BET inhibition silences expression of MYCN and BCL2 and induces cytotoxicity in neuroblastoma tumor models

Abstract

BET family proteins are epigenetic regulators known to control expression of genes involved in cell growth and oncogenesis. Selective inhibitors of BET proteins exhibit potent anti-proliferative activity in a number of hematologic cancer models, in part through suppression of the MYC oncogene and downstream Myc-driven pathways. However, little is currently known about the activity of BET inhibitors in solid tumor models, and whether down-regulation of MYC family genes contributes to sensitivity. Here we provide evidence for potent BET inhibitor activity in neuroblastoma, a pediatric solid tumor associated with a high frequency of MYCN amplifications. We treated a panel of neuroblastoma cell lines with a novel small molecule inhibitor of BET proteins, GSK1324726A (I-BET726), and observed potent growth inhibition and cytotoxicity in most cell lines irrespective of MYCN copy number or expression level. Gene expression analyses in neuroblastoma cell lines suggest a role of BET inhibition in apoptosis, signaling, and N-Myc-driven pathways, including the direct suppression of BCL2 and MYCN. Reversal of MYCN or BCL2 suppression reduces the potency of I-BET726-induced cytotoxicity in a cell line-specific manner; however, neither factor fully accounts for I-BET726 sensitivity. Oral administration of I-BET726 to mouse xenograft models of human neuroblastoma results in tumor growth inhibition and down-regulation MYCN and BCL2 expression, suggesting a potential role for these genes in tumor growth. Taken together, our data highlight the potential of BET inhibitors as novel therapeutics for neuroblastoma, and suggest that sensitivity is driven by pleiotropic effects on cell growth and apoptotic pathways in a context-specific manner.

Conflict of interest statement

Competing Interests: The authors have read the journal's competing interest policy and have the following conflicts to disclose: all authors are current or former employees of GlaxoSmithKline. This does not alter the authors' adherence to all the PLOS ONE policies on sharing data and materials.

Figures

Figure 1. I-BET726: a novel selective inhibitor of BET family proteins.

(a) Chemical structure of GSK1324726A (I-BET726). (b) Crystal structure of I-BET726 (magenta) bound to the acetyl-binding pocket of BRD4-BD1 (resolution: 1.6 Å). (c) Concentration response curves for determination of binding affinity of I-BET726 to BRD2, BRD3, and BRD4 bromodomains by ligand displacement detected using Time Resolved Fluorescence Resonance Energy Transfer (TR-FRET). IC50 values for BRD2, BRD3, and BRD4 are indicated. (d) Selectivity profile of I-BET726 showing average temperature shifts (delta Tm) in degrees Celsius for a panel of bromodomain proteins using a fluorescent thermal shift assay. N= 2 for all proteins except CREBBP (n= 4).

Figure 2. I-BET726 treatment results in potent growth inhibition and cytotoxicity in neuroblastoma cell lines.

(a) gIC50 values observed for I-BET726 in a panel of neuroblastoma cell lines obtained from a 6 day growth-death assay. (b) Concentration response curves for SK–N-AS and CHP-212 from 6 day growth-death assay. Black horizontal line indicates growth in DMSO-treated controls. Red line indicates T0 value (100%). gIC50 and Ymin-T0 values are indicated. Data presented as the average of two independent curves from a single experiment, and is representative of data from three independent biological replicates. (c) Ymin-T0 values observed for I-BET726 in the panel of neuroblastoma cell lines obtained from a 6 day growth-death assay. Gray bars indicate a cytotoxic response, defined by a Ymin-T0 value ≤ -50, with evidence of net cell death (Y < T0) along the growth curve at concentrations less than 6 µM. (d) Graph summarizing number of cells in G1 phase (as a percent of total cell population) in the indicated neuroblastoma cell lines following treatment with a titration of I-BET726 (5 nM-20 000 nM) for 2 days based on propidium iodide staining. V represents vehicle (DMSO) control sample. (e) Histograms generated from cell cycle analysis in the CHP-212 cell line following 4 days treatment with the indicated concentration of I-BET726. Percentage of cells in G1 phase and sub-G1 phase are indicated. (f) Caspase induction in the indicated neuroblastoma cell lines following treatment with a titration of I-BET726 for one, two, or three days. Data is presented as fold induction over DMSO controls, following normalization to total cell number as measured by CellTiter-Glo.

