Characterization of restriction nuclease prepared chromatin by electron microscopy

Chromatin was solubilized from rat liver nuclei by digestion with the restriction nuclease EcoRI or HaeIII in the presence or absence of EDTA and sodium chloride. The samples were investigated by electron microscopy after positive and negative staining with uranyl acetate under a number of conditions. Depending on the salt concentration during solubilization the chromatin appeared as beads on the string or in more compact form. Solenoid- and superbead-like structures were seen as had been reported for chromatin solubilized with micrococcal nuclease.     

Endogenous polymers of ADP-ribose are associated with the nuclear matrix

The metabolism of nuclear polymers of ADP-ribose has been implicated in several chromatin-associated processes. However, the distribution of endogenous ADP-ribose polymers in the nucleus or within different fractions of chromatin has not been studied. Using a procedure which allowed the radiolabeling and detection of endogenous polymers of ADP-ribose, we have analyzed the nuclear distribution of these polymers in untreated cells and in cells subjected to hyperthermia, N-methyl-N'-nitro-N-nitrosoguanidine, or both. When isolated nuclei from cells subjected to any of these conditions were digested with micrococcal nuclease such that 80% of the DNA was released, 90% of the total poly(ADP-ribose) remained with the micrococcal nuclease resistant chromatin fraction. When nuclear matrix fractions were prepared by exhaustive DNase I digestion in combination with three different salt extraction procedures (2 M NaCl, 300 mM (NH4)2SO4 or 25 mM lithium diiodosalicylate), the matrices contained less than 1% of the total nuclear DNA but 50 to 70% of the total poly(ADP-ribose). These data suggest that the nuclear matrix may be a major site of poly(ADP-ribose) metabolism.     

Altered chromatin structure of cerebral nuclei in experimental diabetes mellitus.

To determine whether diabetes alters chromatin structure in vivo, micrococcal nuclease digestion kinetics were analyzed in cerebral cortical and hepatic nuclei of streptozotocin-induced diabetic rats. Cerebral nuclei of diabetic rats maintained for 6 weeks were less susceptible to micrococcal nuclease digestion compared with control rats. Insulin treatment reversed diabetes-related changes in nuclease digestion kinetics. There were no changes in the kinetics of digestion in hepatic nuclei. The reduced digestibility of cerebral DNA in diabetes could not be attributed to altered DNA fluorescence spectra, or altered distribution of most abundant chromatin proteins that were either solubilized or that remained insoluble immediately following nuclease digestion. It is concluded that chronic, uncontrolled hyperglycemia can alter chromatin structure of some tissues in vivo, and this change is probably related to subtle alterations in DNA-protein interactions.     

Undermethylation of DNA in mononucleosomes solubilized by micrococcal nuclease digestion of HeLa cell nuclei

To examine the distribution of 5-methylcytosine in chromatin DNA, DNA of HeLa cells was labeled with [3H-methyl]methionine and [14C] thymidine and analyzed after extensive digestion of the nuclei with micrococcal nuclease. When the chromatin solubilized with the nuclease was fractionated on a sucrose density gradient, DNA in mononucleosomes was considerably depleted in 5-methylcytosine, as compared with polynucleosomes. Electrophoretic separation of DNA from the chromatin also revealed the depletion of 5-methylcytosine in the mononucleosomal size of DNA. This was confirmed by the chromatographic analysis of 5-methyldeoxycytidine after enzymatic digestion of the DNA to nucleosides. Thus the DNA in mononucleosomes solubilized by extensive micrococcal nuclease digestion is depleted in 5-methylcytosine, suggesting that 5-methylcytosine is preferentially missing from the DNA in the nucleosome core particles.     

Nuclear association states of rat uterine oestrogen receptors as probed by nuclease digestion

The solubilization of oestrogen receptors from uterine nuclei by micrococcal nuclease and deoxyribonuclease I was examined after the injection of oestradiol or Nafoxidine into castrated female rats. At 1h after an injection of oestradiol, 30% (0.18pmol/mg of DNA) of the nuclear oestrogen receptors was solubilized by 5 min of mild digestion with either nuclease. No further receptor release occurred, although DNA hydrolysis continued throughout a 20min interval. The limitation in receptor solubilization was not due to an artifact of digestion conditions or insufficient nuclease concentrations. Similar patterns of receptor solubilization and DNA hydrolysis were obtained with both nucleases whether the animals had been injected with oestradiol 1h before death or if the uteri from uninjected animals were incubated with [(3)H]oestradiol for 1h in vitro. When uterine nuclei were digested with these enzymes 12h after the animal was injected with oestradiol there was little change in the quantity of nuclease-sensitive sites (0.11pmol/mg of DNA); however, the quantity of nuclease-resistant sites decreased 10-fold. These values correspond quantitatively to the changes in salt-resistant and salt-extractable sites observed over a 12h interval after oestradiol treatment. Nuclease digestion of uterine nuclei obtained 16h after Nafoxidine treatment gave a pattern qualitatively and quantitatively similar to that observed 1h after oestradiol treatment, a result consistent with the agonist/antagonist action of this compound. An analysis by sucrose-density-gradient centrifugation of the time course of nuclease-dependent receptor solubilization indicated that the solubilized receptors were not associated with discrete nucleosomal fragments. We believe that these data indicate that only a portion of the receptors translocated to the nucleus become associated with chromatin, and this association may occur on regions of chromatin that are preferentially susceptible to nucleolytic cleavage.     

