Fig. 3

TFs controlling gastrulation and organogenesis promote the cis-regulatory activity of evolutionarily young TE subfamilies activated during pluripotency. A TE subfamily cis-regulatory activities (color: activity coefficients; area: statistical significance) estimated from dox-induced transgene overexpression experiments at 48 h in primed hESCs [47] using N computed with L = 250 kb. The number of replicates for each condition varies. Experiments were clustered using complete linkage hierarchical clustering on Euclidean distances computed from activity coefficients. Selected TE subfamilies were ordered by evolutionary age in millions of years, as previously estimated [5]. The number of protein-coding genes with total cis-regulatory weights > 0.13 (weight obtained at a distance of 2L, see Fig. S2) is shown for each subfamily. The color labeling of the estimated activities was saturated at \(|\Delta A| < 0.1\). B Top binding enrichment at selected evolutionarily young TEs by selected TFs controlling germ layer development. TEs (rows) were ordered as in A. ChIP-seq experiments (columns) were ordered by developmental stage or germ layer lineage. Color: number of peaks overlapping with subfamily-specific integrants, normalized for subfamily size. Area: statistical significance. Left: ChIP-seq peaks obtained from the ChIP-Atlas [62]. Right: ChIP-seq peaks obtained from [86] and [70]. C Top: Estimated differences in TE-dependent cis-regulatory activities between hESC-derived EpCAM\(^+\)/INTEGRIN\(\alpha\)6\(^+\) double positive (DP) hPGCLCs and double negative (DN) somatic cells at day 6 of differentiation, replicate #1 \(n=2\) [86]. Bottom: SOX15 KO DP hPGCLCs vs. DP hPGCLCs, day 6, \(n=2\). D Left: hESC-derived differentiating endoderm, 48 h vs 24 h, \(n=3\) [70]. Middle: iPSC-derived GATA6 KO mesendoderm vs. iPSC-derived mesendoderm, \(n=2\) [69]. Right: GATA6 rescue in iPSC-derived GATA6 KO mesendoderm vs. iPSC-derived GATA6 KO mesendoderm, \(n=2\)