referred to as heterochromatic, because the fibers are darkly stained, highly condensed and usually .... ultraviolet scanner. We estimated the proportions of ...
Volume 7 Number 4 1979
Nucleic Acids Research
Separation of satellite DNA chromatin and main band DNA chromatin from mouse brain
J.A.Mazrinas, Rod Balhorn and F.T.Hatch Lawrence Livermore Laboratory, University of California, P.O.Box 5507, Biomedical Sciences Division, livermore, CA 94550, USA
Received 25 June 1979 ABSTRACT
Using restriction endonucleases which preferentially digest mouse main band DNA and leave satellite DNA intact, we have isolated highly purified chromatin fractions containing only mouse satellite or main band DNA. Following the digestion of mouse brain nuclei with EndoR Alu I, main band DNA chromatin is selectively extracted with lOmM Tris, 10mM EDTA. Satellite DNA chromatin is subsequently extracted from the nuclear pellet with Tris-3M urea and further purified on sucrose gradients. Chromatin extracted from digested nuclei with Tris-EDTA contains only main band DNA and has a molecular weight lower than 2 X 106. Chromatin fractions obtained from the lower regions of sucrose gradients of the Tris-Urea extracts contain 40-95% satellite DNA and have a molecular weight of 6 to 8 X 106. Both the satellite DNA and main band DNA chromatins contain all five histones and have a protein to DNA ratio of 1.3 to 1. INTRODUCTION When the interphase nucleus of a rodent cell is properly stained for chromatin and viewed microscopically, two physical states of the chromatin may be distingushed. One state is called euchromatic, because the fibers are lightly stained and diffused throughout the nucleus. The other state is referred to as heterochromatic, because the fibers are darkly stained, highly condensed and usually located in specific areas adjacent to the perinuclear membrane and surrounding the nucleolilb3. Cytological evidence indicates that euchromatin contains the genetically active and potentially active chromatin that replicates early in most eukaryotic cells. The heterochromatin fraction is quantitatively smaller, transcriptionally inactive, and it usually replicates late. An enrichment of highly repetitive sequences such as mouse satellite DNA has been seen in this fraction using several different techniques4'5. Recent studies show that restriction enzymes may be used to digest specific types of chromatin preferentially6 1. The experiments indicate that restriction nucleases cleave chromatin predominantly within the linker C Information Retrieval Limited 1 Falconberg Court London Wl V 5FG England
935
Nucleic Acids Research region between nucleosome cores, and these enzymes can be used to selectively digest chromatin fractions containing or lacking the satellite or highly repetitive DNAs of mouse, rat, calf, and African green monkey. The satellite and nonsatellite chromatin fractions are then separated on sucrose gradients. Bernstine reported a four- to five-fold enrichment of mouse satellite chromatin in one fraction. However, these digestions with EndoR Eco RI were only for 30 minutes and required a large quantity of enzyme (4,000 U/mg DNA); the source of nuclei was liver, a tissue that contains extremely active endogenous nucleases and proteases. To minimize damage to native nucleosomal structure, we devised a method to fractionate chromatin that avoids mechanical homogenization or sonication and gently solubilizes chromatin with nucleases12. In this method we use a tissue (brain) containing virtually no endogenous nuclease or protease activities,13 a gentle enzymatic digestion, and extractions at low salt concentrations aided by urea. Chromatin fractions containing satellite DNA and main band DNA are isolated in high yield and purity and exhibit native physicochemical properties, except that their molecular weight distributions are different.