Figure 3. Global transcript profiling in neuroblastoma cell lines treated with I-BET726 reveals gene expression changes in apoptotic and signaling pathways.

(a) Hierarchical clustering of statistically significant probes that were differentially expressed in 100 nM or 1 µM treatments of I-BET726 relative to vehicle in CHP-212 and SK–N–SH. (b) Venn analysis for up-regulated and down-regulated probes described in (a) in the CHP-212 cell line. (c) Venn analysis for overlap of up-regulated and down-regulated probes in the SK–N–SH and CHP-212 cell lines. (d) Functional analyses of expression changes were performed by GO and canonical pathway enrichment at the gene level. A subset of statistically significant categories for GO Biological Process (>100 genes) and canonical pathways (>20 genes) from KEGG and BioCarta that were common among the two cell lines are shown. (e) qRT-PCR confirmation of a subset of genes selected from the functional analyses described in (d). Data represent mean value ± standard deviation for three independent biological replicates. Asterisks indicate statistical significance as measured by t-test (p <0.05).

Figure 4. Global transcript profiling reveals gene expression changes in MYC-family pathways.

(a) qRT-PCR confirmation of changes in MYCN and MYC expression following treatment with the indicated concentration of I-BET726. Data represent mean value ± standard deviation for three independent biological replicates. Asterisks indicate statistical significance as measured by t-test (p <0.05). (b) GSEA enrichment plots showing the down-regulation of gene sets associated with Myc/Max and N-Myc binding motifs in I-BET726-treated CHP-212 cells. Normalized enrichment scores (NES) and FDR q values are indicated. (c) A MYCN transcriptional regulation network was constructed (see Materials and Methods) to depict N-Myc pathway genes that were modulated by I-BET726 treatment. Red and blue circles represent increased and decreased expression changes, respectively. Green, red and grey edges are shown for activation, inhibition and unspecified interaction types, respectively. (d) qRT-PCR confirmation of a subset of genes selected from the N-Myc network analysis described in (c). Data presented as described in (a).

Figure 5. MYCN expression is directly regulated by BRD4 and repressed by treatment with I-BET726.

(a) Left: Concentration response curve for MYCN RNA expression following 24 hour treatment with I-BET726 in the CHP-212 cell line. Data was normalized to GAPDH and is presented as expression relative to DMSO-treated controls. Data presented as the average of two independent curves from a single experiment, and is representative of data from three independent biological replicates. Right: Table of IC50 and percent inhibition values for MYCN suppression following 24 hour treatment with I-BET726 in the indicated cell lines. (b) BRD4 ChIP in the non-MYCN-amplified cell line SK–N–SH. Binding of BRD4 to the MYCN promoter or to an intergenic region on Chromosome 12 following treatment with vehicle or 1µM I-BET726 for six hours. Data is presented as fold enrichment over signal generated from IgG control immunoprecipitations. Data shown is from a single experiment representative of typical results. (c) BRD4 ChIP data at the MYCN promoter, presented as percent of vehicle control signal in the non-MYCN-amplified cell line SK–N–SH (left) and the MYCN-amplified cell line CHP-212 (right). SK–N–SH data represents the mean value ± standard deviation for three independent biological replicates. Asterisk indicates statistical significance as measured by t-test (p= 0.005). CHP-212 data represents the mean value ± standard deviation for two independent biological replicates. (d) Western blot analysis of N-Myc expression in the MYCN-amplified cell lines CHP-212 and IMR32 following 24 or 48 hour treatment with vehicle or 1µM I-BET726. Actin expression included as a loading control. (e) gIC50 values obtained from CHP-212 cells overexpressing GFP or MYCN following treatment with I-BET726 in a 6 day growth-death assay. Data represents the mean value ± standard deviation from four independent experiments. (f) Concentration response curves for GFP or MYCN–overexpressing CHP-212 cells from a 6 day growth-death assay. Horizontal line indicates T0 measurement (normalized to 100%). Data shown was from a single experiment representative of typical results. (g) Left: Histograms generated from cell cycle analysis in GFP- or MYCN-overexpressing CHP-212 cells following 4 days treatment with 5 µM I-BET726. Right: Percentage of cells in subs G1, G1, S, and G2 phases from the cell cycle experiment.

