DNase 1 (deoxyribonuclease 1) is an enzyme that cuts DNA with relatively little specificity. The enzyme is reasonably large (>30,000 Da), so that assembing DNA into nucleosomes and chromatin impedes its enzymatic activity by limiting access to the DNA double helix. Sensitivity to DNase I is commonly believed to reflect the degree to which DNA is compacted in the chromosome. However other variables, such as proteins that like to bind DNase I and the actual structure of the DNA, might also influence enzymatic activity. DNase I likes to bind to DNA across the minor groove of the double helix.
Early experiments demonstrated the selective association of nonspecific DNA-binding proteins (prokaryotic RNA polymerases) with transcriptionally active chromatin. Following these advances, Weintraub, Felsenfeld, and colleagues showed that a comparable general accessibility to nucleases is associated with transcriptional activity (1, 2). This general sensitivity to nucleases includes the coding region of a gene and may extend several kilobases to either side of it, potentially defining a chromosomal domain . DNase I normally introduces double-strand breaks into transcriptionally active chromatin more than 10 times more frequently than into inactive chromatin. However, the exact structural basis of this generalized sensitivity is unknown. Careful analysis reveals that the untranscribed regions are just as sensitive to DNase I digestion as the transcribed regions, provided the last-cut approach to the measurement of DNase I sensitivity is used. This is defined as digestion by DNase I to fragments so small (smaller than 50 bp) that the DNA no longer hybridizes efficiently to complementary strands after denaturation (3). A certain length of DNA is necessary to allow specific recognition (hybridization) of two separated single-stranded regions. An important question not yet completely resolved is whether transcription is required to generate generalized nuclease sensitivity in certain instances or whether sensitivity always precedes transcription.
A useful example of the regulated appearance of DNase I sensitivity of a gene is provided by experiments on the action of mitogens on quiescent cells. In response to mitogens, a subset of genes, called the immediate-early genes , is rapidly induced (4). The most studied examples of such genes are c-myc and c-fos. These two proto-oncogenes are transcriptionally activated within minutes. Coincident with transcription, the chromatin structure of the proto-oncogenes becomes more accessible to nucleases. Once proto-oncogene transcription ceases, preferential nuclease accessibility is lost (5). Possible conformational changes in nucleosome structure might account for such effects. It has been proposed that histone H3 cysteine residues might become accessible in nuclease-sensitive chromatin, reflecting conformational changes within the nucleosome. However, the recruitment of RNA polymerase or other components of the transcriptional machinery may also provide the thiol groups that bind to the columns retaining active chromatin (6). Nevertheless, the rapidity of the changes in nuclease sensitivity (<90 s) and their propagation in both directions 5′ and 3′ to the promoter means that transcription and hence RNA polymerase or HMG proteins cannot account for all of the observed changes. This is consistent with some retention due to histone H3 in higher eukaryotes. In fact, the speed of the response suggests that changes in nuclease sensitivity precede transcription, and they may play a role in regulating c-fos expression.
It is interesting that one of the earliest mitogen-induced nuclear signaling events coincident with proto-oncogene induction is the rapid phosphorylation of histone H3 on serine residues within its highly charged, basic, amino-terminal domain. Acetylation of the amino-terminal domains of the core histones is also likely to be a component of transcriptional activation. Whether these changes are localized to chromatin regions containing either c-fos, c-myc or the other immediate-early genes has not yet been determined (7). An additional component contributing to the prior sensitization of the proto-oncogenes to nucleases may come from the existence of trans-acting factors already associated with the promoter (8). Such interactions are responsible for the second landmark in chromatin: DNase I hypersensitive sites. These sites are the first places where DNase I introduces a double-strand break in chromatin. They usually involve small segments of DNA sequences (100 to 200 bp) and are two or more orders of magnitude more accessible to cleavage than in inactive chromatin. 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. Cleavage at these sites might, however, preferentially solubilize chromatin, leading to an increase in the general accessibility of a chromatin domain to DNase I. The results emphasize the utility of DNase I as a probe for both actively transcribed genes and for detecting of regulatory DNA in the chromosome.