MATERIALS AND METHODS Preparation of Nuclei. The brains and livers were dissected from SWR/J mice (Jackson Labs., Bar Harbor, Maine), quickly frozen on dry ice, and stored at -200C for short periods of time before being used. Nuclei were isolated as described by Panyim et al.,I4 except that the grinding and washing media contained 10 mM MgCl2 and the washing media included 1% Triton X-100. The purified nuclear pellets were extensively washed with Tris-sucrose-MgCl2 buffer after the last homogenization step to remove the detergent. We confirmed the purity of the nuclei by inspecting them for cytoplasmic remnants using light microscopy. Digestion with Restriction Nuclease. Nuclei were centrifuged and resuspended in 10 mM EDTA, 10 mM Tris, pH 8.0, which increases the nuclear diameter two- to three-fold. The nuclei were then centrifuged and resuspended in buffer containing 100 mM Tris, pH 7.4, 50 mM NaCl, 5 mM MgCl2, 5 mM dithiothreitol (DTT), and 0.02% Nonidet P-40. The concentration of chromatin DNA in the nuclear suspension was determined by dissolving an aliquot in 10 mM Tris, pH 8.0, containing 1% sodium dodecyl sulfate (SDS) and measuring the absorbance at 260 nm (1 mg/ml DNA = 20.0 absorbance units per mg DNA), unless otherwise noted. The samples were 936
Nucleic Acids Research incubated for 7 hr at 370C with 30 units EndoR Alu I per mg DNA (New England Biolabs, Beverly, MA; 7200 units/ml; one unit digests one pg lambda DNA in one hour) along with a control sample lacking enzyme. To monitor the kinetics of the digestion, the nuclei were centrifuged at intervals for 30 sec at low speed in a clinical centrifuge, a 20 1l aliquot of the supernatant was.dissolved in the Tris-SDS buffer, and the absorbance was measured at 260 nm. The nuclei were then resuspended and digestion resumed. Lysis and Extraction of Chromatin. After the digestion was complete, as revealed by a plateau in the release of soluble nucleotides, the nuclear suspension was cooled in ice. The lower molecular weight, soluble chromatin was extracted in a buffer solution containing 10 mM Tris and 10 mM EDTA, pH 8.0, at 40c. The suspension was centrifuged and the EDTA extract removed, leaving a white, gelatinous pellet. The pellet was then resuspended, sheared mildly, and extracted at 40C with 2 mM Tris, pH 8.0, 0.1 mM EDTA, 3 M urea (Schwarz-Mann, Ultra Pure Reagent) by passing it through an 18 gauge needle. The extraction was repeated twice with additional Tris-EDTA-urea buffer to achieve complete removal of high molecular weight chromatin from the nuclear residue. The amount of DNA in each fraction was measured spectrophotometrically using the Tris-SDS method described above. The final Tris-EDTA-urea extract was centrifuged at 2000 x g for 2 min to remove membranous debris and incompletely lysed nuclei. Sucrose Gradient Centrifugation. The Tris-EDTA-urea extract was layered on a 10 to 40% linear sucrose gradient containing 3 M urea to separate the chromatin digested by the restriction nuclease from the higher molecular weight, enzyme-resistant fraction. Chromatin containing approximately 2 mg DNA was added to each 12-ml gradient tube. The tubes were centrifuged in a Beckman SW4OTi rotor at 28,000 rpm for 15 hr; then 40 fractions were collected from the bottom of each tube. The absorbance of each fraction was measured at 260 nm; all appropriate tubes containing DNA were pooled and dialyzed against 10 mM Tris, pH 8.0, 0.25 mM EDTA. Histone Isolation. After briefly sonicating the pooled chromatin fractions, H2S04 was added to a final concentration of 0.4 N and the samples were held at 40C overnight. The DNA was pelleted by centrifugation at 12,000 X g. The supernatant contained the histones and was mixed with six volumes of 95% ethanol. The histones were allowed to precipitate at -150C. The histone precipitate was centrifuged, and the pellet was dried in vacuo and dissolved in 20% sucrose, 0.9 N acetic acid, 0.5 M 2-mercaptoethanol. 937
Nucleic Acids Research Agarose- and Polyacrylamide-Gel Electrophoresis. Agarose-gel electrophoresis was performed using 0.75%, 1% and 1.4% agarose gels (Sigma Chem. Co., low electroendosmosis type) in a vertical slab electrophoresis apparatus with a buffer solution containing 90 mM Tris base, 90 mM boric acid and 2.5 mM EDTA at pH 8.3. Ethidium bromide was added to the buffer at 0.5 ig/ml to make the DNA fragments visible. A partial EndoR EcoR II (Bethesda Research Labs., Rockville, MD) digest of pure mouse satellite DNA was used as a marker of DNA molecular weights with a monomer length of 241 base pairs and higher multiples. The DNA was extracted from chromatin fractions by treatment with 0.5% SDS, 10 mM Tris, 1 mM EDTA, and aliquots were subjected to electrophoresis at 5V/cm. The DNA fragments were examined under two 20-W, short-wave UV
