Chromatin fine structure of the histone gene complex of Drosophila melanogaster

volume 11 Number 2 1983 Nucleic Acids Research Chromatln fine structure of the histone gene complex of Drosophila melanogaster Abraham Worcel, Giuseppe Gargiulo, Bret Jessee. Andor Udvardy1, Christos Louis' and Paul Schedl1 Department of Biology. Hutchison Hall, University of Rochester, Rochester, NY 14627, and 'Department of Biology, Princeton University. Princeton, NJ 08544, USA Received 28 September 1982; Revised and Accepted 8 December 1982 INTRODUCTION The relationship between chromatin structure and the functional organization of DNA may have important implications for the mechanisms controlling gene expression in eukaryotes. In previous papers, we reported experiments which suggest that the 5 S RNA (1) and the histone (2) gene repeat units of the fruit fly, D. melanogaster, are organized into characteristic chromatin arrangements. In the case of the 5.0 kb histone repeat unit, for example, we found that each DNA segment was packaged into a distinct structure: the non-transcribed spacers are assembled into ordered nucleosome arrays with core particles aligned with respect to the underlying DNA sequence; the 5' ends of all five histone genes are in an exposed configuration and are sensitive to nuclease attack; the genes themselves have an "altered" chromatin structure perhaps reflecting a multiphasic nucleosome distribution. Other workers have reported similar results for a variety of different eukaryotic DNA sequences (3-6). Our experiments, and those of others, rely on the enzymes micrococcal nuclease and to a lesser extent DNase I to analyze the chromatin organization £ IRL Press Limited, Oxford, England. O3O5-1O48/83/1102-O421»2.0O/0 421 SUMMARY We have used salt extractions of nuclei and long agarose gels to dissect the chromatin fine structure of the histone gene repeat of Drosophila melanogaster. Extraction of nuclei with 0.35 M KC1 removes many non-histone chromosomal proteins but does not significantly disturb the overall nucleosome arrangement of the repeat unit. After extraction of nuclei with 0.55 M KC1, which also removes histone HI, the basic arrangement of nucleosome core particles in the repeat unit is not greatly disturbed and the exposed DNA segments near the 5' ends of the histone genes are also retained. Extraction of nuclei with 0.75 M or higher KC1 concentrations causes extensive nucleosome sliding and rearrangement with accompanying changes in the nucleoprotein organization of the histone gene complex and loss of the 5' hypersensitive sites. Our results indicate that the histone gene repeat displays a highly organized chromatin structure in vivo. Nucleic Acids Research Our data indicate that in spite of the DHA sequence preference snown Dy both micrococcal nuclease and DNase I, the enzymes can be successful^ used as probes for chromatin structure provided they are employed with adequate controls and the resulting DNA fragments are displayed on long gels. Moreover, the discrete changes detected by these nucleases as the various chromatin proteins are removerd from the DNA offer a unique way to dissect the chromatin fine structure of specific genes. 422 of the DNA. The former enzyme, micrococcal nuclease, has been a useful probe for chromatin structure (7). It cleaves in the linker DNA between nucleosomes generating initially the well-known ONA ladder and then, upon further digestion, nudeosome cores with 146 bp of DNA (8-10). In our studies, we used this enzyme to map the position of "linker" DNA sequences in the histone repeat with the indirect end-labeling technique (3-4). Recently, however, several papers have appeared which question the experimental approach and the conclusions which have been drawn from both our studies and those of others (7,11-13). These papers argue that the high degree of sequence specificity observed in micrococcal nuclease digests of free DNA makes it difficult, if not impossible, to identify cleavage products which are determined by the nucleoprotein organization of DNA in chromatin. Moreover, in contrast to our findings, several of these papers report that the cleavage products obtained in chromatin digests are indistinguishable from those observed in eauivalent digests of naked DNA. In view of this controversy, we decided to reinvestigate the chromatin organization of the histone repeat. In this report, we demonstrate that the distribution and yield of micrococcal nuclease cleavage products across the histone repeat unit clearly reflects the chromatin organization of the DNA rather than simply the sequence specificity of this enzyme. This conclusion is confirmed in studies using the enzyme DNase I. We have extended this analysis by using progressive salt extractions of nuclei to analyze the chromatin fine structure of the histone repeat. These salt extractions were designed to selectively remove first non-histone chromosomal proteins, then the non-histone proteins together with histone HI, and finally randomize and completely disrupt the nucleoprotein organization of the Drosophila chromosome. At each step specific and reproducible changes in the chromatin of the histone repeat are observed and thus the approach provides some new insights into the formation of specific chromatin architectures. Nucleic Acids Research RESULTS Chromatin of_ the Histone Repeat Unit We have carefully analyzed the distribution of micrococcal nuclease cleavage products across the histone repeat unit in chromatin and in free DNA, after resolving the DNA fragments on 45 cm long agarose gels. The experiment in Figure I shows a time course of micrococcal nuclease digestion of the histone repeat chromatin. Nuclei were prepared using standard conditions (see Methods) and incubated for increasing time with micrococcal nuclease. The digested DNA was isolated, restricted with Bgl II which cleaves once in this histone repeat at a site near the 31 end of the histone HI gene, and electrophoresed on a 45 cm flat-bed agarose gel. After blotting, the filter was probed with a 473 bp Bgl II-Bam HI subclone which includes the 3' end of the HI gene (2). As a control, lanes 2 and 3 contain micrococcal nuclease 423 MATERIALS AND METHODS Isolation of nuclei from Drosophila melanogaster tissue culture cells and embryos (14), nuclease digestions (2) and indirect end label analyses (3) were performed as previously described. The salt extractions of nuclei were carried out at 0-4° as follows: isolated nuclei in the standard isotonic buffer were adjusted to the required salt concentration by the addition of an equal volume of concentrated KC1. After gentle mixing, the suspension was kept on ice for 30 minutes and then layered on top of a 30 ml column of 5% sucrose containing KC1 at the same concentration as the overlayered nuclear suspension. This column was in turn layered on top of a cushion made up of 20% sucrose and the standard isotonic buffer. Nuclei were spun through the KC1 column and into the cushion by centrifugation at 6,000 rp (...truncated)