Chromatin and transcription: where do we go from here?

Abstract

Not long ago, the very relevance of chromatin structure to transcription was in doubt. The structure of promoters and regulatory sequences could be manipulated and the consequences for transcription determined in vivo without regard to the packaging of the DNA in nucleosomes. Naked DNA could be employed in studies of transcription in vitro, reconstituting most aspects of transcription and its control, without the inclusion of histones or other chromatin proteins. It has since emerged that the role of chromatin is pervasive but could, nonetheless, be neglected in many studies on transcription for at least two reasons. First, numerous chromatin-remodeling complexes make DNA available in vivo, permitting DNA–protein interactions important for transcription despite packaging in nucleosomes and possibly without the complete disruption of chromatin structure. Second, there appear to be at least two levels of transcriptional regulation: one involving chromatin ‘unfolding’ and another involving the transcription machinery at RNA polymerase II promoters. The first level may be a prerequisite for the second, which can then proceed independently; that is, chromatin unfolding may expose promoters for assembly of the transcription machinery, explaining why transcription could be studied so successfully in the past with the use of naked DNA. Chromatin structure and modification Most research and thinking about the transcription of chromatin is based on the nucleosome, the conserved unit of chromatin organization. In reprising what is known, as a point of departure for discussing the outstanding issues in the field, it therefore seems most appropriate to begin with the nucleosome. X-ray structure determination of the histone octamer and nucleosome core particle has revealed a division between globular ‘histone fold’ and extended ‘histone tail’ regions [9,10]. The histone folds are responsible both for the interactions of the histones with one another to form the octamer and for the interactions with DNA that constrain it in a superhelical path around the octamer. The histone tails protrude, making no apparent contribution to either octamer or core particle structure. Inasmuch as the tails contain almost all sites of histone modification — including sites of acetylation, phosphorylation, methylation, and ubiquitination — it seems reasonable to conclude that these modifications do not affect the structure or stability of the core nucleosome directly. Effects on tail conformation, as well as indirect effects on core nucleosome structure, through the agency of specific tail-binding proteins, are possible but have not yet been demonstrated. Histone modifications, featured in the reviews of this issue, are thought to impart functional diversity to an otherwise monotonous array of nucleosomes. Much of the work performed to date is descriptive, detailing the variety of modifications, the proteins and multiprotein complexes responsible, and the functional consequences. The results are impressive and exciting, showing direct connections with gene regulation, through histone-modifying enzymes that serve as transcriptional regulatory factors, and giving significant insight into the basis for epigenetic effects on gene expression. For the most part, however, the molecular mechanisms involved, and thus a full solution of the problem, remain elusive. Two types of mechanism have been considered: histone modifications as determinants of ‘higher order’ chromatin structure, and histone modifications as tags for recruiting further proteins to specific chromosomal regions. These mechanisms may be related, as in the case of histone acetylation and chromosome condensation.