Harnessing BET Inhibitor Sensitivity Reveals AMIGO2 as a Melanoma Survival Gene

Cell culture

501MEL (Female), SKmel147 (Female), SKmel2 (N/A), SKmel28 (N/A), SKmel239 (Female), SKmel197 (N/A), SKmel94 (N/A), SKmel100 (N/A), SKmel173 (N/A), SKmel29 (Male), SKmel103 (N/A), A375 (Female), SKmel192 (N/A), SKmel5 (Female) and SKmel187 (N/A) were grown in DMEM supplemented with 10% FBS, penicillin and streptomycin. Low passage melanoma short-term cultures (STCs), WM793B (Male), WM278 (Female), WM1361A (Male) and 451Lu (Male), a kind gift of the Osman laboratory, NYU, were grown in MCDB 153 supplemented with 10% FBS, penicillin and streptomycin (de Miera et al., 2012). MeWO (Male) were grown in EMEM supplemented with 10% FBS, penicillin and streptomycin. Normal human melanocytes (NHM1-5) were extracted and cultured as follows: Neonatal foreskins were collected in PBS, 1:100 Anti-Anti (100X Antibiotic-Antimycotic (GIBCO, Cat# 15240-062)), and kept at 4°C for no longer than 48 hr. Skins were then washed once in 70% EtOH and 3x in PBS, 1:100 Anti-Anti, cut into 3-4 pieces in HBSS (GIBCO, Cat# 24020-117) and placed in Dispase II solution (Roche, Cat# 04942078001) at 4°C O/N. Epidermis then separated from the dermis in HBSS, floated in 1 mL of 0.25% trypsin (GIBCO, Cat# 15400-054), and incubated in 37°C for 3 min with agitation. Cells were then collected into 4 mL of DMEM with 10% FBS and spun at 1200 rpm for 5 min twice. Media was then removed from the first and the second cell pellets, which were then consolidated and resuspend in Melanocyte Growth Media 254 supplemented with Human Melanocyte Growth Supplement (HMGS, Life Technologies, Cat# S-016-5) and 0.2 μM CaCl₂, for 3-5 days. Cells were then rinsed in PBS, 1:100 Anti-Anti, and trypsinized in 1 mL 0.05% Trypsin for 3 min to separate melanocytes from fibroblasts and keratinocytes, which take longer to detach. Cells were then collected into 4 mL of DMEM with 10% FBS, and spun at 1200 rpm for 5 min and resuspended and cultured in Melanocyte Growth Media 254 supplemented with HMGS, 0.2 μM CaCl₂, and 10 ng/mL PMA (Sigma, Cat# P8139).

Mice

NOD/SCID/IL2γR^−/− female mice (Jackson labs, Cat# 05557) were used for in vivo studies. Experiments were conducted following protocols approved by the NYU Institutional Animal Care Use Committee (IACUC) (protocol number, S16-00051).

IC₅₀ curves

Cells were seeded at 2×10⁴ cells per well on a 96-well plate and treated with varying concentrations of JQ1 (JQ1(+)) for 72 hr. Cell viability was measured as described below and results were plotted relative to the lowest concentration.

Reverse transfection

Cells were seeded at 3×10⁴ cells per well on a 96-well plate and transfected with Lipofectamine 2000 (Invitrogen) and siRNA ON-target plus SMARTpools for each of the indicated candidate genes and with siSCR control, at a final concentration of 20 nM (n = 3/siRNA). At 48 hr post-transfection, cells were seeded for cell proliferation screens.

Cell viability assays

Transfected, transduced or JQ1-treated cells were seeded at 3×10³ or 2×10³ cells per well in 96 well plates. The following day (day 0) and every 24 hr after (up to 4 days), cells were fixed in 0.1% glutaraldehyde and stored in PBS at 4°C. Cells were then stained with 0.5% crystal violet, dissolved with 15% acetic acid. Optical density was read at 590nm. For normalization and control purposes, cells transfected, transduced with non-targeting scramble controls (siSCR) were present on each experiment. Viability data are relative to the siSCR-transfected cells.

