Recent reports reinforce the notion that nucleosomes are highly dynamic in response to the process of transcription. Nucleosomes are displaced at promoters during gene activation in a process that involves histone modification, ATP-dependent nucleosome remodeling complexes, histone chaperones and perhaps histone variants. During transcription elongation nucleosomes are acetylated and transferred behind RNA polymerase II where they are required to suppress spurious transcription initiation within the body of the gene. It is becoming increasingly clear that the eukaryotic transcriptional machinery is adapted to exploit the presence of nucleosomes in very sophisticated ways.
Nucleosome-free regions at gene promotersRemoval of nucleosomes has been a long-standing hypothesis for how genes might become activated, but only recently have experiments begun to rigorously address this topic. For some time after the discovery of DNasehypersensitive sites (DHSs) in chromatin (for review, see Elgin 1981), it was anticipated that these nuclease sensitive regions would correspond to segments along the chromosome that are free of nucleosomes. With the common occurrence of DHSs at known transcriptional regulatory regions such as enhancers and promoters, mapping these exposed sites in genomic DNA became a powerful tool for identifying novel regulatory sites such as locus control regions (LCRs) and other functional genomic sequences (for review see, Gross and Garrard 1988). However, the question of whether these sites were actually histone free or simply did not look like nucleosomes to nucleases remained unresolved. The initial idea that DHSs must be nucleosome-free was attractive since the structure and stability of nucleosomes did not seem likely to accommodate the cobinding of transcription factors (Brown 1984;Kornberg and Lorch 1991). Early support for the notion that at least some of these regions are free of histones came from McGhee and Felsenfeld (McGhee et al. 1981), who examined the chicken adult -globin DHS by restriction endonuclease digestion. They showed the release of a 115-base-pair (bp) fragment from the DHS, of which one-third migrated on gels like naked DNA. In contrast, studies of an inducible DHS at the MMTV promoter suggested that it was still occupied by core histones following induction (Bresnick et al. 1992;Truss et al. 1995), thereby demonstrating the co-occupancy of histones and transcription factors. Moreover, a DHS could be reconstituted by the binding of transcription factors to the HIV-1 enhancer within a nucleosome array without the loss of histones (Steger and Workman 1997;Angelov et al. 2000). Thus, while DHSs do not appear to contain canonical nucleosomes as defined by nuclease digestion, an altered nucleosome or alternative histone DNA complex may still exist at these sites in some instances. Nevertheless, remembering the caveat that DHSs might still be bound by histones, we will refer to these sites as nucleosome-free for the purposes of this review.
Nucleosome distribution in Saccharomyces ce...