Portal de Periódicos da CAPES

Nucleosome Structural Transition during Chromatin Unfolding Is Caused by Conformational Changes in Nucleosomal DNA

1998; Elsevier BV; Volume: 273; Issue: 4 Linguagem: Inglês

10.1074/jbc.273.4.2429

1083-351X

Igor M. Gavin, Sergei I. Usachenko, S.G. Bavykin,

DNA and Nucleic Acid Chemistry

We have recently reported that certain core histone-DNA contacts are altered in nucleosomes during chromatin unfolding (Usachenko, S. I., Gavin I. M., and Bavykin, S. G. (1996) J. Biol. Chem. 271, 3831–3836). In this work, we demonstrate that these alterations are caused by a conformational change in the nucleosomal DNA. Using zero-length protein-DNA cross-linking, we have mapped histone-DNA contacts in isolated core particles at ionic conditions affecting DNA stiffness, which may change the nucleosomal DNA conformation. We found that the alterations in histone-DNA contacts induced by an increase in DNA stiffness in isolated core particles are identical to those observed in nucleosomes during chromatin unfolding. The change in the pattern of micrococcal nuclease digestion of linker histone-depleted chromatin at ionic conditions affecting chromatin compaction also suggests that the stretching of the linker DNA may alter the nucleosomal DNA conformation, resulting in a structural transition in the nucleosome which may play a role in rendering the nucleosome competent for transcription and/or replication. We have recently reported that certain core histone-DNA contacts are altered in nucleosomes during chromatin unfolding (Usachenko, S. I., Gavin I. M., and Bavykin, S. G. (1996) J. Biol. Chem. 271, 3831–3836). In this work, we demonstrate that these alterations are caused by a conformational change in the nucleosomal DNA. Using zero-length protein-DNA cross-linking, we have mapped histone-DNA contacts in isolated core particles at ionic conditions affecting DNA stiffness, which may change the nucleosomal DNA conformation. We found that the alterations in histone-DNA contacts induced by an increase in DNA stiffness in isolated core particles are identical to those observed in nucleosomes during chromatin unfolding. The change in the pattern of micrococcal nuclease digestion of linker histone-depleted chromatin at ionic conditions affecting chromatin compaction also suggests that the stretching of the linker DNA may alter the nucleosomal DNA conformation, resulting in a structural transition in the nucleosome which may play a role in rendering the nucleosome competent for transcription and/or replication. The compaction of DNA into chromatin provides many obstacles for its functioning in eukaryotic nuclei (1Kingston R.E. Bunker C.A. Imbalzano A.N. Genes Dev. 1996; 10: 905-920Crossref PubMed Scopus (403) Google Scholar). The current model of transcription and replication suggests that chromatin must be remodeled in at least two distinct steps (2Kornberg R.D. Lorch Y. Cell. 1991; 67: 833-836Abstract Full Text PDF PubMed Scopus (109) Google Scholar, 3van Holde K.E. Lohr D.E. Robert C. J. Biol. Chem. 1992; 267: 2837-2840Abstract Full Text PDF PubMed Google Scholar, 4Paranjape S.M. Kamakaka R.T. Kadonaga J.T. Annu. Rev. Biochem. 1994; 63: 265-297Crossref PubMed Scopus (319) Google Scholar, 5Felsenfeld G. Boyes J. Chung J. Clark D. Studitsky V. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 9384-9388Crossref PubMed Scopus (155) Google Scholar). The first step involves decondensation of chromatin fibers facilitating the access of trans-acting factors to DNA. The second step implies either a reversible dissociation of the histone octamer from DNA or a structural transition of the nucleosome allowing transcription or replication to proceed (3van Holde K.E. Lohr D.E. Robert C. J. Biol. Chem. 1992; 267: 2837-2840Abstract Full Text PDF PubMed Google Scholar, 6Thoma F. Trends Genet. 1991; 7: 175-177Abstract Full Text PDF PubMed Scopus (59) Google Scholar, 7Gruss C. Sogo J. Bioessays. 1992; 14: 1-8Crossref PubMed Scopus (44) Google Scholar, 8Morse R.H. Trends Biochem. Sci. 1992; 17: 23-26Abstract Full Text PDF PubMed Scopus (56) Google Scholar, 9Adams C.C. Workman J.L. Cell. 1993; 72: 305-308Abstract Full Text PDF PubMed Scopus (139) Google Scholar, 10Travers A.A. Curr. Biol. 1994; 4: 659-661Abstract Full Text Full Text PDF PubMed Scopus (4) Google Scholar, 11Krude T. Curr. Biol. 1995; 5: 1232-1234Abstract Full Text Full Text PDF PubMed Scopus (46) Google Scholar). A number of reports have suggested that the nucleosome structure in active chromatin is changed (12Prior C.P. Cantor C.R. Johnson E.M. Littau V.C. Allfrey V.G. Cell. 1983; 34: 1033-1042Abstract Full Text PDF PubMed Scopus (173) Google Scholar, 13Nacheva G.A. Guschin D.Y. Preobrazhenskaya O.V. Karpov V.L. Ebralidse K.K. Mirzabekov A.D. Cell. 1989; 58: 27-36Abstract Full Text PDF PubMed Scopus (222) Google Scholar, 14Lee M.