Figure 6. Suppression of BCL2 expression by I-BET726.

(a) Left: Concentration response curve for BCL2 RNA expression following 24 hour treatment with I-BET726 in the CHP-212 cell line. Data was normalized to GAPDH and presented as expression relative to DMSO-treated controls. Data presented as the average of two independent curves from a single experiment, and is representative of data from two independent biological replicates. Right: Table of IC50 values and percent inhibition of BCL2 expression following 24 hour treatment with I-BET726. (b) BRD4 ChIP in the non-amplified neuroblastoma cell line SK–N–SH. Binding of BRD4 to the BCL2 promoter or to an intergenic region on Chromosome 12 following treatment with vehicle or 1 µM I-BET726 for six hours. Data is presented as fold enrichment over signal generated from IgG control immunoprecipitations. Data shown was from a single experiment representative of typical results. (c) BRD4 ChIP data at the BCL2 promoter, presented as percent of vehicle control signal. Data represent the mean value ± standard deviation for three independent biological replicates. Asterisk indicates statistical significance as measured by T-test (p= 0.002). (d) Western blot analysis of Bcl-2 expression in the MYCN-amplified cell lines CHP-212 and IMR32 following 48 hour treatment with vehicle or 1 µM I-BET726. Tubulin expression included as a loading control. (e) gIC50 values obtained from CHP-212 or LA-N-2 cells overexpressing GFP or BCL2 following treatment with I-BET726 in a 6 day growth-death assay. Data represents the mean value ± standard deviation from three independent experiments. (f) Ymin-T0 values for CHP-212 or LA-N-2 cells overexpressing GFP or BCL2. Data represents the mean value ± standard deviation from three independent experiments. Asterisk indicates statistical significance as measured by t-test (p= 0.02). (g) Western blot analysis of Bcl-2 expression in CHP-212 or LA-N-2 cells overexpressing GFP or BCL2 following 48 hour treatment with DMSO or 1 µM I-BET726. Actin expression included as a loading control.

Figure 7. Analysis of I-BET726 activity in vivo.

(a) Mean absolute body weight ± SD for mice in the SK–N-AS (left) and CHP-212 (right) xenograft studies treated with vehicle, 5 mg/kg, or 15 mg/kg I-BET726. (b) Mean absolute tumor volumes ± SEM for SK–N-AS subcutaneous xenografts following treatment with 5 mg/kg or 15 mg/kg I-BET726. Asterisks indicate statistical significance as measured by t-test (p <0.05). Tumor growth inhibition (TGI) for the 15 mg/kg group was 58% on day 14 (n= 9; p= 0.0060). (c) Mean absolute tumor volumes ± SEM for CHP-212 subcutaneous xenografts following treatment with 5 mg/kg or 15 mg/kg I-BET726. Asterisks indicate statistical significance as measured by T-test (p <0.05). TGI for 5 mg/kg was 50% on Day 42 (n= 8; p= 0.1816). TGI for 15 mg/kg was 82% on Day 42 (n=5; p =0.0488). (d) Pharmacodynamic analysis in CHP-212 and SK–N-AS xenografts 8 hours after initial dose of I-BET726. qRT-PCR analysis of MYCN, MYC, and BCL2 expression following I-BET726 treatment in the indicated models. Data is presented as fold induction compared to vehicle treated controls, and represents the average ± SD of data from three animals. (e) qRT-PCR analysis of apoptotic pathway and N-Myc pathway genes in CHP-212 xenografts 8 hours following treatment with I-BET726 on day 8 of study. Data is presented as described in (d).