Colony formation assay

Cells were seeded at 1×10³ cells per well on a 6-well plate (n = 2/condition) and media was changed every 3 days. Fourteen days post-seeding, cells were fixed in 0.1% glutaraldehyde and 0.5% crystal violet. After washing off excess crystal violet, plates were scanned and colony number was manually assessed.

Plasmids, lentiviral vector production and cell transfection

To generate AMIGO2-GFP, the human AMIGO2 cDNA was purchased from Dharmacon (pBluescript, MHS6278-202804749) and subcloned into the pLenti-CMV-GFP-Puro vector (Addgene, Cat# 17448) using XbaI (5′-GCGGATCCGCagtggacgccacaaaaggtgtgtcagaaaa-3′) and BamHI (5′-CTAGTCTAGAATGTCGTTACGTGTACACACTCTGCCCACCCT-3′). For lentiviral vectors production HEK293T cells were seeded in 10 cm tissue culture dish and incubated at 37°C and 5% CO₂. When cells reached 80% confluency they were co-transfected with 12 μg of lentiviral expression constructs, 8 μg of psPAX2 and 4 μg pMD2.G vectors using Lipofectamine 2000 (Invitrogen) following manufacturer’s recommendations. At 48 hr post transfection supernatants were collected, filtered (0.45 μm) and stored at −80°C (Segura et al., 2013). Melanoma cells were infected with lentiviral supernatant supplemented with polybrene at a final concentration of 4 μg/mL. siRNA ON-target plus SMARTpools were purchased from Dharmacon. 50 nM of the corresponding siRNA were transfected using Lipofectamine 2000 (Invitrogen) following manufacturer’s protocol. Transfection efficiency was monitored using 50 nM BLOCK-iT Fluorescent Oligo (Invitrogen).

RNA extraction, Reverse Transcription and qRT-PCR

Total RNA was extracted using RNeasy QIAGEN extraction kit according to the manufacturer’s protocol. 1 μg of RNA was subjected to DNase I treatment and reverse transcription. Real-time PCR of the gene of interest was performed using SYBR green (Applied Biosystems or Roche). Expression levels are shown relative to PPIA or GAPDH as indicated. For NHM, RNA was further purified using OneStep PCR Inhibitor Removal columns (ZYMORESEARCH, Cat# D6030) to remove melanin prior to qRT-PCR.

Flow cytometry, apoptosis, and cell cycle analysis

Cells were seeded at 1.5×10⁵ cells per well on 6-well plates (n = 2/condition). For apoptosis measurements, non-adherent and adherent cells were collected and stained with AnnexinV-Pacific Blue (BioLegend, Cat# 640917) and PI, following manufacturer’s protocol. For cell cycle analysis, cells were pulsed with EdU (ThermoFisher, Click-IT EdU Pacific Blue Flow Cytometry Assay Kit) for 1 hr, brought to suspension and fixed following manufacturer’s protocol. Pulsed cells were co-stained with PI at a final concentration of 10 μg/mL. A BD LSRII cytometer was used for data acquisition and Flowjo software for analysis.

Caspase activity assay

Cells were treated with or without a pan-caspase inhibitor (10 nM, QVD) 24 hr post-lentiviral infection. Cells were then rinsed 1x with PBS and resuspended in 1X lysis buffer (20 mM HEPES/NaOH, pH 7.2, 10% sucrose, 150 mM NaCl, 10 mM DTT, 5 mM EDTA, 1% Igepal CA-630, 0.1% CHAPS, and 1X EDTA-free Complete protease inhibitor mixture (Roche)). Lysates were cleared by centrifugation at 16,000×g for 5 min and supernatants quantified by the Lowry method (Bio-Rad). Assays were performed in triplicate using 25 μg of protein in lysis buffer supplemented with 25 μM of the fluorogenic substrate Ac-DEVD-afc. Plates were read in a fluorimeter using a 400 nm excitation filter and a 505 nm emission filter.

