Plasmids and recombinant adenovirus
The search for EZH2-related proteins was performed by comparing the human EZH2 SET domain protein sequence (GenBank: BC010858) against the Expressed Sequence Tag database using the BLAST and the UniGene programs from the National Center for Biotechnology Information (National Institutes of Health, Bethesda, MD, USA). This comparison indicated the presence of the EZH2β-encoding sequence (NCBI: NM_152998.2). The exact sequences matching this entry as well as other PRC2 proteins, such as SUZ12 (GenBank: BC015704) and EED (GenBank: BC068995), were verified by sequencing and analysis of publically deposited cDNAs. Standard molecular biology techniques were used to clone full-length EZH2α, EZH2β, SUZ12 and EED into pcDNA3.1/HIS (Invitrogen, Carlsbad, CA, USA). All constructs were verified by sequencing at the Mayo Clinic Molecular Biology Core Facility. QuickChange Site-Directed Mutagenesis was performed as suggested by the manufacturer (Agilent Technologies, Santa Clara, CA, USA). Silent mutations were made to delete endogenous HindIII and XbaI restriction enzyme sites to permit passage of EZH2α and EZH2β cDNAs into pacAd5 CMV K-N pa shuttle vector. Epitope-tagged (6XHis-Xpress) EZH2α and EZH2β were generated as recombinant adenoviruses by the Gene Transfer Vector Core at the University of Iowa. Empty vector (pacAD5 CMV) was used as the experimental control.
Human tissue RNA panel
Human total RNA for 22 major organs and tissues was commercially obtained from Ambion (Austin, TX, USA) and Stratagene (Agilent). cDNA was generated from 1 μg RNA using SuperScriptT III enzyme (Invitrogen) according to manufacturer’s instructions. cDNA concentrations were assessed via internal housekeeping gene glyceraldehyde-3-phosphate dehydrogenase or hypoxanthine phosphoribosyltransferase. PCRs were performed with the following cycle conditions: 30 to 35 cycles of 94°C for 15 s, 50°C for 30 s, and 72°C for 2 min using 1 to 2 μl of cDNA product. Amplified products were electrophoresed on 1.5% agarose gels, digitally imaged, and quantified with ImageJ (National Institutes of Health, Bethesda, MD, USA). Primers were synthesized by Integrated DNA Technologies (Coraville, IA, USA). PCR primers may be found in Additional file3: Table S1.
Western blot analysis
Samples were run on 4% to 20% (Lonza, Walkersville, MD, USA), 6% or 10% SDS-PAGE gels and electroblotted onto polyvinylidene difluoride membranes (Millipore, Billerica, MA, USA). The membranes were blocked in 5% bovine serum albumin or milk in Tris buffered saline with Tween (TBST) for 1 h at room temperature. The blots were incubated overnight at 4°C with primary antibody. After repeated washes in TBST, horse radish peroxidase -conjugated anti-rabbit or mouse IgG secondary antibody (1:2,000 to 5,000) was added for 1 h at room temperature. Blots were developed by Pierce ECL Chemiluminescent Substrate (Thermo Scientific, Rockford, IL, USA). Human tissue lysates were procured from Calbiochem (Millipore) as a ready-to-probe INSTA-blot. Approximately 20 μg of lysate was loaded per tissue with loading controlled via amido black straining by the manufacturer. The blot was incubated overnight with EZH2β (purified, 1:2,000) and subsequently stripped and re-incubated with β-actin (1:1,000; Sigma, St. Louis, MO, USA).
