Cell culture and reagents
LNCaP, C4-2, 22Rv1, and HEK293FT cells were purchased from American Type Culture Collection (ATCC). LNCaP, C4-2, and 22Rv1 cells were maintained in phenol red-free RPMI1640 medium (Gibco) that is supplemented with 9% fetal bovine serum (FBS) (Gibco) and 1% antibiotic-antimycotic (Gibco). LNCaP-abl (58) was a generous gift from Dr. Myles Brown’s lab at Dana-Farber Cancer Institute and were maintained in phenol red-free RPMI1640 medium (Gibco) that is supplemented with 9% Charcoal-stripped fetal bovine serum (CSS) (Gibco) and 1% antibiotic-antimycotic (Gibco). HEK293FT cells were grown in phenol red containing Dulbecco’s modified Eagle’s medium (high glucose, Life Technologies) that is supplemented with 9% FBS and 1% antibiotic-antimycotic. All cell lines were kept in a 37°C incubator at 5% CO2. Cells that were used in transfection studies were seeded in medium without antibiotics. NEPC organoids, including WCM1078, WCM154, WCM155 and WCM1262 used for this study have been described previously (14). The NEPC organoids were grown with growth-factor-reduced Matrigel (Corning). Five droplets of around 50 µl cell suspension/Matrigel mixture was pipetted onto each well of a six-well cell suspension culture plate (Greiner). The six-well plate was placed into a cell culture incubator at 37°C and 5% CO2 for 30 min to solidify the droplets before 2 ml prostate-specific culture medium was added to each well. The organoids were grown in prostate-specific culture medium consisting of Advanced DMEM/F12 (Invitrogen) with GlutaMAX (1×, Invitrogen), HEPES buffer (1 M, Invitrogen), 1% antibiotic-antimycotic (Gibco), B27 supplement (Gibco), N-acetylcysteine 1.25 mM (Sigma-Aldrich), mouse recombinant protein EGF 50 ng/ml (Peprotech), human recombinant FGF-10 20 ng/ml (Peprotech), recombinant human FGF-basic 1 ng/ml (Peprotech), A-83-01 500 nM (Tocris), SB202190 10 µM (Sigma-Aldrich), nicotinaminde 10 mM (Sigma-Aldrich), PGE2 1 µM (Tocris), NRG 100 µl/ml (GenScript), Y-27632-2HC1 10 µM (Selleck), Noggin conditioned medium (10%) and R-spondin conditioned medium (5%). Tazemetostat (xcessbio; cat. no. M60122-10S), valemetostat (selleckchem; cat. no. S8926), GSK126 (xcessbio; cat. no. M60071-2S), PF-06821497 (A21388), CIR7-2512 (CirclePharma), CIR7-2724 (CirclePharma), Tagtociclib (sellckchem; cat. no. S9878) were dissolved in DMSO (sigma) for in vitro studies.
Clinical samples and H3K27me3 CUT&Tag
Tumor specimens were secured adhering to approved protocols by the DFCI and WCM Institutional Review Boards (nos. 19–883 and 1305013903, respectively). Clinical and information and pathology were collected by medical record review. Tumors histology was classified as castration resistant PRAD or NEPC by board certified pathologist. Upon pathological confirmation of a phenotype, punches were made in frozen OCT section and processed for CUT&Tag sequencing at the Centre for Functional Cancer Epigenetics at DFCI. Sequenced reads were mapped to hg38 using bowtie2 version 2.2.5 and visualized in the figures using IGV, ChIPSeeker package in R, and deepTools (59). Broad peak calling was conducted employing MACS2 (60), with a q-value (FDR) < 0.01 set as the threshold for each CUT&Tag dataset. A union set of peaks was generated using BEDTools (61). The read counts mapped to each peak in each sample were computed from BAM files using chromVAR (62). Subsequently, read counts for each peak were normalized through variance stabilizing transformation from DESeq2 (63). PCA and UMAP for Supplementary Fig. 2 were generated using R.
Western blotting and immunohistochemnistry
Cells were washed twice with ice-cold phosphate-buffered saline (PBS, Gibco) and harvested in whole-cell lysis buffer (110 mM sodium dodecyl sulfate (SDS), 100 mM dithiothreitol, 80 mM Tris-HCl (pH 6.9), 10% glycerol). Cell lysates were boiled at 95°C for 5min. Equal amounts of protein was estimated by Lowry protocol (RCDC Protein Assay, Bio-Rad) and were subjected to NuPAGE Novex gel electrophoresis (Life Technologies). Based on the predicted molecular weight of the protein of interest, Gel electrophoresis was performed on 4–12% or 10% Bis-Tris NuPAGE gels or 3–8% Tris-Acetate NuPAGE gels. Proteins were transferred onto nitrocellulose membranes (NuPAGE) and blocked with 5% milk or bovine serum albumin prepared in tris-buffered saline with 0.1% Tween-20. Primary antibodies against H3K27me3 (cell signaling technology; cat. no. 9733S; 1:2000 dilution) or total-histone-H3 (cell signaling technology; cat. no. 3638S; 1:5000 dilution) or β-actin (cell signaling technology; cat. no. 8457S or 3700S; 1:5000 dilution) or EZH2 (cell signaling technology; cat. no. 5246S; 1:1000 diltuion), or HA-Tag (cell signaling technology; cat. no. 8457S or 3700S; 1:2000 dilution) were used as indicated in the figures. Near Infra-Red Fluorescent detection was performed using LI-COR secondary antibodies matching the host species of the primary antibodies. LI-COR Odyssey M and imaging software were used to visualize the western blots. Blots were stripped with stripping buffer (Restore Plus Western blot, ThermoFisher Scientific) and reprobed with additional primary antibodies as required.
