Figure 3. Enhancer transcription mediates target gene expression.

(A) Genomic loci showing circularized chromatin conformation capture (4C) from the viewpoint of four identified regulatory elements (near Atp2a2, Wrn, Ryr2, and Ubc). Total reads (top) and significant contacts (bottom, orange) are plotted. Annotated genes are shown in blue and red (sense and antisense, respectively). The viewpoint is with a dashed red line and the nearest TBX5-dependent target is labeled. (B) Sparse scatterplot for combined Pi-scores calculated from Tbx5-activated enhancers and Tbx5-activated genes. The X-axis is the pi-score using the ‘nearest’ Tbx5-dependent gene, and the Y-axis is the pi-score using all potential TF-dependent gene targets within a 2-Mbp window. The orange line indicates y = x. The dashed lines mark the standard deviations. Example enhancers were labeled with the names of their nearest Tbx5-dependent gene. Red indicates enhancers with TBX5 occupancy, and blue indicates enhancers without TBX5 occupancy as determined by ChIP-seq. (C) Scatterplot in hexagonal binning for the 3067 identified TF-dependent ncRNAs. The X-axis is the differential expression fold change on a log₂ scale of these ncRNAs, and the Y-axis is the differential expression fold change on log₂ scale of any nearest gene targets. The fold change with the maximum absolute value is used when both neighbor genes are TF-dependent.

Figure 3—figure supplement 1. A TBX5-dependent enhancer at Sln that was not identified by TF ChIP-seq.

(A) Genomic view of a Tbx5-activated ncRNA region adjacent to gene Sln. This locus shows significant Tbx5-dependence as determined by transcription expression in mouse atrium, open chromatin detected by ATAC-seq in HL1 cells, DHS signature in adult mice, but not by TBX5 occupancy or H3K27Ac in HL1 cells (GSE21529). (B) ChIP-qPCR of TBX5 comparing this Sln locus with the Gapdh locus. The Sln locus shows enrichment for TBX5 occupancy (n = 8, p=0.036). Two-sided t-test assuming unequal variances was performed.