7 Number 8 1979 Volume Volume 7 Number 8 1979

Nucleic Acids Research

Nucleic Acids Research

Nucleosome dissociation and transfer in concentrated salt solutions

Pamela C.Stacks and Verne N.Schumaker Department of Chemistry and Molecular Biology Institute, University of California, Los Angeles, CA 90024, USA

Received 14 August 1979

ABSTRACT We have examined the dissociation of nucleosomes into histones and free, 4.5S DNA over a range of sodium chloride concentrations between 0.25 and 1 M. We have also studied this dissociation as a function of nucleosome concentration at two salt concentrations, 0.8 M and 0.9 M. In addition, we have measured the kinetics of transfer of histone cores from nucleosomes onto recipient bacteriophage T7 DNA in 0.6, 0.7 and 0.8 M NaCl solutions. Al though the mechanism of nucleosome transfer is unknown, the data presented here are consistent with either a reversible dissociation of the nucleosome or DNA strand displacement by another DNA.

INTRODUCTION Electrostatic interactions between DNA and histones are weakened with increasing ionic strength. Thus, as the ionic strength is raised between 0.55 M to 0.65 M, histones Hi and H5 become dissociated from chromatin1l3. At about 0.6 M, nucleosome sliding comnences4-7, although one report indicated sliding occurs in 0.15 M NaCl8. Between 0.7 M and 1.2 M NaCl histones H2A and H2B are eluted from chromatin using zonal techniques.2'28 Nucleosome histone cores are successfully transfered to exogenous DNA at an ionic strength of 0.85 M NaCl and this ionic strength has been employed to initiate the reannealing of completely dissociated histones and DNA. 9 At ionic strengths higher than 1.2 M NaCl histones H3 and H4 are also eluted from chromatin.2'28 In the absence of DNA, complex equilibria are established between various oligomeric forms of histones which are dependent upon ionic strength and pH10-15. In the presence of nucleosonies, however, a tight association is thought to be formed between histone core octamer and intact nucleosomes which stabilizes the octamer, at least at pH 8 and 0.6 M ionic

strength16. In this communication we report the results of our studies on the C Information Retrieval Umited 1 Falconberg Court London Wl V 5FG England

2457

Nucleic Acids Research extent of dissociation of nucleosomes over a range of ionic strengths between 0.25 and 1.0 M, and upon the equilibria obtained as a function of nucleosome concentration. The results of studies on the kinetics of transfer of histone core from nucleosomes to recipient DNA molecules are also presented.

MATERIALS AND METHODS

Nucleosome preparation: Chicken erythrocyte nucleosomes were prepared from 75 ml of chicken blood as described by Olins and coworkers17, except that: (a) The STM and STMN buffers were maintained at pH 8; (b) Worthington micrococcal nuclease was added to a ratio of 2.86 A260 units of nuclei/unit of nuclease. Incubation for 90 minutes at 370 C resulted in a 12.8% digestion of the DNA, calculated as described in reference 18; (c) After the lysed nuclei were clarified by centrifugation, 2 ml of the supernatent fluid was diluted 1:1 to a final concentration of 50 A260 units/ml in 0.2 mM EDTA, pH 6.8, and then 2 ml were layered on top of each 7.5% to 25% (w/v) linear sucrose gradient in a cellulose nitrate SW 27 tube (Beckman). The gradients also contained 2 mM EDTA, pH 6.8, 0.1 nM phenylmethylsulfonyl fluoride (PMSF), and 0.2% isopropanol. The gradients were centrifuged for 16 hours at 27,000 rpm and 50 C using a SW 27 rotor. The main peak in the sucrose gradient contained 25 A260 units. It was isolated, dialyzed into 0.2 mM EDTA, pH 6.8, and made 0.1 mM PMSF, 0.2% isopropanol. The nucleosomes were examined by analytical ultracentrifugation and by gel electrophoresis for DNA19 and histones20. A single sedimenting peak was observed, with a s20,w of 11.7 S in lbmM Tris, 0.25 mM EDTA pH8, containing less than 5% dimer. The distribution of nucleosomal DNA sizes ranged from 130-180 base pairs as indicated by gel electrophoresis mobility. Equimolar quantities of the core histones appeared to be present and free from proteolytic digestion. Although, trace amounts of Hi and H5 were present, no other protein bands were observed. Bacteriophage T7 DNA preparation: After growth the bacteriophage were concentrated with polyethylene glycol 6000 and by centrifugation in a 70 Ti rotor (Beckman) for 40 minutes at 40,000 rpm and 50 C21. The bacteriophage were then banded in a CsCl step gradient. T7 DNA was prepared by phenol extraction22 followed by a chloroform:isoamyl alcohol (24:1) extraction. The DNA was dialyzed into 10 mM NaCl, 10 mM Tris, 0.1 mM EDTA, pH 8. Nucleosome dissociation studies: In order to achieve the desired salt concentrations, a stock solution of 4.4 M NaCl, 10 mM Tris, pH 8 was 2458