lamps and photographed through Kodak ultraviolet (Wratten 2A) and orange (Wratten 21) filters with Kodak Plus X, 4- by 5-in cut film. The negatives were scanned with a Zeineh soft laser microdensitometer. Histone samples were subjected to electrophoresis in 10 cm Panyim-Chalkley acid urea gels containing 2.5 M urea at 130 V for 3.5 hr. The gels were stained in 0.1% Naphthol blue black, 0.9 N acetic acid, 30% ethanol and were destained electrophoretically. Analytical Centrifugation. Samples were centrifuged to equilibrium in neutral CsCl buoyant density gradients in an AnG rotor at 40,000 rpm for 24 hr with a Beckman Model E ultracentrifuge equipped with a multiplexer and ultraviolet scanner. We estimated the proportions of mouse satellite and main band DNA sequences in the enriched fractions by scanning with a microdensitometer and cutting out the areas under the peaks and weighing them. Protein Assay. Protein content was assayed with the dye reagent Coomassie Brilliant Blue G250 (Bio-Rad). This is a microassay procedure15 in which 50 iI of dye reagent is added to 200 pl of protein solution. After samples are mixed and held for 2 min for maximum stable dye binding, we measured the absorbance of the samples at 595 nm against a reagent blank. The reference standard was bovine gamma globulin (Bio-Rad) at a concentration of 1.36g/ml. The absorbance of gamma globulin was compared to that of purified mixed histones and the chromatin data were calculated in terms of the amount of histone. Fixation with Formaldehyde and Glutaraldehyde. We dialyzed the chromatin fraction for 24 hr at 40C against 5mM sodium phosphate buffer, pH 6.8, 0.2mM EDTA with 1% formaldehyde (Baker) added. We continued the
938
Nucleic Acids Research dialysis for
additional 24 hr against a large volume of formaldehyde-free buffer.16 A similar regimen was followed using glutaraldehyde in place of formaldehyde. E.M. grade glutaraldehyde (Polysciences, Inc.) was diluted to 1% from the 8% stock solution. Gradient centrifugation of the fixed chromatin samples in CsCl was performed as described above. The initial density was po = 1.400 g/ml, and we calculated peak densities by the isoconcentration point method using equations described by Ifft et a117 and a a-value of 1.346. To obtain accurate differences in density between chromatins containing satellite DNA and main band DNA, the two fractions were added to the same cell in mass ratios sufficiently different to identify the peaks unequivocally. From the densities of the chromatin peaks we calculated the weight ratios of protein to DNA in the chromatin samples by the method of Brutlag et al.18 an
RESULTS DNA Digestion.
Three type-II restriction enzymes with tetranucleotide restriction sites containing all four bases are commercially available. All three enzymes, EndoR Alu I, EndoR Mbo I, and EndoR Taq I, digest main band DNA extensively. Undigested main band DNA from mice forms fairly sharp bands in CsCl gradients (Fig. 1B); but after incubation with these restriction endonucleases, the molecular weight decreases considerably and the width of the band in CsCl increases. Only EndoR Alu I, with the restriction site 5'-AGCT-3', failed to digest the satellite DNA (Fig. lC). Prolonged incubation (17 hr) of isolated satellite DNA with excess enzyme failed to increase significantly the width of DNA banded in CsCl. EndoR Mbo I, which recognizes the sequence 5'-GATC-3', digested the satellite DNA only slightly; EndoR Taq I, with the restriction site 5'-TCGA-3', digested satellite DNA to a moderate extent. Agarose-gel electrophoretograms showed that after digestion with EndoR Alu I, the main band DNA fragments varied in size from 180 to 2300 base pairs, with a median length of 450 base pairs; however mouse satellite DNA (data not shown) had a molecular weight greater than 8500 base pairs (5.5 x 106 daltons). Chromatin Digestion and Fractionation. The procedures used to digest mouse nuclei and extract chromatin fractions after incubation with EndoR Alu I are described in the Materials and Methods. Mouse brain nuclei were used in all experiments because they do not contain significant levels of endogenous nucleases (Fig. 1A) or proteases.13 By pretreating nuclei with
939
Nucleic Acids Research
B
Figure 1
Analytical CsCl buoyant density centrifugation of mouse DNA and fractions. A. Whole mouse DNA after incubation of brain nuclei at 370C for 17 hr without enzyme. B. Isolated main band DNA incubated with 7 units EndoR AluI enzyme for 17 hr. C. Isolated satellite DNA incubated with 7 units EndoR AluI enzyme for 17 hr.