Synthesis, purification and validation of EZH2α and EZH2β antibodies
A 21-mer peptide bridging across the large insert region missing from EZH2β compared to EZH2α was synthesized, high performance liquid chromatography-purified and conjugated to keyhole limpet hemocyanin by the Mayo Clinic Protein Core. For the EZH2α antibody, a 21-mer peptide was synthesized that localized to the insert region. Subsequently, a rabbit was immunized with the peptide, and test and final bleeds were performed by Cocalico Biologicals (Reamstown, PA, USA). For the antibody that recognizes both EZH2α and EZH2β, a 21-met peptide in a region conserved between the two proteins was synthesized. The anti-serum was affinity purified using the Protein A IgG Purification Kit according to the manufacturer’s protocol (Pierce Biotechnology, Rockford, IL, USA). To test the specificity of the antibodies, Chinese hamster ovary epithelial cells were transfected with a histidine-tagged (HIS)/EZH2α and HIS/EZH2β. Whole cell lysates (30 μl) and pancreatic cell lines (30 μg) were resolved on 4% to 20% SDS-PAGE gels, and probed with whole sera of EZH2α (1:200), EZH2β (1:200) and EZH2αβ (1:200). Blots were stripped and re-probed with Omni-probe (D-8) (1:1,000; Santa Cruz Biotechnology, Santa Cruz, CA, USA) to ensure equal loading.
Panc1 epithelial cells were plated at a cell density of 1 × 106 cells/100 mm dish and transduced with epitope-tagged (6XHis-Xpress) EZH2α, EZH2β or empty vector at multiplicity of infection (MOI) 150. Subconfluent cells were lysed in a buffer containing 20 mM Tris-Cl at pH 8.0, 100 mM NaCl, 1 mM EDTA, 0.5% Nonidet P-40 and a protease inhibitor tablet (Roche, San Francisco, CA, USA). Proteins were immunoprecipitated as previously described using 10 μg of Omni-probe (D-8) (Santa Cruz Biotechnology). Resulting complexes were resolved on a 6% or 10% SDS-PAGE gels, using antibodies against SUZ12 (1:1,000; Cell Signaling, Beverly, MA, USA) and EED (1:1,000; Cell Signaling). Membranes were stripped and incubated with Omni-probe (D-8) (1:1,000; Santa Cruz), to ensure equal loading of precipitated EZH2 proteins. A 5% input control of whole cell lysates under all conditions was included to ensure the presence of uniform levels of the proteins of interest.
Cell culture, immunofluorescence and confocal microscopy
Cell lines were obtained from the American Type Culture Collection (ATCC, Rockville, MD, USA) and maintained according to their recommendations. Immunofluorescence and confocal microscopy were performed as previously described. Panc1 cells were plated in eight-chamber glass slides at a density of 5 × 104 cells/chamber and transduced with epitope-tagged (6XHis-Xpress) EZH2α, EZH2β or empty vector at MOI 150. Primary antibodies were used at the following dilutions: EZH2α (1:50; described above), EZH2β (1:50; described above), EZH2 (1:200; Cell Signaling) and Omni-probe (D-8) (1:250; Santa Cruz). Images were obtained at 100× magnification. Frozen cryosections of human testis (5 μm) were purchased from Zyagen (San Diego, CA, USA). Sections were fixed in ice-cold acetone for 10 min and rehydrated in PBS for 3 min. Endogenous peroxidase activity was quenched using a 3% hydrogen peroxide in methanol for 20 min (Sigma). Avidin/Biotin blocking was performed using a kit from Vector Laboratories (Burlingame, CA, USA). Tissues were blocked in CAS Block for 1 h (Invitrogen) prior to overnight incubation at 4°C in primary antibody. Dilutions were as follows: EZH2β (1:200; described above) and EZH2 (1:200; Cell Signaling). Sections were subsequently washed in PBS and incubated in biotinylated goat anti-rabbit secondary antibody (Vector Laboratories) for 30 min. Samples were incubated in Alexa Fluor-488-streptaviding conjugate (Invitrogen). Sections were counterstained with Hoescht. Images were obtained at 10× magnification.