Nucleic Acids Research slowly added to nucleosomes in 10 rrM Tris, pH 8.0. A 1:100 addition of 100 mM PMSF in isopropanol was made to inhibit any endogeneous proteolytic activity. The samples were incubated at 220 C for 16 hours and subsequently examined in the analytical ultracentrifuge without dilution. Transfer studies: In order to achieve the desired salt concentrations a stock solution of 4.4 M NaCl , 10 mM Tris was slowly added to a mixture of T7 DNA and isolated chicken erythrocyte nucleosomes. The final incubation concentrations of T7 DNA and nucleosomes were 100 and 200 micrograms/ml, respectively. The samples were incubated at 220 C. To terminate the incubations, aliquots were diluted slowly with gentle mixing to 0.25 M NaCl, using 10 nM Tris, pH 8.0. Analytical ultracentrifugation: The Beckman Model E photoelectric scanning system at 265 n, was used to follow the sedimenting boundaries. For the more concentrated solutions, a wavelength between 280 to 300 nm The temperature of the was selected to keep the absorbancy below 0.8. centrifugation was monitored but not regulated, and was routinely kept The obserbetween 20 and 220 C with a variation of less than 0.3° C. ved sedimentation coefficients were corrected to s20,w and observed concentrations were corrected to initial concentrations using the square law of radial dilution23.

RESULTS In Table 1 are listed the results Nucleosome dissociation studies: of the ultracentrifuge analyses of two separate, 16 hour, 220 C incubations of nucleosomes at a concentration of 50 micrograms/ml over a range of ionic strengths between 0.01 and 1.0 M. At the lowest ionic strength in 10 mM Tris, 0.25 mM EDTA, pH 8.0, the nucleosomes sedimented as a single boundary in the analytical ultracentrifuge, and as indicated by the flat baseline, no free DNA was found. In 0.25 M NaCl, a sloping base line appeared in the solvent region, which comprised less than 7% of the A265 material in Table 1. At all higher ionic strengths between 0.5 to 0.9 M, two sedimenting boundaries appeared with increasing amounts of A265 material present in the slow boundary with increasing ionic strengths. The slower boundary had an average sedimentation coefficient of 4.5S, consistent with free nucleosomal DNA. Figure 1 shows a typical derivative plot of the two sedimenting boundaries, generated by computer from the photoelectric scanner data. The sedimentation coefficients of the fast boundaries, on the other 2459

Nucleic Acids Research TABLE I

Effect of Ionic Strength on the Sedimentation and Dissociation Behavior of Nucleosomes

NaCl a

S20,w of

(mol ar)

slow boundary

% DNA in

S20,w of fast boundary

0.0

11.7

0.25

9.9

slow boundary -b

0.5

4.3 ± 0.05

10.0 ± 0.06

16.5 ± 0.7

0.6

4.3 ± 0.40

9.5 ± 0.01

16.5 ± 2.1

0.7

4.2 ± 0.04

9.3 ± 0.02

20.5 ± 0.07

0.8

4.5 ± 0.9

8.9 ± 0.45

29.5 ± 2.1

0.9

4.6 ± 0.32

8.5 ± 0.15

40.5 ± 6.4

1.0 -

4.8

8.1

51

a

10 mM Tris, pH

8.0, was also present.

b A sloping baseline was observed in the solvent region.

hand, clearly decreased with increasing ionic strength, from 10.OS at 0.5 M NaCl to 8.1S at 1.0 M NaCl. The extent of nucleosome dissociation as monitored by the release of 4.5S DNA as a function of nucleosome concentration, is shown in Figure 2.

