Synergistic re-activation of epigenetically silenced genes by combinatorial inhibition of DNMTs and LSD1 in cancer cells

Abstract

Epigenetic gene silencing, mediated by aberrant promoter DNA hypermethylation and repressive histone modifications, is a hallmark of cancer. Although heritable, the dynamic nature and potential reversibility through pharmacological interventions make such aberrations attractive targets. Since cancers contain multiple epigenetic abnormalities, combining therapies that target different defects could potentially enhance their individual efficacies. 5-Aza-2'-deoxycytidine (5-Aza-CdR), FDA-approved drug for the treatment of myelodysplastic syndrome, can inhibit DNA methyltransferases (DNMTs) upon incorporation into the DNA of dividing cells, resulting in global demethylation. More recently, the first histone demethylase, lysine specific demethylase 1 (LSD1), which demethylates both histone and non-histone substrates, has become a new target for epigenetic therapy. Using, clorgyline, an LSD1 inhibitor (LSD1i) to treat cancer cell lines, we show that clorgyline employs two mechanisms of action depending on the cell type: it can either induce global DNA demethylation or inhibit LSD1-driven H3K4me2 and H3K4me1 demethylation to establish an active chromatin configuration. We also investigate the therapeutic efficacy of combining 5-Aza-CdR with clorgyline and determine that this combinatorial treatment has synergistic effects on reactivating aberrantly silenced genes by enriching H3K4me2 and H3K4me1. Many of the reactivated genes are categorized as cancer testis antigens or belong to the interferon-signaling pathway, suggesting potential implications for immunotherapy. Together, our results demonstrate that combinatorial treatment consisting of a DNMT inhibitor (DNMTi) and an LSD1i have enhanced therapeutic values and could improve the efficacy of epigenetic therapy.

Figure 1. Combinatorial treatment shows the greatest efficacy in inhibiting cell growth and limiting colony formation.

A. The expression levels of LSD1 in bladder cancers, T-cell acute lymphoblastic leukemia, colon adenoma and their normal counterparts were obtained from ONCOMINE. Numbers in the parentheses indicate number of samples used to generate box plots. B-D. Population doubling times for control (black), clorgyline treated (yellow), 5-Aza-CdR treated (red) and combinatorial treatment treated (green) in T24, HL60 and HCT116 cells. The corresponding numbers of hours are listed in the table below each graph. E. Quantification of number of colonies produced by colony formation assays in T24 and HCT116 cells after indicated treatments.

Figure 2. Combinatorial treatment elicits a synergistic effect in up-regulating genes in T24 cells

. A. Gene expression log2 difference is plotted on the x-axis, and the –log10 (p-value) is plotted on the y-axis. Probes that are identified as significantly different between two groups are colored in red. B. Venn intersects of the number of genes that are up-regulated by the indicated treatment. C. Network enrichment diagram containing genes that are synergistically up-regulated upon combinatorial treatment. Synergistically upregulated genes are marked by red bars. The other symbols used are as seen on the Metacore website.

Figure 3. Combinatorial treatment elicits a synergistic effect in up-regulating genes in HL60 cells.

A. Gene expression log2 difference is plotted on the x-axis, and the –log10 (p-value) is plotted on the y-axis. Probes that are identified as significantly different between two groups are colored in red. B. Venn intersects of the number of genes that are up-regulated upon the indicated treatment. C. Expression status obtained from ONCOMINE of 20 genes, whose expression was synergistically reactivated in HL60 cells, in peripheral blood mononuclear and chronic lymphocytic leukemia.

Figure 4. Gene up-regulation induced by combinatorial treatment is due to histone modifications in T24 cells.

A. Density plots for each treatment across all 450,000 CpG sites. The x-axis represents beta values ranging from 0 (not methylated) to 1 (highly methylated). B. Heatmap of CpG probes belonging to genes which are synergistically up-regulated upon combinatorial treatment. The level of DNA methylation for each probe in each sample is represented by using the color scale shown in the legend. C. ChIP results of histone modifications after normalization to input. Error bars represent standard deviation from 3 independent experiments.

Figure 5. Clorgyline induces demethylation in HCT116 cells.

A. Venn intersects of the genes that are up-regulated upon the indicated treatment. B. Density plots for each treatment across all 450,000 CpG sites. The x-axis represents beta values ranging from 0 (not methylated) to 1 (highly methylated). C. Venn intersect of the probes that are demethylated by the indicated treatment. The overlap of two circles represents the common probes that are demethylated by both treatments.