Cell culture

MCF10A MII cells were obtained from Dr Peter ten Dijke (Leiden University, The Netherlands) and maintained at 37 °C and 5% CO2 in DMEM/F12 (Gibco, Life Technologies Ltd, Paisley, UK), supplemented with 5% fetal bovine serum (FBS) (Gibco, Life Technologies Ltd, Paisley, UK), 20 ng/ml EGF (PeproTech, EC Ltd, London, UK), 100 ng/ml cholera toxin (Sigma-Aldrich AB, Stockholm, Sweden), 0.5 µg/ml hydrocortisone (Sigma-Aldrich AB, Stockholm, Sweden), 10 µg/ml insulin (Sigma-Aldrich AB, Stockholm, Sweden and Gibco, Life Technologies Ltd, Paisley, UK). HCC1954 breast cancer cells, obtained from Dr Andrew J. G. Simpson (Ludwig Cancer Research, New York, USA), were maintained in RPMI-1640 (Sigma-Aldrich AB, Stockholm, Sweden and Gibco, Life Technologies Ltd, Paisley, UK), supplemented with 10% FBS and 100 U/ml penicillin and 100 mg/ml streptomycin (Sigma-Aldrich Sweden AB, Stockholm, Sweden) and 20 ng/ml EGF. The cell lines were frequently tested for the absence of mycoplasma and were authenticated by short tandem repeat analysis.

TGFβ treatment and inhibitors

Recombinant human TGFβ1 (denoted TGFβ in this study) was purchased from PeproTech (EC Ltd, London, UK). Cells were starved overnight in a medium containing 0.2% serum (MCF10A MII) or 3% serum (HCC1954) before treatment with 5 ng/ml TGFβ. The following small molecule inhibitors were utilized at the indicated concentrations: TGFβRI kinase inhibitors (ALK5i) SB505124 (2.5 µM; Sigma-Aldrich AB, Stockholm, Sweden) and LY2157299 (2.0 µM; Sigma-Aldrich AB, Stockholm, Sweden), MEK1/2 inhibitor (MEKi) AZD6244 (0.25 µM; Selleckchem, Houston, TX 77230, USA), Jun N-terminal kinase inhibitor (JNKi) SP600125 (10 µM; Calbiochem, Merck, Stockholm, Sweden), p38 MAP-kinase inhibitor (p38i) SB203580 (10 µM; Tocris Bioscience, Bio-techne, Bristol, UK), and EZH2 inhibitor GSK343 (5 µM; Sigma-Aldrich AB, Stockholm, Sweden). All kinase inhibitors were dissolved in DMSO and added to the cells 20 min before TGFβ treatment. Protein synthesis was blocked by cycloheximide (CHX; C1988, Sigma-Aldrich AB, Stockholm, Sweden), administered to the cells at the same time as TGFβ treatment at a final concentration of 20 µg/ml.

siRNA transfections

The ΔNp63 specific On-target plus SMART siRNA (sense sequence, 5´-GGACAGCAGCAUUGAUCAAUU; antisense sequence, 5´-UUGAUCAAUGCUGCUGUCCUU),

the On-target plus Non-Targeting Control siRNA (Cat no: D-001810-01-20), the SUZ12 On-target plus siRNA pool (Cat no: L-006957-00-0005), the SMAD2 On-target plus siRNA pool (Cat no: L-003561-00-0005) and the SMAD3 On-target plus siRNA pool (Cat no: L-020067-00-0005) were purchased from Dharmacon (Horizon Discovery, Cambridge, UK). Stealth siRNAs specific for p63 (ID: HSS189462), SMAD2 (ID: VHS41107), SMAD3 (ID: VHS41111) and control siRNAs (Cat No. 12935-300 and 12935-200) were obtained from Invitrogen (Life Technologies, Ltd, Paisley, UK). siRNAs, at 20 nM (stealth) or 25 nM (siRNA pool) final concentration, were transfected using SiLentFect (Bio-Rad Laboratories AB, Solna, Sweden) transfection reagent according to the manufacturer’s instructions.

Sphere formation assay

After siRNA transfection, HCC1954 cells (1 × 104/well) were seeded in 96-well Costar ultra-low attachment plates (Corning, Corning, NY, USA) in RPMI medium supplemented with 20 ng/ml EGF and 10 ng/ml bFGF (Sigma-Aldrich AB, Stockholm, Sweden) and incubated with or without 5 ng/ml TGFβ for 8 days. Total sphere numbers per well (diameter > 50 μm) were counted under a microscope.

