Biochemistry 1982, 21, 4999-5009 Loeb, G. I., & Saroff, H. A. (1964) Biochemistry 3, 18 19-1 826. Markley, J. L. (1975) Biochemistry 14, 3546-3553. Matthew, J. B., Friend, S. H., Botelho, L. H., Lehman, L. D., Hanania, G. I. H., & Gurd, F. R. N . (1978) Biochem. Biophys. Res. Commun. 81, 416-421. Matthew, J. B., Hanania, G. I. H., & Gurd, F. R. N. (1979a) Biochemistry 18, 1919-1928. Matthew, J. B., Hanania, G. I. H., & Gurd, F. R. N. (1979b) Biochemistry 18, 1928-1936. Matthew, J. B., Friend, S . H., & Gurd, F. R. N. (1981a) Biochemistry 20, 57 1-580. Matthew, J. B., Friend, S . H., & Gurd,-F. R. N. (1981b) Interactions between Iron and Protein in Oxygen and Electron Transport, ElsevierjNorth-Holland, New York. Matthews, C. R., & Westmorland, D. G. (1973) Ann. N.Y. Acad. Sci. 222, 240-253. Meadows, D. H., Jardetsky, O., Epand, R. M., Ruterjans, H. H., & Scheraga, H. A. (1968) Proc. Natl. Acad. Sci. U.S.A. 60, 766-772. Nagasawa, M., & Holtzer, A. (1964) J . Am. Chem. SOC.86, 5 3 1-5 38. Niu, C. H., Matsuura, S., Shindo, H., & Cohen, J. S. (1979) J . Biol. Chem. 254, 3788-3796.
4999
Richards, F. M. (1982) Brookhaven Symp. Biol. No. 32. Richards, F. M., & Wyckoff, H. W. (1973) Atlas of Molecular Structure I . Ribonuclease-S, Oxford University Press, Ely House, London. Richarz, R., & Wiithrich, K. (1978) Biochemistry 1 7 , 2263-2269. Saroff, H. A., & Carroll, W. R. (1962) J . Biol. Chem. 237, 33 84-3 387. Shindo, H., & Cohen, J. S . (1976) J . Biol. Chem. 251, 2648-2652. Shire, S . J., Hanania, G. I. H., & Gurd, F. R. N. (1974) Biochemistry 13, 2967-2974. Tanford, C., & Kirkwood, J. G. (1957) J . Am. Chem. SOC. 79, 5333-5339. Tanford, C., & Roxby, R. (1972) Biochemistry 11, 2 192-2 195. Wall, F. T., & Berkowitz, J. (1957) J . Chem. Phys. 26, 114-122. Walters, D. E., & Allerhand, A. (1980) J . Biol. Chem. 255, 6200-6 204. Wyckoff, H. W., Tsernoglou, D., Hanson, A. W., Knox, J. R., Lee, B., & Richards, F. M. (1970) J . Biol. Chem. 245, 305-328.
Torsional Motion and Elasticity of the Deoxyribonucleic Acid Double Helix and Its Nucleosomal Complexest I. Hurley, P. Osei-Gyimah, S. Archer, C. P. Scholes, and L. S. Lerman*
ABSTRACT:
Torsional thermal oscillations of the DNA double helix within the electron paramagnetic resonance (EPR) time scale (10-10-10-3 s) as indicated by a rigid, intercalating probe are much smaller in the spacer segment between nucleosomes in chromatin than in long, free DNA molecules. Still smaller DNA oscillation is indicated in intact nuclei and yet smaller if the nuclei have been treated with glutaraldehyde. The values of EPR measurements are not affected by the loading density of probe. If the probe were capable of substantial oscillations or movement different from that of the helix, those oscillations would be expected to dominate the spectra when movement of the helix is restrained. We conclude that the correlation time for torsional movement of free DNA inferred from EPR spectra is characteristic of the double helix and that there is
no significant independent motion of the probe. The correlation time for the DNA double helix in molecules longer than approximately 500 base pairs is close to 30 ns, corresponding to an elastic constant of 1.5 X ergs cm for deformation by twisting. The motions observed in chromatin are consistent with a model in which spheres of 50-60-A radius are connected by simple elastic rods with the length of spacer DNA and the same elastic constant. The spin-labeled ethidium probe has been characterized in detail by nuclear magnetic resonance, infrared, fluorescence, and visible light spectroscopy. The binding equilibria are consistent with the hypothesis that strongly immobilized probe molecules are preferentially bound to spacer DNA.
