Portal de Periódicos da CAPES

Histone Acetylation Is Required to Maintain the Unfolded Nucleosome Structure Associated with Transcribing DNA

1998; Elsevier BV; Volume: 273; Issue: 23 Linguagem: Inglês

10.1074/jbc.273.23.14516

1083-351X

Harminder Walia, Hou Yu Chen, Jian-Min Sun, Laurel T. Holth, James Davie,

MicroRNA in disease regulation

Nucleosomes associated with transcribing chromatin of mammalian cells have an unfolded structure in which the normally buried cysteinyl-thiol group of histone H3 is exposed. In this study we analyzed transcriptionally active/competent DNA-enriched chromatin fractions from chicken mature and immature erythrocytes for the presence of thiol-reactive nucleosomes using organomercury-agarose column chromatography and hydroxylapatite dissociation chromatography of chromatin fractions labeled with [3H]iodoacetate. In mature and immature erythrocytes, the active DNA-enriched chromatin fractions are associated with histones that are rapidly highly acetylated and rapidly deacetylated. When histone deacetylation was prevented by incubating cells with histone deacetylase inhibitors, sodium butyrate or trichostatin A, thiol-reactive H3 of unfolded nucleosomes was detected in the soluble chromatin and nuclear skeleton-associated chromatin of immature, but not mature, erythrocytes. We did not find thiol-reactive nucleosomes in active DNA-enriched chromatin fractions of untreated immature erythrocytes that had low levels of highly acetylated histones H3 and H4 or in chromatin of immature cells incubated with inhibitors of transcription elongation. This study shows that transcription elongation is required to form, and histone acetylation is needed to maintain, the unfolded structure of transcribing nucleosomes. Nucleosomes associated with transcribing chromatin of mammalian cells have an unfolded structure in which the normally buried cysteinyl-thiol group of histone H3 is exposed. In this study we analyzed transcriptionally active/competent DNA-enriched chromatin fractions from chicken mature and immature erythrocytes for the presence of thiol-reactive nucleosomes using organomercury-agarose column chromatography and hydroxylapatite dissociation chromatography of chromatin fractions labeled with [3H]iodoacetate. In mature and immature erythrocytes, the active DNA-enriched chromatin fractions are associated with histones that are rapidly highly acetylated and rapidly deacetylated. When histone deacetylation was prevented by incubating cells with histone deacetylase inhibitors, sodium butyrate or trichostatin A, thiol-reactive H3 of unfolded nucleosomes was detected in the soluble chromatin and nuclear skeleton-associated chromatin of immature, but not mature, erythrocytes. We did not find thiol-reactive nucleosomes in active DNA-enriched chromatin fractions of untreated immature erythrocytes that had low levels of highly acetylated histones H3 and H4 or in chromatin of immature cells incubated with inhibitors of transcription elongation. This study shows that transcription elongation is required to form, and histone acetylation is needed to maintain, the unfolded structure of transcribing nucleosomes. Acetylation of the core histones (H2A, H2B, H3, and H4) is a dynamic process catalyzed by histone acetyltransferases and histone deacetylases (1Davie J.R. J. Cell. Biochem. 1996; 62: 149-157Crossref PubMed Scopus (43) Google Scholar, 2Davie J.R. Mol. Biol. Rep. 1997; 24: 197-207Crossref PubMed Scopus (86) Google Scholar). In chicken immature erythrocytes, 4% of the modifiable lysine sites participate in dynamic histone acetylation. These core histones are rapidly acetylated (t½ = 12 min for monoacetylated H4) and rapidly deacetylated (t½ = 5 min for the tetraacetylated isoform of H4) (3Zhang D.-E. Nelson D.A. Biochem. J. 1988; 250: 233-240Crossref PubMed Scopus (55) Google Scholar, 4Zhang D.-E. Nelson D.A. Biochem. J. 1988; 250: 241-245Crossref PubMed Scopus (41) Google Scholar). Histones undergoing rapid acetylation and deacetylation are associated with transcriptionally active chromatin (5Ip Y.T. Jackson V. Meier J. Chalkley R. J. Biol. Chem. 1988; 263: 14044-14052Abstract Full Text PDF PubMed Google Scholar, 6Boffa L.C. Walker J. Chen T.A. Sterner R. Mariani M.R. Allfrey V.G. Eur. J. Biochem. 1990; 194: 811-823Crossref PubMed Scopus (43) Google Scholar, 7Hendzel M.J. Delcuve G.P. Davie J.R. J. Biol. Chem. 1991; 266: 21936-21942Abstract Full Text PDF PubMed Google Scholar). The recent findings that histone acetyltransferases and deacetylases are transcriptional coactivators and corepressors have increased our understanding of how the process of dynamic histone acetylation is established on transcriptionally active chromatin (2Davie J.R. Mol. Biol. Rep. 1997; 24: 197-207Crossref PubMed Scopus (86) Google Scholar, 8Wade P.A. Wolffe A.P. Curr. Biol. 1997; 7: R82-R84Abstract Full Text Full Text PDF PubMed Google Scholar). Transcriptionally active chromatin has a soluble and insoluble nature (9Gross D.S. Garrard W.T. Trends Biochem. Sci. 1987; 12: 293-297Abstract Full Text PDF Scopus (84) Google Scholar). Transcribed DNA is found in chromatin fragments that are soluble in 0.15 m NaCl and/or 2 mm MgCl2and in chromatin fragments associated with the low salt-insoluble residual nuclear material (nuclear skeletons) (for review, see Davie (10Davie J.R. Int. Rev. Cytol. 