 5 min at 4℃ to isolate the nuclei. Nuclei were suspended in 500 μl of 0.5% SDS lysis buffer (0.5% SDS, 10 mM EDTA, 50 mM Tris–HCl, pH 8.0) and subjected for sonication to shear chromatin fragments to 200—500 bp. Fragmented chromatin was centrifuged at 16,100 g for 10 min at 4℃. 450 mL of supernatant was transferred to a new tube and NaCl solution was added to the final concentration of 300 mM. Supernatant was then incubated with 10 μl of Dynabeads MyOne streptavidin C1 (Thermo-Fisher Scientific) at 4 ℃ overnight. Then the Dynabeads were washed twice with 2% SDS, twice with RIPA buffer containing 0.5 M NaCl, twice with LiCl buffer (250 mM LiCl, 0.5% NP-40, 0.5% sodium deoxycholate, 1 mM EDTA and 10 mM Tris–HCl, pH 8.0), and twice with TE buffer (10 mM Tris–HCl, 1 mM EDTA, pH 8.0). The chromatin was eluted in SDS elution buffer (1% SDS, 10 mM EDTA, 50 mM Tris–HCl, pH 8.0) followed by incubation at 65℃ overnight. DNA was treated with RNase A (5 mg/ml, Sigma) and protease K (0.2 mg/ml, Biosharp) at 37℃ for 30 min, and extracted with DNA purification kit (TIANGEN). The purified DNA was assayed by quantitative PCR with quantitative PCR. Primers for ChIP-qPCR were listed in Additional file 3: Table S2.

Library preparation for ChIP-sequencing

ChIP-seq libraries were constructed with ChIP and input DNA using VATHS Universal DNA Library Prep Kit for Illumina (Vazyme ND606). Briefly, 50 μL of DNA (8–10 ng) was end-repaired for dA tailing, followed by adaptor ligation. Each adaptor was marked with a barcode of 8 bp DNA. Adaptor-ligated DNA was purified by AMPure XP beads (1:1) and then amplified by PCR of 9 cycles with the primer matching with adaptor universal part. Amplified DNA was purified again using AMPure XP beads (1:1) in 35 μL EB elution buffer. For multiplexing, libraries with different barcodes were mixed with equal molar quantities (30–50 million reads per library). Libraries were sequenced by Illumina Nova-seq platform with pair-end reads of 150 bp.

RNA-sequencing

RNA extraction was performed using Ultrapure RNA Kit (CWBIO, CW0581M). Briefly, around 40 mg tissues were triturated for 30 s in 1 mL TRIzon provided in the kit, incubated at room temperature for 5 min, added with 200 μL chloroform and shaken drastically. After centrifugation at 12,000 rpm, 4 °C for 10 min, the upper water phase was moved into an adsorption column provided by the kit. The column was then eluted with 50 μL RNase-free water. RNA-seq libraries were constructed by NEBNext Poly(A) mRNA Magnetic Isolation Module (NEB E7490) and NEBNext Ultra II Non-Directional RNA Second Strand Synthesis Module (NEB E6111). mRNA was purified with poly-T magnetic beads and first and second strand cDNA was synthesized. The resulted cDNA was purified by AMPure XP beads (1:1) and eluted in 50 μL nucleotide-free water. The subsequent procedures were the same as described in ChIP-seq library construction, except that the sequencing depth was 20 million reads per library. RNA-seq libraries were sequenced by Illumina Nova-seq platform with pair-end reads of 150 bp.

ChIP-seq data analysis

ChIP-seq raw fastq data were quality controlled using FastQC (version 0.11.9, https://www.bioinformatics.babraham.ac.uk/projects/fastqc/). Clean data were obtained by removing adapter with fastp (version 0.21.0, https://github.com/OpenGene/fastp, parameters were “-f 5 -t 20 -F 5 -T 20 -l 30”). Cleaned reads were aligned in paired-end mode to homo sapiens UCSC reference genome hg19 with BWA mem (version 0.7.12, http://bio-bwa.sourceforge.net) [57]. Duplicate reads were removed by picard Markduplicates (version 2.18.29, https://broadinstitute.github.io/picard/). All mapped reads were used in the analysis. Peaks were called by MACS2 (version 2.1.1, https://github.com/taoliu/MACS, parameters “–nomodel -p 1E-7 -B –broad –extsize 147”) [58]. For the comparison in enhancers, signal of modification in each enhancer was normalized by reads per kilobase per million mapped reads (RPKM).

