American Society of Hematology

Figure 6

$Figure 6. ChIP-Seq reveals UNC1999-induced loss of H3K27me3 and concurrent gain of H3K27ac in MLL-AF9–transformed leukemia progenitors. (A) Summary of H3K27me3 peaks showing loss in the UNC1999- (red) vs mock-treated (blue) samples. X-axis indicates all H3K27me3 peaks (left) or those associated with promoters, enhancers, or intergenic regions. (B) The fractions of H3K27me3 peaks showing reduction in ChIP-Seq signals by 1.5-fold or more in UNC1999-treated (red circle) or UNC2400-treated (cross) samples in comparison with mock treatment. The H3K27me3 peaks and their densities shown on x-axis were first defined and then grouped by the number of ChIP-Seq reads identified in the mock-treated sample; y-axis represents the fraction in each group of H3K27me3 peaks that show reduction by >1.5-fold in the compound- vs mock-treated samples, after normalization of ChIP-Seq reads to the sequencing depths and peak sizes. (C) ChIP-qPCR detects H3K27me3 at the TSS of several Hox-A genes in MLL-AF9–transformed leukemia progenitors after treatment with 3 μM UNC2400 or UNC1999 for 4 days. ChIP signals (y-axis) were normalized to 5% of input and presented as mean ± SD. TSS of β-actin was used as negative control. *P < .05; **P < .01; ***P < .001. (D) Plot showing a global reduction in the H3K27me3 peak sizes following UNC1999 treatment (red). X-axis shows the ratios (in their Log2 values) of peak sizes following UNC1999 (red) or UNC2400 (black) treatment in comparison with mock; y-axis shows the relative fraction of peaks at each individual ratio. The dashed vertical lines mark the mean value of peak size ratios. (E) IGB view showing the distribution of input (black), H3K27me3 (red) and H3K27ac (blue) ChIP-Seq read densities (normalized by the ChIP-seq read depths) at the Smoothened (Smo) gene in MLL-AF9 leukemia progenitors after treatment with 3 μM UNC2400 or UNC1999 for 4 days. (F) Boxplots showing a significantly greater reduction of ChIP-Seq enrichment at the non-CpG- than the CpG-contained promoter associated H3K27me3 peaks after UNC1999 treatment in comparison with mock treatment. (G) Plot showing the relative size of SUZ12 peaks after compound treatments. X-axis shows the ratios (in their Log2 values) of peak sizes following UNC1999 (red) or UNC2400 (black) treatment in comparison with mock; y-axis shows the relative fraction of peaks at each individual ratio. The dashed vertical lines mark the mean value of peak size ratios. (H) Heatmap showing the ChIP-Seq read densities of H3K27me3 (red), H3K27ac (blue), and SUZ12 (brown) across the TSS (±20 kb) of upregulated genes following UNC1999 vs mock treatment (Figure 4A). Color represents the degree of ChIP-Seq signal enrichment, with the lowest set to white. The data indicate that a large majority (top of the heatmaps) of the UNC1999-derepressed genes contains H3K27me3 across TSS prior to compound treatment, and following UNC1999 treatment, H3K27me3 peaks become narrower and sharper. (I) IGB profiles showing the distribution of ChIP-Seq read densities (normalized by the ChIP-seq read depths) for input (black), H3K27me3 (red), and H3K27ac (blue) at p16Ink4a and p19Arf. Black bars under the track of H3K27me3 (UNC1999) mark the regulatory regions showing loss or reduction of H3K27me3 after UNC1999 treatment in comparison with mock or UNC2400. (J-K) ChIP-qPCR of H3K27me3 (J) and H3K27ac (K) across the Cdkn2a locus in MLL-AF9–transformed leukemia progenitors after treatment with 3 μM UNC2400 (blue) or UNC1999 (red) for 4 days. The genomic organization of p16Ink4a and p19Arf and positions of each ChIP PCR amplicon (labeled alphabetically, not drawn to scale) are depicted at the bottom of panel I. ChIP signals (y-axis) from independent experiments were normalized to input and presented as mean ± SD. *P < .05; **P < .01; ***P < .001. IGB, Integrated Genome Browser.$

