Figure 7.
Functional antagonism between AP1, an early response complex, and Mi-2β in keratinocytes.(A) Venn diagram depicting the overlap of c-JUN peaks identified in CTR NHKs (blue) or Mi2KD NHKs (red) and JUNB peaks identified in CTR NHKs (green). (B) Heatmap of K-means clustering analysis of ChIP-seq data for JUNB in CTR, c-JUN in Mi2KD, Mi-2β in CTR and HaCaT, and H3K27Ac from CTR NHKs and Mi2KD NHKs, centered on de novo c-JUN peaks detected after Mi2KD. (C) Occupancy (read count per million mapped reads) of TFs (JUNB, JUN, and Mi-2β) and H3K27Ac at de novo c-JUN peaks (±2 kb) as described in B is shown by histogram format. (D) Pathway analyses of up-regulated genes associated with cJUN enrichment peaks in Mi2KD NHKs. Select GO pathways and associated P values (−log10) are shown. (E) Heatmap of K-means clustering analysis for a subset of significantly induced genes after Mi2KD (|log2 fold change| ≥1, FDR <0.05), which revert to normal levels after concomitant knockdown of c-JUN (|log2 fold change| <0, P < 0.05). See also Fig. S5 D for all significantly induced genes in Mi2KD. Normalized expression is shown in CTR, Mi2KD, c-JUN knockdown (c-JUNKD), and Mi2KD and c-JUNKD (DKD) NHKs. Data from two experiments each with two or three replicates in each are shown. Expt, experiment. (F) Among the genes de-repressed in Mi2KD NHKs (days 3–5), the frequency of genes bound by Mi-2β in CTR NHKs (blue) and bound by de novo c-JUN peaks in Mi2KD NHKs (pink stripes) and genes bound by both Mi-2β in CTR NHKs and c-JUN in Mi2KD NHKs (pink) are depicted as bar graphs. (G) Statistical evaluation of genes induced by loss of Mi-2β, positively regulated by c-JUN and bound by c-JUN. Induced genes that were reverted by concomitant knockdown of c-JUN were significantly associated (P = 0.001) with c-JUN peaks compared with induced genes that were not reverted. Data from χ2 analysis (two-sided) are shown. DF, degrees of freedom. (H) Statistical evaluation of association of genes induced by loss of Mi-2β with c-JUN and Mi-2β binding. Induced genes that were bound by c-JUN were also significantly enriched for Mi-2β binding (P value 0.0001) compared with induced genes targets that were not bound by c-JUN. Data from χ2 analysis (two-sided) are shown. (I) Genome browser tracks of normalized ChIP-seq reads for TFs (JUNB, JUN, Mi-2β, and RNApII) and histone modifications (H3K27Ac and H3K4me3) generated from NHKs (CTR, black histogram), Mi2KD NHKs (red histogram) and HaCaT cells (black histogram) are shown at the COL7A1 locus. ChIP-seq for JUNB and JUN was performed on one set of CTR and Mi2KD primary NHKs. RNA-seq for CTR, Mi2KD, JUNKD, and DKD was performed twice with duplicates or triplicate samples in each experiment.

Functional antagonism between AP1, an early response complex, and Mi-2 β in keratinocytes. (A) Venn diagram depicting the overlap of c-JUN peaks identified in CTR NHKs (blue) or Mi2KD NHKs (red) and JUNB peaks identified in CTR NHKs (green). (B) Heatmap of K-means clustering analysis of ChIP-seq data for JUNB in CTR, c-JUN in Mi2KD, Mi-2β in CTR and HaCaT, and H3K27Ac from CTR NHKs and Mi2KD NHKs, centered on de novo c-JUN peaks detected after Mi2KD. (C) Occupancy (read count per million mapped reads) of TFs (JUNB, JUN, and Mi-2β) and H3K27Ac at de novo c-JUN peaks (±2 kb) as described in B is shown by histogram format. (D) Pathway analyses of up-regulated genes associated with cJUN enrichment peaks in Mi2KD NHKs. Select GO pathways and associated P values (−log10) are shown. (E) Heatmap of K-means clustering analysis for a subset of significantly induced genes after Mi2KD (|log2 fold change| ≥1, FDR <0.05), which revert to normal levels after concomitant knockdown of c-JUN (|log2 fold change| <0, P < 0.05). See also Fig. S5 D for all significantly induced genes in Mi2KD. Normalized expression is shown in CTR, Mi2KD, c-JUN knockdown (c-JUNKD), and Mi2KD and c-JUNKD (DKD) NHKs. Data from two experiments each with two or three replicates in each are shown. Expt, experiment. (F) Among the genes de-repressed in Mi2KD NHKs (days 3–5), the frequency of genes bound by Mi-2β in CTR NHKs (blue) and bound by de novo c-JUN peaks in Mi2KD NHKs (pink stripes) and genes bound by both Mi-2β in CTR NHKs and c-JUN in Mi2KD NHKs (pink) are depicted as bar graphs. (G) Statistical evaluation of genes induced by loss of Mi-2β, positively regulated by c-JUN and bound by c-JUN. Induced genes that were reverted by concomitant knockdown of c-JUN were significantly associated (P = 0.001) with c-JUN peaks compared with induced genes that were not reverted. Data from χ2 analysis (two-sided) are shown. DF, degrees of freedom. (H) Statistical evaluation of association of genes induced by loss of Mi-2β with c-JUN and Mi-2β binding. Induced genes that were bound by c-JUN were also significantly enriched for Mi-2β binding (P value 0.0001) compared with induced genes targets that were not bound by c-JUN. Data from χ2 analysis (two-sided) are shown. (I) Genome browser tracks of normalized ChIP-seq reads for TFs (JUNB, JUN, Mi-2β, and RNApII) and histone modifications (H3K27Ac and H3K4me3) generated from NHKs (CTR, black histogram), Mi2KD NHKs (red histogram) and HaCaT cells (black histogram) are shown at the COL7A1 locus. ChIP-seq for JUNB and JUN was performed on one set of CTR and Mi2KD primary NHKs. RNA-seq for CTR, Mi2KD, JUNKD, and DKD was performed twice with duplicates or triplicate samples in each experiment.

This Feature Is Available To Subscribers Only

or Create an Account

Close Modal
Close Modal