Massively parallel characterization of regulatory dynamics during neural induction

Anat Kreimer ^1,2,3 Fumitaka Inoue ^2,3 Tal Ashuach ¹ Nadav Ahituv ^2,3 Nir Yosef ¹

¹Department of Electrical Engineering and Computer Sciences and Center for Computational Biology, University of California Berkeley, USA
²Department of Bioengineering and Therapeutic Sciences,, University of California San Francisco, USA
³Institute for Human Genetics, University of California San Francisco, USA

The temporal interplay between gene regulation and gene expression during cell differentiation remains largely unknown. Using neural induction, we set out to explore these dynamics. We performed RNA-seq, ChIP-seq (H3K27ac, H3K27me3) and ATAC-seq on human embryonic stem cells at seven early neural differentiation time points (0-72 hours). We found that DNA accessibility precedes H3K27ac, which is followed by gene expression changes. Using massively parallel reporter assays (MPRAs) to test the activity of 2,464 candidate regulatory sequences at all seven time-points, we show that many of these sequences have a temporal pattern that correlates with their respective cell-endogenous gene expression and chromatin changes. Development of a prioritization method that incorporated all genomic and MPRA data identified key transcription factors involved in temporal function. Combined, our results provide a resource of genes and regulatory elements that orchestrate neural induction and illuminate the temporal framework that is needed to obtain this differentiation.