-S. Garrard W. EMBO J. 1991; 10: 607-615Crossref PubMed Scopus (100) Google Scholar, 15Bazett-Jones D.P. Mendez E. Czarnota G.J. Ottensmeyer F.P. Allfrey V.G. Nucleic Acids Res. 1996; 24: 321-329Crossref PubMed Scopus (65) Google Scholar). Since only a few experiments have distinguished between transcriptionally competent and actively transcribed chromatin (3van Holde K.E. Lohr D.E. Robert C. J. Biol. Chem. 1992; 267: 2837-2840Abstract Full Text PDF PubMed Google Scholar, 8Morse R.H. Trends Biochem. Sci. 1992; 17: 23-26Abstract Full Text PDF PubMed Scopus (56) Google Scholar), very little is known about the structural details of these changes. Transcriptionally active chromatin is unfolded and significantly depleted of linker histones (16Zlatanova J. van Holde K. J. Cell Sci. 1992; 103: 889-895Crossref PubMed Google Scholar), which play a key role in chromatin compaction (17Thoma F. Koller Th Klug A. J. Cell. Biol. 1979; 83: 403-427Crossref PubMed Scopus (1173) Google Scholar, 18Widom J. Annu. Rev. Biophys. Biophys. Chem. 1989; 18: 365-395Crossref PubMed Scopus (145) Google Scholar). Recently, we have reported that certain histone-DNA contacts in the nucleosome core are heavily attenuated during chromatin unfolding (19Usachenko S.I. Gavin I.M. Bavykin S.G. J. Biol. Chem. 1996; 271: 3831-3836Abstract Full Text Full Text PDF PubMed Scopus (24) Google Scholar), indicating that the nucleosome structure is changed at the first step of chromatin activation, chromatin decondensation. These changes may facilitate the second step, which involves either remodeling or removal of the nucleosome. The binding of histone H1 to linker DNA considerably reduces its electrostatic free energy which determines linker DNA bending during chromatin folding into the higher order structure (20Clark D.J. Kimura T. J. Mol. Biol. 1990; 211: 883-896Crossref PubMed Scopus (189) Google Scholar). In contrast, in linker histone-depleted chromatin, the distance between the points where the linker DNA enters and leaves the nucleosome core may be increased by repulsion between adjacent linker DNA segments (17Thoma F. Koller Th Klug A. J. Cell. Biol. 1979; 83: 403-427Crossref PubMed Scopus (1173) Google Scholar,21Goulet I. Zivanovic Y. Prunell A. Revet B. J. Mol. Biol. 1988; 200: 253-266Crossref PubMed Scopus (49) Google Scholar, 22Zivanovic Y. Duband-Goulet I. Schultz P. Stofer E. Oudet P. Prunell A. J. Mol. Biol. 1990; 214: 479-495Crossref PubMed Scopus (51) Google Scholar, 23Bednar J. Horowitz R.A. Dubochet J. Woodcock C.L. J. Cell Biol. 1995; 131: 1365-1376Crossref PubMed Scopus (136) Google Scholar, 24Hamiche A. Schultz P. Ramakrishnan V. Oudet P. Prunell A. J. Mol. Biol. 1996; 257: 30-42Crossref PubMed Scopus (156) Google Scholar) and thereby can affect the conformation of the nucleosome core DNA. To address the question of whether or not the alterations in histone-DNA contacts during chromatin unfolding, which we observed earlier (19Usachenko S.I. Gavin I.M. Bavykin S.G. J. Biol. Chem. 1996; 271: 3831-3836Abstract Full Text Full Text PDF PubMed Scopus (24) Google Scholar), are caused by the changes in the nucleosome core DNA conformation, we have analyzed the histone-DNA contacts in isolated core particles under ionic conditions affecting DNA stiffness. Several physical and biochemical studies have demonstrated that a reduction of the monovalent ion concentration below 10 mmhas a significant effect on the nucleosome core structure (reviewed in Ref. 25van Holde K.E. Chromatin. Springer-Verlag, NY1989Crossref Google Scholar). It has also been well documented that the loss of counter-ions at low ionic strengths increases the electrostatic repulsion of unneutralized DNA phosphate groups, resulting in an increase in DNA stiffness (26Manning G.S. Q. Rev. Biophys. 1978; 11: 179-246Crossref PubMed Scopus (2602) Google Scholar). This increase in the stiffness may cause stretching of the nucleosomal DNA that in turn may change the conformation of the nucleosome (27Marky N.L. Manning G.S. Biopolymers. 1991; 31: 1543-1557Crossref PubMed Scopus (48) Google Scholar) and affect histone-DNA contacts. In the present paper, we show that the alterations in histone-DNA contacts in isolated core particles induced at low ionic strength are caused by a decrease in the neutralization of negative DNA charges and are identical to those observed in nucleosomes during chromatin unfolding (19Usachenko S.I. Gavin I.M. Bavykin S.G. J. Biol. Chem. 1996; 271: 3831-3836Abstract Full Text Full Text PDF PubMed Scopus (24) Google Scholar). This suggests that the conformational changes in nucleosome core particles at low ionic strength and in nucleosomes in chromatin during chromatin unfolding are due to the stretching of the nucleosome core DNA. Based on these observations, we propose a model for the reversible structural transition yielding a new nucleosome conformation in an unfolded chromatin. Nucleosome core particles were isolated from chicken erythrocyte nuclei as described previously (28Shick V.V. Belyavsky A.V. Bavykin S.G. Mirzabekov A.D. J. Mol. Biol. 1980; 139: 491-517Crossref PubMed Scopus (109) Google Scholar, 29Bavykin S.G. Usachenko S.I. Lishanskaya A.I. Shick V.V. Belyavsky A.V. Undritsov I.M. Strokov A.A. Zalenskaya I.A. Mirzabekov A.D. Nucleic Acids. Res. 1985; 13: 3439-3459Crossref PubMed Scopus (40) Google Scholar). Protein-DNA cross-linking under various ionic conditions, purification, and histone-labeling of cross-linked complexes with 125I were performed as described (30Usachenko S.I. Bavykin S.G. Gavin I.M. Bradbury E.M. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 6845-6849Crossref PubMed Scopus (74) Google Scholar, 31Mirzabekov A.D. Bavykin S.G. Belyavsky A.V. Karpov V.L. Preobrazhenskaya O.V. Shick V.V. Ebralidse K.K. Methods Enzymol. 1989; 170: 386-408Crossref PubMed Scopus (59) Google Scholar). Linker histone-depleted chromatin prepared from chicken erythrocyte nuclei as described earlier (19Usachenko S.I. Gavin I.M. Bavykin S.G. J. Biol. Chem. 1996; 271: 3831-3836Abstract Full Text Full Text PDF PubMed Scopus (24) Google Scholar) was digested with micrococcal nuclease (3 μg/1 mg of DNA) in 10 mm Tris-Cl, pH 8.0, and either 0.3 mm CaCl2 or 2 mm CaCl2 for the time course of 20, 40, and 80 min at 37 °C. The reaction was stopped by adding EDTA to a final concentration of 4 mm. DNA from nucleosomes released by the micrococcal nuclease digestion of linker histone-depleted chromatin was analyzed in a 9% denaturing gel containing 7m urea following digestion with Pronase (0.5 mg/ml) for 30 min at 37 °C (32Maniatis T. Jeffrey A. van deSande H. Biochemistry. 1975; 14: 3787-3794Crossref PubMed Scopus (806) Google Scholar). Histone-DNA contacts were mapped by the “protein version” of two-dimentional gel electrophoresis of 125I-histone-labeled cross-linked complexes as described earlier (31Mirzabekov A.D. Bavykin S.G. Belyavsky A.V. Karpov V.L. Preobrazhenskaya O.V. Shick V.V. Ebralidse K.K. Methods Enzymol. 1989; 170: 386-408Crossref PubMed Scopus (59) Google Scholar). For the qualitative analysis of relative intensities of certain signals, the autoradiographs of two-dimensional gels from at least three sets of experiments were scanned in a SL-504-XL Zeineh soft laser scanning densitometer (Biomed Instruments Inc., Fullerton, CA). To study the histone-DNA interactions in isolated core particles under different ionic conditions, we used the method of chemically induced zero-length protein-DNA cross-linking (31Mirzabekov A.D. Bavykin S.G. Belyavsky A.V. Karpov V.L. Preobrazhenskaya O.V. Shick V.V. Ebralidse K.K. Methods Enzymol. 1989; 170: 386-408Crossref PubMed Scopus (59) Google Scholar,33Pruss D. Bavykin S.G. Methods. 1997; 12: 36-47Crossref PubMed Scopus (10) Google Scholar). Cross-linking causes a single-stranded nick of the nucleosomal DNA at the site of cross-linking such that only the 5′-terminal DNA fragment becomes attached to a histone molecule (33Pruss D. Bavykin S.G. Methods. 1997; 12: 36-47Crossref PubMed Scopus (10) Google Scholar). Therefore, the length of the DNA fragment cross-linked to a particular histone is the precise distance of a protein cross-linking site from the 5′-end of the nucleosomal DNA and can be assessed by the two-dimensional gel electrophoresis. To map the histone-DNA contacts in nucleosome core particles under different ionic conditions, we have used a protein version of the two-dimensional gel electrophoresis system (31Mirzabekov A.D. Bavykin S.G. Belyavsky A.V. Karpov V.L. Preobrazhenskaya O.V. Shick V.V. Ebralidse K.K. Methods Enzymol. 1989; 170: 386-408Crossref PubMed Scopus (59) Google Scholar, 33Pruss D. Bavykin S.G. Methods. 1997; 12: 36-47Crossref PubMed Scopus (10) Google Scholar). In the denaturing first dimension SDS gel, the mobility of125I-labeled histone-DNA cross-linked complexes depends on the molecular weight of the histones and the size of the cross-linked DNA fragment. After separation of cross-linked complexes in the first dimension, the DNA is hydrolyzed directly in the gel according to a modified Burton's procedure (31Mirzabekov A.D. Bavykin S.G. Belyavsky A.V. Karpov V.L. Preobrazhenskaya O.V. Shick V.V. Ebralidse K.K. Methods Enzymol. 1989; 170: 386-408Crossref PubMed Scopus (59) Google Scholar, 33Pruss D. Bavykin S.G. Methods. 1997; 12: 36-47Crossref PubMed Scopus (10) Google Scholar, 34Burton K. Methods Enzymol. 1967; 12: 222-224Crossref Scopus (75) Google Scholar), and released125I-labeled histones are separated in the second dimension SDS gel (35Laemmli U.K. Nature. 1970; 227: 680-685Crossref PubMed Scopus (205523) Google Scholar) according to their size. In the first dimension, the mobility of histones is decreased proportionally to the size of DNA fragments cross-linked to a particular histone (33Pruss D. Bavykin S.G. Methods. 