R e v i o u s work in this laboratory (Robinson et al., 1979) has shown the rotational correlation time ( T J reported by electron paramagnetic resonance (EPR) spectra of bound spin-labeled ethidium depends upon the size of the DNA molecule with which it is associated. The inferred T , was found to increase with increasing DNA length up to a limiting value of 30 ns
for double helices 500 base pairs (bp) or longer. These results were seen to be in agreement with the predictions of a model of DNA torsional motion in which the double helix is modeled as a set of coaxial disks linked with torsional springs immersed in a viscous medium (Robinson et al., 1980b). An alternative conjectural explanation for the observed dependence upon length might be that the correlation time increases with increasing length up to a limit imposed by hypothetical independent motion of spin-labeled ethidium relative to the DNA base pairs to which it is bound and does not truly reflect the internal motion of long DNA helices. We will refer to the conjectural independent motion as wobble. We have now obtained EPR spectra of spin-labeled ethidium bound to DNA in systems in which the motion of each DNA segment is restricted. An approximate upper limit to the extent
From the Department of Biological Sciences, State University of New York at Albany, Albany, New York 12222 (I.H. and L.S.L.), the Department of Chemistry, Rensselaer Polytechnic Institute, Troy, New York (P.0.-G. and S.A.), the Department of Physics, State University of New York at Albany, Albany, New York 12222, and the Center for Biological Macromolecules, State University of New York at Albany, Albany, New York 12222 (I.H., C.P.S., and L.S.L.). Received March 16, 1982. This work was supported by Grants PCM 772558304 and PCM 81 11321 from the National Science Foundation.
0006-2960/82/0421-4999$01.25/0
0 1982 American Chemical Society
5000 B I o c H E M I s T R Y of wobble is established in this paper, much smaller than the motion characterized by the 30-11s correlation time. It is shown that the motional and elastic properties of DNA inferred from our earlier measurements (Robinson et al., 1980a) are essentially correct.
HURLEY ET AL.
molecular weight chicken erythrocyte DNA (ceDNA) was prepared by a conventional procedure from chicken erythrocytes obtained by cardiac puncture. The A200/A260 ratio of solutions of DNA was typically 1.8 in 0.1 M NaC10,-1 mM sodium phosphate, pH 6.8, indicating the absence of complexed proteins. DNA solutions exhibited recoil when swirled and Experimental Procedures would not penetrate a 60 mg/mL polyacrylamide gel under ( I ) Preparation of Compounds. Melting points were deelectrophoresis. DNA was stored in high-salt buffer (2 M termined with a Laboratory Devices Melt-temp apparatus and NaCl or 2 M NaClO,) and dialyzed against the sample buffer corrected. Infrared spectra were taken on a Perkin-Elmer 621 immediately before use. Concentrations of ceDNA solutions spectrophotometer. The chemical analyses were performed were determined by absorbance at 260 nM. by Instranal Laboratory, Rensselaer, NY. Analytical results ( 6 ) Chicken Erythrocyte Chromatin. Soluble HI- and for the elements indicated were within 0.4% of the theoretical H5-depleted chromatin was prepared from chicken erythrocytes values. 