0 N0

I

I

7.0

6.4

RADIUS (cm)

2460

Figure 1: A derivative plot of two sedimenting boundaries which were formed when nucleosomes at a concentration of 50 micrograms/ml in 0.9 MNaCl pH8and at 220 C were studied in the analytical ultracentrifuge. The data were obtained after 36 minutes of centrifugation at 52,000 rpm.

Nucleic Acids Research 49 z 60 o

60 B ~~~~A

in40

40

20

20

200

400

200

400

[NUCLEOSOME] ug/mi

Figure 2: Effect of nucleosome concentration on the amount of free nucleosomal DNA present at 220 C in (a) 0.8 M NaCl or (b) 0.9 M NaCl as measured by analytical ultracentrifugation.

Nucleosomes were incubated at 220 C for 16 hours in 0.8 M NaCl (Fig. 2a) or in 0.9 M NaCl (Fig. 2b) over a range of nucleosome concentrations between 25 to 500 micrograms/ml, at pH 8. Then, the samples were analyzed in 0.8 or 0.9 M NaCl in the ultracentrifuge. The percent of DNA present in- the slow moving boundary is plotted in Figure 2 as a function of nucleosome concentration. Nucleosome transfer studies: Chicken erythrocyte nucleosomes and T7 DNA of 25 million daltons molecular weight were uised to examine the kinetics of transfer of histone cores as a function of ionic strength. Since the T7 DNA has an S20,w of about 30S, it could be separated from the 5S DNA and the 11S nucleosome in the analytical ultracentrifuge. The extent of transfer was estimated both from the amount of free nucleosomal DNA liberated and also from the enhanced sedimentation rate of the T7 DNA. The kinetics of transfer at 220 C are shown in Figure 3 for ionic strengths of 0.6, 0.7 and 0.8 M NaCl. At various times, aliquots were diluted to 0.25 M NaCl, and the amount of 4.5S DNA present determined by ul tracentri fugati on. In the absence of T7 DNA, nucleosomes incubated in 0.6 M NaCl for 1 hour,, and then gently diluted to 0.25 M NaCl, contained 15.2% free nucleosomal DNA. Similar controls for the 0.7 M and 0.8 M incubations showed 17.6% and 19% free DNA, respectively. Controls performed with a mixture of nucleosomes and nucleosomal DNA at concentrations of 200 and 100 micrograms/ml, respectively, at 0.6, 0.7 and 0.8 M NaCl demonstrated that little or no histone loss occurred upon 2461

Nucleic Acids Research 60

0

OI 0

I00

s0

z

A

10

B

: 20

40 2

I

I

Io 26 TIME (hr)

_

a

I

10

20

Figure 3: Kinetics of histone core transfer onto T7 DNA at 22 C in (a) 0.8 M (b) 0.7 M, or (c) 0.6 M NaCl, as measured by the amount of free 4.5S nucleosomal DNA present after dilution to 0.25 M NaCl.