1995; 162A: 191-250PubMed Google Scholar)). Chromatin engaged in transcription is thought to be retained by the nuclear skeleton by multiple dynamic attachments between the nuclear matrix and transcribed chromatin; hence rendering the transcribing chromatin insoluble (11Andreeva M. Markova D. Loidl P. Djondjurov L. Eur. J. Biochem. 1992; 207: 887-894Crossref PubMed Scopus (12) Google Scholar, 12Gerdes M.G. Carter K.C. Moen Jr., P.T. Lawrence J.B. J. Cell Biol. 1994; 126: 289-304Crossref PubMed Scopus (146) Google Scholar). As histone acetyltransferase and deacetylase activities are associated with the nuclear matrix (7Hendzel M.J. Delcuve G.P. Davie J.R. J. Biol. Chem. 1991; 266: 21936-21942Abstract Full Text PDF PubMed Google Scholar,13Hendzel M.J. Sun J.-M. Chen H.Y. Rattner J.B. Davie J.R. J. Biol. Chem. 1994; 269: 22894-22901Abstract Full Text PDF PubMed Google Scholar), we proposed that these nuclear matrix-bound enzymes may mediate some of the dynamic attachments between active chromatin and nuclear matrix (13Hendzel M.J. Sun J.-M. Chen H.Y. Rattner J.B. Davie J.R. J. Biol. Chem. 1994; 269: 22894-22901Abstract Full Text PDF PubMed Google Scholar, 14Davie J.R. Hendzel M.J. J. Cell. Biochem. 1994; 55: 98-105Crossref PubMed Scopus (74) Google Scholar). Most information on the structure and composition of transcriptionally active nucleosomes is from studies that analyze soluble transcriptionally active chromatin. However, most of the transcribed chromatin fragments partition with the low salt-insoluble nuclear material (nuclear skeleton) (7Hendzel M.J. Delcuve G.P. Davie J.R. J. Biol. Chem. 1991; 266: 21936-21942Abstract Full Text PDF PubMed Google Scholar, 15Stratling W.H. Dolle A. Sippel A.E. Biochemistry. 1986; 25: 495-502Crossref PubMed Scopus (51) Google Scholar, 16Stratling W.H. Biochemistry. 1987; 26: 7893-7899Crossref PubMed Scopus (19) Google Scholar). We presented evidence that dynamically acetylated histones are associated with the nuclear matrix-bound transcriptionally active chromatin (7Hendzel M.J. Delcuve G.P. Davie J.R. J. Biol. Chem. 1991; 266: 21936-21942Abstract Full Text PDF PubMed Google Scholar). Otherwise, little is known about the structure and composition of transcribing nucleosomes attached to the nuclear skeleton. Allfrey and co-workers demonstrated that nucleosomes in the transcribed regions of soluble chromatin of mammalian cells unfold exposing the cysteinyl-thiol groups of histone H3 (17Allfrey V.G. Chen T.A. Methods Cell Biol. 1991; 35: 315-335Crossref PubMed Scopus (19) Google Scholar, 18Bazett-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). The unfolding of the nucleosome was dependent upon ongoing transcription. Exploiting this feature of transcribing nucleosomes, a procedure to isolate soluble transcriptionally active chromatin by organomercury-agarose affinity chromatography was developed. The transcribing chromatin was associated with highly acetylated histones (18Bazett-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, 19Chen-Cleland T.A. Boffa L.C. Carpaneto E.M. Mariani M.R. Valentin E. Mendez E. Allfrey V.G. J. Biol. Chem. 1993; 268: 23409-23416Abstract Full Text PDF PubMed Google Scholar, 20Sterner R. Boffa L.C. Chen T.A. Allfrey V.G. Nucleic. Acids. Res. 1987; 15: 4375-4391Crossref PubMed Scopus (29) Google Scholar). However, current evidence argues that histone acetylation is not involved in the generation of the unfolded nucleosome. Reconstitution of nucleosomes with highly acetylated histones did not result in the formation of a thiol-reactive nucleosome (21Imai B.S. Yau P. Baldwin J.P. Ibel K. May R.P. Bradbury E.M. J. Biol. Chem. 1986; 261: 8784-8792Abstract Full Text PDF PubMed Google Scholar). Further, treating mammalian cells with the histone deacetylase inhibitor, sodium butyrate, did not increase the level of thiol-reactive nucleosomes (6Boffa L.C. Walker J. Chen T.A. Sterner R. Mariani M.R. Allfrey V.G. Eur. J. Biochem. 1990; 194: 811-823Crossref PubMed Scopus (43) Google Scholar). Analysis of chicken mature erythrocyte salt-soluble polynucleosomes highly enriched in transcriptionally competent DNA and highly acetylated histones (22Rocha E. Davie J.R. Van Holde K.E. Weintraub H. J. Biol. Chem. 1984; 259: 8558-8563Abstract Full Text PDF PubMed Google Scholar, 23Ridsdale J.A. Rattner J.B. Davie J.R. Nucleic Acids Res. 1988; 16: 5915-5926Crossref PubMed Scopus (27) Google Scholar) showed that this chromatin fraction lacked thiol-reactive nucleosomes. 