- Email: [email protected]
Interactions of the histone octamer with single-stranded DNA Sedimentation analysis and low-angle X-ray diffraction
Biochimica et Biophysica Acta, 739 (1983) 235-243
235
Elsevier Biomedical Press BBA 91187
I N T E R A C T I O N S O F T H E H I S T O N E O C T A M E R W I T H S I N G L E - S T R A N D E D DNA S E D I M E N T A T I O N ANALYSIS AND L O W - A N G L E X-RAY D I F F R A C T I O N E. CAFFARELLI a, p. DE SANTIS b, L. LEONI a,., M. SAVINO a and E. TROTTA a a Centro di Studio per gli Acidi Nucleici, C.N.R., Istituto di Fisiologia Generale and b lstituto di Chimica Fisica, Universit~ di Roma, 00100 Roma (Italy)
(Received June 15th, 1982) (Revised manuscript received November 30th, 1982)
Key words." DNA-histone binding," Electron microscopy," X-ray diffraction," Chromatin; Nucleosome," Histone
A complex between 140-160 nucleotide single-stranded DNA and the octamer of histones was formed and analyzed by electron microscopy and X-ray low angle diffraction. The morphology of the complex is very similar to that of the nucleosome; the diffraction pattern appears less defined than for chromatin showing broader maxima in the same positions. These results strongly suggest that this particle has a geometry very similar to that of the fundamental subunit of chromatin. The possibility of artifacts due to renaturation reaction promoted by histones is ruled out by the analysis of the complex with S I nuclease and by the formation of a 'nucleosome like' particle using poly(dT) instead of DNA. Association of the histone octamer with either the 140-160 nucleotide single-stranded DNA or the 140-160 bp double-stranded DNA was evaluated at different h i s t o n e / D N A input ratios. In both cases, the formation of the complex appears to be regulated by comparable association constants, and in both cases the trend of the complexation reaction in function of the temperature is almost the same. These results suggest that an alternative binding of the histone octamer to double-stranded or to single-stranded DNA requires low energy charge and may be involved in the processes of replication and transcription of the 'active chromatin'.
Introduction It is now well established that the basic repeating unit in the chromatin is the nucleosome, which consists of an octamer of histones (H2A)2 (H2B)2 (H3)z (H4)2 around which approx. 140 base pairs of D N A are wrapped [1]. In order to get some understanding of different chromatin structures which might be functionally related to differential gene expression, the ex-
* To whom correspondence should be addressed. Abbreviations: PMSF, phenylmethylsulfonylfluoride; DNAss , single-stranded DNA 0167-4781/83/0000-0000/$03.00 © 1983 Elsevier Science Publishers
istence of complexes between single-stranded D N A and histones has been investigated [2]. In 1979 Palter and Alberts [3] suggested, on the basis of differential salt dissociation experiments, that the binding interactions which stabilize the normal nucleosome appear to be present when histones are complexed with single-stranded DNA. Further characterization of the interactions between these two components [4] showed that interaction of histones with a small defined length of singlestranded D N A leads to the formation of a '9 S' nucleosome-like complex. This '9 S' subunit appears to contain an octamer of the four nucleosomal histones and to have approximately the same D N A mass of a normal nucleosome. These
236
data have been interpreted [4] in terms of non-random histone segregation during both DNA replication and RNA transcription processes, which require histone to be transiently bound to singlestranded DNA during polymerases interaction. In this communication we report a low-angle X-ray diffraction analysis of the single-stranded DNA-histone complex performed with the aim of further characterization of its structural aspects. We consider useful a comparison between the association features of the histone octamer with double-stranded and single-stranded DNA to establish whether this complex may be considered a possible local state of chromatin involved in the transcription and replication processes. The existence of a similar binding in the two cases could stabilize the strand separation and could be relevant in the analysis of active chromatin. Great interest has recently developed on the problem of active chromatin and it seems established that a local nucleosomal structure exists in the active genes different, at least in limited regions, from that of bulk chromatin [5]. It has recently been shown [6] that histones isolated from several sources, either single or in combination, promote the renaturation of complementary single strands of DNA; this evidence has been obtained by measurement of resistance to S 1 nuclease. We have therefore tested the complexes with $1 nuclease in different salt solutions to rule out the possibility that partial renaturation may be involved in the formation of complex between single-stranded DNA and histones. Poly(dT) is a typical single-stranded polynucleotide unable to form double-helical structures in a large range of physicochemical conditions [7]. A complex between the octamer of histones and poly(dT) was therefore prepared and analyzed. Materials and Methods
