Analysis of functional BRD4 inhibition and resulting cellular

4.1.3.1 Cell cycle distribution analysis and determination of apoptosis after BET inhibition

To further characterize the effect of BET inhibitors on the cellular phenotype a cell cycle distribution assay was performed using a flow cytometry EdU staining assay. EdU is a nucleotide analog that gets incorporated into the genomic DNA during the synthesis (S) phase of the cell. EdU can be specifically stained and used, together with the DNA content stained by DAPI, to determine the cell cycle phases of the cell.

In both sensitive and resistant cell lines the cell cycle distribution was determined after 24 h of treatment with the BET inhibitors JQ1, OTX-015 or I-BET762. H1373 and DV90 cells showed a dose-dependent shift from S phase towards G0/G1-population upon treatment (Figure 19A, B), with a stronger effect in H1373 compared to DV90 cells.

Figure 19: Cell cycle distribution analysis of two sensitive cell lines H1373, DV90 and two insensitive cell lines A549 and H460 after BET inhibitor treatment. A: Representative cell cycle distribution of viable H1373 or A549 cells 24 h after 1 μM JQ1 treatment. The S-phase cell population is shown in green, G0/G1 in blue and G2 in purple. B: Cell cycle analysis of sensitive DV90 and H1373 and resistant A549 and H460 cells following 24 h of JQ1, OTX-015 or I-BET762 treatment. Data are shown as the mean (n=2) for JQ1.

Additionally, induction of programmed cell death was determined in H1373, DV90, H460 and A549 cells. Dose-dependent induction of apoptosis after 48 h of JQ1 or OTX-015 treatment was observed in sensitive DV90 and H1373 cells but not in resistant A549 and H460 cells, irrespective of the p53 status of the cells (Figure 20). Interestingly, in sensitive cell lines, the induction of apoptosis was stronger in DV90 compared to H1373 cells, suggesting that the predominant response in DV90 cells is induction of apoptosis, while in H1373 it is primarily cytostatic. It is noteworthy that both DV90 and H1373 cells were quite sensitive towards detachment from the culture surface and re-suspension, a procedure needed for flow cytometer analysis, which explains the higher apoptotic rates of DMSO control samples.

Figure 20: Analysis of apoptosis in two sensitive cell lines H1373, DV90 and two insensitive cell lines A549 and H460 after BET inhibition. Flow cytometry results showing percentage of viable and apoptotic cells of H1373 and H460 cells after JQ1 or OTX-015 treatment for 48 h followed by AV-FITC and PI staining. AV- positive only population (red: early apoptotic), AV/PI- double positive population (green: late apoptotic), AV/PI negative population (black: viable) and PI positive only (yellow: necrotic).

4.1.3.2 Functional analysis of BET inhibition in the cellular context

BET inhibitors are competitive binders of the bromodomain pockets and interfere with BET – chromatin interactions. In order to investigate the cellular function of BET inhibitors in removing BRD4 from chromatin of the cell, H1373 cells were treated with 1 μM of JQ1 24h post transfection with a pcDNA6.2 plasmid overexpressing N-EmGFP-BRD4. Expression of BRD4-GFP was visually confirmed using microscopy. The unbound BRD4 was washed out of the nucleus using a detergent buffer containing Triton X-100 and PIPES (3.2.12.2). Cells were then fixed using paraformaldehyde and counterstained with DAPI in PBS before visualization using microscopy. BRD4-GFP was localized only in the nucleus, as seen by the overlap of DAPI and GFP. In the DMSO-treated control, nuclei still containing BRD4-GFP were observed, while no BRD4-BRD4-GFP could be detected in the nuclei of JQ1-treated cells after the washout was performed (Figure 21).

Figure 21: Chromatin-unbound washout experiment of BRD4-GFP overexpression plasmid transfected H1373 cells. H1373 cells were transfected with N- terminal tagged BRD4 full-length overexpression plasmid 24 h prior treatment with 1 μM JQ1 and chromatin-unbound washout using Detergent washout buffer 1 before fixation using para-formaldehyde and DAPI counter staining (3.2.12.2). Fluorescence was visualized using microscopy and 500-fold magnification. White bar indicates 20 μm.

While most small molecules can penetrate into the cells by passing the cellular membrane without active transport, cancer cells have been described to export small molecules actively by transporters like P-glycoprotein or Multidrug-Resistance-Protein-1 (MDR-1) and thereby acquiring resistance to chemotherapy. To investigate whether the differential response of the different cell lines was linked to a reduced intracellular function of JQ1, endogenous BRD4

was washed out after treatment of sensitive H1373 cells and insensitive A549 cells. Similar to the BRD4-GFP fusion, the endogenous BRD4 was strongly removed by the washout detergent when cells were treated with 1 μM JQ1 2 h before the washout (3.2.12.2). To be noted that harsher washout conditions were needed (3.2.12.3). Potentially due to the stronger interaction of the native form compared to the BRD4-GFP fusion protein , even though it was taken care of adding GFP at the N terminal site in order to maintain the interactions of the extra-terminal domain and the C terminal motif (Figure 6). JQ1 was able to reduce BRD4 chromatin interaction in sensitive H1373 cells and insensitive A549 cells (Figure 22).

Figure 22: Washout experiments of chromatin-unbound endogenous BRD4 after JQ1 treatment of sensitive H1373 cells and insensitive A549 cells. H1373 cells were treated with 1 μM JQ1 for 2 h and chromatin-unbound washout using Detergent washout buffer 2 before fixation using Methanol and stained with primary BRD4 antibody, alexa 488-conjugated tubulin antibody and DAPI (3.2.12.3). Fluorescence was visualized using microscopy and 630-fold magnification. White bar indicates 20 μm.

To confirm this finding, a chromatin separation assay was used to separate the chromatin-bound fraction from the chromatin-unchromatin-bound and cytosolic fractions. Sensitive DV90 and H1373 cells and insensitive H2030 cells were treated with varying doses of JQ1 (0.01 μM – 5 μM) for 24 h followed by complete lysis of the cells using M-PER buffer containing protease inhibitor and centrifugation (3.2.12.1). Chromatin free lysates were analyzed using western

blot staining for BRD4 and GAPDH loading control. In all 3 cell lines JQ1 treatment led an increase of chromatin-unbound BRD4 in a dose-dependent manner. In DV90 cells a low level of BRD4 was chromatin-unbound even in the DMSO control, but increasing doses of JQ1 led to an even higher unbound fraction (Figure 23). The effective removal of BRD4 from the chromatin by JQ1 was seen in all 3 cell lines and was independent of the anti-proliferative activity of the compound.

Figure 23: Western blot analysis of chromatin-unbound BRD4 after JQ1 treatment of sensitive DV90, H1373 cells and insensitive H2030 cells. For the separation of chromatin-bound and chromatin-unbound/cytoplasmic fraction DV90, H1373 and H2030 cells were treated with varying doses of JQ1 24 h before lysis using M-PER-Buffer containing protease inhibitor. Equal amounts of lysate were analyzed using western blot.

It is therefore assumed that the differential cellular response to BET inhibition might depend on gene regulatory changes defined by the epigenetic state of the cells. To follow up on this the effect of BET inhibition on gene expression was examined in a genome-wide transcription study of the sensitive DV90 cells.