Fig 2: Quiescent state of differentiated cells entails the assembly of RBL2-E2F4-HDAC1-BRM repressor complex at PARP1 promoter and the HDAC1-BRM induced chromatin compaction. The effect of RBL2 silencing on PARP1 expression was determined with western blot (a) and real-time PCR (b) in human blood-derived monocytes 72 h after cell transient transfection with siRNA. Transcription regulating epigenetic marks (c,d,e) as well as HDAC1 (f), histone H3 density (g) and BRM (h) at PARP1 promoter were analyzed with ChIP. The interaction among repressive complex components was confirmed with immunoprecypitation (i) by pulling down HDAC1 and detection of co-immunoprecipitated proteins with western blot. In order to verify the contribution of HDACs (j,k) and BRM (l,m) in PARP1 repression blood-derived monocytes were incubated with the corresponding inhibitors (iHDAC – sodium butyrate – 0.5 mM and iBRM/iBRG1 – PFI-3 – 1 µM) for 24 h. PARP1 expression was monitored at the protein level with western blot (j,l)), while PARP1 mRNA was determined with real-time PCR (k,m). The effect of iHDAC on HDAC1, RBL2 and BRM association with chromatin (n,o,p) and the chromatin compaction (H3 density) in differentiated cells was assayed with ChIP (r).

Fig 3: CBX2 inhibits expression of the tumour suppressor gene RBL2 to promote cell growth in TNBC. (a) RNA-seq analysis of RBL2-DREAM complex target genes in MDA-MB-231 cells transfected with siCBX2#2 and #3 compared to siSCR transfected cells (n = 3). Data are presented as log2 fold change in expression. The red line represents a 1.5-fold increase or decrease in expression compared to siSCR transfected cells. All genes are significantly differentially expressed in CBX2 depleted cells compared to siSCR transfected cells (padj < 0.05). (b) MDA-MB-468 cells were transfected with either a non-silencing control siRNA (siSCR) or siCBX2#2/#3 and grown for 72 h, followed by RNA extraction and qPCR analysis of RBL2-DREAM complex target genes. qPCR data are an average of 3 repeats ± SEM and are expressed relative to gene expression in siSCR transfected cells. p-values were determined by Student’s t-test comparing CBX2 depleted cells to siSCR transfected cells (* denotes p < 0.05). (c) MDA-MB-231 and MDA-MB-468 cells were transfected with siRNA and grown for 72 h, followed by protein extraction and Western blot analysis using antibodies specific to RBL2, CBX2 and α-tubulin. α-tubulin was used to compare protein loading between samples. (d) CUT&RUN qPCR analysis of the promoter region of RBL2 in MDA-MB-231 cells using antibodies specific to CBX2 and isotype control (IgG). Data are an average of 2 independent experiments ± SEM and are expressed relative to the level of enrichment measured in the IgG control. (e) Correlation of CBX2 and RBL2 gene expression in the TCGA, PanCancer Atlas Breast Invasive Carcinoma data set. Value = Spearman co-efficient. p < 0.05. (f) ChIP-qPCR analysis of RBL2-DREAM complex target gene promotors in MDA-MB-231 and MDA-MB-468 cells transfected with either siSCR or a siCBX2 targeting pool using an antibody specific to RBL2. Data are an average of 3 independent experiments ± SEM and are expressed relative to the level of enrichment measured in the siSCR transfected cells. p-values were determined by Student’s t-test comparing CBX2 depleted cells to siSCR transfected cells (* denotes p < 0.05).

Fig 4: Repression of PARP1 transcription is associated with cell cycle exit to G0 and the recruitment of RBL2 protein to PARP1 promoter. Panel (a) shows the graphic representation of PARP1 regulatory region adjacent to TSS, which spans POLR2A and H3K27ac according to the ENCODE. Small black boxes indicate the binding sites of transcription factors, which are considered as possibly involved in the regulation of PARP1 transcription in hematopoietic stem and progenitor cells and monocytes. The association of E2F1 and E2F4 transcription factors as well as RB1 and RBL2 proteins with PARP1 promoter in undifferentiated and differentiated THP-1 cells was determined with ChIP (b). The effect of LPS (100 ng/ml, 24 h) and cell cycle inhibitors (phase specific arrest in G1 and G2 was induced by cell treatment with 200 μM mimosine (Mim) and 0.5 μM nocodasol (Noco) for 48 h) on PARP1 protein level was monitored with western blot (upper c). In order to eliminate the impact of caspases, likely activated by cell cycle inhibitors, on the level of PARP1 protein LPS, mimosine and nocodasol were added in a mixture with caspase-3 and -7 inhibitor Z-DEVD-FMK (10 μM) (lower c). PARP1 mRNA was quantified in cells treated with LPS, Mim and Noco using real-time PCR (d). The recruitment of E2F1, E2F4, RB1 and RBL2 to PARP1 promoter in cells arrested in G1 phase was determined with ChIP (e).

Fig 5: CBX2 is associated with DREAM complex activity in other cancer types. MCF-7 and T47D cells were transfected with either a non-silencing siRNA control (siSCR) or CBX2 targeting siRNAs (siCBX2#1/#2/#3) and grown for 72 h followed by protein expression analysis (a), cell growth analysis (b) and cell cycle analysis (c). (a) CBX2 and RBL2 protein expression was assessed by Western blot. α-tubulin protein expression was assessed to compare protein loading between samples. (b) Cell growth was assessed by cell counts. Cell count data are the mean of 3 repeats ± SEM and are expressed relative to cell counts for siSCR transfected cells. (c) The effect on the cell cycle was assessed by propidium iodide flow cytometry. Data show the % of cells in the sub-G1 phase of the cell cycle and are an average of 3 repeats ± SEM. p-values for (b,c) were determined by Turkey’s multiple comparisons test (* denotes p < 0.05). (d) RNA-seq analysis of RBL2-DREAM complex target genes in MCF-7 cells transfected with siCBX2#3 compared to siSCR transfected cells. Data are presented as log2 fold change in expression. The red line represents a 1.5-fold increase or decrease in expression compared to siSCR transfected cells. All genes are significantly differentially expressed (padj < 0.05). (e) GSEA of the CBX2-regulated transcriptome (defined as genes significantly (padj < 0.05) up (n = 2973) or down (n = 2011) regulated 1.5-fold after knockdown compared to siSCR controls) against the “Hallmark” curated data sets identified significant negative enrichment of genes associated with HALLMARK_E2F_TARGETS. NES = normalised enrichment score. p = nominal p-value. (f) GSEA of genes significantly positively and inversely correlated (Spearman’s p < 0.05) with CBX2 expression from TCGA PanCancer Database for Lung Adenocarcinoma and GBM. NES = normalised enrichment score. p = nominal p-value.