'H N M R spectra of the diluter samples (containing by the method of Lutter (1978) as modified in Olins' labo8 mg or less of solute) were obtained with a Bruker WH-90 ratory (A. E. Paton, personal communication). After the MHz FT NMR and those of more concentrated samples with multinucleosomal fraction from Sepharose 4B column chroa Varian EM 360 60-MHz NMR. matography was concentrated, the concentrate was dialyzed (a) 1 -Oxy-2,2,5,5-tetramethyl-3-pyrroline-3-carboxylic Acid against 0.2 mM Na2EDTA-0.1 mM phenylmethanesulfonyl Chloride. This compound was prepared according to the fluoride (PMSF), pH 6.8. Chromatin fragments were sized method of Rozantzev (1970). An ice-cold stirred suspension on a 25 pg/mL polyacrylamide gel after proteinase K diof 1-oxy-2,2,5,5-tetramethyl-3-pyrroline-3-carboxylic acid gestion. Traces were found of material as small as 3-somes, (0.368 g, 0.002 mol) in dry benzene containing excess pyridine but the bulk of the DNA ran together as a single band near was reacted for 1 h with 0.22 mL (0.003 mol) of freshly the gel top, with a mobility 10% of that of monosomal DNA, distilled thionyl chloride added dropwise. The filtrate from as expected of DNA from long chromatin fragments (210the reaction mixture was evaporated in vacuo below 40 "C. somes) (No11 et al., 1975). Chromatin samples recovered from The residual oil, 0.41 g, was used without purification. solutions used for EPR spectroscopy were redigested and an( b ) 3-Amino-8-[ [ ( I -oxy-2,2,5,5-tetramethyl-3-pyrrolin-3- alyzed by gel electrophoresis. A ladder pattern characteristic yl)carbonyl]amino]-6-phenyl-5-ethylphenanthridinium of extended chromatin was obtained, indicating that the maChloride (Spin-Labeled Ethidium). To a mixture of dry terial was well-ordered. Concentrations of chromatin solutions ethidium bromide (0.55 g, 0.0014 mol) and 1.2 mL (0.0015 were determined spectrophotometrically. mol) of dry pyridine in 5 mL of dry N,N-dimethylformamide (c) Chicken Erythrocyte Nuclei. Chicken erythrocyte nuclei (DMF), a solution of the acid chloride (0.41 g, 0.002 mol) in were isolated following the protocol of Olins & Olins (1979), 2 mL of DMF was added dropwise at room temperature. avoiding the use of reagents such as nonionic detergents, high After the mixture was stirred overnight (16 h) the solvent was salt, and EDTA, believed to disrupt nuclear structure (Agutter removed by vacuum distillation. The oily residue was dissolved & Richardson, 1980). Phase-contrast micrographs of isolated in 8 mL of 1:l MeOH-H20 and refluxed with AgCl 4 h to nuclei confirmed the presence of detailed structures as reported replace Br- with C1-. The residual red oil obtained by vacuum by Olins & Olins (1979). After separation of intact nuclei evaporation of filtrate was chromatographed on preparative in a sucrose step gradient and washing with TKM buffer (20 TLC plates (silica gel) by using EtOH-CHCl, (2:l) as the mM Tris-HC1-20 mM KC1-5 mM MgC12-250 mM sucrose, eluant. In order of decreasing Rf,distinguishing color bands pH 7) a portion of the nuclear pellet was fixed in glutarof yellow, peach, red, and pink appeared on the plates. The aldehyde (3%, in 20 mM sodium cacodylate-5 mM Mgred band, which consisted of the desired compound, was recl2-250 mM sucrose, pH 7, for 2 h) and washed 4 times with moved from the plate, powdered, and extracted with dry a Tris-containing buffer (20 mM Tris-HC1-5 mM MgC12-250 MeOH. The purification procedure using preparative TLC mM sucrose, pH 7) to react with any remaining aldehyde plates was repeated until the red band appeared as a single groups. The