9276
Biochemistry 1994, 33, 9276-9284
Effects of Various Salts and pH on the Stability of the Nucleosome in Chromatin Fragmentst Xirong Ni and R. David Cole' Department of Molecular and Cell Biology, University of California, Berkeley, California 94720 Received April 1 1 , 1994; Revised Manuscript Received May 27, 1994'
ABSTRACT: The stability of nucleosomes in long chromatin fragments was observed by differential scanning
calorimetry over a wide range of solution conditions. The thermal denaturation of chromatin was characterized in general as three major transitions, although the process clearly is more complex. The three major transitions were (1) denaturation of the nucleosome, (2) base unstacking of DNA in the resulting denatured nucleoprotein, and (3) base unstacking of naked DNA. In very low salt concentrations (e.g., 2 mM sodium cacodylate), these three processes were essentially coincident (near 76 "C), but in medium salt concentrations (e.g., 100 mM NaC1) the nucleosome denaturation occurred first at about 69 OC and then base unstacking occurred at 85 OC. As [NaCl] was increased, all three processes were resolved with the observation of increasing amounts of naked DNA being melted, until at 2000 mM NaCl the calorimetric profile showed mainly the melting of DNA. The transition temperature for nucleosome denaturation decreased from 76 to 63 OC as the salt concentration increased from 1 to 600 mM. Destabilization of the nucleosome by increasing [NaCl] was also evident above 100 mM as a decrease in enthalpicchange attributable to nucleosome denaturation. Similarly, as [NaCl] was increased above 100 mM, less and less denatured nucleoprotein was evident as more and more of the DNA melted as naked DNA. The fatty acid salts, sodium valerate and sodium caproate, destabilized the nucleosome but not the denatured nucleoprotein that resulted from the collapse of the nucleosome. In the series acetate, butyrate, valerate, caproate, it was clear that destabilization of the nucleosome increased as hydrophobicity (chain length) increased. Pimelate, with the same number of carbon atoms as caproate but with an extra negative charge, did not destabilize the nucleosome as caproate did. The nucleosome was substantially stabilized by MgCl2 within the range 0.5-2 mM and by spermidine in the range 0.1-3 mM. The transition temperature for DNA in the denatured nucleoprotein was unaffected by spermidine, but was lowered by MgCl2. Spermidine was more effective than MgClz at displacing DNA from the denatured nucleoprotein.
Previously, we reported (Jin & Cole, 1986; Guo & Cole 1989a,b) that the condensation of chromatin was strongly affected by particular ions and by pH in the physiological range. This made clear the importance of paying attention to buffer composition when comparing reports of chromatin function and dynamics. Such factors might also have substantial effects on the structure of the nucleosome when it is contained within chromatin, but only modest attention seems to have been given to that possibility. The work to be presented here was an exploration of salt and pH effects on the stability of the nucleosome in large fragmentsof chromatin. The technique used to observe stability was differential scanning calorimetry because it is applicable to chromatin even in the physiological range of pH and ionic conditions where precipitation occurs. Studies of chromatin at salt concentrations below 10 mM by differential scanning calorimetry (Reczek et al., 1982; Riehm & Harrington, 1987) or other thermal denaturations have been reported before, as well as similar studies on core particles (Weischet et al., 1978; Simpson & Shindo, 1979; Bryan et al., 1979; Bina et al., 1980; Cowman & Fasman, 1980; McGhee & Felsenfeld, 1980), nucleosomes (Simpson, 1978), and whole nuclei. At higher salt concentrations,in the physiological range, we investigated whole nuclei (Touchette & Cole, 1992) and chromatin (Almagor & Cole, 1989). Our This work was supported in part by a research grant (GMS 20338) from the National Institutes of Health and by the Agricultural Research Station. "Abstract published in Advance ACS Abstracts. July 15, 1994.