' Go

z

0

i)

c

8>-20

1808

26 10 TIME (hr)

lowering the salt concentrations to 0.25 M NaCl. Thus, the amount of 4.5S DNA present after dilution to 0.25 M NaCl was a measure of the amount of transfer to the T7 DNA. In the presence of T7 DNA, nucleosomes incubated in 0.6 M NaCl did not show any increase in the amount of 4.5S nucleosomal DNA with increasing times of incubation, as shown in Figure 3C. There appeared to be very little transfer of histones to the T7 DNA during incubation at 0.6 M NaCl, 10 nM Tris, pH 8.0. For the 0.7 nM NaCl, 10 nmM Tris, pH 8.0 incubations, the zero time point was obtained by inmnediately diluting an aliquot after raising the salt concentration to 0.7 M NaCl. At this zero point, 15% of the input nucleosomes were present as free nucleosomal DNA. Subsequent time points showed a gradual increase in the amount of nucleosomal DNA which by 24 hours had leveled off at 51.4% of the input nucleosome concentration, as shown in Figure 3B. In contrast with the gradual release at 0.7 M, the 0.8 M NaCl samples showed a moderately rapid release of free nucleosomal DNA. The "zero" time incubation mixture contained 23% free nucleosomal DNA. By one hour, 48.5% of the input nucl eosomal DNA was present as free 4.5S DNA, and the amount quickly leveled off at 58.5% of the material, as shown in Figure 3A. The sedimentation behavior of the T7 DNA after the transfer reaction 2462

Nucleic Acids Research function of the number of nucleosomes acquired. Formation of T7 DNAnucleosome complexes resulted in an increase in molecular weight, a decrease in particle density, and a decrease in the frictional coefficient of the T7 DNA due to a seven-fold compaction of each section of DNA associated with a The appropriate equation allowing a conversion between the nucleosome24. sedimentation coefficient of the T7 DNA-nucleosome complex and the number of bound nucleosomes was derived in reference 25, and it is plotted in Figure 4. It has been reported that at least histones H3 and H4 are required for the observed increased sedimentation of reconstituted DNA26929. In Table 2 are listed the observed sedimentation coefficients of the T7 DNA-nucleosome complexes after 16 hours of incubation and dilution to 0.25 M NaCl. The number of histone cores transferred to the T7 DNA was estimated from the release of 4.5S DNA (Table 2, under the heading of "5 4.5S DNA. The number of histone cores transferred was also estimated from the values of S20,w measured for the T7 DNA-nucleosome complexes and the conversion curve of Figure 4. These values also are listed in Table 2. There is good agreement between the two sets of estimates for the amount of histone core transfer. That the transferred histones were present on the T7 DNA in the form of intact nucleosomes was strongly suggested by the results of digestion experiments with DNase I. An autoradiogram of a gel of DNA resulting from digestion of uniformly labelled 32P-T7 DNA-nucleosome complexes displayed a ladder pattern with DNA fragments, separated by 10 base pairs, ranging in size from 20 to 130 nucleotides (not shown). This was indicative of nucleosomal structure present on the T7 DNA18. A control mixture of 32P-T7 DNA and unlabelled nucleosomes which had been maintained at low ionic strength resulted in a smear of radioactive DNA of no specific length after DNase I digestion.

was a

_ /-

so

Figure 4: S20,w

Of

The calculated T7

DNA-nucleo-

some complexes is plotted as a function of the number

50so

/of nucleosomes associated /with the T7 DNA.

IC;

NUCLEOSOMES/T7 DNA

2463

Nucleic Acids Research TABLE 2 Nunber of Nucleosomes per T7 DNA Predicted from the Sedimentation Rates of the Complexes or the Release of 4.5S DNA

(oNaCr

SO2O,wa

n/T7 DNAb

% 4.5S DNAC

n/T7 DNAd

0.6

35.0

16

15

35

0.7

61.5

104

50

116

0.8

68.6

119

58

134

a

The T7 DNA concentration during the centrifugation was 25 jAg/ml. A correction of +13% was used to correct the observed sedimentation rates to values of s020,w.

b Values obtained by use of Fig. 4 and the values of listed in the preceding column. c

s°20,w

Values of % 4.5S DNA released obtained from Fig. 3.

d Values predicted from the percent release of 4.5S DNA, as listed in the preceding column.