remaining portion of the nuclei was washed with spot on the TLC plate. The compound was obtained as a solid TKM buffer and stored in a 1:1 TKM buffer-glycerol mixture red residue, 80 mg: mp 267-269 OC; IR (KBr) 3400 (NHat -20 "C until used. Unfixed nuclei were always used within C=O), 1660 cm-' (amide carbonyl). Anal. Calcd for 4 days of preparation. Concentrations of DNA in nuclear C30H3202N4C1~H20: C, 67.46; H, 6.04; N, 10.49. Found: C, preparations were estimated spectrophotometrically following 67.83; H, 6.11; N, 10.20. lysis and treatment with proteinase K. (c) 3-Amino-8-[[ ( l-hydroxy-2,2,5,5-tetramethyl-3( 3 ) Nuclease Digestion. Freshly prepared chicken erythp y r r o l i n - 3 - y l ) c a r b o n y l ]a m i n o ] - 6 - p h e n y l - 5 - e t h y l rocyte nuclei were washed several times with buffer (20 mM pkenanthridinium Chloride (Reduced Spin-Labeled EthidiTris-HC1-5 mM MgC1,-250 mM sucrose, pH 7). Most of mol) of spin-labeled um). A total of 12.9 mg (2.4 X the pellet was resuspended by pipetting a few times after each ethidium in 2 mL of dry DMF was added to 1 mL of butyl wash. The clumped portion of nuclei was discarded. A total mercaptan, and the reaction mixture was stored at room of 300 units of micrococcal nuclease (Worthington) was added temperature for 5 days. Volatile components were removed to the final nuclear suspension (2 mL containing 5.1 mg of by vacuum evaporation. The oily residue was taken up in 0.5 DNA), and the suspension was incubated for 1 h at 0 O C . At mL of dry Me,SO and reevaporated. The red solid was disthe end of the hour, 20 pL of a 1 mg/mL solution of spinsolved in 0.5 mL of dry MeOH and chromatographed on an labeled ethidium was added, and the nuclei were pelleted by analytical TLC plate (silica gel) as before. The principal band centrifugation. A portion of the moist nuclear pellet was was removed, powdered, and extracted with dry methanol. The loaded into an EPR tissue cell; the rest was lysed immediately extract was evaporated to dryness in vacuo: 6 mg of red by resuspension and then addition of Na,EDTA, NaDodS04, powder; IR (KBr) 3320 (NOH), 3400 cm-' (NH=CO); and proteinase K. The course of the room temperature diN M R 6 -1.0-1.6 (m, 13 H), -6.6 (8, 1 H), -8.3-8.8 (m, -1 gestion of the sample in the tissue cell was followed by EPR. H). The DNA in samples of nuclei before and after digestion was analyzed by UV spectroscopy and gel electrophoresis. (2) Preparation of Biological Materials. ( a ) DNA. High
TORSIONAL MOTION A N D ELASTICITY OF D N A
~1 l
-10,O -9.0
,
l
,
l
-8.0 -70
,
l
-6.0
,
l
-5.0
,
-4.0
l
DMSO
~
-3.0
l
,
l
,
-2.0 -1.0
l
,
l J ,
0.0 1.0
PPM
FIGURE 1: Digitized 'HN M R spectra of (a) ethidium bromide, (b) reduced spin-labeled ethidium chloride, (c) 8-acetylethidium chloride in dimethyl-d, sulfoxide (DMSO). Spectrum a was taken by using a 60-MHz conventional NMR; spectra b and c were taken by using a 90-MHz Fourier transform NMR. The large peak at -2.4 ppm is due to proton impurities in Me2S0. The large peak between -3.0 and -4.0 ppm is due to water.
Results (1) Characterization of Spin-Labeled Ethidium. Spectral studies to confirm the structure of spin-labeled ethidium were essential because of the poor yield of product (-10%). Spin-labeled ethidium's NMR spectrum is uninterpretable due to the unpaired electron. Reduction by butyl mercaptan (see Experimental Procedures, section IC) lowered unpaired spin concentration below detectability by EPR (