Whole cell protein extractions and western blot analysis

Cells were harvested with PathScan Sandwich ELISA 1X lysis buffer (Cell Signaling Cat# 7018) supplemented with protease and phosphatase inhibitor cocktail (Roche). Cell lysates (15-40 μg of protein) were resolved in 4%–12% Bis-Tris gels (Invitrogen) and transferred to nitrocellulose membranes using semi-dry transfer. After blocking (1 hr to overnight with 5% nonfat milk), membranes were probed with primary antibodies overnight at 4°C. Membranes were developed using Clarity western ECL blotting substance (Biorad, Cat# 1705060). For AMIGO2, primary antibody was used at 1:100 (R&D systems). Secondary detection was performed using rabbit anti-goat at 1:20,000 (Sigma, Cat# A5420). Immunoblots were done on whole-cell lysates unless otherwise indicated.

Immunohistochemistry (IHC)

Unconjugated anti-human AMIGO2, raised against full-length human AMIGO2 was used for IHC. Antibody optimization was performed on formalin-fixed, paraffin-embedded 4-micron composite tissue microarrays containing normal and tumor: skin, stomach, prostate, liver, and pancreas. IHC parameters were validated on formalin-fixed, paraffin-embedded 4-micron human hippocampus with cells clearly expressing cytoplasmic-membrane staining considered positive. Chromogenic IHC was performed on a Ventana Medical Systems Discovery XT instrument with online deparaffinization using Ventana’s reagents and detection kits unless otherwise noted (Ventana Medical Systems Tucson, AZ USA). AMIGO2 was antigen-retrieved in Ventana Cell Conditioner 1 (Tris-Borate-EDTA) for 20 min. Antibody was diluted 1:50 in PBS and incubated for 6 hr at room temperature. Primary antibody was detected with goat anti-mouse, horseradish peroxidase conjugated multimer incubated for 8 min. The complex was visualized with alpha-naphthol pyronin incubated for 8 min. Slides were washed in distilled water, counterstained with Hematoxylin, dehydrated and mounted with permanent media. Negative controls were incubated with diluent instead of primary antibody. Staining score was performed in a blinded manner. The IHC score was calculated using staining intensity (1-4) × percentage of positive cells.

Immunofluorescence

SKmel147 cells were grown on glass coverslips (SKmel147 cells stably expressing AMIGO2-GFP were used in Figure 4A), fixed in 4% PFA for 15 min and permeabilized with 100 μM saponin for 30 min. Primary antibodies were used at the following dilutions: AMIGO2 1:500 (Figure 4A), and AMIGO2 1:200 and PTK7 1:200 (Figure S4C). Secondary detection was performed using Alexa Fluor conjugated antibodies diluted 1:1000, nuclei were counterstained with Hoechst 33342 and coverslips were mounted in ProLong Diamond Antifade (ThermoFisher). Images were acquired on a LSM 780 confocal microscope controlled by ZEN software (Carl Zeiss) using a 40× immersion objective, 2× internal magnification and optimal voxel size. Single Z slices are shown.

Affinity purification and mass spectrometry (MS) analysis

SKmel147 cells stably expressing AMIGO2-GFP or GFP were fixed with 1% formaldehyde for 10 min at RT. Fixing reaction was quenched with 2.5 μM glycine. Whole cell extracts were harvested with 1X lysis buffer (10 mM Tris pH 7.5, 150 mM NaCl, 0.5 mM EDTA, 0.5% NP-40) supplemented with phospatase and protease inhibitors (Roche). 1 mg of protein was incubated with 25 μL of washed beads O/N, washed 3x with wash buffer (1 mM Tris pH 7.5, 150 mM NaCl, 0.1% NP-40, supplemented with phosphatase and protease inhibitors), and eluted in 2X sample buffer. Samples were then run on NuPAGE 4%–12% Bis-Tris Gel 1.0 mm (Life Technologies) and desired fractions were extracted and subsequently digested with trypsin. Peptides were extracted using Poros beads (Life Technologies) for MS analysis. LC separation was conducted with the auto sampler of an EASY-nLC 1000 (ThermoFisher). Peptides were gradient eluted from the column directly to QExactive mass spectrometer using a 90 min gradient (ThermoFisher). For identification of AMIGO2 unique interactors, we determined the peptides uniquely found in AMIGO2-GFP pull downs versus GFP (n = 3). We then filtered out potential contaminants using a contaminant repository (CRAPome (Mellacheruvu et al., 2013)) with a threshold of > 3/411 (Table S5).