1997; 12: 36-47Crossref PubMed Scopus (10) Google Scholar). As a result, in the gel of the second dimension, the 125I-labeled histones are arranged as spots on different horizontal lines (Figs.1 and 2). The position of these spots indicates the location of the DNA cross-linking site for a particular histone in the nucleosome (33Pruss D. Bavykin S.G. Methods. 1997; 12: 36-47Crossref PubMed Scopus (10) Google Scholar).Figure 2The effect of divalent cation concentrations on the strength of histone H2B/H4-DNA contacts attenuated at low ionic strength. The concentrations and type of divalent cations are indicated in the upper right corner of each gel. For other details see text and legend to Fig. 1.View Large Image Figure ViewerDownload Hi-res image Download (PPT) The qualitative analysis of two-dimensional gels shown in Fig. 1revealed that the position of signals corresponding to histone-DNA contacts in core particles cross-linked in the range of decreasing ionic strengths from 100 to 2 mm NaCl is the same. However, relative intensities of some signals vary. In this work, using densitometric analysis, we qualitatively assessed the variation in histone H2B and H4 signal intensities representing the extent to which a particular histone is cross-linked to a particular site of the nucleosomal DNA. The main cross-linking sites on the nucleosomal DNA for histone H2B are around nucleotides 109, 119, and 129 and for histone H4 are around nucleotides 57, 66, and 93 from the 5′-end of the nucleosomal DNA (19Usachenko S.I. Gavin I.M. Bavykin S.G. J. Biol. Chem. 1996; 271: 3831-3836Abstract Full Text Full Text PDF PubMed Scopus (24) Google Scholar, 29Bavykin S.G. Usachenko S.I. Lishanskaya A.I. Shick V.V. Belyavsky A.V. Undritsov I.M. Strokov A.A. Zalenskaya I.A. Mirzabekov A.D. Nucleic Acids. Res. 1985; 13: 3439-3459Crossref PubMed Scopus (40) Google Scholar, 30Usachenko S.I. Bavykin S.G. Gavin I.M. Bradbury E.M. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 6845-6849Crossref PubMed Scopus (74) Google Scholar). To estimate the extent of variation in these contacts, we compared their relative intensity within the same histone line and within the same gel because, as shown previously, such a comparison prevents ambiguity caused by choosing the reference spot at different exposures of different gels (19Usachenko S.I. Gavin I.M. Bavykin S.G. J. Biol. Chem. 1996; 271: 3831-3836Abstract Full Text Full Text PDF PubMed Scopus (24) Google Scholar, 30Usachenko S.I. Bavykin S.G. Gavin I.M. Bradbury E.M. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 6845-6849Crossref PubMed Scopus (74) Google Scholar). The data represented in Fig. 1 demonstrate that by decreasing the concentration of Na+ from 100 to 2 mm, the relative intensities of the above-mentioned signals vary, showing a gradual attenuation of contacts H2B(129) and H4(66) compared with contacts H2B(109) and H4(57), respectively (numbers in parentheses indicate the distance in nucleotides from the 5′-end of one strand of nucleosome core DNA to the particular histone contact). It should be mentioned that the same alterations in the same histone-DNA contacts were observed recently as a result of the stretching of a linker DNA during chromatin unfolding in linker histone-depleted chromatin (19Usachenko S.I. Gavin I.M. Bavykin S.G. J. Biol. Chem. 1996; 271: 3831-3836Abstract Full Text Full Text PDF PubMed Scopus (24) Google Scholar). This indicates that an increase in the nucleosomal DNA stiffness, induced by low ionic strength, affects the analyzed histone-DNA contacts in isolated core particles much the same as the stretching of linker DNA during chromatin unfolding and suggests similar changes in the conformation of the nucleosomal DNA, which may result in the structural transition of the nucleosome (27Marky N.L. Manning G.S. Biopolymers. 1991; 31: 1543-1557Crossref PubMed Scopus (48) Google Scholar). This suggestion is supported by the recent observation of an alteration in the shape of core particles from the oblate to the prolate form at the same low salt concentrations (36Czarnota G.J. Ottensmeyer F.P. J. Biol. Chem. 1996; 271: 3677-3683Abstract Full Text Full Text PDF PubMed Scopus (24) Google Scholar). The elevation of salt concentration causes electrostatic blocking of DNA phosphate groups, which depends on the charge of cations (26Manning G.S. Q. Rev. Biophys. 