Enrichment of transcribed and newly replicated DNA in soluble chromatin released from nuclei by mild micrococcal nuclease digestion

A chromatin fraction solubilized from mouse myeloma nuclei under near-physiological ionic conditions by very mild micrococcal nuclease digestion at 0°C is enriched at least 7-fold in DNA complementary to total myeloma polyadenylated mRNA, and 15-fold in DNA originating near the replication fork (labeled within 30 s). Newly replicated DNA recovered in solubilized chromatin after brief labeling was incorporated mainly into particles sedimenting with, or faster than, mononucleosomes. A rapid decrease in enrichment of newly replicated DNA in readily released, soluble chromatin with increasing labeling times indicated that newly replicated chromatin matured within 90 s to a form that was partitioned similarly to bulk chromatin by this fractionation method. Previous studies showed that chromatin readily solubilized from myeloma nuclei is enriched in high-mobility-group (HMG) and other non-histone proteins, RNA and single-stranded DNA; and depleted in H1 and 5-methylcytosine, relative to bulk chromatin (Jackson, J.B., Pollock, J.M., Jr., and Rill, R.L. (1979) Biochemistry 18, 3739–3748). Mild digestion of chicken erythrocyte nuclei with micrococcal nuclease yielded a soluble chromatin fraction (1–2% of the total DNA) with similar properties. This fraction was enriched at least 6-fold in DNA complementary to chicken globin mRNA, relative to total erythrocyte DNA.     

Analysis of the nuclear estrogen receptor from MCF-7 cells by limited proteolysis

The proteolytic fragments of the nuclear estrogen receptor in the MCF-7 cell line were characterized following limited digestion with chymotrypsin and trypsin. Nuclei were isolated from cells previously exposed to 10 nM [3H]estradiol. The proteolytic digestion was performed either on the micrococcal nuclease hydrolysate or on intact nuclei. The molecular weights (Mr) were calculated from the sedimentation coefficients determined on a sucrose gradient and from the Stokes radii estimated by gel filtration. Digestion of the nuclei with micrococcal nuclease solubilized a receptor form of Mr = 151,000. This receptor form was degraded by chymotrypsin to a receptor of Mr = 33,000 and by trypsin to a receptor of Mr = 60,000. Digestion of intact nuclei with chymotrypsin solubilized a receptor form of Mr = 62,000 which dissociated in 0.4 M KCl to a receptor of Mr = 32,000. Digestion of intact nuclei with trypsin followed by micrococcal nuclease solubilized a receptor form of Mr = 75,000 which was further dissociated by 0.4 M KCl to a receptor form of Mr = 60,000. The ability of the receptor forms to bind DNA was tested using DNA-cellulose column chromatography. About 40% of the micrococcal nuclease solubilized receptor form, compared to about 7% of the chymotrypsin degraded receptor and to about 13% of the trypsin degraded receptor forms, all bound to the column and could be eluted by high salt concentrated buffer. We conclude that the nuclear estrogen receptor in the MCF-7 cell line can be partially degraded either in the micrococcal nuclease hydrolysate or in intact nuclei by chymotrypsin or trypsin generating protein moieties, probably receptor fragments of Mr = 33,000 and 60,000 respectively. Both fragments retain their estradiol binding domain and it may be hypothesized that the heavier fragment retains its chromatin binding domain.     