Microarray, validation and subnetwork constructions
BxPC3 epithelial cells were plated at a density of 1 × 106 cells/100 mm dish and transduced with empty vector, EZH2α or EZH2β (Ad5CMV) at an MOI of 150. RNA was prepared as previously described 48 h after transduction. Experiments were performed from pooled biological triplicates in technical duplicates. The transduction efficiency of these cells at MOI 150 is 81.3 ±1.99% as determined by transduction with GFP adenovirus. Global gene expression profiling was carried out at the Microarrays Facility of the Research Center of Laval University CRCHUL using the Affymetrix Human Gene 1.0 ST arrays (28,869 well-annotated genes and 764,885 distinct probes). Intensity files were generated by Affymetrix GCS 3000 7 G and the GeneChip Operating Software (Affymetrix, Santa Clara, CA, USA). Data analysis, background subtraction and intensity normalization was performed using robust multiarray analysis. Genes that were differentially expressed along with false discovery rate were estimated from t test (>0.005) and corrected using Bayes approach[46, 47]. Data analysis, hierarchical clustering and ontology were performed with the OneChanelGUI to extend affylmGUI graphical interface capabilities and Partek Genomics Suite, version 6.5 (Partek Inc., St. Louis, MO, USA) with analysis of variance analysis. A cutoff of expression log2 fold change of two and P <0.05 was set to identify molecules whose expression was significantly differentially regulated. EZH2β and EZH2α baseline transcript levels were assessed compared to overexpression by qPCR to assure that each isoform was expressed at approximately equivalent levels (Additional file4: Figure S3A). Additionally, a small subset of targets was validated by qPCR (Additional file4: Figure S3C).
Selected probes and their fold changes were loaded into IPA Software (Ingenuity Systems. Each identifier was mapped to its corresponding object in the Ingenuity Knowledge Base. These molecules, called Network Eligible molecules, were overlaid onto a global molecular network developed from information contained in the Ingenuity Knowledge Base. For the purposes of network reconstruction, a log2 fold change of two was used. Networks of Network Eligible molecules were then algorithmically generated based on their connectivity. The functional analysis of a network identified the biological functions and/or diseases that were most significant to the molecules in the network. The network molecules associated with biological functions and/or diseases in the Ingenuity Knowledge Base were considered for the analysis. Right-tailed Fisher’s exact test was used to calculate a P-value determining the probability that each biological function and/or disease assigned to that network was due to chance alone.
Flp-in system, transfection and luciferase assays
The human FOXP3 core promoter containing −511 bp from transcription start site was amplified by PCR using FOXP3 promoter sequence-specific primers from position −511 to +176. The genomic DNA extracted from CD4+ T cells of a healthy donor was used as a template. The PCR product was subcloned in the pGL3 basic vector (Promega, Madison, WI, USA). Similarly, the FOXP3 core promoter plus the first enhancer (E1) containing −511 bp to +2,738 was also amplified by PCR and subcloned in the pGL3 basic vector (Promega). The Flp-In system (Invitrogen) was used for the generation of a stable human FOXP3 core promoter and FOXP core +E1 promoter Flp-In-Jurkat. Flp-In-Jurkat cells (Invitrogen) were co-transfected with FOXP3 core or FOXP3 core + E1 in a pcDNA5/ FLP recombination target (FRT) vector and a FLP-recombinase vector (pOG44) (pOG44:FOXP3 core or FOXP3 core + E1/pcDNA5/FRT ratio = 9:1), resulting in a stable integration of the gene of interest at the FRT-site in the genome. For the selective growth test, individual cells were grown in 24-well plates. The culture medium was supplemented with hygromycin at 250 μg /ml or 100 μg/ml. Two million FOXP3 core and FOXP3 core + E1 Flp Jurkat cells were transfected using the Amaxa Cell Line Nucleofector Kit V for Jurkat cells according to the optimized protocol provided with the kit. Two micrograms of plasmid DNA for EZH2α, EZH2β, SUZ12 and EED were used in the nucleofection procedure. Luciferase assays were done following the manufacturer’s recommendations (Promega). Data represent the mean and SD of three independent experiments (*P <0.05).