Mouse experiments

SKmel147 cells were infected with shSCR or shAMIGO2 lentivirus (shA2 #2) for 2 days and then injected at 5×10⁵ cells per mouse in the flanks of female NOD/SCID/IL2γR^−/− mice (n = 8/group). Tumor volume was measured every 2-3 days over a period of 18 days and tumor weight was measured at endpoints.

Reverse phase protein array (RPPA)

501MEL and SKmel147 cells where transfected with shSCR or shAMIGO2 (shA2 #2) and 48 hr post-transfection protein extracts where collected as follows: cells were washed 2x with cold PBS and incubated for 20 min on ice with 1X lysis buffer (1% Triton X-100, 50 mM HEPES, pH 7.4, 150 mM NaCl, 1.5 mM MgCl₂, 1 mM EGTA, 100 mM NaF, 10 mM Na-pyrophosphate, 1 mM Na₃VO₄, 10% glycerol, with freshly added phosphatase and protease inhibitors (Roche)). Lysates were cleared by centrifugation at 14,000×g for 5 min, and supernatants were quantified by the Lowry method (Bio-Rad). Samples were diluted with 4X SDS sample buffer (40% Glycerol, 8% SDS, 0.25 M Tris pH 6.8, supplemented with fresh 2-mercaptoethanol). Equal amounts of proteins per sample were loaded onto the array (Park et al., 2010). Unpaired Student’s t test was performed to determine statistically significant changes (p < 0.05) between shSCR and shA2 #2. Heatmaps for significant changes (Figure S3B, S3D) were generated using Cluster 3.0 and Java TreeView.

RNA-sequencing

Total RNA was extracted from two NHM cultures, SKmel2, SKmel239, SKmel147 and 501MEL melanoma cell lines treated with enan-tiomer (JQ1[−]) or 10 μM JQ1 (JQ1(+)) for 6 or 24 hr, in technical replicates. RNA was extracted using QIAGEN miRNeasy minikit. RNA was then processed with Ribo-Zero rRNA Removal Kit (Illumina) to remove rRNA, and further processed into sequencing libraries using Illumina ScriptSeq Complete Gold kit, following manufacturer’s protocol. All libraries were sequenced on Illumina HiSeq2500 (~150M, 50bp paired-end). Reads were aligned to the GRCh37/hg19 using TopHat (version 2.1.0) (Trapnell et al., 2012). Normalized transcriptome assemblies in FPKM, and differential expression ratios were computed with Cufflinks (version 2.1.1) (Trapnell et al., 2012). Coverage tracks were generated using deepTools (version 2.4.1) with parameters bamCoverage –normalizeUsingRPKM –bin-size 10 (Ramírez et al., 2014). In all samples, genes were called ‘expressed’ if normalized FPKM ≥ 1.5. Genes were called downregulated upon JQ1 treatment if FPKM ≥ 1.5; log2FoldChange ≤ −0.625; and Padj < 0.05. Genes were called overexpressed over NHM if nFPKM ≥ 1.5 (in the melanoma sample); log2FoldChange ≥ 2; and Padj < 0.05. Genes were called downregulated upon AMIGO2 depletion if FPKM ≥ 1.5 (prior to depletion); log2FoldChange ≤ −1.2; and Padj < 0.05. log2FoldChange ≥ 1.2 for upregulated genes. Scatterplots were generated using R (gplots). Volcano plots were generated using R (CummeRbund). Heatmap for 78 putative pro-tumorigenic genes (Figure 1D) was generated using R (ggplot2). All FPKM values are normalized.