1978; 11: 179-246Crossref PubMed Scopus (2602) Google Scholar). The experimentally measured persistence length of DNA increases rapidly when the concentration of monovalent cations is decreased below 10 mm, whereas the same effect for divalent cations was observed at concentrations below 0.1 mm (37Hagerman P.J. Annu. Rev. Biophys. Biophys. Chem. 1988; 17: 265-286Crossref PubMed Scopus (925) Google Scholar). Consistent with this observation, our data, represented in Fig. 2, demonstrate a similar change in the relative intensities of the analyzed contacts after the addition of divalent cations at 2 orders of magnitude less concentrations than monovalent cations. The weak intensity of contacts H2B(129) and H4(66) in 5 mm HEPES (Fig. 2 A) is noticeably increased upon addition of 0.1 mm CaCl2 (Fig. 2 B). A further increase in the concentration of divalent cations up to 2 mm causes an additional increase in the intensities of contacts H2B(129) and H4(66) (Fig. 2 C). The relative intensities of the analyzed contacts at 100 and 30 mmmonovalent cations (Fig. 1, A and B) resemble those at 2 and 0.1 mm divalent cations, respectively (Fig.2, B and C) confirming previous observations that monovalent cations affect the nucleosome structure at approximately 2 orders of magnitude higher concentrations than divalent cations (36Czarnota G.J. Ottensmeyer F.P. J. Biol. Chem. 1996; 271: 3677-3683Abstract Full Text Full Text PDF PubMed Scopus (24) Google Scholar,38Libertini L.J. Small E.W. Biochemistry. 1982; 21: 3327-3334Crossref PubMed Scopus (32) Google Scholar, 39Wu H.-M. Dattagupta N. Hogan M. Crothers D.M. Biochemistry. 1979; 18: 3960-3965Crossref PubMed Scopus (57) Google Scholar). We also tested various monovalent and divalent cations, such as K+, Li+, Cs+, Mg2+, Zn2+ for their ability to change histone H4/H2B contacts, and as expected, different cations that have the same charge had an identical effect on the analyzed contacts at the same concentrations. Cross-linking analysis also reveals that anions have no impact on the conformational transition induced by low ionic strength (not shown). Obtained results suggest that the conformational transition, observed as alterations in relative intensities of histone H2B/H4-DNA contacts at low ionic strength, is caused by the repulsion of adjacent unneutralized negative charges of DNA phosphate groups resulting in an increase in DNA stiffness. The level of chromatin compaction induced by 1–2 mm divalent cations is similar to that observed at physiological concentrations of monovalent cations (17Thoma F. Koller Th Klug A. J. Cell. Biol. 1979; 83: 403-427Crossref PubMed Scopus (1173) Google Scholar, 18Widom J. Annu. Rev. Biophys. Biophys. Chem. 1989; 18: 365-395Crossref PubMed Scopus (145) Google Scholar). An increase in the salt concentration also results in folding of linker histone-depleted chromatin and bending of linker DNA (17Thoma F. Koller Th Klug A. J. Cell. Biol. 1979; 83: 403-427Crossref PubMed Scopus (1173) Google Scholar, 40Hansen J.C. Ausio J. Stanik V.H. van Holde K.E. Biochemistry. 1989; 28: 9129-9136Crossref PubMed Scopus (194) Google Scholar, 41Yao J. Lowary P.T. Widom J. Biochemistry. 1991; 30: 8408-8414Crossref PubMed Scopus (57) Google Scholar, 42Garcia-Ramirez M. Dong F. Ausio J. J. Biol. Chem. 1992; 267: 19587-19595Abstract Full Text PDF PubMed Google Scholar, 43Schwarz P.M. Hansen J.C. J. Biol. Chem. 1994; 269: 16284-16289Abstract Full Text PDF PubMed Google Scholar). To investigate how this salt-induced chromatin condensation might change the nucleosome conformation, we have analyzed the pattern of micrococcal nuclease digestion of linker histone-depleted chromatin at different concentrations of divalent cations (Fig.3). The first time point in the course of digestion at 0.3 mm Ca2+ yields nucleosome core particles with 146-bp 1The abbreviations used are: bp, base pair(s). DNA represented in the gel by a single band (Fig. 3 A, lane 1). These nucleosomes are further digested to the level of subnucleosomes (Fig. 3 A, lanes 2 and3). In contrast, the first time point at 2 mmCa2+ yields a set of core particles and particles with 157- and 168-bp DNA (Fig. 3 B, lane 1), which then turn into particles with 146-bp DNA (Fig. 3 B, lane 3). The bending of linker DNA induced by 2 mm divalent cations (17Thoma F. Koller Th Klug A. J. Cell. Biol. 1979; 83: 403-427Crossref PubMed Scopus (1173) Google Scholar, 43Schwarz P.M. Hansen J.C. J. Biol. Chem. 1994; 269: 16284-16289Abstract Full Text PDF PubMed Google Scholar) facilitates the wrapping of an additional ∼20-bp DNA around the histone octamer (21Goulet I. Zivanovic Y. Prunell A. Revet B. J. Mol. Biol. 1988; 200: 253-266Crossref PubMed Scopus (49) Google Scholar, 22Zivanovic Y. Duband-Goulet I. Schultz P. Stofer E. Oudet P. Prunell A. J. Mol. Biol. 1990; 214: 479-495Crossref PubMed Scopus (51) Google Scholar, 24Hamiche A. Schultz P. Ramakrishnan V. Oudet P. Prunell A. J. Mol. Biol. 1996; 257: 30-42Crossref PubMed Scopus (156) Google Scholar). This renders linker DNA regions less accessible to micrococcal nuclease yielding nucleosomes with 157- and 168-bp DNA (Fig. 3 B). The interaction of core histones with the end regions of nucleosomal DNA in particles with 155- and 165-bp DNA at monovalent salt concentrations above 30 mm has been demonstrated by protein-DNA cross-linking in isolated chromatin (44Belyavsky A.V. Bavykin S.G. Goguadze E.G. Mirzabekov