Nucleosome binding by the bromodomain and PHD finger of the transcriptional cofactor p300.

The PHD finger and the bromodomain are small protein domains that occur in many proteins associated with phenomena related to chromatin. The bromodomain has been shown to bind acetylated lysine residues on histone tails. Lysine acetylation is one of several histone modifications that have been proposed to form the basis for a mechanism for recording epigenetically stable marks in chromatin, known as the histone code. The bromodomain is therefore thought to read a part of the histone code. Since PHD fingers often occur in proteins next to bromodomains, we have tested the hypothesis that the PHD finger can also interact with nucleosomes. Using two different in vitro assays, we found that the bromodomain/PHD finger region of the transcriptional cofactor p300 can bind to nucleosomes that have a high degree of histone acetylation. In a nucleosome retention assay, both domains were required for binding. Replacement of the p300 PHD finger with other PHD fingers resulted in loss of nucleosome binding. In an electrophoretic mobility shift assay, each domain alone showed, however, nucleosome-binding activity. The binding of the isolated PHD finger to nucleosomes was independent of the histone acetylation levels. Our data are consistent with a model where the two domains cooperate in nucleosome binding. In this model, both the bromodomain and the PHD finger contact the nucleosome while simultaneously interacting with each other.

Multiple epigenetic maintenance factors implicated by the loss of Mll2 in mouse development.

Epigenesis is the process whereby the daughters of a dividing cell retain a chromatin state determined before cell division. The best-studied cases involve the inheritance of heterochromatic chromosomal domains, and little is known about specific gene regulation by epigenetic mechanisms. Recent evidence shows that epigenesis pivots on methylation of nucleosomes at histone 3 lysines 4, 9 or 27. Bioinformatics indicates that mammals have several enzymes for each of these methylations, including at least six histone 3 lysine 4 methyltransferases. To look for evidence of gene-specific epigenetic regulation in mammalian development, we examined one of these six, Mll2, using a multipurpose allele in the mouse to ascertain the loss-of-function phenotype. Loss of Mll2 slowed growth, increased apoptosis and retarded development, leading to embryonic failure before E11.5. Using chimera experiments, we demonstrated that Mll2 is cell-autonomously required. Evidence for gene-specific regulation was also observed. Although Mox1 and Hoxb1 expression patterns were correctly established, they were not maintained in the absence of Mll2, whereas Wnt1 and Otx2 were. The Mll2 loss-of-function phenotype is different from that of its sister gene Mll, and they regulate different Hox complex genes during ES cell differentiation. Therefore, these two closely related epigenetic factors play different roles in development and maintain distinct gene expression patterns. This suggests that other epigenetic factors also regulate particular patterns and that development entails networks of epigenetic specificities.

The many colours of chromodomains.

Local differences in chromatin organisation may profoundly affect the activity of eukaryotic genomes. Regulation at the level of DNA packaging requires the targeting of structural proteins and histone-modifying enzymes to specific sites and their stable or dynamic interaction with the nucleosomal fiber. The "chromodomain", a domain shared by many regulators of chromatin structure, has long been suspected to serve as a module mediating chromatin interactions in a variety of different protein contexts. However, recent functional analyses of a number of different chromodomains revealed an unexpected diversity of interaction targets, including histones, DNA and even RNA. The chromodomains of today seem to have evolved from a common ancestral fold to fulfill various functions in different molecular contexts. Combining information gained from recent functional and structural studies of chromodomains with a bioinformatic classification of their structure could lead to the definition of sequence motifs with predictive quality for chromodomain function.