(a) Preparation of core protein. Nuclei were prepared from chick erythrocytes according to the method of Hewish and Burgoyne [8] and adjusted to 50 A260 units/ml (read in 0.1 M NaOH) in 0.34 M sucrose/buffer A (15 mM Tris-HC1 (pH 7.4)/15 mM NaCI/15 mM 2-mercaptoethanol/60 mM KCI/0.5 mM spermidine/0.15 mM sper-
mine). These nuclei were then digested with 50 U micrococcal nuclease/ml (Worthington) in the presence of 1 mM CaC12 for 30 min at 37°C, after which time the digestion was stopped by addition of N a 2 E D T A to 0.2 mM. The nuclei were then centrifuged 5 min at 8000 x g, resuspended in an equivalent volume of 0.2 mM Na2EDTA (pH 7) and lysed by leaving for 12 h at - 2 0 ° C . Histone H1 as well as non-histone proteins were then stripped from the soluble long chromatin. The stripping procedure involved adding 2 M NaC1 dropwise at 0°C (final concentration 0.55 M) and PMSF 10% in dioxane (final concentration 0.25 mM). The sample was applied to a Sepharose 4B column which had been equilibrated with 0.5 M NaC1/5 mM Tris-HC1 (pH 7.5)/0.2 mM 2-mercaptoethanol/0.25 mM PMSF. This column separated the stripped long chromatin from released H1 and non-histone proteins. Fractions containing A 260 absorbing material were pooled and concentrated to approx. 50 A260 u n i t / m l by pressure concentration through an Amicon XM-50 membrane [9]. Alternatively, the H1 and non-histone proteins were separated from soluble chromatin by layering 5 ml onto a 5-20% linear sucrose gradient in 10 mM Tris-HC1 (pH 8)/0.25 mM P M S F / 1 mM E D T A / 0 . 5 M NaC1 in an SW28 rotor at 27000 rev./min for 32 h at 4°C in a Beckman L2-65B ultracentrifuge. All the fractions corresponding to 11 S (this position was marked by a nucleosome solution) were pooled and concentrated [10] to approx. 50 A260 units, as described. The ionic strength was raised to 2 M NaC1 by adding solid NaC1 and the solution was shaken for 1 h at 0°C. The solution was then made 50 mM in potassium phosphate (pH 7.0). The sample was then passed through hydroxyapatite column equilibrated in 2 M NaC1/50 mM potassium phosphate (pH 7.0)/0.25 mM PMSF. In 2 M NaC1 the DNA dissociated from the octamer of histone and binds to hydroxyapatite. The histones elute off in the exclusion volume as a simple peak. The peak was concentrated to approx. 4 A23o units/ml by pressure concentration through an Amicon XM-50 membrane [11]. The absorption coefficients for a typical preparation were [10,12]: "'230A[1%c = 4.2; A2601% I% = 2.1; A280 = 4.2. (b) SDS-polyaerylarnide gels. In order to check
237
the quality of preparation by analysis in SDS-gels, a small sample was precipitated with an equal volume of 50% trichloroacetic acid and washed twice with acetone. SDS-18% polyacrylamide slab gels were run according to Thomas and Kornberg [131. The core protein contained the four main histones H3, H4, H2A and H2B in the same relative amount and was devoid of H1 as revealed by overload chromatography (Fig. lb). Polynucleotides. Sperm salmon D N A was purchased by Serva. Its molecular weight, as determined by CsC1 analytical density-gradient ultracentrifugation, was around 1 • 10 6. Poly(dT) was purchased by Miles. Its physicochemical characteristics were: $20,5.3; Ep (0.1 M PO 4 buffer, pH 7.0) at 260 nm, 8.1 • 103; .4 . . . . 264 nm; Az5 o nm/.4260 nm: 0.69:0.68. DNA preparations. A mixture of DNA fragments ranging between 140-160 bp was prepared by deproteinization of 11 S mononucleosomes obtained after incubation of 50 A260 units/ml chick erythrocyte nuclei with 150 U / m l micrococcal
bases
587--
i Ill
_~H1
434--
267_
.,..
184_
m
124_
w,,
.~H3 --H2B
~H2A
--H4
89_
a
b
Fig. 1. (a) Polyacrylamide gel electrophoresis profile of D N A fragments isolated from chick erythrocyte nuclei digested with micrococcal nuclease. BspRl endonuclease digests of pBR322 were used as size markers. The numbers on the left side refer to the size of the marker bands. (b) SDS-polyacrylamide gel electrophoresis profiles of octamer histones isolated from chick erythrocyte nuclei. The bands on the fight side refer to total histone standard.
