Human samples
We obtained cadaveric pancreatic tissues with informed consent for research purposes from heart-beating, brain-dead donors through the National Islet Transplantation Programme at Westmead Hospital in Sydney and the St. Vincent’s Institute in Melbourne, Australia. These tissues were collected from individuals both with and without diabetes and were processed to isolate islet, acinar, and ductal tissues. Our research received institutional approval from the Human Research Ethics Committee at St. Vincent’s Hospital in Melbourne, under HREC Protocol number: 011/04.
Preparation and culture of isolated human pancreatic cells
Exocrine tissue was isolated as a by-product of islet isolation by intraductal perfusion and digestion of the pancreas with collagenase AF-1,44 (SERVA/Nordmark, Germany), followed by purification using Ficoll density gradients.45 The acinar and ductal tissue obtained from high-density fractions were cultured in Miami Media 1 A (Mediatech/Corning 98–021, USA) supplemented with 2.5% human serum albumin (Australian Red Cross, Melbourne, VIC, Australia), in a 37 °C, 5% CO2 incubator.
Ex vivo stimulation of human pancreatic progenitors with GSK126 and Tazemetostat
Human pancreatic exocrine cells were either left untreated or exposed to 10 μM GSK126 (S7061, SelleckChem) or 1 μM Tazemetostat (S7128, SelleckChem) at a density of 1 × 106 cells per well for 24 h. After the initial 24 h incubation, fresh Miami Media was added, and the cells were cultured for an additional 24 h with either 10 μΜ GSK126 or 1 μM Tazemetostat. All incubations occurred in Miami Media 1 A (Mediatech/Corning 98-021, USA) supplemented with 2.5% human serum albumin (Australian Red Cross, Melbourne, VIC, Australia) in a cell culture incubator at 37 °C with 5% CO2 for a total of 48 h, using non-treated six-well culture plates (Corning). Because of low cell numbers isolated from the adult T1D donor, harvests were prioritised for gene expression and ChIP analyses, thus data for immunofluorescent staining and GSIS have not been provided.
Molecular modelling of GSK126 and Tazemetostat bound to EZH2
Molecular dynamic simulations were performed for the PRC2 complex with GSK126 and Tazemetostat bound to the SET domain of EZH2 in triplicate for 200 ns with a time-step of 2 fs using GROMACS 2018.2 and the CHARMM36 force field.46,47 Binding free energy of the ligands was calculated on the final 20 ns of trajectories at 10 ps snapshots and decomposed on a per-residue basis using molecular mechanics Poisson-Boltzmann surface area (MM-PBSA).48
RNA isolation and mRNA-seq
Stimulated human ex vivo pancreatic cells and human pancreatic ductal epithelial cells were isolated using TRIzol (Invitrogen), and Total RNA was extracted from the cells using RNeasy Kit (QIAGEN) including a DNase treatment. Bulk RNA-seq sequence reads underwent quality and adaptor trimming with fastp (v0.20.0). Trimmed reads were mapped to human genome build 38 (hg38) using STAR aligner (v2.7.9a) and sorted with samtools (v1.9) before counting of mapped reads against Ensembl gene level annotations (GRCh38.104) using FeatureCounts (subread/2.0.1) to generate a raw gene-sample count matrix. Gene-sample count matrix counts were normalised and analysed using edgeR (v3.42.4) to generate differential gene expression.
Gene expression analysis
Stimulated human ex vivo pancreatic cells and human pancreatic ductal epithelial cells were isolated using TRIzol (Invitrogen), and Total RNA was extracted from the cells using RNeasy Kit (QIAGEN) including a DNase treatment. A high-capacity cDNA Reverse Transcription Kit (Applied Biosystems) was used on a total of 1000 ng of RNA to perform first-strand cDNA synthesis according to the manufacturer’s instructions. The oligoperfect designer (Thermo Fisher Scientific) was used to obtain primers against specific genes, as shown in Table 1. Quantitative real time PCR analyses were performed with the PrecisionFast 2× qPCR Master Mix (Primerdesign) and primers using Applied Biosystems 7500 Fast Real-Time PCR System. The mix for each qPCR reaction comprised: 5 μl qPCR Master Mix, 0.5 μl of forward and reverse primers, 2 μl nuclease-free water, and 2 μl of the pre-synthesized cDNA, diluted 1/20. Expression levels of specific genes were tested using the 2-ΔΔCT method with test CT values normalised to H3F3A housekeeping gene.