Immunohistochemistry was performed as described previously (14).
Cell viability assays
Cell viability was measured at indicated time points after treatment using CellTiter-Glo assay (Promega) in 96-well plate format as per the manufacturer’s protocol.
Flow Cytometry Analysis
After live/dead staining with Zombie NIR Fixable Viability Kit (cat. 423106; Biolegend) per the manufacturer's instructions, single-cell suspensions were stained with fluorophore-conjugated primary antibodies (HLA-ABC-BV510, clone W6/30, cat. 311436; Biolegend) in PBS containing 2% v/v FBS. After washing, cells were resuspended in PBS containing 2% FBS and analyzed on a LSRFortessa flow cytometer (Becton Dickinson). Expression levels were compared with isotype control antibody (mouse IgG2a κ-BV510 Isotype Ctrl, clone MOPC-173; cat. 400268; Biolegend). Data analyses were performed using FlowJo software (TreeStar).
Generation of cells with stable over-expression
For EZH2 CRISPR knockout: One million cells were resuspended in 20 µl of electroporation buffer (P3 solution, Lonza) and mixed ribonuclear protein complexes (RNP) of Cas9 protein with bound gRNAs. Assembly of RNPs was performed with 100 µM of tracrRNA (IDT) with 100 µM crRNAs (sgCtrl: AAA AAT GCC AAA AGT TAC CC, sgEZH2 #1: TGC GAC TGA GAC AGC TCA AG, sgEZH2 #2: TAT GAT GGG AAA GTA CAC G) at 1:1 ratio, incubated at 95°C for 5min followed by cooling to room temperature followed by addition of 20 µM Cas9 protein incubated for 15mins at 37°C. The mixtures were transferred to a 16-well Nucleocuvette Strip and nucleofection was performed in a 4D-Nucleofector (Lonza). Following nucleofection, cells were grown in 6-well plate coated with 1% collagen (ThermoFisher). Single cell selection was performed and colonies were selected based on knockout of EZH2 in western blot analysis.
For rescue of EZH2
Vector containing EZH2 cDNA was purchased from Addgene (#81742) and site directed mutagenesis (Agilent) was performed to add a stop codon (Fwd 5’-3’, AGA AAG TTG GGC ATC AAG GGA TTT CCA TTT CTC TTT CGA TG; Rev 5’-3’, CAT CGA AAG AGA AAT GGA AAT CCC TTG ATG CCC AAC TTT CT) and remove the start codon (Fwd 5’-3’, TTC TTC CCA GTC TGG CCG CCA ACT TTT TTG TAC A; Rev 5’-3’, TGT ACA AAA AAG TTG GCG GCC AGA CTG GGA AGA A). Gateway cloning (NEB) LR reaction was performed using the modified construct and destination vector (Addgene #91797) to produce pLEX_305-N-dTAG-HA-EZH2. The destination vector was a generous gift from Dr. William Kaelin lab. This plasmid was co-transfected with pMD2.G and psPAX2 using Lipofectamine 3000 (Invitrogen) into HEK293FT cells to generate viral particles. Lentiviral supernatants and 10 µg/ml Polybrene were added and incubated with cell lines mentioned in the figures. Infected cell lines were selected with antibiotics as described in the constructs. Over-expression of the desired proteins was confirmed by western blot analysis as described above.
Epigenetic and transcriptomic studies
ChIP-qPCR
ChIP was performed using 5–7 million cells in a 10cm dish. Cells were washed with ice-cold PBS containing EDTA-free protease inhibitor (Roche) and 5mM sodium butyrate. Crosslinking was performed using 1% methanol-free formaldehyde (invitrogen) for 5 mins, quenched by adding 0.125M glycine for 5 mins. Cells were scraped and centrifuged to pellet down at 2000rpm for 2 mins. Pelleted cells were lysed using sarkosyl lysis buffer (0.1%SDS, 1x TX-100, 10mM Tris-HCl pH = 8, 1 mM EDTA pH = 8, 0.1% Sodium Deoxycholate, 0.25% sarkosyl, 0.3 mM NaCl, EDTA-free protease inhibitor (Roche) and 5mM Sodium butyrate). Sonication was performed using Covaris machine in 1ml tubes (Covaris) at 5% duty factor, 140 PIP, 200 cycles per burst, for 5 mins. Chromatin immunoprecipitation was performed overnight using antibodies conjugated with dynabeads (invitrogen) against IgG control (cell signaling technology, cat. no. 3900S, 1 µg/reaction) or H3K27me3 (cell signaling technology, cat. no. 9733S, 1 µg/reaction) or H3K4me3 (cell signaling technology, cat. no. 9751S, 1 µg/reaction). After 5 washes with wash buffer (HEPES-based RIPA buffer), elution buffer (1%SDS, 0.1M NaHCO3) was added to the beads. RNAse treatment was performed for an hour at 37°C followed by 16 hours of Proteinase K treatment at 65°C overnight. DNA was purified (Qiagen) and qPCR was performed using primers mentioned in the table below along with 0.1% input.