nuclease (Worthington Biochemical Co.) for 1 h at 37°C (Fig. la). The DNA was made singlestranded by boiling for 15 min, followed by immersion in an ice-water bath. (For all samples showing a hyperchromic effect less than 20%, the procedure was repeated.) The reassociated fraction (approx. 30%) was eliminated using an hydroxyapatite column according to Tapiero et al. [14]. Reconstitution procedures. Varying amounts of core complex histone in 2 M NaCI/0.25 mM PMSF were added to a constant amount of either 140-160 bp or 140-160 nucleotide single-stranded D N A (500/~g) and the mixture was brought to 500 ttl with 5 mM Tris-HCi (pH 7)/1.0 mM EDTA/0.25 mM PMSF at final NaCI concentration of 1.0 M. The sample was then allowed to stay for 1 h on ice with vigorous stirring. The contents were then brought to 0.6 M NaCI by gradual addition of 50 ~tl aliquots of 5 mM TrisHC1 (pH 7.0)/1.0 mM EDTA/0.25 mM PMSF and then allowed to stay on ice for 5 h. Finally the reconstitution mixture was brought to 3.0 ml and 0.17 M NaC1 by the addition of the same buffer. The final concentration of 140-160 DNA in the 1 ml reconstitute was 5 . 1 0 -4 M. The final concentration of the histone octamer in the 1 ml reconstitute ranged from 1.5 to 5 • 10 - 4 M. Electron microscopic visualization. DNA-histone or poly(dT)-histone complexes were prepared for electron microscopy visualization using the protocol of Thoma and Koller [15]. The DNA-histone complex was fractionated on sucrose gradient, isolated and brought in 0.2 mM EDTA by using pressure concentration through an Amicon XM-50 membrane and diluted to 10 /~g/ml DNA in a freshly prepared solution containing 0.1% glutaraldehyde/10 mM sodium acetate/5 mM triethanolamine acetate (pH 7.9). The preparations were lightly rotary-shadowed with platinum at 7°C in an Edwards 4A vacuum evaporator. Specimens were examined in a Philips EM400 electron microscope. Protein analysis. Mononucleosome and singleand double-stranded DNA-histone reconstitutes were all purified through a sucrose gradient. DNA concentration was estimated by absorbance at 260 nm using extinction coefficients of 6 500 and 9 600 M - l for double- and single-stranded DNA, respectively. Protein concentration was determined
238
by fluorescence measurement after reaction with fluram (fluorescamine, Roche) [16] in presence of 2 M NaC1 to dissociate the histones from DNA. S 1 nuclease analysis. S I nuclease reactions (Sl from Calbiochem) were performed according to the method of Case et al. [17] at 40°C in a reaction mixture containing 0.1 M NaC1/30 m M sodium acetate buffer (pH 4.5)/0.3 m M ZnSO 4. Each reaction mixture, before addition of enzyme, was preincubated at 40°C for 10 min after addition of substrate DNA. After addition of the enzyme, 0.5 ml samples were taken into 1.0 ml of ice-cold 7% perchloric acid; the samples were centrifuged at 8000 r e v . / m i n and the absorbance of the supernatant at 260 nm was determined. The reaction was alternatively carried out in presence of 1 M NaC1 to dissociate histone from DNA. In this last case the reference was obviously single-stranded D N A in 1 M NaCI and the reaction time was 30 min. X-ray diffraction analysis. Gels pulled in Lindeman capillaries were mounted in a Searle X-ray camera, fitted with Elliot toroidal optics and equilibrated to 98% relative humidity by standing over a saturated solution of potassium chlorate prior to exposure to X-ray beam. Helium gas, initially bubbled through saturated KC103 solution, was passed continuously through the system during exposure. Unfiltered CuK~ radiation was produced from the copper target of a Philips semi-microfocus generator. Diffraction patterns from approx. 15 h exposure were recorded on Ilford-G film and the ring spacings initially measured with dividers. The equivalent Bragg spacings were obtained from the densitometer maxima. The best conditions of recombination for X-ray measurements were established as direct mixing of high molecular weight D N A (1. 106), either native or denaturated (in this latter case the denaturation procedure was the same as previously described for D N A fragments 140-160 bp), in distilled water at 1 m g / m l concentration and an equal weight of histone octamer at a 1 m g / m l concentration in 2 M NaCI/2.5.10-4M E D T A (pH 7.0)/2.510- 4 M PMSF. After mixing, by dropwise addition at 0°C during energic stirring, the complex was directly diluted with 2. 10 - 4 M E D T A / 2 . 5 • 10 - 4 M PMSF so as to reach the final NaC1 concentration of
0.3 M. Each dilution step requires approx. 2 h. The final complex was concentrated in an SW41 Ti rotor at 36000 r e v . / m i n for 16 h at 4°C in Beckman L2-65B ultracentrifuge. Results Formation of a nucleosome like complex between the octamer with 140-160 bp double-stranded D N A or 140-160 n ucleotides single-stranded DNA Fig. 2a, b, c, d shows the sedimentation profiles obtained when complexes of the histone octamer with a 140-160 bp long double-stranded D N A (at histone to D N A ratios of 0.4, 0.5, 0.7 and 0.9) are spun through a 5-20% sucrose gradient. Part of the D N A sediments as a peak at 5 S (as determined by velocity sedimentation in an analytical ultracentrifuge); in addition, there is a second peak of D N A at 11 S, in a position coincident with that of the nucleosome. The two bands of the gradients were analyzed for their histone composition by SDS-polyacrylamide gel electrophoresis: the 11 S fraction
A260
2O ~5 o
1.0-
v 10 ~
0.5-
i
1'o
3O Fraction n u m b e r
Fig. 2. 5-20% sucrose gradient sedimentation patterns of 140-160 bp double-stranded D N A complexes with histone octamer. Histone: D N A ratio ( I ) 0.4 g / g ; (A) 0.5 g / g ; (O) 0.7 g / g ; ( O ) 0.9 g / g . The arrow marks the position where nucleosome sediments.
239
A260 ~.~e~some
- 20
-10 ~ -5
0.5c
B
10
20 Froction number
30
Fig. 3. 5-20% sucrose gradient sedimentation patterns of 140-160 nucleotide single-stranded DNA complexes with histone octamer. Histone : DNA ratio (11) 0.4 g/g; (A) 0.5 g/g; (e) 0.7 g/g. The arrow marks the position where nucleosome sediments.
contains all four histones in equimolar amounts, whereas histones are absent in the 5 S fraction (not shown). Fig. 3a, b, c shows the sedimentation profiles obtained when complexes of the histones octamer with a 140-160 nucleotide long single-stranded D N A (at histone to D N A ratios of 0.4, 0.5 and 0.7) are spun through a 5-20% sucrose gradient. Experiments at higher histone-DNA ratios cannot be performed since the solubility of these complexes is much lower than that of the doublestranded D N A complexes. Also in this case part of the D N A sediments as a band at 4 S and there is also a second band at approx. 10 S. In addition, a fraction corresponding to 13 S (as determined in the analytical ultracentrifuge) appears at the bottom of the gradient as an edge. Using the procedure previously adopted by Palter et al. [4], we carried out, on every band, some characterizations: histone composition, histone to D N A ratio, electron microscopic visualization. In agreement with Palter et al. [4], the l0 S band results in a nucleosome like complex, in which the four histones are present in equimolar
.