Adenoviral transduction and flow cytometry
The CAR transgenic mouse was obtained through the NIAID Exchange Program, NIH: Balb/cJ[Tg]CARdelta1-[Tg]DO11.10 mouse line #4285[49, 50]. Murine naïve CD4+ splenocytes were isolated using a combination of magnetic separation kits (Miltenyi Biotec, Auburn, CA, USA). Sequential use of the CD4+CD25+ regulatory T cell isolation kit and the CD4+CD62L+ T cell isolation kit resulted in naïve FOXP3-negative T cells used for in vitro induction of FOXP3. Naïve T cells were isolated from the CAR transgenic Balb/cJ[Tg]CARdelta1-[Tg]DO11. Cells were activated for 48 h with empty vector, EZH2α or EZH2β at an MOI of 250. The transduction efficiency of these cells as determined by flow cytometry with propidium iodide exclusion using GFP adenovirus is 89.4 ±2.1%. Cells were activated under the typical stimulation conditions for 3 days and processed for ChIP and qPCR to determine methylation of H3K27me3 marks at the FOXP3 core promoter and levels of FOXP3 expression, respectively. Flow cytometry was used to look at levels of FOXP3 expression within the CD4+ population across four biological replicates. Intracellular staining procedures for FOXP3 were followed using the application notes from Alexa Fluor 488 anti-mouse/rat/human FOXP3 (BioLegend, San Diego, CA, USA). For qPCR analysis, biological triplicates were pooled and analyses performed in technical duplicate. Data represent the mean and SD of four independent experiments (*P <0.05).
T cell stimulation
In vitro activation of the isolated T cells followed similar conditions among the different cell types. Anti-CD3, OKT3 (eBioscience, San Diego, CA, USA) for the Jurkat cells, 145-2C11 (BD Biosciences, San Jose, CA, USA) for the mouse T cells, and UCHT1 (BD Biosciences) for the human T cells was platebound at 2 μg/ml. Soluble anti-CD28 (BD Biosciences) at 2 μg/ml plus 100 units/ml IL-2 was added to the cultures throughout the incubation period. Human transforming growth factor beta-1 recombinant (AUSTRAL, San Romano, CA, USA) at a concentration of 5 ng/ml was used to generate adaptive Treg cells.
Chromatin immunoprecipitation assays
ChIP assays were performed as previously described using H3-27me3 (Cell Signaling) and Omni-probe (D-8) (Santa Cruz) antibodies. Primers used to analyze the FOXP3 promoter are listed in Additional file3: Table S1. For the Polycomb target screen, mRNA and ChIP samples were processed from BxPC3 epithelial cells as described above and used with the Human Polycomb and Trithorax Target Genes ChIP PCR Array (SA Biosciences, Valencia, CA, USA). ChIP were performed in biological duplicate and of the 84 targets present on the array, 78.6% (66 out of 84) were occupied by EZH2α, serving as an internal positive experimental control. Expression profiling was performed in biological triplicate with the averaged values reported.
3H-thymidine incorporation proliferation assay
Naïve T cells from a CAR D011.10 mouse were isolated and transduced with empty vector, EZH2α and EZH2β as described above. Cells were plated at 6.6 × 105/ml in complete Roswell Park Memorial Institute medium containing αCD28 at 2 μg/ml plus 100 units/ml IL-2, and 200 μl was added per well to a 96-well round bottom plate coated with αCD3 at a concentration of 2 μg/ml. Five days after plating, 20 μl of 3H-thymidine (6.7 Ci/mmol NET-027) at a 1:20 dilution in complete Roswell Park Memorial Institute medium (1.0 μCi) was added to each well and incubated for approximately 18 h. Cells were harvested and counted on the microtiter plate counter.
Bioinformatics and statistical analysis
Bioinformatics-assisted splice-mapping of the human EZH2 locus was performed using AceView. An evolutionary dendrogram of common invertebrate and vertebrate EZH2 isoforms was created using the Geneious Tree Builder with a BLOSUM62 matrix, free end global alignment with a gap open penalty of 12 and a gap extension penalty of 3 (no outbound group selected). Predicted EZH2 splice variant sequences were curated from National Center for Biotechnology Information. Statistical analyses were performed using Graphpad Prism (La Jolla, CA, USA). Descriptive analyses including means and SDs were performed in normally distributed data. One-way analysis of variance with Tukey’s post-hoc test was utilized to determine statistically significant observations. A P-value of <0.05 was considered as statistically significant.