A.D. J. Mol. Biol. 1980; 139: 519-536Crossref PubMed Scopus (85) Google Scholar), in nuclei (45Bavykin S.G. Usachenko S.I. Zalensky A.O. Mirzabekov A.D. J. Mol. Biol. 1990; 212: 495-511Crossref PubMed Scopus (68) Google Scholar), and in H1-depleted nucleosomes (46Davies N. Lindsey G.G. Biochem. Biophys. Acta. 1991; 1129: 57-63Crossref PubMed Scopus (8) Google Scholar). In contrast, at concentrations of divalent cations below 0.3 mm, linker DNA is stretched (17Thoma F. Koller Th Klug A. J. Cell. Biol. 1979; 83: 403-427Crossref PubMed Scopus (1173) Google Scholar) such that the end regions of the nucleosomal DNA become susceptible to micrococcal nuclease. As a result, nucleosomes with 157- and 168-bp DNA are not observed under these conditions (Fig.3 A). It has been shown that micrococcal nuclease activity is only slightly reduced in 0.3 mm Ca2+ compared with 2 mm Ca2+ (47Cuatrecasis P. Fuchs S. Anfinsen C.B. J. Biol. Chem. 1967; 242: 1541-1547Abstract Full Text PDF PubMed Google Scholar), whereas the specificity of the enzyme remains the same (48Afinsen C.B. Cuatrecasas P. Tanguchi H. Boyer P.D. 3rd Ed. The Enzymes. IV. Academic Press, NY1971: 177-204Google Scholar). This indicates that the observed difference in micrococcal nuclease digestion pattern reflects the change in the chromatin structure rather than the activity of the micrococcal nuclease, which is supported by early studies where a similar effect was observed at various concentrations of monovalent cations (49Weischet W.O. Allen J.R. Riedel G. van Holde K.E. Nucleic Acids Res. 1979; 6: 1843-1862Crossref PubMed Scopus (60) Google Scholar, 50Ruiz-Carrillo A. Jorcano J.L. Eder G. Lurz R. Proc. Natl. Acad. Sci U. S. A. 1979; 76: 3284-3288Crossref PubMed Scopus (81) Google Scholar). In the present work, we have demonstrated that a structural transition in isolated core particles induced by low concentrations of monovalent cations is caused by the change in the nucleosomal DNA conformation that alters the analyzed histone H2B/H4-DNA contacts. These altered contacts are located at sites ± 5.5 and ± 1 (Fig.4), which are very close to the sharply bent regions of the nucleosomal DNA (51Luger K. Mäder A.W. Richmond R.K. Sargent D.F. Richmond T.J Nature. 1997; 389: 251-260Crossref PubMed Scopus (6725) Google Scholar, 52Hogan M.E. Rooney T.F. Austin R.H. Nature. 1987; 328: 554-557Crossref PubMed Scopus (84) Google Scholar, 53Uberbacher E.C. Bunick G.J. J. Biomol. Struct. Dyn. 1989; 7: 1-18Crossref PubMed Scopus (26) Google Scholar, 54Struck M.-M. Klug A. Richmond T.J. J. Mol. Biol. 1992; 224: 253-264Crossref PubMed Scopus (29) Google Scholar, 55Richmond T.J. Rechsteiner T. Luger K. Cold Spring Harbor Symp. Quant. Biol. 1993; 53: 265-272Crossref Scopus (32) Google Scholar). Histone H4 cross-links to the sites around ± 1 mainly through the single highly basic domain (30Usachenko S.I. Bavykin S.G. Gavin I.M. Bradbury E.M. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 6845-6849Crossref PubMed Scopus (74) Google Scholar, 56Ebralidse K.K. Mirzabekov A.D. FEBS Lett. 1986; 194: 69-72Crossref PubMed Scopus (21) Google Scholar, 57Ebralidse K.K. Grachev S.A. Mirzabekov A.D. Nature. 1988; 331: 365-367Crossref PubMed Scopus (97) Google Scholar), which may alter DNA structure in this region (57Ebralidse K.K. Grachev S.A. Mirzabekov A.D. Nature. 1988; 331: 365-367Crossref PubMed Scopus (97) Google Scholar,58Manning G.S. Ebralidse K.K. Mirzabekov A.D. Rich A. J. Biomol. Struct. Dyn. 1989; 6: 877-889Crossref PubMed Scopus (88) Google Scholar). On the other hand, the change in nucleosomal DNA conformation induced by low ionic strength in isolated core particles or by the stretching of linker DNA during chromatin unfolding can in turn affect histone interactions with these DNA sites. Since the analyzed contacts are located close to the sharply bent DNA regions, they would be expected to be very sensitive to conformational changes in the nucleosomal DNA. Identification of nucleotide-tagged histone H4 peptides (56Ebralidse K.K. Mirzabekov A.D. FEBS Lett. 1986; 194: 69-72Crossref PubMed Scopus (21) Google Scholar, 57Ebralidse K.K. Grachev S.A. Mirzabekov A.D. Nature. 1988; 331: 365-367Crossref PubMed Scopus (97) Google Scholar) and selective proteolysis of cross-linked core particles by trypsin and clostripain (30Usachenko S.I. Bavykin S.G. Gavin I.M. Bradbury E.M. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 6845-6849Crossref PubMed Scopus (74) Google Scholar) have demonstrated that histone H4 cross-links via His-18 to three major sites, at nucleotides 57, 66, and 93, which are extended over 1.2 helical turns of the nucleosomal DNA (Fig. 4). The cross-linking domain of histone H4 has divergent structure within the nucleosome core (51Luger K. Mäder A.W. Richmond R.K. Sargent D.F. Richmond T.J Nature. 