,o
"-
.L,-.
Fig. 4. Electron microscopic images of nucleosome and DNAhistone complexes rotary-shadowed with platinum. (A) mononucleosomes isolated after micrococcal nuclease digestion of chromatin; (B) 140-160 nucleotides single-stranded DNA histone-complex (10 S); (C) 140-160 nucleotides single-stranded DNA-histone complex (13 S). Final magnification x 17300.
amounts (data not shown), with a p r o t e i n / D N A ratio of 1.2 very similar to the h i s t o n e / D N A ratio (1.3) found for the nucleosome and for the double-stranded 140-160 bp DNA-histone complex. The morphology, also, is very similar to that of the nucleosome as shown in Fig. 4B.
240
A26o
A260
0.5-
-20
1.Q-
-30
-~5 o
-20
2 -~o =~
~
03
¢ -10
5
O° -18o_37 o 0.5
110
~ 20 Frsction number
i 30
Fig. 5. 10-30% sucrose gradient sedimentation patterns of 140-160 nucleotides single-stranded D N A complexes with histone. H i s t o n e : D N A ratio is 0.5 g / g . The arrow marks the position where nucleosome sediments.
Sv 10
20
30
FrQction number
The 13 S broad band corresponding to the edge of the gradient shows characteristics very similar to those of the 11 S band; also in this case the h i s t o n e / D N A ratio is 1.2. When this fraction was visualized by electron microscopy (Fig. 4C) many aggregate particles, very similar to associate nucleosomes, are observed. Trying to further characterize this band we spun complexes of singlestranded DNA on a 10-30% sucrose gradient. Fig. 5 shows that, alongside the 10 S band, a continuous sequence of poorly resolved bands is present; these bands probably represent a continuous
Fig. 6. Electron microscopic images of poly(dT)-histone octamer complexes. Rotary-shadowed with platinum. Final magnification x 17300.
Fig. 7. 5-20% sucrose gradient sedimentation patterns of 140-160 bp double-stranded D N A complexes with histone at different temperatures.
aggregation process due to the poor solubility of the complex. Association equilibria as a function of the temperature for double- and single-stranded DNA were also studied. Figs. 7 and 8 report the profiles of 5-20% sucrose gradient of the reconstitutes at different temperatures. In both cases the association process appears almost independent from the temperature, although at 55°C the complex with the single-stranded DNA has such a low solubility that it precipitates immediately from the solution. Since renaturation of DNA in the presence of histones has been reported [6], analyses of S t nuclease were performed. The results reported in Table I rule out renaturation in our experimental conditions, at least in the range (10%) of experimental error. It is, moreover, possible to carry out complexation with the single-stranded polynucleotide poly(dT) and, although it is difficult to isolate the complex on sucrose gradient owing to its low solubility, we have found, also in this case, the formation of 12 S complex by analysis of the reconstitute at the analytical centrifuge (not shown). This complex can be visualized with elec-
241
tron microscopy (Fig. 6) and shows a morphology very similar to that of the nucleosome.
A260 Nucl.~,~osome
-2O
1.0"
-15
P
-10
-5
1'o
2'o
3'0
Froction number
Fig. 8. 5-20% sucrose gradient sedimentation patterns of 140-160 nucleotides single-stranded DNA complexes with histone at different temperatures.
X-ray diffraction analysis The complex between single-stranded DNA and the octamer of histones was characterized by lowangle X-ray diffraction analysis and compared with chromatin. The diffraction pattern from chromatin has long been described by a series of low-angle diffraction reflections of spacings of about 55, 38 and 26 ,~ [18]. A quite similar diffraction pattern characterizes reconstituted nucleohistones obtained from double-stranded DNA, and the four histones H2A, H2B, H3 and H4, as well as the complex between DNA and H3 and H4 fractions alone. It is now clear that such a typical pattern arises mainly from the overall shape of the nucleosome which resembles a flat cylindroid particle with a diameter of approx. 110 A and a height of 55 ,~, corresponding to the envelope of DNA supercoiled around the histone octamer core [13]. Since the electron microscopy visualization shows that particles, with a morphology very simi-
i
TABLE I MEASUREMENT OF NUCLEASE S I DIGESTION IN DOUBLE- AND SINGLE-STRANDED DNA-HISTONE COMPLEXES The digestion percentages are calculated assuming that the absorbance for a fully digested DNA is l A-times larger than initial absorption (as determined from the digestion of singlestranded DNA at different times). DNA-histone complexes
l(sl>
% digestion 0.1 M NaCI
140-160 nucleotides single-stranded DNA-histone complex (10 S) 50 140-160 nucleotides single-stranded DNA 100 140-160 bp double-stranded DNA 0 140-160 nucleotides single stranded DNA (4 S) 100 2. l03 nucleotides single-stranded DNA-histone complex (X-ray) 50 2.103 nucleotides single-stranded DNA 100
1.0 M NaCI
70 70 0 70
|
I
I
I
0
0.02
0.04
0.06
s ( A ) -1
70 70
Fig. 9. Densitometer profiles of X-ray diffraction patterns of chromatin (lower) and single-stranded DNA-histone complex (upper).
