Abstract

The folding of chromosomes and the structural organization of the genome impacts human health and disease. Long-range physical contacts between non-coding regulatory regions and their target genes regulate gene expression. In dividing cells, chromatin contacts are established when cells exit mitosis to dissolve upon re-entry into mitosis. We set out to investigate the functional relevance of chromatin contacts in terminally differentiated cells, such as neurons, which retain their physiology for years in living animals. We applied a novel ligation-free technique to map chromatin contacts genome-wide Genome Architecture Mapping (GAM) which is ideally suited to study rare cell types.

GAM extracts spatial information by sequencing the DNA content from a large collection of randomly orientated, thin nuclear sections, before quantifying the frequency of locus co-segregation. By applying GAM to mouse embryonic stem cells, we had previously identified specific chromatin contacts enriched for interactions between active genes and enhancers spanning large genomic distances. We have now develop faster and more affordable versions of GAM , which are compatible with their application in specific rare cell types without tissue disruption. As a proof-of principle, we have used GAM to map chromatin contacts genome-wide in specific neuronal subtypes directly from mouse brain. Our work shows that genome architecture is highly cell-type specific and reflects cell-type specific gene expression patterns at both short and long genomic distances.