the EDTA buffer we could reduce the amount of enzyme required to digest chromatin; perhaps because the nuclear volume expands and surrounding membrane stretches. After we added the incubation solution and enzyme, the nuclear volume seemed to return to normal. Nuclei were then digested with restriction endonuclease for 7 hr at 370C. The chromatin fractions were isolated from incubated nuclei by sequential extraction with Tris-EDTA and Tris-EDTA-urea, as described in the Methods section. The DNA of the 940
Nucleic Acids Research chromatin fractions was characterized using both analytical CsCl buoyant
density gradients and agarose gel electrophoresis. Approximately two-thirds of the chromatin of brain nuclei is extracted with 10 mM Tris, 10 mM EDTA. In CsCl (Fig. 2A) and Cs2SO4 buoyant density gradients, this fraction contained only main band DNA. In agarose gels the DNA ranged in size from 250 to 3600 base pairs with a median size of 1900 base pairs (1.2 x 106 daltons). Further extractions of the nuclei with 2 mM Tris, 0.1 mM EDTA, 3 M urea, and mild shearing through an 18-gauge needle, removed an additional 15 to 25% of the chromatin. The DNA extracted from this fraction contained from 24 to 80% of the satellite DNA (Fig. 2B), depending on the amount of restriction enzyme added (this varied from 28 to 224 units per mg DNA). In agarose gels, the DNA fragments in this fraction ranged in size from 2200 to greater than 10,000 base pairs.
The total recovery of chromatin ranged from
80 to 90%. The chromatin fraction extracted with Tris-EDTA-urea was fractionated further on a linear sucrose gradient. Over 90% of the chromatin DNA that
Agarose gels (0.7%)
Neutral CsCI gradients EDA xtract
A. TRIS-UREA extract
B.
Mouse sat.
ECORI Figure 2
1
1 110
Characterization of the EDTA extract and Tris-urea extract of EndoR Alu I digested mouse brain nuclei by agarose electrophoresis and analytical buoyant density centrifugation in CsCl. A. DNA extracted from chromatin soluble in the EDTA extract.
B.
DNA extracted from chromatin soluble in the
Tris-urea extract. DNA was extracted from the chromatin fractions and electrophoresed in 0.7% agarose gels as described in Materials and Methods. The gels were photographed and the negatives scanned in a microdensitometer. A standard EcoRII digest of mouse satellite DNA was used as a molecular weight marker. Analytical CsCl gradients of the DNA were centrifuged as described in Materials and Methods and scanned at 260 nm. 941
Nucleic Acids Research was placed on the gradient was recovered. A pellet at the bottom of the tube contained the remaining DNA along with membrane fragments and a few unbroken nuclei. The DNA in this pellet cannot be recovered except under degrading conditions. Representative sucrose-gradient fractions were analyzed in CsCl gradients and agarose gels (Fig. 3). Near the bottom of Bottom
2.0
Top
F
0
1o 1.1
A
0 )
Figure 3
942
5
10
15
20 : Fractions
35
40
Sucrose gradient sedimentation of Tris-urea extract on a sucrose-urea gradient: analysis of isolated DNA in agarose gels and in CsCl gradients. DNA isolated from pooled chromatin fractions (shown by the different degrees of shading) was submitted to gel electrophoresis in 1.7% agarose and buoyant density analysis as described in Materials and Methods. The agarose gels were photographed and the resulting negatives were scanned for the DNA profiles. The black-shaded scan is a marker DNA of an EndoR Hae III digest of guinea pig satellites 2 & 3. Analytical CsCl buoyant density centrifugation was used to detect the fraction most enriched in satellite DNA.