RNA-seq data analysis

RNA-seq clean data were obtained same as ChIP-seq data (parameters were “-f 5 -t -20 -F 5 -T -20 -l 30”). Quality control was done with FastQC (version 0.11.9, https://www.bioinformatics.babraham.ac.uk/projects/fastqc/) and Multiqc (version 1.10.1, https://multiqc.info). Clean reads were aligned to homo sapiens UCSC hg19 genome with STAR (version 2.7.9, https://github.com/alexdobin/STAR) [59]. Gene annotation was done using UCSC hg19 genome. The gene expression level was normalized as fragments per kilobase of bin per million mapped reads (FPKM).

Identification of VELs (Variant enhancer loci)

To identify the significant variant enhancer loci between native and tumor tissues, we first identified all VELs in paired native and tumor tissues, defined as enhancers whose H3K27ac fold change (FC) >  = 2 between native and tumor tissues. We merged all VELs into one single coordinate file, and calculated the recurrence and significance (Benjamini & Hochberg corrected p-value) for all VELs. We used recurrence of 8 and 9 as significance threshold for gain and lost VELs, respectively, because gain and lost VELs achieved the significant percentage cut-off (0.90) when recurrence larger than these numbers.

Identification of H3K4me3-enriched enhancers

To identify the significant v3mEs between native and tumor tissue, we first identified active enhancers based on H3K27 peaks 2 kb away from transcription start site (TSS), and then overlapped them with H3K4me3 peaks 2 kb away from TSS.

These enhancers were defined as m3Es. Then we identified all v3mEs in paired native and tumor tissues. V3mEs of Individual samples were defined as m3Es whose H3K27ac fold change (FC) >  = 2 between native and tumor tissues. We merged all m3Es into one single coordinate file, and calculated the mean H3K27ac signal and normalized to RPKM. We used fold change (FC) of 1.5 as significance threshold for Vm3Es.

Identification of VSELs

For variant super-enhancer loci (VSEL), the identifying procedure was similar as described above in “Identification of VELs”. We used recurrence of 5 as significance threshold for both gain and lost VELs.

Transcript factor enrichment

We used the findMotifsGenome.pl module in HOMER (version 4.11, http://homer.ucsd.edu/homer/, parameters were -size 200 -mask) to identify transcript factor motifs.

Human disease ontology and Gene Ontology analysis

The coordinate file of gain and lost Vm3Es were submitted to GREAT website (version 3.0.0) and the results of human disease ontology and GO analysis (biological process) were obtained for plotting.

ChIP-Seq and RNA-Seq data visualization

UCSC genomic track for histone marks, RNA expression and chromatin state beyond the RefSeq gene model were drew by karyoploteR [60], using the alignment file of chromatin markers and state annotation bed files produced by ChromHMM. Histone marker’s signal density panel across TSS and gene body were plotted by in house R script, using the density matrix data produced by deeptools (version 3.3.2) [61].

Reverse transcription and quantitative PCR

Cells were scraped down and collected with centrifugation. For tissue RNA extraction, 20 mg of tissues were homogenized and collected with centrifugation. Total RNA was extracted with RNA extraction kit (Aidlab or CWBIO) according to the manufacturer’s manual. Approximately 1 μg of total RNA was used for reverse transcription with a first strand cDNA synthesis kit (Toyobo). The resulted cDNA was assayed with quantitative PCR, β-actin for normalization. The sequences of primers are in Additional file 3: Table S2. Assays were repeated at least three times. Data were shown as average values ± SD or SEM of at least three representative experiments. P value was calculated using student’s t test.