0006-2960/94/0433-9276$04.50/0
earlier studies on isolated chromatin revealed two major structural transitions as the temperature was raised from 25 to 95 OC. The first transition was clearly shown to be the collapse of the nucleosome. The second transition was at the temperature of base unstacking of naked, relaxed DNA, but as was pointed out then, it was not clear whether the DNA was free or still bound to proteins. The present results clarified that issue by using a wide range of ionic conditions, to which nucleoprotein and naked DNA responded differently. METHODS AND MATERIALS Cell Culture. HeLa cells, strain S3, were maintained in suspension culture at cell densities of (2-8) X 105/mL in Joklik's modified spinner medium supplemented with 5% calf serum. Cells were harvested at cell densities of (5-6) X lo5/ mL. Isolation of Nuclei. All steps were done at 0-4 "C, and all centrifugation was at 300g for 3 min. HeLa cells were pelleted and washed in swelling buffer (0.1 M hexylene glycol, 1 mM CaC12, and 0.06 mM PIPES,' pH 6.8) with 0.25 mM PMSF, which was added just before use. Then the cells were resuspended in swelling buffer with 1 mM PMSF for 10-15 min at a density of (1.3-2.6) X 107/mL. Cells were disrupted with 10 strokes of a loose-fitting Dounce homogenizer. The nuclei were washed three times in the same swelling buffer, two times in buffer A (50 mM Tris, 25 mM KCl, 0.3 M Abbreviations: PIPES, 1,4-piperazinediethanesulfonicacid; PMSF, phenylmethanesulfonyl fluoride; SDS, sodium dodecyl sulfate.
0 1994 American Chemical Society
Scanning Calorimetry of Chromatin sucrose, and 5 mM MgC12, pH 6.5) with 0.1% (v/v) Triton X- 100 (Boehringer, purified for membrane research), and three times in buffer A without Triton. All washing buffers contained 1 mM PMSF, which was added just before use. The water used was Milli-Q water from Ultra-Pure Water System (resistance above 15 MSZ-cm). Isolation of Chromatin. For large chromatin fragments, nuclei were resuspended in buffer A with 1 mM PMSF at A260 = 150. After preincubation at 37 OC for 5 min, micrococcal nuclease (Worthington) was added at 200 units/ mL and CaCl2 was added to 1 mM for additional incubation at 37 OC for 225 s. The reaction was quenched with cold 200 mM EDTA (pH 6.5) (final concentration, 10 mM) and kept on ice. The pellets were resuspended in 2 mM EDTA (pH 7.0) on ice and allowed to lyse for 2 h, with occasional gentle mixing. Chromatin was recovered in the supernatant after centrifugation, and its concentration was measured assuming A260 = 270 for 1% DNA in 0.1 M NaOH. The solution was stored at -80 "C. For medium or short chromatin fragments, the procedure was the same except that nuclease digestion was done at an A26o of 100, and the enzyme incubation was micrococcal nuclease of 100 units/mL. Digestion was for 30 or 60 min for medium or short fragments, respectively. Extraction of DNA and DNA Electrophoresis. Chromatin samples diluted 6-fold weredeproteinized in an aqueous phase containing 30 mM Tris (pH 7.8), 1% SDS, 0.01% proteinase K, and 1 M NaCl and then extracted twice with equal volumes of chloroform/isoamyl alcohol (24:1, v/v). The DNA was precipitated for 30 min in 2.5 vol of ethanol at -20 "C, centrifuged at 23000g for 15 min, and washed with 80% cold ethanol. The DNA was dried under vacuum for 10 min and then dissolved in electrophoresis buffer (0.04 M Tris, 0.02 M sodium acetate, and 1 mM Na2EDTA, pH 7.2) to make the DNA concentration 5 pg/pL. DNA fragments were separated in electrophoresisbuffer on 2%, 1.5%, and0.75% (w/v) agarose gels for short, medium, and long sizes, respectively. The gels were stained with ethidium bromide and photographed. Alternatively, the DNA gels were stained with Stains-All (Kodak) solution and scanned with an LKB 2202 ultroscan laser densitometer. For the Stains-All procedure, it is necessary to do pre-electrophoresis; 54 V for 3 h before loading the DNA samples. Relative amounts of fragments of different size were determined from the areas of peaks observed by scanning. Typical preparations of long, medium, and short chromatin fragments are represented, respectively, by Figure 1A,B,C. The percentages of fragments that were pentamers of nucleosomesor shorter were 14%for Figure 1A, 40% for Figure lB, and 76% for Figure 1C. Naked DNA used for calorimetry was prepared from chromatin by phenol extraction as above but scaled up, and the DNA was finally dissolved in water instead of electrophoresis buffer. Protein Electrophoresis. Chromatin samples were analyzed for histonesby 13.5% SDS-polyacrylamide gel electrophoresis, essentially according to Laemmli (1970). Quantitative measurement of histones was done by scanning densitometry of Coomassie-stained gels using color factors for the individual histones as published previously (Ring & Cole, 1979). The amount of H1 histone ranged from 0.75 to 0.88 molecules per octamer of core histones. Differential Scanning Calorimetry. Samples used for calorimetry were dialyzed against buffer (1 5 mM PIPES and 1.5 mM NaN3, pH 7.0) at 4 OC for 4 h. The volume ratio
Biochemistry, Vol. 33, No. 31, 1994 9277
Migration