Native and crosslinked ChIP and next-generation sequencing

For native ChIP (H3K4me3, H3K4me1, H3K27ac, macroH2A1, macroH2A2 and H3K27me3), 3–4×10⁷ cells were collected and resuspended in 2 mL of ice cold buffer I (0.32 M Sucrose, 60 mM KCl, 15 mM NaCl, 5 mM MgCl₂, 0.1 mM EGTA, 15 mM Tris pH 7.5, 0.5 mM DTT, 0.1 mM PMSF, 1:1000 protease inhibitor cocktail). 2 mL of ice-cold buffer I supplemented with 0.05%–0.2% IGEPAL (Sigma) was then added. The resulting 4 mL of nuclei was placed on ice for 10 min and then gently layered on top of 8 mL of ice-cold buffer III (1.2 M Sucrose, 60 mM KCl, 15 mM NaCl, 5 mM MgCl₂, 0.1 mM EGTA, 15 mM Tris pH 7.5, 0.5 mM DTT, 0.1 mM PMSF, 1:1000 protease inhibitor cocktail) and centrifuged at 10,000×g for 20 min at 4°C with no brake. Pelleted nuclei were resuspended in 200 μL buffer A (0.34 M sucrose, 15 mM HEPES pH 7.4, 15 mM NaCl, 60 mM KCl, 4 mM MgCl₂, 1 mM DTT, 0.1 mM PMSF, 1:1000 protease inhibitor cocktail) and further diluted to 600 ng/μL. MNase (Affymetrix) digestion reactions were carried out on 100 μg or more chromatin using 0.05–0.2 U/μg chromatin in buffer A supplemented with 3 mM CaCl₂, 1 mM DTT, 0.1 mM PMSF and 1:1000 protease inhibitor cocktail, for 10 min at 37°C. The reaction was quenched with 5 mM EGTA on ice and centrifuged at 13,500×g for 10 min at 4°C. The chromatin was resuspended in 10 mM EDTA pH 8.0, 1 mM DTT, 1 mM PMSF and 1:1000 protease inhibitor cocktail, and rotated at 4°C for 2 hr. The mixture was adjusted to 500 mM NaCl, allowed to rotate for another 45 min and then centrifuged at 13,500×g for 10 min yielding nucleosomes in the supernatant. 100 μg or more of chromatin was diluted to 100 ng/μL with buffer B (20 mM Tris pH 8.0, 5 mM EDTA, 500 mM NaCl, 0.2% Tween 20, 1 mM DTT, 1 mM PMSF, 1:1000 protease inhibitor cocktail) and pre-cleared with protein A (Invitrogen) or A+G (Millipore) magnetic beads pre-conjugated with rabbit or mouse IgG (15 μL slurry and 15 μg of IgG per 25 μg chromatin), for 2 hr at 4°C. 1–2 μg of the pre-cleared supernatant (bulk nucleosomes) was saved for further processing. To the remaining supernatant, pre-conjugated A or A+G magnetic beads;antibodies (10 μL slurry and 1 μg antibody per 10 μg chromatin) was added and rotated overnight at 4°C. Beads were washed 3x with buffer B, and 1x with buffer B without Tween, and DNA was eluted 2x in 100 μL fresh elution buffer (1% SDS, 100mM NaHCO₃) at 65°C for 30 min. For the input fraction, an equal volume of input recovery buffer (0.6 M NaCl, 20 mM EDTA, 20 mM Tris pH 7.5, 1% SDS) and 1 μL of RNase A (10 mg/mL) was added followed by incubation for 1 hr at 37°C. 100 μg/mL Proteinase K (ThermoFisher) was then added and was incubated for another 3 hr at 37°C. For the ChIP fraction, 12 μL of 5M NaCl and 1 μL of RNase A (10 mg/mL) was added followed by incubation for 1 hr at 37°C. 500 μg/mL Proteinase K (ThermoFisher) was then added and was incubated for another 3 hr at 37°C. Samples were then subjected to a phenol-chloroform extraction followed by purification using AMPure XP beads (Beckman Coulter, Cat# A63882). DNA was diluted to 4-6 ng/μL and run on an Agilent High Sensitivity DNA chip using Agilent Technologies 2100 Bioanalyzer (Hasson et al., 2013).