Chromatin immunoprecipitation
Assessment of chromatin status in human exocrine and pancreatic ductal epithelial cells using immunoprecipitation assays were performed as previously described.49,50 Following stimulation, 1% formaldehyde was used to fix cells for 10 min, followed by quenching of the reaction with 0.125 M glycine for 10 min. Fixed cells were resuspended in sodium dodecyl (lauryl) sulfate (SDS) lysis buffer (1% SDS, 10 mM EDTA, 50 mM Tris-HCl pH 8.1) including a protease inhibitor cocktail (Roche Diagnostics GmBH, Mannheim, Germany) and homogenised followed by incubation on ice for 5 min. Next, sonication was performed on all samples to obtain fragments of 200–600 bp. The chromatin was then resuspended in ChIP Dilution Buffer (0.01% SDS, 1.1% Triton X-100, 1.2 mM EDTA, 16.7 mM Tris-HCl pH 8.0, and 167 mM NaCl) and 20 µl of Dynabeads® Protein A (Invitrogen, Carlsbad, CA, USA) was added and pre-cleared. H3K27me3 (Millipore, Cat# 07-449), H3K27ac (Abcam, Cat# ab4729), and H3K4me3 (Abcam, Cat# ab8580) antibodies were used for immunoprecipitation of chromatin and incubated from 4 h to overnight at 4 °C depending on the antibody used as previously described.23 DNA that was immunoprecipitated was collected using magnetic separation, sequentially washed with high and low salt buffers, followed by lithium chloride and TE. DNA was then eluted in a solution of 0.1 M sodium bicarbonate with 1% SDS. To reverse the protein-DNA cross-links, Proteinase K (Sigma, St. Louis, MO, USA) was added, and the mixture was incubated at 62 °C for two hours. H3K27me3, H3K27ac, and H3K4me3 enriched DNA was purified using Nucleospin columns (Machrey-Nagel GmbH&Co, Germany). ChIP primers (shown in Table 2) were designed using the integrative ENCODE resource,51 and used to assess changes in levels of DNA associated with H3K27me3, H3K27ac, and H3K4me3.
Culture and treatment of immortalised human pancreatic ductal epithelial cells
Immortalized human derived pancreatic ductal epithelial cells of normal phenotype and genotype were obtained (AddexBio, Cat#:T008001) and maintained following the manufacturer’s instructions. In brief, cells were cultures at 37 °C in an atmosphere of 5% CO2 in complete Keratinocyte Serum-Free Media (supplemented with human recombinant EGF, Bovine Pituitary Extract and Antibiotic-Antimycotic [Gibco]). Cells were stimulated with EZH2 inhibitors using the same protocol for ex vivo human exocrine tissue. In brief, cells were seeded on Day 0 and allowed to adhere for 24 h. Stimulation was initiated on Day 1 with the first dose of GSK126 or Tazemetostat. Following a further 24 h of culture, the second dose was delivered in fresh K-SFM for a total of 10 μΜ GSK126 or 1 μM Tazemetostat. Cells were then harvested following a total of 48 h incubation in EZH2 inhibitor or DMSO vehicle control. To investigate transient effects of GSK126 and Tazemetostat, cells were cultured in drug free conditions for additional 48 h following which cultures were washed with K-SFM. Cells were incubated for a further 2 days, for a total of 96 h before harvesting.
Culture of immortalised and functionally mature human β-cells
The EndoC- βH5 cell line was purchased from Human Cell Design and cultured according to the manufacturer’s protocol using the provided reagents. In brief, cells were seeded on plates pre-coated in βCOAT diluted in DMEM (ThermoFisher, #11965092) with 100x penicillin/streptomycin (Gibco). Seeded cells βH5 were maintained at 37 °C in an atmosphere of 5% CO2 in the provided ULTIβ medium, with changes performed every 3 days.
Protein blotting
1 × 106 human pancreatic ductal epithelial cells were seeded in treated 6-well plates (Corning) and stimulated with either GSK126 or Taz as detailed above. Histone proteins were examined as previously described,52 using immunoblotting to assess acid purified nuclear proteins. In brief, protein concentrations were determined by incubating samples and standard concentrations of Bovine Serum Albumin (Invitrogen) with Bradford’s Reagent (Sigma). 2 ug of nuclear protein per sample was run on a 4–12% gel (Nu-Page, Invitrogen). Transfer was performed for 2 h using a PVDF membrane (Immobilon-FL; Millipore). Membranes were incubated in primary antibody against H3K27me3 (07-449, Millipore) or H3 (1B1B2, CST) at 4 °C overnight (dilutions listed in Table 2). Membranes were then washed and incubated at r.t.p. for 1 h in fluorescent secondary antibodies against mouse and rabbit (dilutions listed in Table 2). Images of the membranes were obtained using the LiCoR Odyssey infrared system. Quantification of the protein bands was performed using Image Studio with total H3 serving as a loading control.