242 lar to the nucleosome, are formed in the interaction between single-stranded D N A and histones, the low-angle X-ray diffraction of single-stranded D N A and histone octamer complex was tried. While less defined than in reconstituted nucleosomes and in chromatin, the X-ray diffraction patterns reveals, as shown in Fig. 9, the presence of reflections at 55 and 38 ,~ (approx. the same spacings as chromatin). It is, in fact, conceivable that single-stranded D N A is able to wrap around the basic surface of the histone octamer following the best pathway of neutralization of phosphates because of its larger flexibility with respect to double-stranded DNA. In such cases, while a spheroidal shape of the complex is the most plausible to be expected, nervertheless, the scattering pattern would be very similar to that of the chromatin, providing the radial coordinates are of similar dimensions. Attempts to produce low-angle diffraction rings from core octamer prepared identically to the reconstituted samples, except for the addition of DNA, were unsuccessful. This result was expected taking into account that the octamer dissociates in the four histones, going from 2 to 0.3 M NaCI (unless DNA is present), as in a typical reconstitution experiment, and that it was free of DNA as indicated by A230J-,4260 ratio of approx. 20 (see Materials and Methods). Discussion
Specific interactions between single-stranded D N A and the histone octamer, as previously shown by Alberts [3], produce a 'nucleosome like particle'. The possibility that this complex, morphologically and hydrodynamically very similar to nucleosome, is an artifact due to DNA renaturation promoted by histones is ruled out by the analysis with S~ nuclease; no difference is, in fact, observed between the digestion pattern of single-stranded D N A purified on hydroxyapatite and the DNAoctamer complex, when the reaction is carried out in 1 M NaC1. A similar complex, is obtained using poly(dT) instead of DNA; since this polymer is unable to assume double helix structure in a very large range of physicochemical conditions [7], this result clearly confirms that a single-chain polynucleotide can wrap around the histone octamer,
thereby forming a particle with morphological and hydrodynamic properties very similar to those of the nucleosome. This finding is further supported by the low-angle X-ray diffraction pattern which suggests, in analogy with the chromatin pattern, the formation of a particle with geometric features similar to the nucleosome. In this case, however, the broaded maxima indicates that the complex with single-stranded DNA is more disordered. Although X-ray analysis is one of the few methods available to reveal tertiary structures, this method does not give quantitative estimations. This point is extremely important when considering recombination of histones with single-stranded DNA, since a percentage of DNA renaturation could be responsible for prominent features which are observed in the diffraction pattern. However, the data reported in Table I show that the renaturation of DNA, if present, is undetectable with the S~-nuclease assay, and is lower than 10%. These data provide evidence that the low-angle X-ray diffraction from core octamer singlestranded DNA complexes is not due to renaturated DNA since it is directly comparable to the diffraction from equal amounts of core octamer double-stranded DNA complexes exposed for the same lenght of time. X-ray diffraction of poly(dT)-octamer complex could have been a resolutory analysis, but it proved impossible because of the disordered precipitates it forms. From the data reported in Figs. 2 and 3, the percentage of double- or single-stranded DNA complexes with histone octamer versus increasing h i s t o n e / D N A input ratios, can be determined. The resulting values are reported in Fig. 10. In the case of the single-stranded DNA we have, obviously, considered as bound DNA the amount corresponding either to the 10 or 13 S band. The resulting trend is very similar in both cases. It can be observed that the difference between the octamer initially added and the complexed one is almost negligible, thus indicating a very strong binding that prevents a precise estimate of the association constants in both cases. This finding is qualitatively in agreement with the high value reported by Stein [20] for the nucleosome association constant, namely 10 H M-~. A high association constant explains the invariance of association equilibrium in the range of 0-55°C. This
243
Acknowledgements
100
/
@/ =o =
o_
/
/
/
/
/
/
/
We are grateful to Drs. M. Teresa Carri and Gioacchino Micheli for electron microscopy images, and to Mr. Arcangelo Di Francesco for skillful technical aid.
©
References
QQ
50
E
LL
//="
/ ?¢
/
/ ,
/
/
/
/
~
I
l
I
I
,
I
1.0
0.5 R added
Fig. 10. Percenta&e of 140-160 nucleotide (O) or 140-160 bp (e) DNA complexes with histone octamer evaluated by sedimentation analysis (Figs. 2 and 3); the areas of the 11 or 10 S bands were normalized, dividing for the sum of the respective area and the area of the free D N A band. The input ratios between the histone octamer and the DNA are given in molarity (R added).
finding predicts that the stability of the complex between histones and single-stranded D N A will be similar to the one of the nucleosome in physiological conditions. The similarity of the association process in the two cases could suggest that, as proposed by Weintraub et al. [2], in the active chromatin the histone octamer can bind alternatively the doubleor single-stranded D N A with low energy requirement, and this mechanism is certainly necessary for any model which postulates the maintenance of the nucleosome organization during replication and transcription.