CUT&RUN
CUT&RUN studies were performed with 100,000 cells in duplicates using a kit from Epicypher adhering to manufacturer’s protocol using antibodies against IgG or H3K4me3 (Epicypher, cat. no. 13–0041, 0.5 µg per reaction) or H3K27me3 (Invitrogen, cat. no. MA511198, 0.5 µg per reaction). Automated library-preparation and sequencing was performed by Molecular Biology Core Facilities in Dana-Farber Cancer Institute. Sequencing reads were mapped to hg38 human genome reference using bwa and visualized using deepTools as plotProfiles.
RNA-Seq and qRT-PCR
Total RNA isolated (QIAGEN) was submitted to Novogene for paired-end sequencing. Sequencing reads were mapped to hg38 human genome reference using STAR-RSEM. Differential gene expression analysis was performed using the limma and edgeR and GSEA software was used to perform pathway analysis. Gene ontology analysis was performed using hypergeometric overlap statistic tool to calculate the overlap between a gene list and pathways in Molecular signature database. Bar graphs and heatmaps were generated using sample FPKM values.
qRT-PCR
RNA was isolated in biological triplicates (QIAGEN) and complementary DNA (cDNA) was prepared from 500 ng total RNA (Bio-Rad). Quantitative Real-time reverse transcription–PCR (qRT–PCR) was done using SYBR Green PCR mastermix (Bio-Rad) on a Bio-Rad qPCR system. Primers sequences are mentioned in table below. Relative quantification was calculated using the △Ct method normalized to GAPDH levels.
Target
|
Assay
|
Forward primer (5’ to 3’)
|
Reverse primer (5’ to 3’)
|
GAPDH
|
qRT-PCR
|
GAAGGTGAAGGTCGGAGTC
|
GAAGATGGTGATGGGATTTC
|
ASCL1
|
qRT-PCR
|
GAAGATGGTGATGGGATTTC
|
CAAAGCCCAGGTTGACCA
|
SYPL2
|
qRT-PCR
|
GCTGGGCTTCATCAAAGTTC
|
GAAGGGATAGCCAAATGCAA
|
TMEFF2
|
qRT-PCR
|
CTGCATGCAAACAGCAGAGT
|
CTGCACCAAACTGGCAAATA
|
SYPL2
|
ChIP-qPCR
|
CTTAGTGGCAGGAGGGTGAA
|
ACCAAGGAGTGGTATGTGCAG
|
TMEFF2
|
ChIP-qPCR
|
GGCGTTTGGCAGTCACTTA
|
GTCATGGTGCTGTGGGAGT
|
Data and code availability
All sequencing data have been deposited in the public repository (Gene Expression Omnibus) and code used to generate figures are available upon request.
Animal studies
All animal studies were approved by the Dana-Farber Cancer Institute (DFCI) or Weill Cornell Medicine Institutional Animal Care and Use Committee (IACUC).
WCM12 PDX tumors were subcutaneously implanted were injected to right flank of 6–8 weeks-old NSG male mice (NOD.Cg-PrkdcscidIl2rgtm1Wjl/SzJ, the Jackson laboratory). Tumor volumes defined as 0.5 x length x width2 and the body weights were measured every 3 days. Mice were randomly assigned into vehicle or tazemetostat group once their tumors reach 100mm3. Mice were treated with vehicle (0.5% Methylcellulose (400 cps), 0.1% Tween80, pH4.0) or tazemetostat (250 mg/kg) b.i.d. p.o. for indicated number of days. The mice were sacrificed in a CO2 chamber adhering IACUC guidelines. At necropsy, tumor size, tumor weight, and body weight were measured to evaluate the drug effect and toxicity.
MSKPCa4 organoids were dissociated into single cells and 4 million cells were mixed 50:50 with Matrigel and injected into CB17 SCID mice for subcutaneous growth. Once the tumors were obtained, they were subsequently passaged in vivo for the experiment. GSK126 (150mg/kg) or vehicle was administered q.d. i.p. at a dose volume in 20% captisol adjusted to pH 4–4.5 with 1 N acetic acid.
Statistical analysis
Standard two-tailed Student t-tests and nonparametric tests were performed to analyze statistical significance of two groups. Two-way ANOVA tests were performed to examine the statistical significance of datasets with grouped analyses. Statistical significance was set at p < 0.05 in this study. All graphs and statistical analyses were completed using GraphPad Prism or ggplot package in R or deepTools. Pearson correlation was used for nearest neighbor analysis and pairwise-correlation.