DNase 1 hypersensitive sites are the first place this nuclease introduces a double-strand break in chromatin (see DNase 1 Sensitivity). These sites usually involve small segments of DNA sequences (100 to 200 bp) and are two or more orders of magnitude more accessible to DNase 1 cleavage than inactive chromatin (1, 2). DNase 1 hypersensitive sites result from the assembly of specific nucleoprotein complexes that generally contain transcription factors flanked by positioned nucleosomes. DNase I hypersensitive sites also reflect the stable association of a transcription factor on the surface of a nucleosome (3). DNase I hypersensitive sites and nucleosome arrays are usually detected by DNase 1 or micrococcal nuclease digestion, followed by indirect end-labeling methodologies to map the hypersensitive site relative to a site of restriction enzyme cleavage (Fig. 1). As the most accessible regions of chromatin to non-histone DNA-binding proteins, DNase I hypersensitive sites generally denote DNA sequences with important functions in the nucleus.
Figure 1. A scheme for mapping the position of DNase I hypersensitive sites in chromatin relative to restriction endonuclease cleavage sites.
DNase 1 hypersensitive sites were first detected in the SV40 minichromosome, at the region that functions as an origin of replication, and in Drosophila chromatin (4). In general, these sites are accessible to all enzymes or reagents that cut duplex DNA. These sites have been mapped to a large number of functional segments of DNA, including promoters, enhancers, locus control regions, transcriptional silencers, origins of replication, recombination elements, and structural sites within or around telomeres (5).
These sites often fall into a hierarchy of patterns around regulated genes. Twelve DNase I hypersensitive sites are found in the chicken b-globin locus that contains four globin genes (5′-s-bH-bA-e-3′) and covers more than 65 kbp of DNA (Fig. 2). One site is present in all cells, independent of whether the genes are transcriptionally active or not. Three sites, upstream of the s-globin gene, were present only in erythroid cells destined to express the globin genes. These sites were initially without clear functional significance. However, a similar site was found between the bA and e genes that corresponds to an enhancer element. Four sites were found over the promoters of each gene, depending on whether the gene was transcriptionally active, and three sites were found downstream of the genes, corresponding to transcription termination elements (the b A gene excluded). It is important to note that the formation of DNase I hypersensitive sites at the promoters of the globin genes is a relatively late step in the commitment of these genes to become transcriptionally active. However, it is clear that the formation of such sites precedes the actual initiation of transcription by evidence for transcriptional activation mediated by this nucleosome is yet to be established. It is clear, however, that the positioning of a nucleosome in this particular way allows key transcription factors to obtain access to essential regulatory elements in spite of the assembly of the gene into chromatin.
Figure 2. DNase I hypersensitive sites of the chicken b-globin locus. Contiguous genes are indicated as open arrows, the three distinct types of DNase I hypersensitive sites present in all cells specific to erythroid cells, and those that are developmentally regulated within the erythroid lineage are indicated.
One of the most thorough dissections of a DNase I hypersensitive site was carried out by Elgin and colleagues (4). The Drosophila heat-shock protein hsp26 gene is very rapidly activated transcriptionally by raising the temperature of a fly to a stressful level (a heat shock of 34°C). Two DNase I hypersensitive sites exist at the promoter of the hsp26 gene, including recognition sequences for the promoter-specific, heat-shock transcription factor (HSTF, a leucine zipper protein) (Fig. 3). Following heat shock, HSTF binds to these sites. In contrast, transcription factor TFIID is bound to the TATA box before and after heat shock. TFIID alone is insufficient to cause the hsp26 gene to be transcribed, and the specific association of the HSTF protein is also required. High-resolution analysis revealed that a nucleosome is positioned between the proximal and distal binding sites for HSTF, i.e. between the two DNase I hypersensitive sites. In this case, the exact position of histone-DNA contacts within this nucleosome depends on the DNA sequence to which the histones bind (from -300 to -140 relative to the start site of transcription at +1), and also on adjacent DNA sequences. These are repeats of the type (CT) (GA)n, which bind a specific transacting factor, the GAGA protein (7). The (CT)n.(GA) n repeat regions are located to either side of the positioned nucleosome at -347 to -341 and at -135 to -85. The GAGA factor bound to these repeats may function as a "bookend" to determine exactly where the nucleosome will be positioned. Recent evidence suggests that the GAGA factor might function through ATP-dependent mechanisms to direct the assembly of a particular chromatin architecture actively. Transcription of the gene is regulated through the association of the HSTF with recognition elements at -51,-170,-269, and -340. The sites at -170 and -269 will be wrapped around the core histones in rotational frames that prevent HSTF association. However, the wrapping of DNA around the nucleosome will also bring the HSTF molecules bound to the sites at -340 and -51 into juxtaposition, and the histones may potentially facilitate transcription by causing a clustering of HSTF activation domains. Direct RNA polymerase. Indeed, the generation of these sites may account for a component of the general nuclease sensitivity of a gene (see DNase 1 Sensitivity) (6).
Figure 3. The specific chromatin organization of the Drosophila hsp26 promoter. Key cis-acting elements are indicated relative to the start site of transcription (hooked arrow). The organization of these sites on a specific nucleosomal scaffold is indicated together with the interactions necessary to prevent or activate transcription. A tethered RNA polymerase II molecule is released through events initiated by the binding of HSTF.
For a DNase I hypersensitive site on the vitellogenin genes, nucleosome positioning directed by the DNA sequence occurs between -300 and -140 relative to the start site of transcription at +1. The binding sites for the stimulatory transcription factors, the estrogen receptor and nuclear factor 1, lie outside the region of DNA that is wrapped around the histones, at -300 to -330 and at -120 to -110, respectively (8). When these sites are brought together either by positioning a nucleosome in between them (or artificially by deleting the intervening DNA), transcription is enhanced about 5- to 10-fold. This moderate stimulatory effect is much more significant than it might appear because assembling a nonspecific chromatin structure would normally lead to a >20-fold repression of transcription as the binding sites for transcription factors are occluded by the histones. Thus in the vitellogenin gene example, nucleosome positioning has two roles: (1) to provide a scaffold that allows transcription factors to communicate more effectively; and (2) to prevent the formation of repressive histone-DNA interactions that may prevent any transcription factor from gaining access to a chromatin template.
DNase I hypersensitive sites have proven very useful in defining important DNA sequences. Among these are the four strongly nuclease-sensitive sites located 10 to 20 kbp upstream of the cluster of human b-globin genes. These sites, known as locus control regions (LCRs), represent cis-acting elements that allow genes integrated into a chromosome to be expressed independently of chromosomal position, i.e. position effects are abolished. A consequence of this is that LCRs allow each copy of a gene integrated in multiple copies to be expressed equivalently, so that gene expression is copy-number dependent. Therefore the LCR functions to control gene activity over an entire chromatin domain.