1J. A. Ridsdale, P. Fredette, and J. R. Davie, unpublished observations. To address the question of whether unfolded nucleosomes exist in chicken erythrocytes, we investigated the H3 thiol reactivity of salt-soluble and low salt-insoluble (nuclear skeleton-associated) chromatin from mature (transcriptionally silent) and immature (transcriptionally active) chicken erythrocytes. We report that the thiol-reactive, unfolded nucleosome exists in immature, but not mature, erythrocyte salt-soluble chromatin fragments and chromatin fragments associated with the nuclear skeleton. However, histone deacetylase activity had to be suppressed to detect thiol-reactive nucleosomes in immature erythrocyte chromatin. These studies show that highly acetylated histones maintain the unfolded nucleosome structure formed by transcriptional elongation. Mature and immature erythrocytes were isolated from normal and anemic young adult White Leghorn chickens, respectively, as described previously (24Delcuve G.P. Davie J.R. Biochem. J. 1989; 263: 179-186Crossref PubMed Scopus (50) Google Scholar). Immature and mature erythrocytes were collected in 75 mm NaCl and 25 mm EDTA, while cells to be incubated with sodium butyrate or trichostatin A were collected in 75 mm NaCl, 25 mm EDTA, and 30 mm sodium butyrate. Cells were incubated at 37 °C in Swims S-77 medium (Sigma) with 30 mm sodium butyrate or 100 ng/ml trichostatin A for 90 min to prevent the deacetylation of highly acetylated histones. To inhibit transcription elongation, cells were incubated at 37 °C for 90 min with transcription inhibitors actinomycin D (0.04 μg/ml) (24Delcuve G.P. Davie J.R. Biochem. J. 1989; 263: 179-186Crossref PubMed Scopus (50) Google Scholar), 5,6-dichloro-1-β-d-ribofuranosylbenzimidazole (DRB) 2The abbreviations used are: DRB, 5,6-dichloro-1-β-d-ribofuranosylbenzimidazole; DTT, dithiothreitol. (75 μm) (24Delcuve G.P. Davie J.R. Biochem. J. 1989; 263: 179-186Crossref PubMed Scopus (50) Google Scholar, 25Chadee D.N. Allis C.D. Wright J.A. Davie J.R. J. Biol. Chem. 1997; 272: 8113-8116Abstract Full Text Full Text PDF PubMed Scopus (43) Google Scholar), or camptothecin (20 μm) (26Ljungman M. Hanawalt P.C. Carcinogenesis. 1996; 17: 31-35Crossref PubMed Scopus (65) Google Scholar) followed by a 90-min incubation with or without 30 mm sodium butyrate. The fractionation of chromatin was done as described previously (24Delcuve G.P. Davie J.R. Biochem. J. 1989; 263: 179-186Crossref PubMed Scopus (50) Google Scholar). All buffers contained 1 mm phenylmethylsulfonyl fluoride. Briefly, nuclei (50A 260 units/ml) were digested with micrococcal nuclease (15 A 260 units/ml for 25 min at 37 °C), collected by centrifugation, and then resuspended in 10 mm Tris-HCl, pH 7.5, 1 mm EDTA. Following centrifugation the soluble chromatin fraction (fraction SE) and low salt insoluble chromatin fraction (nuclear skeleton, PE fraction) were isolated. The chromatin fragments of fraction SE were further fractionated by the addition of NaCl to 150 mm. Following centrifugation, chromatin fractions P150 (pellet) and S150 (salt-soluble chromatin) were isolated. Chicken erythrocyte chromatin fraction S150 was dialyzed against buffer A (10 mm Tris-HCl, pH 7.5, 25 mm KCl, 25 mm NaCl, 5 mm sodium butyrate, 0.1 mm phenylmethylsulfonyl fluoride, and 5 mmEDTA, pH 7.5) and then applied to an organomercurial column (Affi-Gel 501; Bio-Rad) that was pre-equilibrated with buffer A. The column (1.5 × 8 cm) was then washed with buffer A (flow rate, 60 ml/h) to remove unbound chromatin fragments until the absorbance at 260 nm returned to base line. This was followed by washing the column with 0.5m NaCl in buffer A (buffer B) until the absorbance at 260 nm returned to base line. The bound nucleosomes were eluted by 10 mm DTT in buffer A. The released material was monitored by measuring absorbance at 260 nm (17Allfrey V.G. Chen T.A. Methods Cell Biol. 1991; 35: 315-335Crossref PubMed Scopus (19) Google Scholar). The fractions containing the unbound material and 0.5 m NaCl eluted material were pooled; the fractions eluted with DTT were pooled. For analysis of histones in the unbound and bound fractions, the fractions were acid-extracted by the addition of 4 n sulfuric acid to a final concentration of 0.4 n. Before lyophilization, the supernatants were dialyzed overnight against 0.1 m acetic acid and then against two changes of double-distilled H2O. Ten μl/ml [3H]iodoacetic acid (NEN Life Science Products, 204.9 mCi/mmol) containing 2.50 μCi was added to the chromatin fraction SE, S150, P150, and PE (2A 260/ml) in buffer E (10 mmTris-HCl, pH 8.2, 1 mm EDTA) and allowed to incubate at room temperature for 1 h in the dark. The chromatin fraction was applied directly to a hydroxylapatite column. The histones were isolated by acid extraction as described above. Histones were electrophoretically resolved on SDS-polyacrylamide gel electrophoresis. Following staining with Coomassie Blue, the gel pieces containing a histone band were disrupted in hydrogen peroxide and then counted in 5 ml of scintillation fluid. The chromatin fraction was mixed with hydroxylapatite HTP gel powder (Bio-Rad) at a ratio of 1 mg of DNA to 0.25 g of hydroxylapatite as described previously (27Li W. Nagaraja S. Delcuve G.P. Hendzel M.J. Davie J.R. Biochem. J. 1993; 296: 737-744Crossref PubMed Scopus (99) Google Scholar). The column was washed with 0.63 m NaCl in 0.1 mpotassium phosphate buffer, pH 6.7, to remove histone H1, H5, and non-histone chromosomal proteins before applying a linear gradient of NaCl (0.63 to 2 m NaCl in 0.1 m potassium phosphate buffer, pH 6.7) at a flow rate of 35 ml/h as described previously (27Li W. Nagaraja S. Delcuve G.P. Hendzel M.J. Davie J.R. Biochem. J. 1993; 296: 737-744Crossref PubMed Scopus (99) Google Scholar). DNA was prepared from the different chromatin fractions as described previously (24Delcuve G.P. Davie J.R. Biochem. J. 1989; 263: 179-186Crossref PubMed Scopus (50) Google Scholar). For electrophoresis, equal amounts of DNA were dissolved in DNA sample loading buffer, and the samples were loaded onto 1% agarose minigels containing 0.5 μg of ethidium bromide/ml. The DNA was transferred to Hybond-N+ nylon transfer membrane and hybridized to radiolabeled probes as described previously (28Brown T. Ausubel F.M. Brent R. Kingston R.E. Moore D.D. Seidman J.G. Smith J.A. Struhl K. Current Protocols in Molecular Biology. John Wiley & Sons, New York1995: 2.9.7-2.10.2Google Scholar). The cloned probes used were pCBG 14.4, a unique intronic sequence of chicken β globin gene; pChV2.5B/H, which contains the gene coding for chicken histone H5 and flanking sequences; and pVTG412, that recognizes the 5′ region of the chicken vitellogenin gene (24Delcuve G.P. Davie J.R. Biochem. J. 1989; 263: 179-186Crossref PubMed Scopus (50) Google Scholar). AUT (acetic acid, 6.7 m urea, 0.375% (w/v) Triton X-100) and SDS-15% polyacrylamide gel electrophoresis were performed as described elsewhere (24Delcuve G.P. Davie J.R. Biochem. J. 1989; 263: 179-186Crossref PubMed Scopus (50) Google Scholar). Antiacetylated H3 and antiacetylated H4 antibodies generously supplied by Dr. D. Allis were used to detect acetylated species of H3 and H4 (29Lin R. Leone J.W. Cook R.G. Allis C.D. J. Cell Biol. 1989; 108: 1577-1588Crossref PubMed Scopus (121) Google Scholar, 30Perry C.A. Dadd C.A. Allis C.D. Annunziato A.T. Biochemistry. 1993; 32: 13605-13614Crossref PubMed Scopus (52) Google Scholar, 31Braunstein M. Sobel R.E. Allis C.D. Turner B.M. Broach J.R. Mol. Cell. Biol. 1996; 16: 4349-4356Crossref PubMed Scopus (330) Google Scholar) in Western blot experiments using a protocol described previously (32Delcuve G.P. Davie J.R. Anal. Biochem. 1992; 200: 339-341Crossref PubMed Scopus (47) Google Scholar). Histone acetylation is rapidly reversible in immature erythrocytes (3Zhang D.-E. Nelson D.A. Biochem. J. 1988; 250: 233-240Crossref PubMed Scopus (55) Google Scholar, 4Zhang D.-E. Nelson D.A. Biochem. J. 1988; 250: 241-245Crossref PubMed Scopus (41) Google Scholar, 7Hendzel M.J. Delcuve G.P. Davie J.R. J. Biol. Chem. 1991; 266: 21936-21942Abstract Full Text PDF PubMed Google Scholar). To find the steady state of acetylated H3 and H4, histones of immature and mature erythrocyte chromatin fractions S150 and PE were electrophoretically resolved on AUT-polyacrylamide gels, transferred to nitrocellulose, and immunochemically stained with either antiacetylated H4 antibodies or antiacetylated H3 antibodies. Both antibodies detect the multiacetylated forms of H4 and H3, with the antibody to acetylated H3 showing a strong preference for the highest acetylated isoforms of H3 (30Perry C.A. Dadd C.A. Allis C.D. Annunziato A.T. Biochemistry. 1993; 32: 13605-13614Crossref PubMed Scopus (52) Google Scholar). We have shown previously that highly acetylated histones are found in chromatin fractions S150 and PE, but not in fraction P150 (7Hendzel M.J. Delcuve G.P. Davie J.R. J. Biol. Chem. 1991; 266: 21936-21942Abstract Full Text PDF PubMed Google Scholar,24Delcuve G.P. Davie J.R. Biochem. J. 1989; 263: 179-186Crossref PubMed Scopus (50) Google Scholar). Thus, this latter fraction was not analyzed. Fig.1 shows that the steady state levels of highly acetylated H3 and H4 isoforms were low in fractions S150 and PE from immature erythrocytes. However, the steady state levels of the highly acetylated H3 and H4 isoforms in fractions S150 and PE were markedly increased when immature erythrocytes were incubated in the presence of sodium butyrate, a histone deacetylase inhibitor, for 60 min (Fig. 1). These results suggest that the rate of deacetylation is so rapid in immature erythrocytes that, once dynamically acetylated H3 and H4 reach a highly acetylated state, they are rapidly deacetylated (4Zhang D.-E. Nelson D.A. Biochem. J. 1988; 250: 241-245Crossref PubMed Scopus (41) Google Scholar); the net result is a low steady state level of highly acetylated histone isoforms in untreated immature cells. The steady state of acetylated H3 and H4 in mature erythrocyte chromatin fractions was higher than that of H3 and H4 in the corresponding fractions of immature erythrocytes (Fig. 1). Incubation of mature cells with sodium butyrate elevated the level of hyperacetylated H3 and H4 isoforms in chromatin fraction S150. The level of highly acetylated H3 and, to a lesser extent, highly acetylated H4 in fraction PE was also increased. The chromatin fraction S150 was isolated from chicken immature erythrocytes that were untreated or incubated with sodium butyrate for 90 min. The S150 chromatin fractions were applied to mercury-agarose columns (Fig.2). Analysis of the proteins released from the mercury column with DTT showed that histones were present in the S150 chromatin fraction from butyrate incubated cells (Fig.2 C), but absent in the S150 fraction from untreated cells (Fig. 2 B). The major proteins in the latter fraction were the cysteine-containing high mobility group proteins 1 and 2 (Fig.2 B). These results suggested that unfolded nucleosomes were absent or at very low levels in the S150 chromatin fraction from untreated immature erythrocytes. However, the unfolded nucleosome appeared to be present in immature erythroid cells