Summary
In vertebrates, genes directing cell-type specific programs and embryonic development are frequently
controlled by cis-acting enhancers located several hundreds of kilobases away from their target promoters.
Disruptions or misregulation of long-distance enhancer-promoter (EP) interactions have been implicated in a
growing number of human pathologies, including developmental abnormalities and cancers. Accordingly, the
mechanisms that regulate EP interaction specificity and quantitative efficiency have a central role in establishing
gene expression patterns: they constitute a new yet largely unexplored level of regulation of genome activities.
The folding of the genome in different 3D structures has been proposed to help organize these distant
regulatory relationships. Notably, the structural partition of vertebrate genomes in distinct 3D domains (TADs)
has been shown to play an important role in defining the effective range of enhancer action. However, within
TADs, enhancers distribute their activity in complex patterns that are not explained by the current models.
Furthermore, because of the lack of proper tools to accurately assess the function of genomic architectural
elements, we barely know anything about the genetic sequences that determine their activity. Thus, despite
significant progress, the specific elements and features that organize and regulate within TADs the functional
interactions that link distant regulatory elements to target genes remain mostly unknown. Consequently, we are
too often unable to predict the consequences of non-coding and structural variants present in human genomes.
To address these critical gaps, we will develop a set of complementary research initiatives combining state-
of-art large-scale genome engineering in human cells and mouse models, innovative genomic screens and
quantitative genomics and transcriptomics analyses to 1) identify the genetic instructions (e.g. insulators,
tethering systems, relay elements) and processes (e.g. looping, non-coding transcription) that fold a linear
genomic locus into a dynamic 3D regulatory ensemble and determine the specificity and efficiency of long-
distance EP interactions 2) determine the regulatory grammar governing the activity of these different
architectural elements 3) measure quantitatively how genetic variants found in these elements may influence
their activity and contribute to phenotypic traits 4) integrate these different data to predict the impact of structural
variants better and in the future design synthetic approaches to regulate EP interactions. Altogether, these
complementary approaches will shed much-needed light on this additional yet poorly characterized layer of
regulation of gene expression and provide unique insights to understand better how non-coding genomic variants
may influence gene expression and lead to human pathologies.