Fig. 6

Rewiring of promoter 3D interactions by RA signaling. (A) Differential analyses of H3K4me3 HiChIP loops between control and atRA treated embryos at 80epi stage (n = 2 biological replicates per condition) at 10-kb resolution. The log2 normalized counts per million (CPM) of control reads versus the log2 fold-change of contacts are plotted. Loops showing a statistically significant differential intensity (FDR < 0.05) are highlighted in red (increased) or blue (decreased). (B) Percentage of loops showing RARαa binding at least in one anchor for increased, stable and decreased loops. (C) Violing plots showing the expression fold-change in atRA treated embryos at 80epi of all DEGs and those associated with increased and decreased loops. (D) From top to bottom, heatmaps showing H3K4me3 HiChIP signal in control and atRA treated embryos, annotated genes, HiChIP loops increased by atRA treatment (FDR < 0.05), total HiChIP loops, tracks with ATAC-seq, RARαa ChIP- seq, and 4C-seq experiments at the indicated viewpoints (VP), in control and atRA treated embryos at 80epi stage, in a 400-kb region of chromosome 17 containing the upregulated gene meis2a. (E) Top, virtual 4C showing contact quantification calculated from H3K4me3 HiChIP data using as viewpoints 10-kb bins containing the gene promoters in the HoxBa cluster, both in control and atRA treated embryos at 80epi stage. A green vertical line indicates the bin used as viewpoint in each case. Bottom, annotated genes and tracks with ATAC-seq, RARαa ChIP- seq, and 4C-seq experiments at the indicated viewpoints (VP), in control and atRA treated embryos at 80epi stage, in a 320-kb region of chromosome 3 containing the HoxBa gene cluster. Boxplots in C show: center line, median; box limits, upper and lower quartiles; whiskers, 1.5× interquartile range; notches, 95% confidence interval of the median. Statistical significance was assessed using a two-sided Wilcoxon's rank sum test in C, and with a two-sided Fisher's exact test in B.