Sulfopin inhibits MYC signaling and reduces viability of DMG cells in a H3-K27M-dependent manner

(A) Functional enrichment analysis on significantly downregulated genes in SU-DIPG13 cells treated with 10uM Sulfopin for 12 hours, compared to DMSO. Enrichr algorithm123 was used to compare downregulated genes against the Molecular Signatures Database (MSigDB) hallmark geneset119. Dashed line denotes adjusted p-value = 0.05. MYC targets are significantly enriched among the Sulfopin downregulated genes.

(B) RT-qPCR analysis of selected MYC target genes in SU-DIPG13 cells treated with 10uM Sulfopin for 12 hours, compared to DMSO. Fold change between Sulfopin and DMSO treated cells was calculated and the mean ± SD of two technical repeats is shown.

(C) H3K27me3 Cut&Run read coverage over MYC target genes (‘MYC Targets V1’ hallmark geneset119, n=200), in SU-DIPG13 cells treated with 10uM Sulfopin for 8 days compared to DMSO. Sulfopin treatment increases H3K27me3 levels on the TSS of MYC target genes.

(D) Functional enrichment analysis of the genes associated with Sulfopin-unique H3K27me3 peaks, in SU-DIPG13 cells treated as in C. Dashed line denotes adjusted p-value = 0.05. MYC target geneset (‘MYC Targets V1’ hallmark geneset119) is mildly enriched among these genes, with adjusted p-value of 0.077.

(E) Cell viability, as measured by CellTiterGlo, of eight DMG cultures (H3.3K27M: SU-DIPG13, SU-DIPG6, SU-DIPG17, SU-DIPG25 and SU-DIPG50. H3.1K27M: SU-DIPG36, SU-DIPG38 and SU-DIPG21), treated with Sulfopin for eight days with pulse at day four, compared to DMSO. Mean±SD of two technical replicates is shown. Logarithmic scale is used for the x-axis. Sulfopin treatment led to a mild reduction in cell viability in all H3-K27M glioma cultures.

(F) Cell viability, as measured by CellTiterGlo, of two isogenic DMG cell lines (SU-DIPG13 and BT245) in which the mutant histone was knocked-out (KO), treated with the indicated concentration of Sulfopin for 8 days, compared to DMSO. For each cell line and concentration, the fold change in viability between Sulfopin and DMSO treated cells is shown. For SU-DIPG13-mean ± SE of at least two independent experiments is shown. For BT245-mean± SD of three technical replicates is shown. H3-K27M glioma cells show higher sensitivity to Sulfopin treatment compared to the KO cells. *P < 0.05; **P < 0.01 (two-sample t-test over all technical replicates). Significance adjusted after Bonferroni correction.

Combination of Sulfopin and Vorinostat elicits robust downregulation of oncogenic pathways

(A) Timeline demonstrating the treatment protocol for the combination of Sulfopin and Vorinostat.

(B) Percentage of cell viability, as measured by CellTiterGlo of SU-DIPG13 cells treated with Sulfopin and Vorinostat at the indicated concentrations, compared to DMSO.

(C) BLISS index measured as the ratio between the observed and the expected effect of the combination of Sulfopin and Vorinostat, for each pair of concentrations, in SU-DIPG13.Synergy: Bliss <1, Additive: Bliss=1, Antagonist: Bliss>1.

(D) Cell viability as measured by CellTiterGlo, of eight DMG cultures treated with Sulfopin (0uM, 2.5uM, 5uM, 10uM, 20uM and 40uM) and Vorinostat (1uM), compared to DMSO. H3.3-K27M and H3.1-K27M cultures are indicated in blue and orange, respectively. Mean±SD of two technical replicates is shown. H3.3-K27M cells showed higher sensitivity to the combined treatment compared to H3.1-K27M cells.

(E) The BLISS index of the combination of Sulfopin (10uM) and Vorinostat (1uM), in the indicated cultures. An additive effect was detected in all the H3.3-K27M cultures at this set of concentrations.

(F) Pearson correlation coefficient matrix of BLISS index of the combined treatment (Sulfopin (10uM) and Vorinostat (1uM)) and mRNA levels of MYC and its target genes, in the eight DMG cultures tested. mRNA levels were measured by RT-qPCR (Fig. S2F-G). Blue and yellow colors indicate negative or positive correlation, respectively. Negative correlation was detected between the BLISS indexes and the expression levels of MYC and its target genes.

(G) Unsupervised hierarchical clustering of expression levels of 620 significantly DE genes detected in SU-DIPG13 cells treated with either Sulfopin (10uM, 8 days), Vorinostat (1uM, 72 hours), the combination of Sulfopin and Vorinostat or DMSO. Gene expression rld values (log2 transformed and normalized) were standardized for each gene (row) across all samples. Color intensity corresponds to the standardized expression, low (blue) to high (red). Clusters 1 and 4 demonstrate additive transcriptional patterns associated with the combined treatment.

(H) Top: Gene Set Enrichment Analysis (GSEA) on SU-DIPG13 treated with combination of 10uM Sulfopin and 1uM Vorinostat compared to DMSO, showing significant downregulation of mTORC1 signaling (‘HALLMARK_MTORC1_SIGNALING’ geneset119) in the combined treatment. NES: Normalized Enrichment Score. FDR: false discovery rate. Bottom: Expression levels of significantly DE genes detected in the combined treatment compared to DMSO that are part of the mTORC1 signaling geneset. MTOR gene was added manually to the heat-map. Heatmaps were generated as described in B.