Cell proliferation assay

The cell proliferation was measured using the CCK-8 assay. Briefly, 1,000 cells were seeded into 96-well plates. The cells were added with Cell Counting Kit-8 (CCK-8) solution (10 μL; Yeasen) in each well at the indicated time points before incubation for 1—4 h at 37℃, followed by analysis via microplate reader. The absorbance was measured at 450 nm. Assays were repeated at least three times. Data were shown as average values ± SD of at least three representative experiments and p value was calculated using student’s t test.

Immunoprecipitation and immunoblot

Cells were harvested and lysed in NP40 lysis buffer (50 mM Tris, pH 7.4, 150 mM NaCl, 0.5% NP40) in the presence of proteinase inhibitors. After removing insoluble particles, the supernatant was incubated with protein G beads (GE Healthcare) and specific antibody at 4℃ for 4 h. The beads were spin down and washed three times with lysis buffer. Then SDS loading buffer was added to the beads to release proteins for SDS-PAGE and Western blotting. To prepare cell lysates for western blotting, cells lysates were prepared with SDS lysis buffer (50 mM Tris–HCl pH 6.8, 4%SDS). Lysates were separated by SDS-PAGE and transferred to nitrocellulose filter membranes blocked with 5% milk and incubated with primary antibody overnight at 4℃. Then the blots were washed three times in TBS-T, incubated with secondary antibodies at room temperature for 1 h, and detected by Clarity Western ECL Substrate (BIO-RAD).

Cell cycle analysis with flow cytometry

Cells at the logarithmic phase were harvested after digestion with 0.05% Trypsin–EDTA, washed twice with cold PBS and immobilized with ice-cold 70% ethanol overnight. Fixed cells were washed twice with PBS and stained in PBS containing propidium iodide (PI, 50 μg/mL, MCE) and RNase (100 μg/mL, Sigma) for 30 min at 37 °C. Data was generated using cytoflex flow cytometer (Beckman Coulter) and analyzed using FlowJo V10.8.1 software.

Transwell migration assay

1 × 105 HCT116 cells were loaded per transwell insert (24-well insert; pore size, 8 µm; BD Biosciences) without serum or growth factors, and medium supplemented with 10% fetal bovine serum was used as a chemoattractant in the lower chamber. After 36 h of incubation, cells at the underside of the membrane were fixed in methanol for 10 min and stained with crystal violet (0.1% in 20% EtOH for 20 min). After washing, remaining cells in the insert were removed by a cotton swab. Migrated cells were visualized by light microscopy and counted with ImageJ. Assays were repeated at least three times. Data were shown as average values ± SD of at least three representative experiments and p value was calculated using student’s t test.

Generation of knockout/knockdown cell line with CRISPR/ Cas9 system

The single guide RNA (sgRNA) sequences were designed by using the CRISPR Design Tool (http://tools.genomeengineering.org), provided by Feng Zhang lab. The target sequences sgRNAs were shown in Supplementary Table 2. The sgRNAs were cloned in lentiCRISPRv2-puro (Addgene, #98,290). To construct knockdown cell lines, the lentiviral particles were generated by transfecting HEK293T cells with CRISPR plasmid. Then the supernatant was used to infect the desired cells, and cells were selected by puromycin.

CRISPR-Cas9-KRAB mediated repression of variant H3K4me3 enhancers

Site-specific sgRNAs targeting enhancer loci were designed with publicly available filtering tools (https://zlab.bio/guide-design-resources) to minimize off-target cleavage. The target sequences sgRNAs were shown in Supplementary Table 2. For CRISPR interference, sgRNAs were cloned into pLH-spsgRNA2 (Addgene, #64,114) through the BbsI site according to the protocol recommended by Addgene. Lentivirus was generated by transfecting HEK293T cells with sgRNA expression cocktails or pHAGE dCas9-KRAB-MeCP2, together with helper plasmids, psPAX and pMD2G. Medium containing virus was collected at 48 or 72 h after transfection, and filtered with 0.45 μm filters (Millipore). Stable cell lines were generated by infecting HCT116 or RKO with lentivirus expressing dCas9-KRAB-MeCP2 and sgRNAs, and then screened with puromycin and hygromycin for 48 h.