that were incubated with butyrate. This result suggested that when deacetylation of dynamically acetylated histones was halted, the transcriptionally active nucleosomes was prevented from reverting to a thiol-nonreactive state. To find if the mercury-agarose-bound nucleosomes from fraction S150 of butyrate-incubated immature erythrocytes had hyperacetylated histones, histones from S150, mercury-agarose-bound, and mercury-agarose-unbound chromatin fractions were analyzed in Western blot experiments with antiacetylated H3 and H4 antibodies. Fig.3 shows that bound nucleosomes were enriched in hyperacetylated H3 and H4 isoforms. DNA isolated from the S150, unbound and bound mercury-agarose chromatin fractions was analyzed by Southern blotting with probes containing DNA sequences to genes that were either expressed or repressed in immature erythrocytes. The chromatin fraction bound to mercury-agarose contained transcriptionally active histone H5 and β-globin (not shown) DNA, but not repressed vitellogenin DNA (Fig. 4). These results show that mercury-bound nucleosomes were associated with transcriptionally active DNA. The characterization of the thiol-reactive, unfolded nucleosome of transcribing chromatin described by Allfrey and colleagues has been done solely with soluble chromatin fragments. However, most transcribing chromatin is associated with the residual nuclear material (fraction PE), the nuclear skeleton. In immature erythrocytes approximately 75% of the transcribed DNA sequences are associated with the nuclear skeleton (7Hendzel M.J. Delcuve G.P. Davie J.R. J. Biol. Chem. 1991; 266: 21936-21942Abstract Full Text PDF PubMed Google Scholar). To test the reactivity of the thiol group (Cys-110) of H3 in chromatin from butyrate-incubated immature erythrocytes, fraction SE and PE chromatin fragments were incubated with [3H]iodoacetic acid. Fig.5, B and C, shows that H3 was labeled in SE and PE chromatin. In contrast to the results obtained with fraction SE and PE chromatin, the H3 of chromatin fraction P150, which contained repressed DNA, was not labeled (data not shown). To monitor the labeling of H3 in the chromatin fractions, hydroxylapatite dissociation chromatography was applied (27Li W. Nagaraja S. Delcuve G.P. Hendzel M.J. Davie J.R. Biochem. J. 1993; 296: 737-744Crossref PubMed Scopus (99) Google Scholar). Hydroxylapatite was added to fraction SE or PE in 0.63 mNaCl, removing non-histone chromosomal proteins and H1 histones from the hydroxylapatite-bound chromatin (27Li W. Nagaraja S. Delcuve G.P. Hendzel M.J. Davie J.R. Biochem. J. 1993; 296: 737-744Crossref PubMed Scopus (99) Google Scholar). Increasing concentrations of NaCl were then applied to the hydroxylapatite column, resulting first in the dissociation of H2A-H2B dimers followed by H3-H4 tetramers from the hydroxylapatite-bound chromatin (Fig.6, A and C) (27Li W. Nagaraja S. Delcuve G.P. Hendzel M.J. Davie J.R. Biochem. J. 1993; 296: 737-744Crossref PubMed Scopus (99) Google Scholar). The interpeak fractions contained H2A, H2B, H3, and H4 (Fig.6 C, lane b). The concentration of NaCl required to dissociate the H2A-H2B dimer or H3-H4 tetramer from the nucleosomal DNA provides a measure of the strength of the interaction between the dimer or tetramer and DNA (27Li W. Nagaraja S. Delcuve G.P. Hendzel M.J. Davie J.R. Biochem. J. 1993; 296: 737-744Crossref PubMed Scopus (99) Google Scholar). For example, highly acetylated H3-H4 tetramers dissociate from hydroxylapatite-bound nucleosomal DNA at a lower ionic strength than do unmodified H3-H4 tetramers (27Li W. Nagaraja S. Delcuve G.P. Hendzel M.J. Davie J.R. Biochem. J. 1993; 296: 737-744Crossref PubMed Scopus (99) Google Scholar, 33Hirose M. J. Biochem. (Tokyo). 1988; 103: 31-35Crossref PubMed Scopus (8) Google Scholar). Chromatin fragments of fractions SE and PE isolated from butyrate-incubated immature erythrocytes were incubated with [3H]iodoacetate and then subjected to hydroxylapatite dissociation chromatography. Fig. 6, A and B, shows that labeled H3 dissociated from the hydroxylapatite-bound chromatin after the H2A-H2B dimers but before the bulk of the H3-H4 tetramers. To determine whether incubation of immature erythrocytes with histone deacetylase inhibitors was required to detect the thiol-reactive H3 in nucleosomes, chromatin fractions SE and S150 were isolated from cells that were untreated or incubated with sodium butyrate. Following incubation with [3H]iodoacetate, the chromatin fractions were applied to hydroxylapatite. Labeled H3 was detected only in fraction SE and S150 when cells were incubated with sodium butyrate (compare Fig. 6 A with 7C and Fig. 7, Awith B). Immature erythrocytes were also incubated with trichostatin A, a specific histone deacetylase inhibitor, instead of sodium butyrate. The dissociation profiles of hydroxylapatite bound SE chromatin fragments from trichostatin A-treated immature erythrocytes were similar to those from butyrate-treated cells (compare Fig.7 D with 6A). Hydroxylapatite dissociation chromatography with PE chromatin was problematic as the addition of insoluble nuclear skeletons to hydroxylapatite reduced the flow rate appreciably. For the following experiments with the PE fraction, the suspension of nuclear skeletons was incubated with [3H]iodoacetate, and then the acid-extracted histones were separated by SDS-gel electrophoresis. The H3 band was excised and counted (see Fig. 5). In accordance with the results with fraction SE and S150, H3 of fraction PE was thiol-reactive in chromatin fragments from butyrate-incubated, but not untreated, immature erythrocytes (Fig. 8). These observations suggest that the thiol reactivity of H3 in immature erythrocyte chromatin fractions SE, S150, and PE is dependent upon the acetylation states of the nucleosomal histones. Transcriptional elongation is arrested in mature erythrocytes. To find if hyperacetylating histones associated with transcriptionally competent chromatin (3Zhang D.