Topologically associating domains (TADs) are genomic regions ("chromosome neighborhoods") used to summarize the three-dimensional nuclear organization of mammalian genomes. A TAD represents a region of DNA within which physical interactions occur relatively frequently, whereas interactions across a TAD boundary occur relatively infrequently. TADs can range in size from thousands to millions of DNA bases.
The mechanisms behind TAD formation are complex and not yet fully elucidated, though a number of proteins are known to be associated with TAD formation: for example, the proteins CTCF and cohesin. Similarly, it is also unknown what components are required at TAD boundaries. However, it has been shown that these boundary regions have comparatively high levels of CTCF binding. In addition, some types of genes (such as transfer RNA genes and housekeeping genes) appear near TAD boundaries more often than would be expected by chance.
TADs are defined as regions whose DNA sequences preferentially contact each other. They have been found using data gathered by the Hi-C technique. They have been shown to be present in fruit flies (Drosophila), mouse and human genomes, but not in the wine yeast Saccharomyces cerevisiae.
To determine TAD locations, or "call TADs", an algorithm is applied to Hi-C data.
TADs are often called according to the method in Dixon et al (2012), using the so-called "directionality index". The directionality index is calculated for individual 40kb bins, by collecting the reads that fall in the bin, and observing whether their paired reads map upstream or downstream of the bin (read pairs are required to span no more than 2Mb). A positive value indicates that more read pairs lie downstream than upstream, and a negative value indicates the reverse. Mathematically, the directionality index is a signed chi-square statistic.
TADs have been reported to be relatively constant between different cell types (in stem cells and blood cells, for example), and even between species in specific cases.
The majority of observed interactions between promoters and enhancers do not cross TAD boundaries. Removing a TAD boundary (for example, using CRISPR to delete the relevant region of the genome) can allow new promoter-enhancer contacts to form. This can affect gene expression nearby - such misregulation has been shown to cause limb malformations (e.g. polydactyly) in humans and mice.
TADs have been reported to be the same as replication domains, regions of the genome that are copied (replicated) at the same time during S phase of cell division. Insulated neighborhoods, DNA loops formed by CTCF/cohesin-bound regions, are proposed to functionally underlie TADs.
Disruption of TAD boundaries can affect the expression of nearby genes, and this can cause disease.
For example, genomic structural variants that disrupt TAD boundaries have been reported to cause developmental disorders such as human limb malformations. Additionally, it has been claimed that disruption of TAD boundaries can provide growth advantages to certain cancers, such as T-cell acute lymphoblastic leukemia(T-ALL), gliomas, and colorectal cancer.