Fractionation of nucleosomes by salt elution from micrococcal nuclease- digested nuclei

The solubilization of nucleosomes and histone H1 with increasing concentrations of NaCl has been investigated in rat liver nuclei that had been digested with micrococcal nuclease under conditions that did not substantially alter morphological properties with respect to differences in the extent of chromatin condensation. The pattern of nucleosome and H1 solubilization was gradual and noncoordinate and at least three different types of nucleosome packing interactions could be distinguished from the pattern. A class of nucleosomes containing 13-- 17% of the DNA and comprising the chromatin structures most available for micrococcal nuclease attack was eluted by 0.2 M NaCl. This fraction was solubilized with an acid-soluble protein of apparent molecular weight of 20,000 daltons and no histone H1. It differed from the nucleosomes released at higher NaCl concentrations in content of nonhistone chromosomal proteins. 40--60% of the nucleosomes were released by 0.3 M NaCl with 30% of the total nuclear histone H1 bound. The remaining nucleosomes and H1 were solublized by 0.4 M or 0.6 M NaCl. H1 was not nucleosome bound at these ionic strengths, and these fractions contained, respectively, 1.5 and 1.8 times more H1 per nucleosome than the population released by 0.3 M NaCl. These fractions contained the DNA least available for micrococcal nuclease attach. The strikingly different macromolecular composition, availability for nuclease digestion, and strength of the packing interactions of the nucleosomes released by 0.2 M NaCl suggest that this population is involved in a special function.     

The variation with age of the structure of chromatin in three cell types from rat liver.

The organization of chromatin in three rat liver nuclear populations, namely diploid stromal, diploid parenchymal, and tetraploid parenchymal nuclei, which were separated by zonal centrifugation, was studied by digestion with micrococcal nuclease and pancreatic deoxyribonuclease in 3-week-old rats in which the parenchymal cells contain diploid nuclei and in 2-and 4-month-old rats with a high proportion of tetraploid nuclei. Digestion by micrococcal nuclease allowed the estimation of DNA-repeat length in chromatin. Parenchymal nuclei have shorter repeat length than stromal nuclei and DNA-repeat length increases with the age in all three nuclei populations. The kinetics of digestion by micrococcal nuclease showed that nuclei with shorter repeat length are more sensitive to micrococcal nuclease and that the sensitivity of chromatin decreases with age for all the types of nuclei in this study. The kinetics of digestion by pancreatic deoxyribonuclease showed that sensitivity of chromatin is related to the repeat length and that the sensitivity decreases with the ages.     

Structure of rat liver nuclei after depletion and reassociation of histone H1

Chromatin in isolated rat liver nuclei was compared with chromatin in (i) nuclei depleted of H1 by acid extraction; (ii) nuclei treated at pH 3.2 (without removal of H1), and (iii) depleted nuclei following reassociation of H1. Electron microscopy and digestion by DNase I, micrococcal nuclease and endogenous Ca/Mg endonuclease were used for this comparative examination. Electron micrographs of H1-depleted nuclei showed a dispersed and finely granular appearance. The rate of DNA cleavage by micrococcal nuclease or DNase I was increased several-fold after H1 removal. Discretely sized intermediate particles produced by Ca/Mg endonuclease in native nuclei were not observed in digests of depleted nuclei. Digestion by micrococcal nuclease to chromatin particles soluble in 60 mM NaCl buffer appeared not to be affected in depleted nuclei. When nuclei were treated at pH 3.2, neither the appearance of chromatin in electron micrographs nor the mode or rate of nuclease digestion changed appreciably. Following reassociation of H1 to depleted nuclei, electron micrographs demonstrated the reformation of compacted chromatin, but the lower rate of DNA cleavage in native nuclei was not restored. Further, H1 reassociation produced a significant decrease in the solubility of nuclear chromatin cleaved by micrococcal nuclease or Ca/Mg endonuclease. In order to evaluate critically the reconstitution of native chromatin from H1-depleted chromatin we propose the use of digestion by a variety of nucleases in addition to an electron microscopic examination.     

Accessibility of some regions of DNA in chromatin (chicken erythrocytes) to single strand-specific nucleases.

The susceptibility of the DNA in chromatin to single strand-specific nucleases was examined using nuclease P1, mung bean nuclease, and venom phosphodiesterase. A stage in the reaction exists where the size range of the solubilized products is similar for each of the three nucleases and is nearly independent of incubation time. During this stage, the chromatin fragments sediment in the range of 30 to 100 S and contain duplex DNA ranging from 1 to 10 million daltons. Starting with chromatin depleted of histones H1 and H5 similar fragments are generated. In both cases these nucleoprotein fragments are reduced to nucleosomes and their multimers by micrococcal nuclease. Thus, chromatin contains a limited number of DNA sites which are susceptible to single strand-specific nucleases. These sites occur at intervals of 8 to 80 nucleosomes and are distributed throughout the chromatin. Nucleosome monomers, dimers, or trimers were not observed at any stage of single strand-specific nuclease digestion of nuclei, H1- and H5-depleted chromatin, or micrococcal nuclease-generated oligonucleosomes. Each of the three nucleases converted mononucleosomes (approximately 160 base pairs) to nucleosome cores (approximately 140 base pairs) probably by exonucleolytic action that was facilitated by the prior removal of H1 and H5. The minichromosome of SV40 is highly resistant to digestion by nuclease P1.     