Nucleic Acids Research the sucrose gradient more than 95% of the DNA was satellite DNA. However, only 8 to 10% of the total satellite DNA present could be recovered with
this high purity. In the middle and upper portions of the gradient the satellite DNA content ranged from 75 to less than 50%, and the remaining DNA consisted of the higher molecular weight portions of the main band DNA chromatin. Protein Composition Total protein and DNA were assayed in pooled fractions taken across a sucrose gradient of a Tris-urea extract of nuclei digested with restriction nuclease (Table 1A). The ratio of protein to DNA in these fractions was 1.3 + 0.1, except at the top of the gradient where extraneous nuclear proteins not bound to chromatin would remain. Thus, the data show that satellite DNA chromatin and main band DNA chromatin isolated from a sucrose gradient have the same protein to DNA ratio. To distinguish tightly bound proteins from those merely associated with the hydrodynamic particle, we crosslinked the chromatin fractions with formaldehyde and glutaraldehyde. After removing excess crosslinking agent by dialysis, the fractions were banded in CsCl gradients at po = 1.40 g mlV1. The protein to DNA ratios were calculated (Table 1B) according to
Table 1 Ratios of protein to DNA (by weight) in chromatin fractions Sucrose gradient fractions of Tris-EDTA-urea extract
A.
Gradient fraction
1 2 3
4 5
B.
Protein Species
DNA
1.38 1.21 1.38 1.21 1.76
Predominantly satellite DNA chromatin Predominantly satellite DNA chromatin Main band DNA chromatin, slight contamination with satellite DNA chromatin Main band DNA protein plus non-associated protein
Chromatin fractions crosslinked with aldehydes.
Chromatin fraction Main band
Crosslinker
Protein DNA
Formaldehyde
0.99 1.00 1.14
Glutaraldehyde
Satellite
Formaldehyde Glutaraldehyde
1.05
943
Nucleic Acids Research the method of Brutlag et al.18 The histone compositions of satellite DNA chromatin and main band DNA chromatin are compared in Fig. 4. Both types of chromatin contain all five major histones and the minor lysine-rich histone H10 found in undigested chromatin. The presence of Hl and the absence of characteristic peptides produced by degradation of HI and the core histones indicate that these chromatin preparations are intact and that histone proteolysis must be minimal.
DISCUSSION We developed a method by which, under gentle conditions, chromatin can be fractionated into satellite DNA-containing and satellite DNA-free fractions. This approach makes use of a common property of many satellite DNAs, i.e., one deoxyribonucleotide base is deficient in each strand. We found that the restriction endonuclease EndoR Alu I, which requires all four
bases in its restriction site, will degrade main band DNA chromatin while leaving satellite DNA chromatin intact. The two types of chromatin are separated by extracting enzyme-treated cell nuclei with specific buffers and then sedimenting these extracts in sucrose gradients. !
d;
) c'I
t i1
Sdtfdlt~? cfh,rron),tin1
Figure 4
944
t
Histones of mouse main band and satellite DNA chromatins. The histones were extracted with sulfuric acid and electrophoresed in acrylamide gels at 130 V for 3.5 hr as described in Methods.
