Figure 6

Effect of G34-substituted histone H3.3 on CGI hypermethylation. (a) Overlap between CGIs in G34-hyper-DMRs and those in KNS-42-hyper-DMRs. p value = 2.4 × 10⁻¹⁵; odds ratio = 8.7 (Fisher’s exact test). (b) Effect of G34V-allele disruption on G34-specific methylomic signature in KNS-42 cells. ICA including the two disruptants identified two IC’s to be G34-specific, one from ICA using CGIs and the other from ICA using promoters (Figure S6d). For each sample, weightings to the former and latter ICs are plotted to horizontal and vertical axes, respectively. Note that both disruptants depart from their parental KNS-42 cells toward the other 13 samples without G34 mutations, or in the opposite direction to four G34R gliomas. (c) Methylation levels of CGIs with differential histone H3.3 deposition patterns in KNS-42 cells. CGIs are classified into four groups based on the presence (+) or absence (−) of overlapping H3p3 and G34V peaks in KNS-42 cells. Each bar indicates mean methylation level of each CGI group. Length indicates the total length of genomic regions exhibiting the indicated ChIP-seq peak patterns. (d) Normalized methylation levels of the four CGI groups in clinical glioma samples. CGIs are classified as in c, and their methylation levels in glioma subgroups are displayed by box plots. Methylation levels are normalized to the genome-wide mean of individual tumor samples. G34, N = 4; K27M, N = 13; WT, N = 8. (e) Effect of G34V-allele disruption on G34-hyper-DMRs. Ladder charts are shown for methylation levels of CGIs in G34-hyper-DMRs in KNS-42 and each disruptant. CGIs in G34-hyper-DMRs are divided into those overlapping with G34V peaks (N = 92) and those without them (N = 334). Red lines indicate the median methylation change.