Developmental genes are highly regulated at the level of transcription and exhibit complex spatial and temporal expression patterns. Key developmental loci are frequently spanned by clusters of conserved non-coding elements (CNEs), referred to as genomic regulatory blocks (GRBs), that have been subject to extreme levels of purifying selection during metazoan evolution. CNEs have been shown to function as long-range enhancers, activating transcription of their developmental target genes over vast genomic distances and bypassing more proximally located unresponsive genes (bystanders). Despite their role in the establishment of cell identity during development, many of these long-range regulatory landscapes remain poorly characterised.
In this thesis, I develop a computational method for the genome-wide identification of regulatory enhancer-promoter associations in human and mouse, based on co-variation of enhancer and promoter transcriptional activity across a comprehensive set of tissues and cell types, in combination with chromatin contact data. Using this method, I demonstrate that previously predicted GRB target genes are amongst the genes with the highest level of enhancer responsiveness in the genome, and are frequently associated with extremely long-range
enhancers. Remarkably, the activity of some previously predicted bystanders is also weakly but significantly associated with enhancer activity, challenging the notion that the promoters of bystanders are unresponsive to enhancers. Next, I systematically annotate human genes with elevated enhancer responsiveness and identify more than 600 putative target genes, associated with the regulation of a wide range of developmental processes, from pattern specification to axonogenesis, as well as with disease. The analysis performed in this thesis has facilitated the identification of hundreds of previously uncharacterised enhancer-responsive genes and their long-range regulatory landscapes, allowing the study of their unique properties.