For BRD2, BRD4, MED1 and CTCF 72×10^6 cells (SKmel147 untreated (JQ1[−]) and treated with 10 μM JQ1 (JQ1(+)) for 6 hr, NHM untreated), were crosslinked with 1% PFA for 10 min at RT. The reaction was quenched with 0.125 M glycine for 5 min at RT and centrifuged at 1100 rpm for 3 min at 4°C. Cells were washed 1x in ice-cold PBS, resuspended at 18×10⁶ in 500 μL ice-cold Cell Lysis Buffer (10 mM Tris pH 8, 10 mM NaCl, 0.2% NP-40, 1 mM PMSF, 1:1000 protease inhibitor cocktail) and incubated on ice for 15 min. Tubes were then centrifuged at 2K rpm for 5 min at 4°C, and pellets were resuspended in 500 μL ice-cold Nuclear Lysis Buffer (50 mM Tris pH 8, 10 mM EDTA, 1% SDS, 1 mM PMSF, 1:1000 protease inhibitor cocktail) and incubated on ice for 10 min. Sonication were carried out for 12-15 cycles, 30 s on, 30 s off, low intensity at 4°C in a Bioruptor sonicator (Diagenode). Chromatin was centrifuged at 13,000×g for 10 min at 4°C, consolidated and diluted 1:4 with IP Dilution buffer (20mM Tris pH 8, 2 mM EDTA, 150 mM NaCl, 1% Triton-X, 0.01% SDS, 1 mM PMSF, 1:1000 protease inhibitor cocktail). Chromatin was pre-cleared with protein A (Invitrogen) or A+G (Millipore) magnetic beads pre-conjugated with rabbit or mouse IgG respectively (240 μL of slurry and 30 μg of IgG), for 2 hr at 4°C. 1–2 μg of the pre-cleared supernatant (bulk nucleosomes) was saved for further processing. To the remaining supernatant, pre-conjugated antibodies (240 μL slurry and 24 μg antibody) was added and rotated overnight at 4°C. Beads were washed 2x with ice-cold High Salt Buffer (20mM Tris pH 8, 2 mM EDTA, 500 mM NaCl, 1% Triton-X, 0.01% SDS, 1 mM PMSF, 1:1000 protease inhibitor cocktail), 1x with ice cold IP Wash I (20 mM Tris pH 8, 2 mM EDTA, 50 mM NaCl, 1% Triton-X, 0.1% SDS, 1 mM PMSF, 1:1000 protease inhibitor cocktail), 1x with ice-cold IP Wash II (10 mM Tris pH 8, 1 mM EDTA, 0.25 LiCl, 1% NP-40, 1% deoxycholic acid, 1 mM PMSF, 1:1000 protease inhibitor cocktail), and 2x with ice-cold TE (5 mM Tris pH 7.4, 1 mM EDTA). DNA was eluted as above (Roe et al., 2015). For TEAD4 and FOSL2, 7-9×10^6 cells (SKmel147 untreated), were washed 3x in PBS, 1 mM MgCl₂, crosslinked with 0.25M DSG (Pierce, Rockford IL) in PBS, 1 mM MgCl₂ for 45 min at RT, washed 3x in PBS, 1 mM MgCl₂, crosslinked in PBS, 1% PFA for 10 min at RT, washed 3x in PBS, and then harvested (Tian et al., 2012). Cells were processed and sonicated as above. ChIP reactions were carried out as above with 100 μL slurry and 10 μg antibody (Roe et al., 2015). Sequencing libraries were generated and barcoded for multiplexing according to Illumina recommendations with minor modifications. Briefly, 2–8 ng Input or ChIP DNA was end- repaired and A-tailed. Illumina Truseq adaptors were ligated, libraries were size-selected to exclude polynucleosomes, and the libraries were PCR-amplified using KAPA HiFi DNA Polymerase with no more than 12-15 cycles. All steps in library preparation were carried out using NEB enzymes (Hasson et al., 2013). Libraries were sequenced on Illumina Hi-Seq2500 or NextSeq500 (50-75bp single-end reads).

Reads were trimmed for Illumina adaptor sequences using in-house scripts and aligned to the GRCh37/hg19 using Bowtie (version 0.12.7) with parameters –l 40-65 –n 2 –S –best –k 1 –m 20/1, and further assessed for quality using NGS-QC (Langmead et al., 2009). Number of aligned reads for each sample is provided in Table S6. For BRD2, BRD4, TEAD4 and FOSL2 ChIP-seq, duplicate reads were removed before downstream analysis. BRD4 significant peaks in SKmel147 were identified using MACS2 (version 2.1.1.2) with parameters -s 100–nomodel —extsize 200 —slocal 1000 -q 0.01. H3K27ac significant peaks in SKmel147 were identified using with parameters–fe-cutoff 1.5 -s 100 -p 0.00001–nomodel–extsize 147–keep-dup 2. (Zhang et al., 2008). Matching Input control was used to call peaks. Coverage tracks were generated from BAM files using deepTools bamCoverage with parameters –normalizeU-singRPKM –binsize 10.