-E. Nelson D.A. Biochem. J. 1988; 250: 233-240Crossref PubMed Scopus (55) Google Scholar, 34Hebbes T.R. Thorne A.W. Crane Robinson C. EMBO J. 1988; 7: 1395-1402Crossref PubMed Scopus (715) Google Scholar) was sufficient to observe a thiol-reactive H3, chromatin fractions S150 and PE from mature cells untreated or incubated with sodium butyrate were labeled with [3H]iodoacetate. Figs. 8 and9 show that labeled H3 was not observed in the mature erythrocyte S150 and PE chromatin fractions. These results and those with immature erythrocyte chromatin suggest that histone acetylation is required but not sufficient for formation and/or stabilization of nucleosomes with thiol-reactive H3. The absence of thiol-reactive nucleosomes in butyrate-treated mature erythrocytes indicates that both transcription elongation and highly acetylated histones are required to form and maintain the unfolded nucleosome conformation. To test whether inhibition of transcription elongation has an effect on the H3 thiol reactivity of immature erythrocyte nucleosomes, immature erythrocytes were incubated with inhibitors of transcription before the addition of sodium butyrate. Camptothecin is an inhibitor of topoisomerase I and has been reported to stimulate initiation but inhibit elongation by RNA polymerase II (26Ljungman M. Hanawalt P.C. Carcinogenesis. 1996; 17: 31-35Crossref PubMed Scopus (65) Google Scholar, 35Stewart A.F. Schutz G. Cell. 1987; 50: 1109-1117Abstract Full Text PDF PubMed Scopus (122) Google Scholar, 36Stewart A.F. Herrera R.E. Nordheim A. Cell. 1990; 60: 141-149Abstract Full Text PDF PubMed Scopus (111) Google Scholar). A thiol-reactive H3 was not detected in the SE and PE chromatin of immature cells treated with DRB or camptothecin followed by butyrate (Figs. 8 and 10) (25Chadee D.N. Allis C.D. Wright J.A. Davie J.R. J. Biol. Chem. 1997; 272: 8113-8116Abstract Full Text Full Text PDF PubMed Scopus (43) Google Scholar). These results show that both highly acetylated histones and elongation are needed to detect the thiol-reactive H3 in immature erythrocyte chromatin. We show that highly acetylated histones are required to maintain the unfolded, thiol-reactive structure of transcribing nucleosomes. The thiol-reactive nucleosome is not detected in the chromatin of transcriptionally active immature erythrocytes where the steady state level of highly acetylated histones is low. But when deacetylation of the highly acetylated histones is prevented by incubating immature erythrocytes with histone deacetylase inhibitors, the thiol-reactive nucleosome is detected. The rate of deacetylation in immature erythrocytes is such that the highly acetylated H3 and H4 isoforms are short lived (4Zhang D.-E. Nelson D.A. Biochem. J. 1988; 250: 241-245Crossref PubMed Scopus (41) Google Scholar). In contrast, the thiol-reactive nucleosome is detected in the chromatin of mammalian cells without the use of histone deacetylase inhibitors (37Allegra P. Sterner R. Clayton D.F. Allfrey V.G. J. Mol. Biol. 1987; 196: 379-388Crossref PubMed Scopus (134) Google Scholar) (unpublished observations). These observations suggest that the net activities of the histone acetyltransferases and deacetylases decide the longevity of the unfolded nucleosome. In transcribing regions where the rate of histone deacetylation exceeds the rate of acetylation, the unfolded nucleosome structure will be short lived and will rapidly revert to a thiol nonreactive state following passage of the RNA polymerase. In yeast the converse is the case. The unfolded nucleosome structure associated with a specific gene persists well after the transcription of that gene has been arrested (38Chen T.A. Smith M.M. Le S.Y. Sternglanz R. Allfrey V.G. J. Biol. Chem. 1991; 266: 6489-6498Abstract Full Text PDF PubMed Google Scholar). The rates of histone acetylation and deacetylation are very slow in yeast, but the high steady state of highly acetylated histones argues that the rate of histone acetylation exceeds the rate of deacetylation (39Davie J.R. Saunders C.A. Walsh J.M. Weber S.C. Nucleic Acids Res. 1981; 9: 3205-3216Crossref PubMed Scopus (54) Google Scholar, 40Nelson D.A. J. Biol. Chem. 1982; 257: 1565-1568Abstract Full Text PDF PubMed Google Scholar). Histone acetylation, however, is not sufficient to generate the unfolded nucleosome structure; transcription elongation is required. Thiol-reactive nucleosomes could not be found in the chromatin of chicken mature erythrocytes. Transcription may be initiated in these mature cells, but RNA polymerases are paused at the 5′ end of the transcribed genes (41Gariglio P. Bellard M. Chambon P. Nucleic Acids Res. 1981; 9: 2589-2598Crossref PubMed Scopus (98) Google Scholar, 42Sun J.