Variation in chromatin structure in two cell types from the same tissue: a short DNA repeat length in cerebral cortex neurons

We have used micrococcal nuclease as a probe of the repeating structure of chromatin in four nuclear populations from three tissues of the rabbit. Neuronal nuclei isolated from the cerebral cortex contain about 160 base pairs of DNA in the chromatin repeat unit, as compared with about 200 base pairs for nonastrocytic glial cell nuclei from the same tissue, neuronal nuclei from the cerebellum and liver nuclei. All four types of nuclei show the same features of nucleosomal organization as other eucaryotic nuclei so far studied: nucleosomes liberated by digestion with micrococcal nuclease give a “core particle” containing 140 base pairs as a metastable intermediate on further digestion and a series of single-strand DNA fragments which are mutiples of 10 bases after digestion with DNAase I. Nuclei from cerebral cortex neurons, which have a short repeat, are distinct from the others in being larger, in having a higher proportion of euchromatin (dispersed chromatin) as judged by microscopy and in being more active in RNA synthesis in vitro.     

Chromatin structure of the chicken lysozyme gene domain as determined by chromatin fractionation and micrococcal nuclease digestion

The chromatin structure encompassing the lysozyme gene domain in hen oviduct nuclei was studied by measuring the partitioning of coding and flanking sequences during chromatin fractionation and by analyzing the nucleosome repeat in response to micrococcal nuclease digestion. Following micrococcal nuclease digestion, nuclei were sedimented to obtain a chromatin fraction released during digestion (S1) and then lysed in tris(hydroxymethyl)aminomethane-(ethylenedinitrilo)tetraacetic acid-[ethylenebis(oxyethylenenitrilo)]tetraacetic acid and centrifuged again to yield a second solubilized chromatin fraction (S2) and a pelleted fraction (P2). By dot-blot hybridization with 14 specific probes, it is found that the fractionation procedure defines three classes of sequences within the lysozyme gene domain. The coding sequences, which partition with fraction P2, are flanked by class I flanking sequences, which partition with fractions S1 and P2 and which extend over 11 kilobases (kb) on the 5'side and probably over about 4 kb on the 3' side. The partitioning of class II flanking sequences, which are located distal of class I flanking sequences, is different from that of class I flanking sequences. Coding sequences lack a canonical nucleosome repeat, class I flanking sequences possess a disturbed nucleosome repeat, and class II flanking sequences generate an extended nucleosomal ladder. Coding and class I flanking sequences are more readily digested by micrococcal nuclease than class II flanking sequences and the inactive beta A-globin gene. In hen liver, where the lysozyme gene is inactive, coding and class I flanking sequences fractionate into fractions S2 and P2.(ABSTRACT TRUNCATED AT 250 WORDS)     

Structural features of the amplified N-myc oncogene detected by micrococcal nuclease digestion of neuroblastoma cell nuclei

The structure of chromatin containing amplified N-myc in neuroblastoma and retinoblastoma cells was investigated using micrococcal nuclease digestion of isolated nuclei. The size distribution of DNA fragments containing N-myc, produced by micrococcal nuclease digestion of nuclei, was determined and compared to that of DNA containing the structural gene for dihydrofolate reductase. A perturbation of the native structure of chromatin containing N-myc was evident from the association of N-myc with more extensively digested DNA when compared with chromatin containing dihydrofolate reductase.     

Expansion of chicken erythrocyte nuclei upon limited micrococcal nuclease digestion. Correlation with higher order chromatin structure

A sensitive method for measuring nuclear volumes with a Coulter counter is described. It has been applied to the digestion of chicken erythrocyte nuclei by micrococcal nuclease and DNase I. Early in digestion, micrococcal nuclease induced a 20% increase in the effective spherical volume of the nuclei, followed by a gradual reduction. At the peak of nuclear swelling, about 17% of the chromatin was soluble after lysis and its average chain length was about 18 kilobase pairs (kb). DNase I digestion did not give rise to a corresponding expansion of the nuclei. Several preparation conditions, including the treatment of nuclei with 0.2% Triton X-100, led to a loss of the expansion effect upon subsequent micrococcal nuclease digestion. The results support the domain theory of higher order chromatin structure. In the context of this model, the observed maximum nuclear expansion correlates with an average of one nuclease scission per domain.