Structural analysis of a reconstituted DNA containing three histone octamers and histone H5

• Published in: Journal of molecular biology Vol. 197; no. 3; pp. 485 - 511
• Main Authors: Drew, Horace R., McCall, Maxine J.
• Format: Journal Article
• Language: English
• Published: England Elsevier Ltd 05.10.1987
• Subjects: Base Sequence; DNA, Bacterial - metabolism; Genes; Histones - metabolism; Methylation; Models, Molecular; Molecular Sequence Data; Nucleic Acid Conformation; Nucleosomes - ultrastructure; Protein Binding; Protein Conformation
• ISSN: 0022-2836; 1089-8638
• DOI: 10.1016/0022-2836(87)90560-2

Previous work has shown that DNA and the histone proteins will combine to form structures of a complex, yet definite nature. Here, we describe three experiments aimed at a better understanding of the interactions of DNA with the histone octamer and with histone H5. First, there has been some question as to whether the methylation of DNA could influence its folding about the histone octamer. To address this point, we reconstituted the histone octamer onto a 440 base-pair DNA of defined sequence at various levels of cytosine methylation, and also onto the unmethylated DNA. The reconstituted structures were probed by digestion with two different enzymes, micrococcal nuclease and DNase I. All samples were found to contain what appear to be three histone octamers, bound in close proximity on the 440 base-pair DNA. The cutting patterns of micrococcal nuclease and DNase I remain the same in all cases, even if the DNA has been extensively methylated. The results show, therefore, that methylation has little, or no, influence on the folding of this particular DNA about the histone octamer. Second, there has been concern as to whether the base sequence of DNA could determine its folding in a long molecule containing several nucleosomes, just as it does within any single, isolated nucleosome core. In order to deal with this problem, we cut the 440 base-pair DNA into three short fragments, each of nucleosomal length; we reconstituted each separately with the histone octamer; and then we digested the reconstituted complexes with DNase I for comparison with similar data from the intact 440 base-pair molecule. The results show that the folding of this DNA is influenced strongly by its base sequence, both in the three short fragments and in the long molecule. The rotational setting of the DNA within each of the three short fragments is as predicted from a computer algorithm, which measures its homology to 177 known examples of nucleosome core DNA. The rotational setting of the DNA in the 440 base-pair molecule remains the same as in two of the three short fragments, but changes slightly in a third case, apparently because of steric requirements when the nucleosomes pack closely against one another. Finally, there has been little direct evidence of where histone H5 binds within a DNA-octamer complex. To learn more about this subject, we reconstituted histone H5 onto the complex of the 440 base-pair DNA containing three histone octamers, and then probed for the locations of histone H5 by digestion with micrococcal nuclease and DNase I. The regions of strongest protection from cleavage upon the addition of histone H5 are found: in the DNA that joins cores 1 and 2, and cores 2 and 3; near the dyads of cores 2 and 3; and at either end of the DNA. When superimposed onto a model of the structure, these patches of protection trace out an image of the histone H5 molecule that is in good agreement with earlier work. It seems plausible that the long carboxy-terminal domain (or “tail”) of each H5 protein associates with the DNA between nucleosome cores, while its central globular domain (or “head”) lies close to a dyad. In the Appendix, the relation between DNase I cutting and the energy of DNA positioning is discussed. In addition, the prospects for assembly of a 300 Å fibre in vitro are considered (1 Å = 0.1 nm).