Abstract
CTA induction but not downstream interferon-downstream stimulation by DAC is correlated with CpG demethylation in glioma cells. A, Heatmap and hierarchical clustering of differentially expressed CTA. Scale represents z-score of DESeq2 counts. B, Violin plot of beta value distribution between treatment conditions for CpG probes corresponding to differentially upregulated CTA (P-value = 0.028; Wilcoxon rank sum test with continuity correction). C, Scatter plot of all CpG island probes containing intersectional DNA methylation log2FoldChange and RNA expression log2FoldChange of annotated genes in black (n = 115,801 CpG island probes). CpG island probes corresponding to CTA with Pearson correlation are highlighted in (red; n = 535 CpG probes; n = 65 CTA genes). D, Enricher analysis of DESEQ2 differentially upregulated genes (Hallmark IFNγ P = 0.00548; Hallmark IFNα P = 0.00548; Hallmark inflammatory response P = 0.0426) and downregulated genes (Hallmark mTORC1 signaling P = 1.59e−3; Hallmark mitotic spindle P = 1.59e−3, Hallmark G2M Checkpoint P = 7.77e−5; Hallmark E2F targets P = 7.77e−5). Hallmark MSigDB signature adjusted P values determined using Fisher exact test adjusted by Benjamini–Hochberg correction. E, Violin plot of beta value distribution between treatment conditions for CpG probes corresponding to differentially upregulated DAC_UP_MSigDB_Hallmark genes (P-value = 0.497; Wilcoxon rank sum test with continuity correction). F, Scatter plot of intersectional DNA methylation log2FoldChange and RNA expression log2FoldChange CpG island probes. CpG probes corresponding to the Hallmark Interferon Alpha Response gene set with Pearson correlation are highlighted in blue (n = 471 probes; n = 53 genes).