-M. Ferraiuolo R. Davie J.R. Chromosoma. 1996; 104: 504-510Crossref PubMed Scopus (5) Google Scholar). The treatment of immature erythrocytes with camptothecin may mimic the mature erythrocyte situation, that is, transcription initiation occurs but elongation does not (26Ljungman M. Hanawalt P.C. Carcinogenesis. 1996; 17: 31-35Crossref PubMed Scopus (65) Google Scholar). Thus, although dynamic histone acetylation and initiation are occurring in these cells at transcriptionally competent/active loci, without elongation the thiol-reactive nucleosome is not formed. Our results suggest that unfolded transcribing nucleosomes are associated with highly acetylated H3-H4 tetramers. In agreement with the studies of Allfrey and colleagues (19Chen-Cleland T.A. Boffa L.C. Carpaneto E.M. Mariani M.R. Valentin E. Mendez E. Allfrey V.G. J. Biol. Chem. 1993; 268: 23409-23416Abstract Full Text PDF PubMed Google Scholar), we found that immature erythrocyte chromatin fragments bound to mercury agarose are enriched in highly acetylated H3 and H4. Furthermore, Sterner et al.(20Sterner R. Boffa L.C. Chen T.A. Allfrey V.G. Nucleic. Acids. Res. 1987; 15: 4375-4391Crossref PubMed Scopus (29) Google Scholar) showed that the thiol-reactive H3 of unfolded mammalian nucleosomes is hyperacetylated. Acetylation of H3 and H4 may maintain the unfolded nucleosome conformation by breaking interactions between the histone N-terminal tail and nucleosomal DNA. The N-terminal tail of H4 is not mobile in nucleosomes, and there is evidence that the H4 N-terminal tail makes intranucleosomal contacts (43Smith R.M. Rill R.L. J. Biol. Chem. 1989; 264: 10574-10581Abstract Full Text PDF PubMed Google Scholar). Indeed, His-18 in the N-terminal region of H4 cross-links to nucleotides 57, 66, and 93 from the 5′ end of nucleosomal DNA (44Ebralidse K.K. Grachev S.A. Mirzabekov A.D. Nature. 1988; 331: 365-367Crossref PubMed Scopus (97) Google Scholar, 45Gavin I.M. Usachenko S.I. Bavykin S.G. J. Biol. Chem. 1998; 273: 2429-2434Abstract Full Text Full Text PDF PubMed Scopus (15) Google Scholar). This position in the nucleosomal DNA corresponds to where the nucleosomal DNA is sharply bent or kinked. In active gene chromatin and in chromosomal domains containing hyperacetylated histones, the cross-linking between His-18 of H4 and nucleosomal DNA in situ is greatly diminished (44Ebralidse K.K. Grachev S.A. Mirzabekov A.D. Nature. 1988; 331: 365-367Crossref PubMed Scopus (97) Google Scholar,46Nacheva 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, 47Ebralidse K.K. Hebbes T.R. Clayton A.L. Thorne A.W. Crane-Robinson C. Nucleic Acids Res. 1993; 21: 4734-4738Crossref PubMed Scopus (15) Google Scholar). Moreover, site 60 from the end of nucleosomal DNA of hyperacetylated nucleosomes has an increased susceptibility to DNase I (48Ausio J. Van Holde K.E. Biochemistry. 1986; 25: 1421-1428Crossref PubMed Scopus (199) Google Scholar). These observations strongly suggest that acetylation at lysines located in N-terminal tail of H4 may have important functions in altering histone-DNA contacts and nucleosome structure. Further, hyperacetylation of the H3-H4 tetramer reduces the linking number change per nucleosome, that is, negative DNA supercoils constrained in unmodified nucleosomes are partially released in nucleosomes with hyperacetylated histones (49Norton V.G. Marvin K.W. Yau P. Bradbury E.M. J. Biol. Chem. 1990; 265: 19848-19852Abstract Full Text PDF PubMed Google Scholar, 50Thomsen B. Bendixen C. Westergaard O. Eur. J. Biochem. 1991; 201: 107-111Crossref PubMed Scopus (20) Google Scholar). The destabilizing effect that histone acetylation has on H3-H4 tetramer-DNA interactions in transcribing nucleosomes is seen in hydroxylapatite dissociation chromatography. The H3 of thiol-reactive nucleosomes dissociated from hydroxylapatite bound S150 or SE chromatin after the dissociation of the H2A-H2B dimers but before the bulk of the H3-H4 tetramers. In a previous study we monitored the dissociation of labeled ([3H]acetate) dynamically acetylated histones from hydroxylapatite-bound chromatin of immature erythrocytes. The dissociation of the labeled ([3H]acetate) highly acetylated H3-H4 tetramers coincided exactly with that of labeled ([3H]iodoacetate) H3 (27Li W. Nagaraja S. Delcuve G.P. Hendzel M.J. Davie J.R. Biochem. J. 1993; 296: 737-744Crossref PubMed Scopus (99) Google Scholar). These observations with SE chromatin from chicken immature erythrocytes suggest that the interaction between highly acetylated H3-H4 tetramer and DNA of transcribing nucleosomes is weaker than that of typical nucleosomes. Analysis of mercury-agarose bound nucleosomes by electron spectroscopic imaging also indicated that the H3-H4 tetramer of unfolded nucleosomes is disrupted (18Bazett-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). The disruption of the tetramer in transcribing nucleosomes may facilitate subsequent rounds of elongation. Transcribing chromatin is associated with the insoluble residual nuclear material (fraction PE) which contains the nuclear matrix. The PE chromatin from butyrate treated immature erythrocytes had 76% of the active DNA and 74% of the acetate-labeled tetraacetylated H4 (7Hendzel M.J. Delcuve G.P. Davie J.R. J. Biol. Chem. 1991; 266: 21936-21942Abstract Full Text PDF PubMed Google Scholar). Further, the PE fraction retained 75–85% of the nuclear histone acetyltransferase and histone deacetylase activity (7Hendzel M.J. Delcuve G.P. Davie J.R. J. Biol. Chem. 1991; 266: 21936-21942Abstract Full Text PDF PubMed Google Scholar, 13Hendzel M.J. Sun J.