DISCUSSION In the presence of high concentrations of sodium chloride, 0.6 to 1.0M, the light absorbing material was observed to sediment as two distinct boundaries (Table I). Since the histones absorb very little light at 265nm, this light absorbing material represents two different, DNA-containing particles. The sedimentation rate of the slow boundary was the same as that of phenolextracted, protein-free, nucleosomal DNA. This was an unexpected observation for it has been reported that while histones H2A and H2B are released at lower ionic strengths, H3 and H4 remain associated with the DNA until about 1.3M28. It does not appear possible that substantial quantities of H3 and H4 could be attached to the 5S DNA since it has been reported that the complex composed of a single H3-H4 tetramer and a single piece of nucleosomal DNA sediments at 9.8S29. Moreover, if a single H3-H4 dimer were to be associated with a 5S DNA, the sedimentation rate of the 5S DNA may be calculated to increase by 25% due to the increase in buoyant molecular weight even if 2464

Nucleic Acids Research the frictional coefficient remained that of 5S DNA. Therefore, we conclude that the 5S boundary represents essentially protein-free, nucleosomal DNA released by salt concentrations in the range of 0.6 to 1.OM. The appearance of a large amount of free, 5S DNA at the high ionic strengths could be due to a reversible dissociation of the nucleosomes with free DNA forming one of the products. In this case, it would be reasonable to expect that the equilibrium constant would be a function of ionic strength, with increasing dissociation occurring as the ionic strength were increased. Moreover, a reversible dissociation in which several products were formed from a single nucleosome would be expected to be a function of nucleosome concentration. This has been tested, as shown in Figures 2A and 2B, and, indeed, a strong concentration dependence has been found to exist. To us, this is convincing evidence for the existence of a reversible reaction in which the nucleosome dissociates to form a nunber of products including free, 5S DNA. However, the curves of Figure 2 appear to approach limiting values of about 16%, which would indicate the existence of a population of nucleosomes not participating in the equilibrium reaction. Therefore, we feel that the best explanation for the release of the free 5S DNA involves both a reversible dissociation of the nucleosomes and nucleosome heterogeneity. More than 80% of the nucleosomes appear to undergo a reversible dissociation over the range between 0.6 to 1.0 M. In addition, there appears to be a subpopulation of about 16% which is more susceptible to dissociation. We would expect the reversible dissociation to be a slow reaction because the two sedimenting boundaries were well resolved. Since the histones are not directly detected during centrifugation, the only product of the dissociation that was detected was the 5S DNA. Because not all of the products of the dissociation are known, it is not possible to make a boundary analysis along the lines suggested by Cann and Kegeles (27) to support our assumption that the dissociation is slow. More direct evidence comes from a study of the half-lives of the nucleosome dissociations as measured by the transfer experiments. The kinetics of nucleosomal transfer at high ionic strengths have been measured by monitoring the appearance of free nucleosomal DNA (Figure 3). In contrast to the experiments just discussed above, in which dissociation was studied during centrifugation at high ionic strengths, in the transfer experiments the centrifuge runs were performed after the ionic strength had been readjusted to 0.25M. The rate of transfer was gradual with time and was dependent upon the ionic strength with half-lives of 6.8 hours in O.7M NaCl 2465

Nucleic Acids Research and 45 minutes in O.8M NaCl. Since an equal weight ratio of donor and recipient DNA was used in these experiments and since the reaction appeared to reach completion at between 50 and 60% free nucleosomal DNA, we conclude there i s no preference for l onger over shorter p1 eces of DNA duri ng nucl eosome formation. The sedimentation rate of the recipient T7 DNA increased in a manner consistant with the appearance of free nucleosomal DNA (Table 2). Since the presence of histones H3 and H4 are required for increased sedimentation (26,29) and DNase I digestion indicated the presence of nucleosomal structures on the T7 DNA after transfer (see text), we have concluded that the transfer of nucleosomes onto the T7 DNA had occurred in these experiments. The mechanism of nucleosome transfer is not known. It could involve the complete dissociation of the histones from the DNA. On the other hand, transfer might be a displacement process, with one piece of DNA gradually displacing a second from the histone core, keeping most of the histone DNA bonds intact at all times. It seems evident from the data presented here that most of the nucleosomes neither dissociate nor transfer until the ionic strength is raised above O.6M. Between 0.7 to O.9M, the particles appear to reversibly di-ssociate. These transitions are not sudden, but rather depend upon the salt concentration and the nucleosome concentrations in a fashion consistent with mass law equilibria. The overall rates of both transfer and of dissociation and reassociation appear to be slow and dependent upon salt concentration. Careful study of this system as a function of temperature, ionic strength and concentration should yield valuable thermodynamic and kinetic information leading to a better understanding of the mechanism of binding of DNA by the core histones.