ATAC-seq

For all ATAC-seq libraries (Buenrostro et al., 2015), 150k cells were harvested, tagmented with 5uL Nextera Tn5 Transposes from Nextera kit (Illumina) for 30 min, amplified up to 10 cycles and then purified using QIAGEN MinElute columns (Buenrostro et al., 2013). Purified libraries were then size-selected on 2% agarose gel (150-700bp), followed by purification using AMPure XP beads (Beckman Coulter, Cat# A63882). DNA was diluted to 4-6 ng/μL and run on an Agilent High Sensitivity DNA chip using Agilent Technologies 2100 Bioanalyzer. Libraries were sequenced on Illumina Hi-Seq2500 (50bp paired-end reads). Reads were trimmed for Illumina adaptor sequences using in-house scripts and aligned to the GRCh37/hg19 using Bowtie2 (version 2.1.0) with parameters –S –X 2000 (Langmead et al., 2009). Reads aligned to mtDNA, with quality value Q < 30, as well as duplicated reads, were removed using in-house scripts and Samtools. The final number of reads used to generate coverage tracks for each sample is provided in Table S6. Significant peaks were identified using MACS2 with parameters–nomodel–nolambada–keepdup all –slocal 10000. Coverage tracks were generated using deepTools bamCoverage with parameters –normalizeUsingRPKM –binsize 10.

Luciferase assay

The pGL3-control vector (Promega) lacking the SV40 enhancer (pGL3-noSV40) was a kind gift of the Wysocka laboratory (Prescott et al., 2015). Intact enhancers E1-E5 and E3 segments E3A-C were amplified from genomic DNA extracted from SKmel147 cells (Table S7). Mutated E3A sequence was amplified from a custom made oligo (Genscript, Table S7). Sequences were then cloned into NheI/XhoI-digested pGL3-noSV40 using In-Fusion HD Cloning Kit (Clontech). Reporter vectors were co-transfected with FuGENE 6 (Promega) at a 1:200 ratio with pRL-SV40 renilla luciferase in SKmel147 cells. Luminescence was measured using the Dual-Luciferase Reporter Assay System at 24 hr post-transfection. Luminescence measurements were performed in triplicate using SpectraMax M5 (Molecular Devices), on at least three biological replicates.

Growth Curves

For TEAD4 and FOSL2 knockdown growth curves in SKmel147 cells, lentiviral vectors were produced in HEK293T cells transfected with standard calcium phosphate method using packaging psPAX2 (Addgene #12260), pMD2.G (Addgene #12259) and plkO.1-human-shRNA (Sigma) (in a ratio 3:1:4). 48 and 72 hr post transfection supernatants were collected, consolidated (12 mL total) and filtered with a 0.45 μm filter. SKmel147 cells were infected with 500-1000 μL supernatant for 12 hr, in the presence of 8 μg/mL poly-brene (Vardabasso et al., 2015), seeded at 15,000 cells per well in a 6-well plate in triplicates, and followed for confluency using In-cuCyte (Essen BioScience). For AMIGO2 knockdown in NHM, cells were infected as above, seeded at 150,000 per well and followed using IncuCyte.

CRISPR-Cas9 genome editing

SKmel147 cells stably expressing Cas9 were generated by infection with the lentiCas9-Blast plasmid (Addgene, Cat# 52962). The sgRNAs (Table S7) were cloned using BbsI site downstream of the human U6 promoter in a lentiviral vector containing either EGFP or mCherry downstream of the human PGK promoter (kind gift of Brown laboratory, ISMMS). Lentiviral vectors were produced as above. SKmel147 cells were first infected with the lentiCas9-Blast plasmid lentiviral supernatant, and then selected with blasticydin (10 μg/mL for 10 days) for cells expressing the Cas9 nuclease. Cells were then infected with pLenti-GFP/mCherry-sgRNAs, and GFP- and/or mCherry-positive cells were sorted 2 days after infection. For clone isolation, positive single cells were sorted into 96 well plates. Clones were grown and expanded for approximately 4 weeks prior to DNA extraction for PCR-based screening (Table S7), and TOPO cloning (Invitrogen, TOPO TA Cloning Kit for Sequencing, Cat# 45-0030). For competition assays, GFP- and/or mCherry-positive cells were counted and mixed at equal numbers of SKmel147 cells expressing Cas9, and examined by FACS at day 0 and day 7.