1 McGee, J.D. and Felsenfeld, G. (1980) Annu. Rev. Biochem. 49. 1115-1156 2 Weintraub, H., Worcel, A. and Alberts, B. (1976) Cell 9, 409 - 417 3 Palter, K.B. and Alberts, B.M. (1979) J. Biol. Chem. 254, 11160-11169 4 Palter, K.B., Foe, V.E. and Alberts, B.M. (1979) Cell 18, 451-467 5 Mathis, D., Oudet, P. and Chambon, P. (1980) Prog. Nucleic Acid Res. Mol. Biol. 24, 1-55 6 Cox, M.M. and Lehman, I.R. (1981) Nucleic Acids Res. 9, 389-400 7 Zmuolzka, B., Bollum, F.J. and Shugar, D. (1969) J. Mol. Biol. 46, 169-183 8 Hewish, D.R. and Burgoyne, L.A. (1973) Biochem. Biophys. Res. Commun. 52, 504-510 9 Lutter, L.C. (1978) J. Mol. Biol. 124, 391-420 10 Thomas, J.O. and Butler, P.J. (1977) J. Mol. Biol. 116, 769- 781 11 Rhodes, D. (1979) Nucleic Acids Res. 6, 1805-1815 12 Chung, S.Y., Hill, W.E. and Doty, P. (1978) Proc. Natl. Acad. Sci. U.S.A. 75, 1680-1684 13 Thomas, J.O. and Kornberg, R.D. (1975) Proc. Natl. Acad. Sci. U.S.A. 72, 2626-2630 14 Tapiero, H., Monier, M.N., Shaol, D. and Hard, J. (1974) Nucleic Acids Res. 1,309-322 15 Thoma, F. and Koller, T. (1977) Cell 12, 101-107 16 Ohba, Y., Moremitsu, Y. and Watarai, A. (1979) Eur. J. Biochem. 100, 285-293 17 Case, S.T., Mongeon, L. and Baker, R.F. (1974) Biochim. Biophys. Acta 349, 1-12 18 Zubay, G., Wilkins, M.W.F. and Blout, E.N. (1962) J. Mol. Biol. 4, 69-72 19 Bosely, P.G., Bradbury, E.M., Butler-Browne, G.S., Carpenter, B.G. and Stephens, R.M. (1976) Eur. J. Biochem. 62, 21-31 20 Stein, A. (1979) J. Mol. Biol. 130, 103-134
235
Elsevier Biomedical Press BBA 91187
I N T E R A C T I O N S O F T H E H I S T O N E O C T A M E R W I T H S I N G L E - S T R A N D E D DNA S E D I M E N T A T I O N ANALYSIS AND L O W - A N G L E X-RAY D I F F R A C T I O N E. CAFFARELLI a, p. DE SANTIS b, L. LEONI a,., M. SAVINO a and E. TROTTA a a Centro di Studio per gli Acidi Nucleici, C.N.R., Istituto di Fisiologia Generale and b lstituto di Chimica Fisica, Universit~ di Roma, 00100 Roma (Italy)
(Received June 15th, 1982) (Revised manuscript received November 30th, 1982)
Key words." DNA-histone binding," Electron microscopy," X-ray diffraction," Chromatin; Nucleosome," Histone
A complex between 140-160 nucleotide single-stranded DNA and the octamer of histones was formed and analyzed by electron microscopy and X-ray low angle diffraction. The morphology of the complex is very similar to that of the nucleosome; the diffraction pattern appears less defined than for chromatin showing broader maxima in the same positions. These results strongly suggest that this particle has a geometry very similar to that of the fundamental subunit of chromatin. The possibility of artifacts due to renaturation reaction promoted by histones is ruled out by the analysis of the complex with S I nuclease and by the formation of a 'nucleosome like' particle using poly(dT) instead of DNA. Association of the histone octamer with either the 140-160 nucleotide single-stranded DNA or the 140-160 bp double-stranded DNA was evaluated at different h i s t o n e / D N A input ratios. In both cases, the formation of the complex appears to be regulated by comparable association constants, and in both cases the trend of the complexation reaction in function of the temperature is almost the same. These results suggest that an alternative binding of the histone octamer to double-stranded or to single-stranded DNA requires low energy charge and may be involved in the processes of replication and transcription of the 'active chromatin'.