ChIP-Seq reveals UNC1999-induced loss of H3K27me3 and concurrent gain of H3K27ac in MLL-AF9–transformed leukemia progenitors. (A) Summary of H3K27me3 peaks showing loss in the UNC1999- (red) vs mock-treated (blue) samples. X-axis indicates all H3K27me3 peaks (left) or those associated with promoters, enhancers, or intergenic regions. (B) The fractions of H3K27me3 peaks showing reduction in ChIP-Seq signals by 1.5-fold or more in UNC1999-treated (red circle) or UNC2400-treated (cross) samples in comparison with mock treatment. The H3K27me3 peaks and their densities shown on x-axis were first defined and then grouped by the number of ChIP-Seq reads identified in the mock-treated sample; y-axis represents the fraction in each group of H3K27me3 peaks that show reduction by >1.5-fold in the compound- vs mock-treated samples, after normalization of ChIP-Seq reads to the sequencing depths and peak sizes. (C) ChIP-qPCR detects H3K27me3 at the TSS of several Hox-A genes in MLL-AF9–transformed leukemia progenitors after treatment with 3 μM UNC2400 or UNC1999 for 4 days. ChIP signals (y-axis) were normalized to 5% of input and presented as mean ± SD. TSS of β-actin was used as negative control. *P < .05; **P < .01; ***P < .001. (D) Plot showing a global reduction in the H3K27me3 peak sizes following UNC1999 treatment (red). X-axis shows the ratios (in their Log2 values) of peak sizes following UNC1999 (red) or UNC2400 (black) treatment in comparison with mock; y-axis shows the relative fraction of peaks at each individual ratio. The dashed vertical lines mark the mean value of peak size ratios. (E) IGB view showing the distribution of input (black), H3K27me3 (red) and H3K27ac (blue) ChIP-Seq read densities (normalized by the ChIP-seq read depths) at the Smoothened (Smo) gene in MLL-AF9 leukemia progenitors after treatment with 3 μM UNC2400 or UNC1999 for 4 days. (F) Boxplots showing a significantly greater reduction of ChIP-Seq enrichment at the non-CpG- than the CpG-contained promoter associated H3K27me3 peaks after UNC1999 treatment in comparison with mock treatment. (G) Plot showing the relative size of SUZ12 peaks after compound treatments. X-axis shows the ratios (in their Log2 values) of peak sizes following UNC1999 (red) or UNC2400 (black) treatment in comparison with mock; y-axis shows the relative fraction of peaks at each individual ratio. The dashed vertical lines mark the mean value of peak size ratios. (H) Heatmap showing the ChIP-Seq read densities of H3K27me3 (red), H3K27ac (blue), and SUZ12 (brown) across the TSS (±20 kb) of upregulated genes following UNC1999 vs mock treatment (Figure 4A). Color represents the degree of ChIP-Seq signal enrichment, with the lowest set to white. The data indicate that a large majority (top of the heatmaps) of the UNC1999-derepressed genes contains H3K27me3 across TSS prior to compound treatment, and following UNC1999 treatment, H3K27me3 peaks become narrower and sharper. (I) IGB profiles showing the distribution of ChIP-Seq read densities (normalized by the ChIP-seq read depths) for input (black), H3K27me3 (red), and H3K27ac (blue) at p16Ink4a and p19Arf. Black bars under the track of H3K27me3 (UNC1999) mark the regulatory regions showing loss or reduction of H3K27me3 after UNC1999 treatment in comparison with mock or UNC2400. (J-K) ChIP-qPCR of H3K27me3 (J) and H3K27ac (K) across the Cdkn2a locus in MLL-AF9–transformed leukemia progenitors after treatment with 3 μM UNC2400 (blue) or UNC1999 (red) for 4 days. The genomic organization of p16Ink4a and p19Arf and positions of each ChIP PCR amplicon (labeled alphabetically, not drawn to scale) are depicted at the bottom of panel I. ChIP signals (y-axis) from independent experiments were normalized to input and presented as mean ± SD. *P < .05; **P < .01; ***P < .001. IGB, Integrated Genome Browser.

This Feature Is Available To Subscribers Only