Chromatin immunoprecipitation (ChIP)

ChIP was performed as previously described [16, 17]. In summary, cells were fixed in 1% formaldehyde, washed with ice-cold PBS, harvested by scraping, pelleted and resuspended in 1 ml of sodium dodecyl sulphate (SDS) lysis buffer (1% SDS, 50 mM Tris-HCl, pH 8.0, 10 mM EDTA, supplemented with Complete EDTA-free protease inhibitors (Roche Diagnostics, Scandinavia AB, Bromma, Sweden)). The cell lysates were subjected to sonication in a water bath using Diagenode Bioruptor sonicator (Diagenode, Bionordika, Stockholm, Sweden), with 30 s pulses for 5–10 min. Following sonication, samples were centrifuged at 14,000 rpm at 4 °C for 10 min. After removal of a control aliquot (whole-cell extract serving as an input), supernatants were diluted in ChIP dilution buffer (1% Triton X-100, 20 mM Tris-HCl, pH 8.0, 150 mM NaCl, 2 mM EDTA), and incubated at 4 °C overnight with antibodies precoupled to anti-mouse IgG or Protein A Dynabeads (Invitrogen, Life Technologies, Ltd, Paisley, UK) in PBS supplemented with 0.5% bovine serum albumin (BSA). The antibodies used for ChIP were raised against p63 (ab124762, Abcam, Cambridge, UK), H3K27ac (39685, Active motif, Carlsbad, CA and ab177178, Abcam, Cambridge, UK), H3K27me3 (61017, Active motif, Carlsbad, CA), p300 (61401, Active motif, Carlsbad, CA, and ab14984, Abcam, Cambridge, UK) and SUZ12 (39357, Active motif, Carlsbad, CA).

The precipitated complexes were washed five times in ChIP washing buffer (50 mM HEPES-KOH, pH 7.0, 0.5 M LiCl, 1 mM EDTA, 0.7% deoxycholate, 1% Igepal CA-630) and once with TE buffer (10 mM Tris-HCl, pH 8.0, 1 mM EDTA). Immunoprecipitated samples were eluted and reverse cross-linked at 65 °C in SDS lysis buffer. Genomic DNA was purified with a PCR purification kit (Qiagen, AB, Sollentuna, Sweden). The immunoprecipitated DNA was analyzed by qRT-PCR using locus-specific primers (the complete primer list can be found in Additional file 1-Table S1 in Supplementary Information) and normalized to the input DNA. The IgG control was included in all the experiments in order to check and confirm the specificity of the antibody used. The quantified relative fold change corresponded to the enrichment in each gene locus under treatment conditions divided by the enrichment in the control condition (control- or the sictrl-condition), as indicated.

RNA isolation, cDNA synthesis and quantitative real-time-PCR

RNA was isolated by Total RNA Purification Kit (Norgen Biotek Corp, Canada). cDNA was prepared using High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, Life Technologies, Ltd, Paisley, UK) utilizing 0.5 µg of total RNA, according to the manufacturer’s instructions. The cDNA samples were diluted 10 times in water. qRT-PCR was performed using 2× qPCR SyGreen Mix (PCR Biosystems, London, UK) and CFX96 real-time PCR detection system (Bio-Rad Laboratories AB, Solna, Sweden), according to the manufacturer’s instructions. Relative gene expression was determined using the ΔΔCt method. The expression was normalized to the GAPDH gene and quantified relative to the control condition. The complete primer list can be found in Additional file1-Table S2 in the Supplementary Information. Normalized mRNA expression levels are plotted in bar graphs that represent average values from triplicate determinations with standard deviations (SD). Each independent experiment was repeated at least three times.

Nuclear/cytoplasmic fractionation

For mass spectrometry and co-immunoprecipitation analyses, nuclear and cytoplasmic fractions of MCF10A MII cells were separated after treatment or not with TGFβ (5 ng/ml) for 6 h. Briefly, cells were rinsed with PBS twice, scraped in PBS and centrifuged at 4oC for 5 min at 450×g. The cell pellet was resuspended in 50 mM Tris-HCl, pH 7.5, 10 mM MgCl2, 15 mM CaCl2, 1.5 M sucrose, complemented with 1% of protease inhibitor and 1% of 0.1 M of dithiothreitol (DTT). Cells were next incubated on ice for 15 min and 10% Igepal CA-630 was added before agitation and centrifugation for 30 s at 11,000×g. The supernatant contained the cytoplasmic fraction. The pellet was next resuspended in 50 µl of nuclear extraction buffer (20 mM HEPES, pH 7.9, 1.5 mM MgCl2, 0.42 M NaCl, 0.2 mM EDTA, 25% glycerol) complemented with 1% of protease inhibitor and 1% of 0.1 M DTT and agitated for 20 min at 4oC. Nuclear fraction was obtained as the supernatant after centrifugation for 5 min at 20,000×g at 4oC. Proteins were then quantified and subjected to SDS-polyacrylamide gel electrophoresis (SDS-PAGE).