ACKNOWLEDGMENTS Our thanks to Anita Wadel for her efforts in the typing of this manuscript. This research was supported by National Institutes of Health Grant to V.N. Schumaker (GM13914). REFERENCES 1 2 3 4 5

2466

Ohlenbusch, H.H., Olivera, B.M., Tuan, D. and Davidson, N., (1967) J. Mol. Biol. 25, 299-315. Bartley, J.A., and Chalkley, R. (1972) J. Biol. Chem. 247, 3647-3655. Tatchell, K. and Van Holde, K.E. (1977) Biochem. 16, 5295-5303. Lohr, D., Tatchell, K. and Van Holde, K.E. (1977) Cell 12, 829-836. Klevan, L. and Crothers, D.M. (1977) Nucleic Acid Research 4, 4077-4089.

Nucleic Acids Research 6 7

8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28

29

Steinmetz, M., Streeck, R.E. and Zachau, H.G. (1978) Europ. J. Biochem. 83, 615-628. Mathis, D.J., Oudet, P., Wasylyk, B. and Chambon, P. (1978) Nucleic Acid Research 5, 3523-3547. Beard, P. (1978) Cell 15, 955-967. Germond, J.E., Bellard, M., Oudet, P. and Chambon, P. (1976) Nucleic Acid Research 3, 3173-3192. Kornberg, R.D. and Thomas, J.0. (1974) Sci. 184, 865-868. Weintraub, H. Palter, K. and Van Lente, F. (1975) Cell 6, 85-110. Thomas, J.0. and Butler, P.J.G. (1977) J. Mol. Biol. 116, 769-781. Chung, S.Y., Hill, W.E. and Doty, P. (1978) Proc. Natl. Acad. Sci. 75, 1680-1684. Eikbush, T.H. and Moudrianakis, E.N. (1978) Biochem. 17, 4955-4964. Butler, A.P., Olins, D.E. Harrington, R.E. and Hill, W.E. (1978) Biophys. J. 21, 66a. Stein, A. (1979) J. Mol. Biol. 130, 103-134. Olins, A.L., Carlson,, R.DD., Wright, E.B. and Olins,, D.E. (1976) Nucleic Acid Research 3. 3271-3291. Sollner-Webb, B., Camerini-Otero, R.D. and Felsenfeld, G. (1976) Cell 9, 179-193. Maniatus, T., Jeffrey, A. and van deSande, H. (1975) Biochem. 14, 3787-3794. Laemnli, U.K. and Favre, M. (1973) J. Mol. Biol. 80, 575-599. Yamamoto, K.R. and Alberts, B.M. (1970) Virol. 40, 734-744. Mandell, J.D. and Hershey, A.D. (1960) Anal. Biochem. 1, 66-77. Schachman, H.K. (1959) Ultracentrifugation in Biochemistry Academic Press New York and London. Griffith, J.D. (1975) Sci. 187, 1202-1203. Stacks, P.C. (1979) Thesis Dissertation, UCLA. Voordouw, G., Kalif, D. and Eisenberg, H. (1977) Nucleic Acid Research 4, 1207-1223. Cann, J.R. and Kegeles, G. (1974) Biochem 13, 1868-1874. Burton, D.R., Butler, M.J., Hyde, J.E., Phillips, D., Skidmore, CJ. and Walker, I.O. (1978) Nucleic Acid Research 5, 3643-3663. Bina-Stein, M. and Simpson, R.T. (1977) Cell 11, 609-618.

2467