Figure 6.
Figure 6. Effect of Chd7 knockout on Cbfb-MYH11–induced gene expression changes in preleukemic cells. (A-I) RNA-seq was performed on c-Kit+ bone marrow cells isolated from poly(I:C)-treated mice (n = 3 for each genotype). (A) Three-dimensional principal component analysis plots showing clear separations among these 4 genotype groups. (B) Relative expression levels of Csf1r in this data set (mean ± SEM). *P < .05; **P < .01. (C) Heatmap representation of unsupervised hierarchical clustering of DEGs between control and Chd7f/fMx1-Cre cells. Numbers on the right indicate DEGs in each of the 2 expression clusters. (D) Heatmap representation of unsupervised hierarchical clustering of DEGs between Mx1-CreCbfb+/56M and Chd7f/fMx1-CreCbfb+/56M cells. Overall, 4 main clusters were identified, with the percentage of DEGs within each cluster labeled to the right. (E) Venn diagrams representing the overlap between DEGs in panel D (after conversion to human equivalent genes; supplemental Table 3) and RUNX1 target genes in ME-1 cells.17,32 (F) GSEA identified the TORCHIA_TARGETS_OF_EWSR1_FLI1_FUSION_UP gene set as being negatively correlated with DEGs upregulated in Chd7f/fMx1-CreCbfb+/56M cells. FDR, false discovery rate; NES, normalized enrichment scores. (G) Heatmap representation of unsupervised hierarchical clustering of overlapped DEGs between control vs Mx1-CreCbfb+/56M and control vs Chd7f/fMx1-CreCbfb+/56M (supplemental Figure 7C). In general, 4 main clusters were identified, with the percentage of DEGs within each cluster labeled to the right. (H) Venn diagrams representing the overlapped DEGs shown in panel G (after conversion to human equivalent genes; supplemental Table 6) with RUNX1 target genes in ME-1 cells.17,32 (I) Canonical pathways and disease functions associated with overlapped DEGs shown in panel G, identified by IPA. DEGs: Padj < .05; absolute fold changes ≥2.

Effect of Chd7 knockout on Cbfb-MYH11–induced gene expression changes in preleukemic cells. (A-I) RNA-seq was performed on c-Kit+ bone marrow cells isolated from poly(I:C)-treated mice (n = 3 for each genotype). (A) Three-dimensional principal component analysis plots showing clear separations among these 4 genotype groups. (B) Relative expression levels of Csf1r in this data set (mean ± SEM). *P < .05; **P < .01. (C) Heatmap representation of unsupervised hierarchical clustering of DEGs between control and Chd7f/fMx1-Cre cells. Numbers on the right indicate DEGs in each of the 2 expression clusters. (D) Heatmap representation of unsupervised hierarchical clustering of DEGs between Mx1-CreCbfb+/56M and Chd7f/fMx1-CreCbfb+/56M cells. Overall, 4 main clusters were identified, with the percentage of DEGs within each cluster labeled to the right. (E) Venn diagrams representing the overlap between DEGs in panel D (after conversion to human equivalent genes; supplemental Table 3) and RUNX1 target genes in ME-1 cells.17,32  (F) GSEA identified the TORCHIA_TARGETS_OF_EWSR1_FLI1_FUSION_UP gene set as being negatively correlated with DEGs upregulated in Chd7f/fMx1-CreCbfb+/56M cells. FDR, false discovery rate; NES, normalized enrichment scores. (G) Heatmap representation of unsupervised hierarchical clustering of overlapped DEGs between control vs Mx1-CreCbfb+/56M and control vs Chd7f/fMx1-CreCbfb+/56M (supplemental Figure 7C). In general, 4 main clusters were identified, with the percentage of DEGs within each cluster labeled to the right. (H) Venn diagrams representing the overlapped DEGs shown in panel G (after conversion to human equivalent genes; supplemental Table 6) with RUNX1 target genes in ME-1 cells.17,32  (I) Canonical pathways and disease functions associated with overlapped DEGs shown in panel G, identified by IPA. DEGs: Padj < .05; absolute fold changes ≥2.

Close Modal

or Create an Account

Close Modal
Close Modal