Introduction It is now well established that the basic repeating unit in the chromatin is the nucleosome, which consists of an octamer of histones (H2A)2 (H2B)2 (H3)z (H4)2 around which approx. 140 base pairs of D N A are wrapped [1]. In order to get some understanding of different chromatin structures which might be functionally related to differential gene expression, the ex-
* To whom correspondence should be addressed. Abbreviations: PMSF, phenylmethylsulfonylfluoride; DNAss , single-stranded DNA 0167-4781/83/0000-0000/$03.00 © 1983 Elsevier Science Publishers
istence of complexes between single-stranded D N A and histones has been investigated [2]. In 1979 Palter and Alberts [3] suggested, on the basis of differential salt dissociation experiments, that the binding interactions which stabilize the normal nucleosome appear to be present when histones are complexed with single-stranded DNA. Further characterization of the interactions between these two components [4] showed that interaction of histones with a small defined length of singlestranded D N A leads to the formation of a '9 S' nucleosome-like complex. This '9 S' subunit appears to contain an octamer of the four nucleosomal histones and to have approximately the same D N A mass of a normal nucleosome. These
236
data have been interpreted [4] in terms of non-random histone segregation during both DNA replication and RNA transcription processes, which require histone to be transiently bound to singlestranded DNA during polymerases interaction. In this communication we report a low-angle X-ray diffraction analysis of the single-stranded DNA-histone complex performed with the aim of further characterization of its structural aspects. We consider useful a comparison between the association features of the histone octamer with double-stranded and single-stranded DNA to establish whether this complex may be considered a possible local state of chromatin involved in the transcription and replication processes. The existence of a similar binding in the two cases could stabilize the strand separation and could be relevant in the analysis of active chromatin. Great interest has recently developed on the problem of active chromatin and it seems established that a local nucleosomal structure exists in the active genes different, at least in limited regions, from that of bulk chromatin [5]. It has recently been shown [6] that histones isolated from several sources, either single or in combination, promote the renaturation of complementary single strands of DNA; this evidence has been obtained by measurement of resistance to S 1 nuclease. We have therefore tested the complexes with $1 nuclease in different salt solutions to rule out the possibility that partial renaturation may be involved in the formation of complex between single-stranded DNA and histones. Poly(dT) is a typical single-stranded polynucleotide unable to form double-helical structures in a large range of physicochemical conditions [7]. A complex between the octamer of histones and poly(dT) was therefore prepared and analyzed. Materials and Methods
(a) Preparation of core protein. Nuclei were prepared from chick erythrocytes according to the method of Hewish and Burgoyne [8] and adjusted to 50 A260 units/ml (read in 0.1 M NaOH) in 0.34 M sucrose/buffer A (15 mM Tris-HC1 (pH 7.4)/15 mM NaCI/15 mM 2-mercaptoethanol/60 mM KCI/0.5 mM spermidine/0.15 mM sper-
mine). These nuclei were then digested with 50 U micrococcal nuclease/ml (Worthington) in the presence of 1 mM CaC12 for 30 min at 37°C, after which time the digestion was stopped by addition of N a 2 E D T A to 0.2 mM. The nuclei were then centrifuged 5 min at 8000 x g, resuspended in an equivalent volume of 0.2 mM Na2EDTA (pH 7) and lysed by leaving for 12 h at - 2 0 ° C . Histone H1 as well as non-histone proteins were then stripped from the soluble long chromatin. The stripping procedure involved adding 2 M NaC1 dropwise at 0°C (final concentration 0.55 M) and PMSF 10% in dioxane (final concentration 0.25 mM). The sample was applied to a Sepharose 4B column which had been equilibrated with 0.5 M NaC1/5 mM Tris-HC1 (pH 7.5)/0.2 mM 2-mercaptoethanol/0.25 mM PMSF. This column separated the stripped long chromatin from released H1 and non-histone proteins. Fractions containing A 260 absorbing material were pooled and concentrated to approx. 50 A260 u n i t / m l by pressure concentration through an Amicon XM-50 membrane [9]. Alternatively, the H1 and non-histone proteins were separated from soluble chromatin by layering 5 ml onto a 5-20% linear sucrose gradient in 10 mM Tris-HC1 (pH 8)/0.25 mM P M S F / 1 mM E D T A / 0 . 5 M NaC1 in an SW28 rotor at 27000 rev./min for 32 h at 4°C in a Beckman L2-65B ultracentrifuge. All the fractions corresponding to 11 S (this position was marked by a nucleosome solution) were pooled and concentrated [10] to approx. 50 A260 units, as described. The ionic strength was raised to 2 M NaC1 by adding solid NaC1 and the solution was shaken for 1 h at 0°C. The solution was then made 50 mM in potassium phosphate (pH 7.0). The sample was then passed through hydroxyapatite column equilibrated in 2 M NaC1/50 mM potassium phosphate (pH 7.0)/0.25 mM PMSF. In 2 M NaC1 the DNA dissociated from the octamer of histone and binds to hydroxyapatite. The histones elute off in the exclusion volume as a simple peak. The peak was concentrated to approx. 4 A23o units/ml by pressure concentration through an Amicon XM-50 membrane [11]. The absorption coefficients for a typical preparation were [10,12]: "'230A[1%c = 4.2; A2601% I% = 2.1; A280 = 4.2. (b) SDS-polyaerylarnide gels. In order to check
237
the quality of preparation by analysis in SDS-gels, a small sample was precipitated with an equal volume of 50% trichloroacetic acid and washed twice with acetone. SDS-18% polyacrylamide slab gels were run according to Thomas and Kornberg [131. The core protein contained the four main histones H3, H4, H2A and H2B in the same relative amount and was devoid of H1 as revealed by overload chromatography (Fig. lb). Polynucleotides. Sperm salmon D N A was purchased by Serva. Its molecular weight, as determined by CsC1 analytical density-gradient ultracentrifugation, was around 1 • 10 6. Poly(dT) was purchased by Miles. Its physicochemical characteristics were: $20,5.3; Ep (0.1 M PO 4 buffer, pH 7.0) at 260 nm, 8.1 • 103; .4 . . . . 264 nm; Az5 o nm/.4260 nm: 0.69:0.68. DNA preparations. A mixture of DNA fragments ranging between 140-160 bp was prepared by deproteinization of 11 S mononucleosomes obtained after incubation of 50 A260 units/ml chick erythrocyte nuclei with 150 U / m l micrococcal
bases
587--
i Ill
_~H1
434--
267_
.,..