Enhancer calling and enhancer-gene associations

Enhancers (TEs) and super-enhancers (SEs) were called based on BRD4 and H3K27ac enrichment separately in SKmel147 using the ROSE algorithm (Rank Ordering of Super-Enhancers) (Lovén et al., 2013; Whyte et al., 2013) (stitching distance 12.5kb and TSS exclusion zone size 2.5kb. BRD4/H3K27ac levels were normalized to Input control; ChIP-Input). Only BRD4/H3K27ac enhancers overlapping with significant ATAC peaks were considered. The ROSE algorithm was also used to extract H3K27ac levels within the AMIGO2 SE in NHM1, NHM2, SKmel147, 501MEL, SKmel239 and SKmel2 (Figure 6D). ChIP-seq occupancy levels for heatmaps and meta-profiles plots were calculated using deepTools. Enhancers were associated to the closest expressed genes (nFPKM ≥ 1.5) using the BEDTools suite (Quinlan and Hall, 2010). All MIR, SNORD, RNVU, SNOR, and RNU genes were excluded. TEs within SEs (Figure 5F, 5G, 5I, S6A) were called based on ATAC-seq significant peaks using the BEDtools suite.

TF motif analysis

Enriched TFs DNA binding motifs (De Novo) were identified with a 1000bp window centered on 569 BRD4 significant peaks within the SKmel147 top 150 BRD4 SEs (Figure 5A, 5B), using HOMER suite (findMotifsGenome.pl; version 4.8) (Heinz et al., 2010). Motifs were considered based on p value (p < 10⁻¹²) and expression levels in SKmel147.

Dendrograms

Normalized total ChIP-seq H3K27ac reads within 464 SKmel147 SEs for NHM1, NHM2, 501MEL and SKmel239 (BRAF^V600E), and SKmel2 and SKmel147 (NRAS^Q61R) melanoma cell lines were calculated using the ROSE algorithm. Correlations were calculated based on distance correlation (R; energy package). Distances were defined as 1-abs(corr) and were calculated in R, and dendrograms were drawn using a neighbor-joining algorithm (R; ape package). Normalized expression levels of 387 SE associated genes in SKmel147 across all cell lines, were used to generate the expression Dendrogram (Figure S6A).

Mosaic plots

Mosaic plots were generated using GEDI (Gene Expression Dynamics Inspector) (Eichler et al., 2003) on H3K27ac ChIP-seq signal within 5142 BRD4 identified SKmel147 TEs and SEs. H3K27ac signal across four melanoma and 2 NHM samples was calculated using ROSE. Maps of size 50×50.

GO analysis

GO terms were obtained using Enrichr (Kuleshov et al., 2016). Combined score of top categories is shown for all plots.

Statistical methodologies

Statistical analysis was performed using unpaired Student’s t test unless otherwise indicated (for Figures 1B and 1C, paired Student’s t test was used). Statistical tests performed as part of the RNA-seq, ChIP-seq, and ATAC-seq analyses are detailed in relevant sections above. Error bars for boxplots represent 10-90 percentile with outliers omitted, unless otherwise indicated. For growth curves (Figure S6H), nonlinear regression (curve fit) was used to determine significance. For competition assay (Figure S7D), Welch’s t test was used. P values are represented as *p < 0.05, **p < 0.01 and ***p < 0.001.

Data deposition

The raw and processed sequencing data have been deposited to GEO under accession number: GSE94488. Raw data for immunoblots and for AMIGO2 Mass Spectrometry have been deposited to Mendeley Data and are available at https://doi.org/10.17632/8g4vcy4ymm.1.