-M. Chen H.Y. Rattner J.B. Davie J.R. J. Biol. Chem. 1994; 269: 22894-22901Abstract Full Text PDF PubMed Google Scholar). The thiol-reactive nucleosome was detected in PE chromatin of butyrate-treated immature erythrocytes, but not in the PE chromatin of untreated immature erythrocytes or mature erythrocytes. Further, inhibition of transcription with camptothecin or DRB prevented the detection of the unfolded nucleosome. Thus, the results obtained with PE chromatin were similar to those observed with S150 or SE chromatin; both hyperacetylated histones and elongation are required to detect the unfolded nucleosome in fraction PE. There is increasing evidence that the transcription machinery is associated with the nuclear matrix and that for chromatin to be transcribed it is spooled through the anchored large RNA polymerase complex (51Iborra F.J. Pombo A. Jackson D.A. Cook P.R. J. Cell Sci. 1996; 109: 1427-1436Crossref PubMed Google Scholar, 52Mortillaro M.J. Blencowe B.J. Wei X.Y. Nakayasu H. Du L. Warren S.L. Sharp P.A. Berezney R. Proc. Natl. Acad. Sci. U. S. A. 1996; 93: 8253-8257Crossref PubMed Scopus (282) Google Scholar, 53Halle J.-P. Meisterernst M. Trends Genet. 1996; 12: 161-163Abstract Full Text PDF PubMed Scopus (41) Google Scholar, 54Maldonado E. Shiekhattar R. Sheldon M. Cho H. Drapkin R. Rickert P. Lees E. Anderson C.W. Linn S. Reinberg D. Nature. 1996; 381: 86-89Crossref PubMed Scopus (307) Google Scholar). We have proposed that histone acetyltransferase and deacetylase are localized in these transcription foci (13Hendzel M.J. Sun J.-M. Chen H.Y. Rattner J.B. Davie J.R. J. Biol. Chem. 1994; 269: 22894-22901Abstract Full Text PDF PubMed Google Scholar, 14Davie J.R. Hendzel M.J. J. Cell. Biochem. 1994; 55: 98-105Crossref PubMed Scopus (74) Google Scholar). Recent studies show that coactivators (CBP/p300, ACTR, and GCN5) and proteins associated with TATA-binding protein (TAFII250) have histone acetyltransferase activity (55Brownell J.E. Zhou J.X. Ranalli T. Kobayashi R. Edmondson D.G. Roth S.Y. Allis C.D. Cell. 1996; 84: 843-851Abstract Full Text Full Text PDF PubMed Scopus (1292) Google Scholar, 56Mizzen C.A. Yang X.J. Kokubo T. Brownell J.E. Bannister A.J. Owen-Hughes T. Workman J. Wang L. Berger S.L. Kouzarides T. Nakatani Y. Allis C.D. Cell. 1996; 87: 1261-1270Abstract Full Text Full Text PDF PubMed Scopus (624) Google Scholar, 57Bannister A.J. Kouzarides T. Nature. 1996; 384: 641-643Crossref PubMed Scopus (1535) Google Scholar, 58Ogryzko V.V. Schiltz R.L. Russanova V. Howard B.H. Nakatani Y. Cell. 1996; 87: 953-959Abstract Full Text Full Text PDF PubMed Scopus (2409) Google Scholar, 59Spencer T.E. Jenster G. Burcin M.M. Allis C.D. Zhou J. Mizzen C.A. McKenna N.J. Onate S.A. Tsai S.Y. Tsai M.J. O'Malley B.W. Nature. 1997; 389: 194-198Crossref PubMed Scopus (1070) Google Scholar, 60Chen H. Lin R.J. Schiltz R.L. Chakravarti D. Nash A. Nagy L. Privalsky M.L. Nakatani Y. Evans R.M. Cell. 1997; 90: 569-580Abstract Full Text Full Text PDF PubMed Scopus (1270) Google Scholar). Histone deacetylases (HDAC-1 and HDAC-2) are associated with corepressors (mSin3A and N-CoR) and the nuclear matrix bound transcription factor YY1 (61Heinzel T. Lavinsky R.M. Mullen T.-M. Soderstrom M. Laherty C.D. Torchia J. Yang W.-M. Brard G. Ngo S.D. Davie J.R. Seto E. Eisenman R.N. Rose D.W. Glass C.K. Rosenfeld M.G. Nature. 1997; 387: 43-48Crossref PubMed Scopus (1086) Google Scholar, 62Laherty C.D. Yang W.-M. Sun J.-M. Davie J.R. Seto E. Eisenman R.N. Cell. 1997; 89: 349-356Abstract Full Text Full Text PDF PubMed Scopus (851) Google Scholar, 63Yang W.-M. Yao Y.-L. Sun J.-M. Davie J.R. Seto E. J. Biol. Chem. 1997; 272: 28001-28007Abstract Full Text Full Text PDF PubMed Scopus (402) Google Scholar, 64McNeil S. Gou B. Stein J.L. Lian J.B. Bushmeyer S. Seto E. Atchison M.L. Penman S. van Wijnen A.J. Stein G.S. J. Cell. Biochem. 1998; 68: 500-510Crossref PubMed Scopus (85) Google Scholar). These studies suggest that the basal transcription machinery and transcription factors recruit histone acetyltransferases and histone deacetylases to sites of transcription at the nuclear matrix. Nucleosome structure will be perturbed when the chromatin fiber is passed through the fixed RNA polymerase (transcriptosome) (53Halle J.-P. Meisterernst M. Trends Genet. 1996; 12: 161-163Abstract Full Text PDF PubMed Scopus (41) Google Scholar). While in a highly acetylated state, the unfolded nucleosome structure will persist, helping subsequent rounds of transcription. Our results are consistent with the idea that the nucleosome is a dynamic structure conforming its structure to facilitate movement of chromatin through the RNA polymerase II elongation complex, with dynamic histone acetylation having a major role in modulating the unfolded structure of transcribing nucleosomes.