Superstructural differences between chromatin in nuclei and in solution are revealed by kinetics of micrococcal nuclease digestion

Quantitative MNase-seq accurately maps nucleosome occupancy levels

Micrococcal nuclease (MNase) is widely used to map nucleosomes. However, its aggressive endo-/exo-nuclease activities make MNase-seq unreliable for determining nucleosome occupancies, because cleavages within linker regions produce oligo- and mono-nucleosomes, whereas cleavages within nucleosomes destroy them. Here, we introduce a theoretical framework for predicting nucleosome occupancies and an experimental protocol with appropriate spike-in normalization that confirms our theory and provides accurate occupancy levels over an MNase digestion time course. As with human cells, we observe no overall differences in nucleosome occupancies between Drosophila euchromatin and heterochromatin, which implies that heterochromatic compaction does not reduce MNase accessibility of linker DNA.

Role of chromatin accessibility in the occupancy and transcription of the insulin gene by the pancreatic and duodenal homeobox factor 1

The pancreatic and duodenal homeobox factor 1 (Pdx-1) is a Hox-like transcription factor that is responsible for the activation of the insulin gene. Previous studies have demonstrated the interaction in vitro of Pdx-1 with short (20-40 nucleotide) DNA fragments corresponding to A boxes of the insulin promoter. Precisely how Pdx-1 binds to DNA in the complex milieu of chromatin, however, has never been studied. In this study, we explored how Pdx-1-DNA interactions might be influenced by chromatin accessibility at the insulin gene in beta-cells (betaTC3) vs. pancreatic ductal cells (mPAC). We demonstrate that Pdx-1 occupies the endogenous insulin promoter in betaTC3 cells but not in mPAC cells, a finding that is independent of the intracellular Pdx-1 protein concentration. Based on micrococcal nuclease protection assays, the difference in promoter binding between the two cell types appears to be secondary to chromatin accessibility at predicted Pdx-1 binding sites between bp -126 to -296 (relative to the transcriptional start site) of the insulin promoter. Binding studies using purified Pdx-1 and reconstituted chromatin in vitro suggest that the positioning of a nucleosome(s) within this crucial region of the promoter might account for differences in chromatin accessibility. Consistent with these observations, fluorescence colocalization studies show that Pdx-1 does not occupy regions of compacted, nucleosome-rich chromatin within the nucleus. Our findings suggest a model whereby insulin transcription in the beta-cell is at least partially facilitated by enhanced chromatin accessibility within a crucial regulatory region between bp -126 to -296, thereby permitting occupancy by transactivators such as Pdx-1.

Cation-dependent solubilization of rat thymocyte chromatin is closely related to decondensation of the nuclei

The cation-dependent solubilization of rat thymocyte chromatin has been compared with decondensation of the nuclei as a function of sodium phosphate-mediated changes in the concentration of Mg2+ and Na+. After digestion of the nuclei with DNase I or Micrococcus nuclease for a time just sufficient to permit extraction of a maximal amount of chromatin (minimum digestion), solubilization of most of the chromatin was found to occur with the same cation dependency as decondensation of untreated nuclei, while further digestion changed the ionic requirements for solubilization. The cation-dependency of the chromatin solubility and of the nuclear decondensation also exhibited the same variations with temperature. The chromatin in the nuclei became up to 4-times more sensitive to DNase I by decondensation, which also induced a shift in the DNase I cleavage mode from a 200 bp to a 100 bp repeat pattern. In contrast, the sensitivity to Micrococcus nuclease appeared to be nearly unchanged. These results suggest that solubilization of chromatin prepared by a mild endonuclease treatment occurs as a direct consequence of structural changes in the chromatin which take place during decondensation of the nuclei.

The estrogen-receptor complex is bound at unusual chromatin regions

The estrogen receptor of MCF-7 cells labeled with high specific activity estradiol was used to mark the chromatin binding sites for this regulatory molecule. Many of these sites are especially sensitive to nuclease, and produce on digestion a series of uniquely sedimenting products. Several of these have been examined in some detail in this paper. These include a form of receptor that sediments in trace digests at 9S but in more extensive digests at 7S, fast mononucleosomes of about 12.5S, and a species at 15S. Two components of digests, fast mononucleosomes and dinucleosomes were isolated and subjected to further digestion. Much of the hormone on these isolated particles was found to be sensitive to additional hydrolysis, although some was nuclease resistant. It appears that a major fraction of the hormone receptor complexes bound to MCF-7 cell chromatin occurs at nucleosome-free regions which can be detected as transient hydrolysis intermediates.

Sensitive quantitation of endonuclease kinetics

Described here is an assay permitting exclusive quantitation of the kinetics of endonucleolysis by a mixed-function nuclease at substrate DNA concentrations much lower than those necessary with other assay methods. It makes possible determination of Michaelis parameters Km and kcat for endonuclease activity of such enzymes. The sensitivity of the assay method to very low DNA concentrations is obtained through use of 32P-labeled DNA as substrate. Digested DNA is electrophoresed, and computerized analysis of an autoradiogram made from the gel gives the extent of digestion. The analysis produces a profile of weight fraction vs. DNA fragment size for each sample taken from a nuclease-DNA reaction mixture. Each experimental profile is compared to a family of theoretical profiles generated by computer using theory of polymer statistics and assuming random cleavage. Theoretical profiles are found whose shapes most closely match those of the experimental profiles. Associated with each best-fitting theoretical profile is a value of the number of cuts made in the DNA sample. These values, plotted against time, give initial reaction velocities. Initial velocities from experiments at different DNA concentrations have been used to obtain Km and kcat for micrococcal nuclease under specific reaction conditions.