1997; 389: 251-260Crossref PubMed Scopus (6725) Google Scholar), suggesting that these contacts may be formed by different histone H4 molecules, which may cross-link to two discrete sites at a given time: to nucleotide 66 around site ± 1 and to nucleotides 57 and/or 93 due to the oscillation of His-18 through the minor groove of the nucleosomal DNA around site ± 2 (Fig. 4). This is supported by our observations that the relative intensities of contacts H4(57) and H4(93) are not changed at low ionic strength (Fig.Figure 1, Figure 2) and during chromatin unfolding (19Usachenko S.I. Gavin I.M. Bavykin S.G. J. Biol. Chem. 1996; 271: 3831-3836Abstract Full Text Full Text PDF PubMed Scopus (24) Google Scholar). The change in the nucleosome conformation induced by low ionic strength or chromatin unfolding may affect the local interaction of the DNA-binding domain of histone H4 with nucleotide 66 and would result in a less intensive cross-linking of its His-18 to this DNA site. The highly basic N-terminal domain of histone H2B also cross-links to nucleosome DNA (59Guschin D.Yu. Ebralidse K.K. Mirzabekov A.D. J. Molec. Biol. 1991; 25: 1400-1411Google Scholar); however, it is still unknown whether the observed contacts are formed by the same domain or by different regions of histone H2B. The crystal structure of the nucleosome core particle recently resolved at 2.8 Å resolution suggests that loop 2 of histone H2B contacts DNA around sites ± 3.5 while loop 1 makes contact around sites ± 4.5 (51Luger K. Mäder A.W. Richmond R.K. Sargent D.F. Richmond T.J Nature. 1997; 389: 251-260Crossref PubMed Scopus (6725) Google Scholar). Since the methodology used in this study detects protein-DNA cross-linking mainly through histidine residues (33Pruss D. Bavykin S.G. Methods. 1997; 12: 36-47Crossref PubMed Scopus (10) Google Scholar), it is possible that histidine 79 at the C terminus of α-helix 2 cross-links to the DNA site at nucleotide 109 while histidine 46 located in the loop 1 cross-links to the site around the 129th nucleotide. Straightening of the nucleosomal DNA at low ionic strength or during chromatin unfolding may affect the interaction of the end fragments of the nucleosomal DNA with loop 1 of histone H2B, resulting in the loss of histone H2B cross-linking to nucleotide 129. Two models for the conformational transition of a nucleosome core caused by the straightening of the nucleosomal DNA have been proposed (27Marky N.L. Manning G.S. Biopolymers. 1991; 31: 1543-1557Crossref PubMed Scopus (48) Google Scholar). In the model for a radial transition, the nucleosomal DNA straightening increases radius of gyration resulting in an oblate conformation of the nucleosome core. In the case of an axial transition, DNA straightens along the superhelical axis and thereby stretching the nucleosome core into the prolate form. Recent observations of the altered core particle shape to a prolate form (36Czarnota G.J. Ottensmeyer F.P. J. Biol. Chem. 1996; 271: 3677-3683Abstract Full Text Full Text PDF PubMed Scopus (24) Google Scholar) that have a decreased (36Czarnota G.J. Ottensmeyer F.P. J. Biol. Chem. 1996; 271: 3677-3683Abstract Full Text Full Text PDF PubMed Scopus (24) Google Scholar, 60Hirai M. Niimura N. Zama M. Mita K. Ichimura S. Tokunaga F. Ishikawa Y. Biochemistry. 1988; 27: 7924-7931Crossref PubMed Scopus (15) Google Scholar) or unchanged (61Sibbet G.J. Carpenter B.G. Ibel K. May R.P. Kneale G.G. Bradbury E.M. Baldwin J.P. Eur. J. Biochem. 1983; 133: 393-398Crossref PubMed Scopus (13) Google Scholar) radius of gyration at concentrations of monovalent cations below 30 mm provide strong support for the second model. This model is also in agreement with the expected repulsion of adjacent superhelical turns of the nucleosomal DNA induced by low salt concentrations, which also tends to stretch the core particle in the direction of the superhelical axis. The axial straightening of DNA may also cause deformation of the histone octamer, which is consistent with the observed alteration in the histone octamer compaction with decreasing ionic strength from 30 to 10 mm NaCl (62Chung D.G. Lewis P.N. Biochemistry. 1985; 24: 8028-8036Crossref PubMed Scopus (13) Google Scholar). Based on the results of this work and the observations of other investigators, we suggest a model for the structural transition of the nucleosome core that may occur during chromatin unfolding (Fig.5). In condensed chromatin, the linker DNA is bent such that the histone octamer interacts with an additional 20 bp of DNA (Fig. 5 A). In contrast, during chromatin decondensation after depletion or rearrangement of linker histones, stretching of the linker DNA may increase DNA tension along the polynucleosomal chain and change the nucleosome conformation in a way similar to that observed for isolated core particles at monovalent salt concentrations below 10 mm. This stretching would cause partial sliding of DNA off the histone octamer and make the