3C maps consistently reveal chromosomal domain structure. Scaling up 3C experiments using large 5C libraries [ 16, 17, 18 and 19••] or combining 3C into open-ended protocols generated comprehensive 3C contact maps encompassing many megabases of chromosomal territories in yeast, Drosophila, Mouse and Human cells [ 6••, 7••, 8•• and 9]. The analysis of such maps first reconfirmed known physical properties of chromosomes, and then proposed selleck products significant genome wide generalization and higher resolution refinements of these properties. The maps confirmed a strong presence of chromosomal territories, clearly distinguishing contacts between elements in the same chromosome and contacts crossing chromosomal

boundaries. Chromosomes were then shown buy CB-839 to divide according to activity patterns, with chromosomal elements harboring actively transcribed genes tending to contact other such elements more often than regions lacking active genes [ 8•• and 20]. Going beyond these coarse grained models of chromosome structure, higher resolution analysis revealed novel modular structures that package genomic regions into domains with strong internal connectivity and limited external interactions. The resulting physical or topological domains ( Figure 1) create an attractive framework for modeling chromosome structure, simplifying (at

least theoretically) the problem into understanding how domains contact each other to form together higher order structures. In Drosophila, about 1000 domains sizing around 100KB each were described. In human and mouse, 2000–3000 domains were described, measuring around 1MB on average, suggesting a modular chromatin organization similar to Drosophila, but with modules of larger size. Interestingly, mammalian genes are also about one order of magnitude larger than their fly counterparts. Whether the conserved ratio between domain and gene sizes is circumstantial or more deeply linked to how domains are established remains unknown. Importantly however, no domain structure was described in yeast [ 21], where a compact and gene-packed genome is divided into chromosomes that are typically in the size of one Drosophila

domain. The epigenomics of 3C domains. The consistent evidence for 3C contact domains in Drosophila and mammals led to many questions nearly regarding the physical structure underlying such domains, and the implication of such structures on genome function. 3C domains were found to correlate strongly with linear epigenetic marks, including histone modification enrichments, active gene density, lamina interaction, replication time, nucleotide and repetitive element composition [ 8••]. The combination of these marks, that were previously studied statistically to extract epigenomic domains and classify them [ 10, 11 and 22••], was shown to distinguish many of the identified 3C domains, allowing their broad classification into groups.