Co-immunoprecipitation (Co-IP) and immunoblotting analysis

For the co-immunoprecipitation assay, MCF10A MII cells treated or not with 5 ng/ml of TGFβ for 45 min or 6 h were lysed in lysis buffer (1% Triton X-100, 20 mM Tris-HCl, pH 7.5, 150 mM NaCl, 10% glycerol) and incubated overnight with anti-mouse IgG or protein A Dynabeads (Invitrogen, Life Technologies, Ltd, Paisley, UK) that had been preincubated with the indicated antibodies or mouse immunoglobulin G1 (IgG), (MAB002, R&D systems, Bio-techne, Minneapolis,MN, USA) or rabbit IgG (SouthernBiotech, Birmingham, AL, USA) in PBS, supplemented with 0.5% BSA. Following precipitation, the complexes were washed three times with lysis buffer and the immunoprecipitated proteins were eluted in 2× SDS Laemmli sample buffer, subjected to SDS-PAGE and blotted onto nitrocellulose membranes (Cytiva, Danaher, Uppsala, Sweden). The chemiluminescent signal was detected using the Immobilon Western kit (Merck Millipore, Stockholm, Sweden). For the immunoblotting analysis of total cell extracts, cells were lysed in 2× SDS Laemmli sample buffer (5% SDS, 25% glycerol, 150 mM Tris-HCl pH 6.8, 0.01% bromophenol blue, 100 mM DTT) prior to SDS-PAGE. The intensities of the bands from the chemiluminescent blot images of p-p63 (Ser66/68), p63 and tubulin from three independent experiments were quantified by Image lab 6.1 software and the intensity values of p63 bands were divided by the values of those of tubulin for the purposes of loading normalization. The normalized p63 values were then used to calculate the ratio of p-p63/p63 presented in the corresponding figures.

The antibodies used for co-IP and/or immunoblotting were raised against: phospho-Ser160/162 p63 (Ser 66/68 in ΔNp63) (#4981), SMAD3 (#9523), ERK1/2 (#4695), phospho-Thr202/Tyr204 ERK1/2 (#4370), phospho-c-JUN (Ser63) (#9261), p38 MAPK (#9212), phospho-p38 MAPK (Thr180/Tyr182) (#9211)  and Histone 3 (#9712) (Cell Signaling Technology, Danvers, MA, USA), SMAD2/3 (#610843) and cJUN (#610327) (BD Transduction Laboratories, Biosciences-Europe, Stockholm, Sweden), phospho-Ser465/467 SMAD2 (home-made [28]), Tubulin (T0198) (Sigma-Aldrich, AB, Stockholm, Sweden), DNMT1 (H-300, sc-20701) (Santa Cruz Biotechnology, California, USA), CHD4 (ab70469) and HDAC2 (ab51832) (Abcam Cambridge, UK) and NCOR2 (PAI-843) and HDAC3 (7G6C5) (Invitrogen, Life Technologies, Ltd, Paisley, UK).

Mass spectrometry analysis

The nuclear fractions of MCF10A MII cells after treatment or not with TGFβ (5 ng/ml) for 6 h were subjected to immunoprecipitation with a p63 antibody immobilized on protein A Dynabeads. The complexes bound to beads were then subjected to mass spectrometry analysis at the Clinical Proteomics Mass Spectrometry Facility, Science for Life Laboratory, Karolinska Institutet, Sweden.

Briefly, on-bead reduction, alkylation and digestion (trypsin, sequencing grade modified, Pierce, Thermo Fischer Scientific, Sweden) was performed, followed by SP3 peptide clean-up of the resulting supernatant [29]. Each sample was separated using a Thermo Scientific Dionex nano LC-system in a 3 h 5–40% ACN gradient coupled to Thermo Scientific High Field QExactive. The software Proteome Discoverer vs. 1.4 including Sequest-Percolator for improved identification was used to search the Homo sapiens Uniprot database for protein identification, limited to a false discovery rate of 1%.

Pathway enrichment analysis

The pathway enrichment analysis of the significantly enriched proteins was performed using the Enrich tool to query the gene ontology molecular function database. The UMAP dimensionality reduction method was applied to visualize the scatter plot of the enriched pathways utilizing the standalone enrichment analysis visualizer Appyter. The top significantly enriched molecular functions are indicated in Fig. 4 [30] .

The area under the curve (AUC) of the integrated signal intensity was used to quantify the relative abundance of each identified protein in the corresponding samples. The scaled AUC values were used for sample clustering and generation of heatmaps utilizing the pheatmap package in R.

Statistical analysis

The figures and figure legends present the number of biological and technical replicates and the assessment of statistical significance. Data are presented as the mean ± SD from at least three independent experiments. Two-experimental group comparisons were performed using two-tailed unpaired Student’s t-test and multiple group comparisons were performed using the two-tailed unpaired Student’s t-test with Bonferroni correction. Statistical significance is represented by p-values *p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001.