Dividing the empire: boundary chromatin elements delimit the territory of enhancers

Chromatin is the most complex supramolecular organization of the cell. It has a dual role: to compact the chromosomal DNA and to ensure a highly efficient regulation of gene expression. In consequence of the very compact packaging of the DNA, chromatin is highly repressive for transcription. Enhancers are key regulatory elements which can relieve the chromatin‐induced repression. The interplay between the chromatin‐mediated repression and the activating function of enhancers ensures that the difference between the induced or repressed level of gene expression in eukaryotic cells is several orders of magnitude higher than that in prokaryotic cells. The ability of enhancers to activate a gene independently of their distance from a promoter poses a topological problem: how is the territory of enhancer action delimited? It is supposed that the higher‐order chromatin structure divides the genome into topologically constrained loops which are attached to the nuclear matrix. This model implies that the loops are both structural and functional units of the chromatin in the sense that genes under common transcriptional regulation are located within the same loop and regulated by the same set of enhancers (Forrester et al ., 1986; Orkin, 1990; Bonifer et al ., 1991). In this attractive model, the monogamy of enhancers is ensured by the attachment of the bases of the chromatin loops to the nuclear matrix, which insulates the promoters located within the loop from the regulatory influence of enhancers present in the neighboring loops. Two distinct approaches have been applied to prove the validity of this model. In the structural approach, the main goals were to demonstrate that chromatin is indeed anchored at defined sites to a nuclear matrix, and to identify and characterize DNA sequences at these attachment sites. Structural studies have led to the discovery of scaffold attachment region (SAR) sequences (Mirkovitch et al., …