Immunohistochemistry

CRC tissues were resected and fixed in 4% paraformaldehyde for overnight, dehydrated, embedded in paraffin, and sectioned. Tissue Sects. (5 μm) were used for staining. For immunohistochemistry (IHC) detection, tissue sections were deparaffinized, rehydrated, and treated for heat-mediated antigen retrieval. Tumor sections incubated with the primary antibodies against: anti-Ki67, anti-CD3, anti-CD4, anti-CD8, anti-F4/80, anti-CD206, and anti-NKp46 over night at 4 °C. After the primary incubation sections were washed and incubated with Alexa lour 488 or Cy3-labelled secondary antibodies for one hour at room temperature, tumor sections were washed three times and counter stained with DAPI and mount with Fluoromount G. Immunofluorescence images were pictured using Leica Thunder Image DMI8 microscope at 20X magnification and analyzed using ImageJ software.

Isolation and culture of organoids

For derivation of tumoroids, surgically resected colorectal tissues were washed with ice-cold PBS-Abs buffer (phosphate-buffered saline with antibiotic, 1% penicillin/streptomycin, and 100 μg/ml Primocin), and then chopped into 1 mm pieces in ice-cold PBS. The fragments were transferred to 15 ml tube and resuspended with 10 ml PBS, then filtered through a 40 μm cell strainer on a 50 ml tube. The tissues from the strainer were transferred to a 15 mL tube, and digested in digestion medium (advanced DMEM/F12 with 2% FBS, collagenase type II (2.5 mg/ml), collagenase type IV (2.5 mg/ml), and DNaseI (0.1 mg/ml)) at 37 °C. After digestion, cells were filtered through 70 μm cell strainer on a 50 ml tube, crypt-containing fraction was collected by centrifugation, and embedded in Matrigel (R&D). After the Matrigel balls were polymerized, human isolation medium (HIM) was added. HIM is composed of advanced DMEM/F-12 (GIBCO) supplemented with Glutamax (GIBCO), HEPES (GIBCO), B-27 (GIBCO), N-2 (GIBCO), 10 mM nicotinamide (Sigma), 1 mM N-acetyl-L-cysteine (Sigma), 2 μM A8301 (Tocris), 10 nM [Leu15]-gastrin (Sigma), Y-27632 (10 μM, MCE), 100 ng/mL human noggin (MCE), 50 ng/mL human EGF (MCE), 500 ng/ml R-Spondin1 (MCE), SB202190 (10 μM, MCE), CHIR99021(5 μM, MCE) and Blebbistatin (10 μM, MCE). After the first 7 days, isolation medium was changed with human expansion medium (HEM), containing HIM but not Noggin, R-Spondin1 and Y-27632. For passage of organoids, organoids were released from the Matrigel by cell-recovery solution (Corning), dissociated into single cells using TrypLE express (Gibco), and resuspended with Matrigel and divided into a 24-well plate (50 μl per well). After the Matrigel balls were polymerized, 500 μl of HEM was added, and media was changed on organoids every 2–3 days.

Xenograft experiments in mice

Five-week-old male BALB/C nude mice were purchased from GemPharmatech Co., Ltd. Colon cancer model was established by injecting subcutaneously 7.5 × 105 HCT116 cells per site into the flank regions of mice. Tumor volumes were measured every 2 days using calipers. Tumor volumes were calculated as V = 0.5 × length × width2. After 14 days of injection, tumors were harvested and weighed. Data were shown as mean ± SEM and p values were calculated by the student’s t test.

Statistics and reproducibility

For experiments other than NGS sequencing, at least three biological replicates for each experiment were performed. Data are presented as mean values ± SEM. Statistical analysis was performed using a two-sided Student t test. p value was either labelled on the corresponding items or listed in the legends.