Nucleic Acids Research The successful application of this procedure requires that the restriction endonucleqse be free of contaminating nucleases and the nuclei contain only minimal endogenous nucleolytic and proteolytic activities. The EndoR Alu I, as commercially available, did not contain significant proteolytic or nonspecific nucleolytic activities, and we could add large amounts of enzyme and prolong the required digestion period. In an extensive survey of mouse tissues we showed that only brain nuclei are sufficiently low in endogenous nucleolytic and proteolytic activities to provide chromatin fractions that are degraded only by the exogenous restriction endonuclease.13 Although brain tissue contains acid DNAse and cathepsin D activities,19 these enzymes are evidently not localized in the nucleus. A further requirement to fractionate chromatin successfully by the enzymatic procedure is to recover chromatin quantitatively from the nuclei after digestion. Two thirds of the total chromatin is readily released into buffer after digestion in the presence of mild nonionic detergent. The remaining chromatin, including all satellite DNA chromatin, is more tightly bound to the nuclear membranes or matrix or both. Almost all of this chromatin was released upon extraction in the presence of 3 M urea. This concentration of urea is sufficient to disrupt hydrogen bonding and apolar interactions between the higher molecular weight portion of main band DNA chromatin or the satellite DNA chromatin and the nuclear residue; further, the urea seems to minimize aggregation of high molecular weight chromatin during sucrose gradient sedimentation.20 Yet 3 M urea is below the concentrations that irreversibly disrupt native chromatin.21 The products of successive digestion by restriction endonuclease, buffer extraction, and sucrose gradient sedimentation are very pure fractions of satellite DNA chromatin and main band DNA chromatin, which differ primarily in their molecular weight distributions. The protein to DNA ratios of the
suggesting that the principal protein constituents are the histones. Crosslinking studies with formaldehyde and glutaraldehyde yielded lower ratios, which were slightly higher for satellite DNA chromatin than for main band DNA chromatin. This is consistent with the hypothesis that only the histones, and possibly small amounts of nonhistone proteins derived from the nuclear membrane or the closely associated with the DNA that they can chromosomal "scaffold", are be bound by crosslinking agents.22 Electrophoretic analyses of whole two fractions are very similar (Table
1),
so
histone isolated from the two chromatins demonstrate that all five histones Additional experiments not are present and protein degradation is minimal.
945
Nucleic Acids Research reported in detail here, however, indicate that there are significant
differences in the histone compositions of the two types of chromatin.
ACKNOWLEDGMENT This work was performed under the auspices of the U.S. Department of Energy by the Lawrence Livermore Laboratory under contract number W-7405-ENG-48. Reference to a company or product name does not imply approval or
recommendation of the product by the University of California or the U.S. Department of Energy to the exclusion of others that may be suitable.
REFERENCES 1 2 3 4 5 6
7 8
9 10 11 12 13
14
15 16 17 18
19 20 21 22
946
Heitz, E. (1928) Jahrb. Wiss. Bot. 69, 762-769. Brown, S.W. (1966) Science 151, 417-425. Gosden, J.R. and Mitchell, A.R. (1975) Exp. Cell Res. 92, 131-137. Frenster, J.H. (1969) Handbook of Molecular Cytology, Lima de Faria, A., Ed., London North Holland Publish Corp., p. 251. Mattoccia, E. and Comings, D. (1971) Nature New Biol. 229, 175-176. Pfeiffer, W., Horz, W., Igo-Kemenes, T., and Zachau, H.G. (1975) Nature 258, 450-452. Lipchitz, L. and Axel, R. (1976) Cell, 9, 355-364. Musich, P.R., Brown, F.L., and Maio, J.J. (1977) Proc. Natl. Adad. Sci. USA 74, 3297-3301. Igo-Kemenes, T., Greil, W., and Zachau, H.G. (1977) Nucleic Acids Res. 4, 3387-3400. Musich, P.R., Brown, F.L., and Maio, J.J. (1977) Cold Spring Harbor Symp. Quant. Biol. 42, 1147-1160. Bernstine, E.G. (1978) Exp. Cell. Res. 113, 205-208. Mazrimas, J.A., Balhorn, R. and Hatch, F.T. (1978) J. Cell Biol. 79, CH661. Balhorn, R., Weston, S., Mazrimas, J.A., and Young, T. (1978) J. Cell Biol. 79, CH669. Panyim, S., Bilek, D. and Chalkley, R. (1971) J. Biol. Chem. 246, 4206-4215. Bradford, M.M. (1976) Anal. Biochem. 72, 248-254. Doenecke, D. (1978) Hoppe-Seyler's Z. Physiol. Chem. 359, 1343-1352. Ifft, J.B., Voet, D.H., and Vinograd, J. (1961) J. Physical Chem. 65, 1138-1145. Brutlag, D., Schlehuber, C., and Bonner, J. (1969) Biochemistry 8, 3214-3218. Koenig, H. (1974) Meth. Ezymol. 31A, 457-477. Chalkley, G.R. and Jensen, R.H. (1968) Biochemistry 7, 4380-4388. Olins, D.E., Bryan, P.N., Harrington, R.E., Hill, W.E., and Olins, A. (1977) Nucleic Acids Res. 4, 1911-1931. Doenecke, D. and McCarthy, B.J. (1975) Biochemistry 14, 1373-1378.
