Topologically associated domains are ancient features that coincide with Metazoan clusters of extreme noncoding conservation

Abstract

In vertebrates and other Metazoa, developmental genes are found surrounded by dense clusters of highly conserved noncoding elements (CNEs). CNEs exhibit extreme levels of sequence conservation of unexplained origin, with many acting as long-range enhancers during development. Clusters of CNEs, termed genomic regulatory blocks (GRBs), define the span of regulatory interactions for many important developmental regulators. The function and genomic distribution of these elements close to important regulatory genes raises the question of how they relate to the 3D conformation of these loci. We show that GRBs, defined using clusters of CNEs, coincide strongly with the patterns of topological organisation in metazoan genomes, predicting the boundaries of topologically associating domains (TADs) at hundreds of loci. The set of TADs that are associated with high levels of non-coding conservation exhibit distinct properties compared to TADs called in chromosomal regions devoid of extreme non-coding conservation. The correspondence between GRBs and TADs suggests that TADs around developmental genes are ancient, slowly evolving genomic structures, many of which have had conserved spans for hundreds of millions of years. This relationship also explains the difference in TAD numbers and sizes between genomes. While the close correspondence between extreme conservation and the boundaries of this subset of TADs does not reveal the mechanism leading to the conservation of these elements, it provides a functional framework for studying the role of TADs in long-range transcriptional regulation.