Immunofluorescent analysis of ex vivo human exocrine tissue
1 × 106 cells from the donor pancreata were stimulated with EZH2 inhibitors or vehicle control for 48 h. Cells were resuspended in 10% FBS diluted in PBS and 0.1 × 106 cells were spun onto slides and fixed using 4% paraformaldehyde. Permeabilization was performed using 0.1% Triton X diluted in PBS for 10 min, followed by blocking using PBG (0.2% w/v gelatin, 2.5% w/v bovine serum albumin in PBS) for 1 h. Cells were co-stained for CK19 (HPA002465 Sigma-Aldrich), and INS (A0564, DAKO) by incubating overnight at 4 °C using human specific primary antibodies diluted in PBG (dilutions listed in Table 3). Fluorescent secondary antibodies against rabbit (Alexa Fluor 488), and guinea pig (IRDye® 680CW) (dilutions in Table 3) were incubated for 1 h at room temperature. Slides were then incubated with 4′,6-diamidino-2-phenylindole (DAPI) (0.10 μg/ml; D8417 Sigma-Aldrich) for 10 min and coverslips were mounted using Prolong Gold Anti-Fade mountant with DAPI (ThermoFisher). Images were obtained using the EVOS (ThermoFisher) with the TagBFP, Cy5, and GFP filters. Processing and analysis of images was performed using the Image J software.
Immunofluorescence analyses
0.5 × 105 human pancreatic ductal epithelial cells were seeded on coverslips and stimulated as previously described,53 prior to immunofluorescence. In brief, 4% paraformaldehyde was used to fix cells following stimulation with EZH2 inhibitors. Permeabilization was performed using 0.1% Triton X diluted in PBS for 10 min, followed by blocking for 1 h using PBG. Cells were co-stained for CK19 (HPA002465 Sigma-Aldrich), and INS (A0564, DAKO) by incubating overnight at 4 °C using human specific primary antibodies diluted in PBG (dilutions listed in Table 3). Fluorescent secondary antibodies against rabbit (Alexa Fluor 488), and guinea pig (IRDye® 680CW) (dilutions in Table 3) were incubated for 1 h at room temperature. Coverslips were rinsed in PBG and mounted using Prolong Gold Anti-Fade mountant with DAPI (ThermoFisher). Images were obtained with the EVOS (ThermoFisher). Images were processed and analysed using Image J.
Glucose stimulated insulin secretion assay
1 × 106 cells were seeded in 6-well plates. Cells were stimulated with 10 μM GSK126 and 1 μM Tazemetostat or vehicle control (DMSO) for 48 h in Miami medium and counted using the Cell Countess II to ensure equal number of cells (1 × 106) were used for the GSIS assay. All subsequent washes and incubations were using Krebs Buffer solution (KRB) made using 25 mM HEPES, 115 mM sodium chloride, 24 mM sodium hydrogen carbonate, 5 mM potassium chloride, 1 mM magnesium chloride heptahydrate, 0.1% bovine serum albumin, and 2.5 mM calcium chloride dihydrate dissolved in deionized water and sterile filtered. Following stimulation, the insulin containing Miami medium was removed and cells underwent two washes with 2.8 mM glucose KRB to reduce the background insulin signal. The low glucose insulin secretion was obtained by incubation in 2.8 mM glucose KRB for 1 h. Cells were then cultured in high (28 mM) glucose Krebs Buffer solution for 1 h to obtain the glucose stimulated insulin secretion. The concentration of insulin secreted into the supernatant was assessed using the Ultrasensitive Insulin ELISA (Mercodia) according to manufacturer’s guidelines. The insulin release by cells in response to hyperglycaemia was calculated as a fold change by adjusting inhibitor stimulated concentrations to control concentrations.
Non-sequential glucose stimulated insulin secretion assay
As recommended by the manufacturer’s protocol, a non-sequential GSIS was performed using mature EndoC-βH5 cells to assay insulin secretion. In brief, 1 × 105 cells were seeded in 96-well plates and cultured for 6 days in UTIβ1 medium provided. The medium was then changed to the starvation medium ULTI-ST the day before the GSIS was carried out. Next, the cells were washed with 0 mM glucose KRB followed by incubation in 2.8 mM or 28 mM glucose KRB for 1 h to obtain the glucose stimulated insulin secretion. The concentration of insulin secreted into the supernatant was assessed using the Ultrasensitive Insulin ELISA (Mercodia) according to manufacturer’s guidelines. The insulin release by cells in response to hyperglycaemia was calculated as a fold change by adjusting inhibitor stimulated concentrations to control concentrations.