additional 20-bp accessible to micrococcal nuclease (Fig. 5 B). This model is also consistent with the observed salt-dependent release of nucleosomal DNA termini resulting in the change in the linking number of DNA in minichromosomes (21Goulet I. Zivanovic Y. Prunell A. Revet B. J. Mol. Biol. 1988; 200: 253-266Crossref PubMed Scopus (49) Google Scholar, 22Zivanovic Y. Duband-Goulet I. Schultz P. Stofer E. Oudet P. Prunell A. J. Mol. Biol. 1990; 214: 479-495Crossref PubMed Scopus (51) Google Scholar) and in the altered nucleosome morphology in linker histone-depleted chromatin (21Goulet I. Zivanovic Y. Prunell A. Revet B. J. Mol. Biol. 1988; 200: 253-266Crossref PubMed Scopus (49) Google Scholar, 22Zivanovic Y. Duband-Goulet I. Schultz P. Stofer E. Oudet P. Prunell A. J. Mol. Biol. 1990; 214: 479-495Crossref PubMed Scopus (51) Google Scholar, 23Bednar J. Horowitz R.A. Dubochet J. Woodcock C.L. J. Cell Biol. 1995; 131: 1365-1376Crossref PubMed Scopus (136) Google Scholar, 24Hamiche A. Schultz P. Ramakrishnan V. Oudet P. Prunell A. J. Mol. Biol. 1996; 257: 30-42Crossref PubMed Scopus (156) Google Scholar). The association of the alterations in the analyzed histone H2B/H4-DNA contacts with the level of chromatin compaction and activity demonstrates the importance of these changes in chromatin structural/functional relationships (19Usachenko S.I. Gavin I.M. Bavykin S.G. J. Biol. Chem. 1996; 271: 3831-3836Abstract Full Text Full Text PDF PubMed Scopus (24) Google Scholar). The alterations in the contacts formed by the highly basic histone domains may result in destabilization of the nucleosomes in unfolded chromatin facilitating nucleosome remodeling or/and removal during transcription and replication. This is strongly supported by the observation that a deletion of the histone H2B cross-linking domain changes the minichromosome topology in vivo, suggesting that this domain is required for the proper folding of DNA in the nucleosome (63Lenfant F. Mann R.K. Thomsen B. Ling X. Grunstein M. EMBO J. 1996; 15: 3974-3985Crossref PubMed Scopus (73) Google Scholar). Deletion or substitution of a single amino acid in the cross-linking domain of histone H4 alters chromatin structure in vivo (63Lenfant F. Mann R.K. Thomsen B. Ling X. Grunstein M. EMBO J. 1996; 15: 3974-3985Crossref PubMed Scopus (73) Google Scholar, 64Roth S.Y. Shimizu M. Johnson L. Grunstein M. Simpson R.T. Genes Dev. 1992; 6: 411-425Crossref PubMed Scopus (121) Google Scholar, 65Fisher-Adams G. Grunstein M. EMBO J. 1995; 14: 1468-1477Crossref PubMed Scopus (65) Google Scholar, 66Ling X. Harkness T.A.A. Schultz M.C. Fisher-Adams G. Grunstein M. Genes Dev. 1996; 10: 686-699Crossref PubMed Scopus (116) Google Scholar), decreases the ability of yeast to mate, increases the duration of S phase (67Kayne P.S. Kim U. Han M. Mullen J.R. Yoshizaki F. Grunstein M. Cell. 1988; 55: 27-39Abstract Full Text PDF PubMed Scopus (310) Google Scholar, 68Johnson L.M. Kayne P.S. Kahn E.S. Grunstein M. Proc. Natl. Acad. Sci. U. S. A. 1990; 87: 6286-6290Crossref PubMed Scopus (276) Google Scholar, 69Megee P.C. Morgan B.A. Mittman B.A. Smith M.M. Science. 1990; 247: 841-845Crossref PubMed Scopus (268) Google Scholar, 70Park E.C. Szostak J.W. Moll. Cell. Biol. 1990; 10: 4932-4934Crossref PubMed Scopus (145) Google Scholar), and affects telomeric repression (71Aparicio O.M. Billington B.L. Gottschling D.E. Cell. 1991; 66: 1279-1287Abstract Full Text PDF PubMed Scopus (606) Google Scholar,72Thompson J.S. Ling X. Grunstein M. Nature. 1994; 369: 245-247Crossref PubMed Scopus (205) Google Scholar) as well as expression of a large number of yeast genes (64Roth S.Y. Shimizu M. Johnson L. Grunstein M. Simpson R.T. Genes Dev. 1992; 6: 411-425Crossref PubMed Scopus (121) Google Scholar, 65Fisher-Adams G. Grunstein M. EMBO J. 1995; 14: 1468-1477Crossref PubMed Scopus (65) Google Scholar,73Durrin L.K. Mann R.K. Kayne P.S. Grunstein M. Cell. 1991; 65: 1023-1031Abstract Full Text PDF PubMed Scopus (255) Google Scholar, 74Megee P.C. Morgan B.A. Smith M.M. Genes Dev. 1995; 9: 1716-1727Crossref PubMed Scopus (150) Google Scholar, 75Edmondson D.G. Smith M.M. Roth S.Y. Genes Dev. 1996; 10: 1247-1259Crossref PubMed Scopus (405) Google Scholar). It should be mentioned also that the cross-linking domain of histone H4 contains sites for post-translational acetylation and phosphorylation (76Wu R.S. Panusz H.T. Hatch C.L. Bonner W.N. CRC Crit. Rev. Biochem. 1986; 20: 201-269Crossref PubMed Scopus (206) Google Scholar) that may also be involved in the regulation of the interaction of this domain with DNA at different levels of chromatin activity and condensation rendering nucleosomes competent for transcription and/or replication. We thank Andrei Mirzabekov for the initial idea of this work. We are also grateful to Hugh Patterton for critical comments and Jodie Usachenko for help in the preparation of this manuscript.