(I) Top: Gene Set Enrichment Analysis (GSEA) on SU-DIPG13 treated as in C, showing significant downregulation of the epigenetic BMI-1 pathway and the oncogenic cAMP pathway in the combined treatment (BMI1_DN.V1_UP; CAMP_UP.V1_UP; MSigDB C6 oncogenic signature120,121). Bottom: Expression levels of significantly DE genes detected in the combined treatment compared to DMSO that are part of the BMI-1 and cAMP genesets. Heatmaps were generated as in described B.

(J) Western blot of SU-DIPG13 treated either with Sulfopin (10uM, 8 days), Vorinostat (1uM, 72 hours), the combination of Sulfopin and Vorinostat or DMSO, using the indicated antibodies. β-tubulin is used as loading control.

The combined treatment attenuates H3K27ac levels on oncogenic targets

(A) Scheme of the single-molecule imaging experimental setup133: cell-derived mono-nucleosomes are anchored in a spatially distributed manner on polyethylene glycol (PEG)-coated surface. Captured nucleosomes are incubated with fluorescently labeled antibodies directed against the H3K27ac modification. Total internal reflection fluorescence (TIRF) microscopy is utilized to record the position and modification state of each nucleosome. Time series images are taken to allow detection of maximal binding events.

(B) Single-molecule imaging quantification of the percentage of H3K27ac nucleosomes, in SU-DIPG13 cells treated with either Sulfopin (10uM, 8 days), Vorinostat (1uM, 72 hours), or the combination of Sulfopin and Vorinostat, normalized to DMSO. Mean fold ± SE of at least two independent experiments is shown. H3K27ac global levels are lower in the combined treatment compared to cells treated solely with Vorinostat. *P < 0.05 (two sample t-test).

(C) SU-DIPG13 cells were treated as in B, and analyzed by western blot using the indicated antibodies.

(D) Left panel: Heatmap shows H3K27ac read coverage around the TSS (+/-5Kb) of the significantly DE genes shown in figure 2B, in SU-DIPG13 cells treated with the combination of 10uM Sulfopin and 1uM Vorinostat versus DMSO. Average coverage is shown on top. Color intensity corresponds to the standardized expression. Clusters 1-4 are indicated. Right panel: The log2 ratio of H3K27ac read coverage in SU-DIPG13 cells treated with the combination of 10uM Sulfopin and 1uM Vorinostat vs. DMSO was calculated. Heatmap shows the ratio around the TSS (+/-5Kb) of the significantly DE genes shown in figure 2B, and average coverage is shown on top. Color intensity corresponds to the ratio between samples, low (red) to high (blue). Clusters 1-4 are indicated, with cluster 1 presenting the strongest local decrease in H3K27ac following the combined treatment compared to DMSO.

(E) IGV tracks of MTOR and SLC7A5 gene promoters, showing H3K27ac coverage in SU-DIPG13 cells treated as indicated.

(F) Functional enrichment analysis of the genes linked to enhancers (top targets of high confident enhancers) marked with H3K27ac exclusively in SU-DIPG13 cells treated with Vorinostat, and not in the combined treatment. gProfiler algorithm126 was used to calculate enrichment against the KEGG pathways DB128. Dashed line denotes adjusted p-value = 0.05. Genes associated with Vorinostat-unique enhancers are enriched for oncogenic signaling pathways.

(G) IGV track of AKT3 and SEC13 linked enhancers, showing H3K27ac coverage in SU-DIPG13 cells treated with 1uM Vorinostat or the combination of 10uM Sulfopin and 1uM Vorinostat.

(H) Normalized expression levels of AKT3 and SEC13 genes in SU-DIPG13 cells treated as in G. Mean ± SD of three technical repeats is shown.*P < 0.05 (two-sample t-test).

The combined treatment with Sulfopin and Vorinostat reduces tumor growth in-vivo

(A-B) SU-DIPG13P* cells were injected to the pons of immunodeficient mice to form tumors. Ten days post injection, mice were treated for 18 days with either DMSO, Sulfopin, Vorinostat, or the combination of Sulfopin and Vorinostat. (A) In-vivo bioluminescent imaging of DMG xenografts following 18 days of treatment. The heat map superimposed over the mouse head represents the degree of photon emission by DMG cells expressing firefly luciferase. (B) DMG xenograft tumor growth as measured by change in bioluminescent photon emission following 15 days of treatment with either DMSO (n=8), Sulfopin (n=4), Vorinostat (n=5) or the combination of Sulfopin and Vorinostat (n=6). Data points represent the fold-change in maximum photon flux between day 3 and day 18 under treatment for each mouse. *P < 0.05 (two-tailed Mann-Whitney U-test).

(C-D) Immunofluorescent staining of brain sections from mice treated with DMSO (n=4) or the combination of Sulfopin and Vorinostat (n=4). (C) Representative fluorescence images of H3-K27M (red) and mTOR (green). (D) Percentage of mTOR positive cells out of the total H3-K27M-positive cells. H3-K27M positive cells show lower levels of mTOR following the combined treatment compared to DMSO. *P < 0.05 (two-tailed t-test).