184_
m
124_
w,,
.~H3 --H2B
~H2A
--H4
89_
a
b
Fig. 1. (a) Polyacrylamide gel electrophoresis profile of D N A fragments isolated from chick erythrocyte nuclei digested with micrococcal nuclease. BspRl endonuclease digests of pBR322 were used as size markers. The numbers on the left side refer to the size of the marker bands. (b) SDS-polyacrylamide gel electrophoresis profiles of octamer histones isolated from chick erythrocyte nuclei. The bands on the fight side refer to total histone standard.
nuclease (Worthington Biochemical Co.) for 1 h at 37°C (Fig. la). The DNA was made singlestranded by boiling for 15 min, followed by immersion in an ice-water bath. (For all samples showing a hyperchromic effect less than 20%, the procedure was repeated.) The reassociated fraction (approx. 30%) was eliminated using an hydroxyapatite column according to Tapiero et al. [14]. Reconstitution procedures. Varying amounts of core complex histone in 2 M NaCI/0.25 mM PMSF were added to a constant amount of either 140-160 bp or 140-160 nucleotide single-stranded D N A (500/~g) and the mixture was brought to 500 ttl with 5 mM Tris-HCi (pH 7)/1.0 mM EDTA/0.25 mM PMSF at final NaCI concentration of 1.0 M. The sample was then allowed to stay for 1 h on ice with vigorous stirring. The contents were then brought to 0.6 M NaCI by gradual addition of 50 ~tl aliquots of 5 mM TrisHC1 (pH 7.0)/1.0 mM EDTA/0.25 mM PMSF and then allowed to stay on ice for 5 h. Finally the reconstitution mixture was brought to 3.0 ml and 0.17 M NaC1 by the addition of the same buffer. The final concentration of 140-160 DNA in the 1 ml reconstitute was 5 . 1 0 -4 M. The final concentration of the histone octamer in the 1 ml reconstitute ranged from 1.5 to 5 • 10 - 4 M. Electron microscopic visualization. DNA-histone or poly(dT)-histone complexes were prepared for electron microscopy visualization using the protocol of Thoma and Koller [15]. The DNA-histone complex was fractionated on sucrose gradient, isolated and brought in 0.2 mM EDTA by using pressure concentration through an Amicon XM-50 membrane and diluted to 10 /~g/ml DNA in a freshly prepared solution containing 0.1% glutaraldehyde/10 mM sodium acetate/5 mM triethanolamine acetate (pH 7.9). The preparations were lightly rotary-shadowed with platinum at 7°C in an Edwards 4A vacuum evaporator. Specimens were examined in a Philips EM400 electron microscope. Protein analysis. Mononucleosome and singleand double-stranded DNA-histone reconstitutes were all purified through a sucrose gradient. DNA concentration was estimated by absorbance at 260 nm using extinction coefficients of 6 500 and 9 600 M - l for double- and single-stranded DNA, respectively. Protein concentration was determined
238
by fluorescence measurement after reaction with fluram (fluorescamine, Roche) [16] in presence of 2 M NaC1 to dissociate the histones from DNA. S 1 nuclease analysis. S I nuclease reactions (Sl from Calbiochem) were performed according to the method of Case et al. [17] at 40°C in a reaction mixture containing 0.1 M NaC1/30 m M sodium acetate buffer (pH 4.5)/0.3 m M ZnSO 4. Each reaction mixture, before addition of enzyme, was preincubated at 40°C for 10 min after addition of substrate DNA. After addition of the enzyme, 0.5 ml samples were taken into 1.0 ml of ice-cold 7% perchloric acid; the samples were centrifuged at 8000 r e v . / m i n and the absorbance of the supernatant at 260 nm was determined. The reaction was alternatively carried out in presence of 1 M NaC1 to dissociate histone from DNA. In this last case the reference was obviously single-stranded D N A in 1 M NaCI and the reaction time was 30 min. X-ray diffraction analysis. Gels pulled in Lindeman capillaries were mounted in a Searle X-ray camera, fitted with Elliot toroidal optics and equilibrated to 98% relative humidity by standing over a saturated solution of potassium chlorate prior to exposure to X-ray beam. Helium gas, initially bubbled through saturated KC103 solution, was passed continuously through the system during exposure. Unfiltered CuK~ radiation was produced from the copper target of a Philips semi-microfocus generator. Diffraction patterns from approx. 15 h exposure were recorded on Ilford-G film and the ring spacings initially measured with dividers. The equivalent Bragg spacings were obtained from the densitometer maxima. The best conditions of recombination for X-ray measurements were established as direct mixing of high molecular weight D N A (1. 106), either native or denaturated (in this latter case the denaturation procedure was the same as previously described for D N A fragments 140-160 bp), in distilled water at 1 m g / m l concentration and an equal weight of histone octamer at a 1 m g / m l concentration in 2 M NaCI/2.5.10-4M E D T A (pH 7.0)/2.510- 4 M PMSF. After mixing, by dropwise addition at 0°C during energic stirring, the complex was directly diluted with 2. 10 - 4 M E D T A / 2 . 5 • 10 - 4 M PMSF so as to reach the final NaC1 concentration of
0.3 M. Each dilution step requires approx. 2 h. The final complex was concentrated in an SW41 Ti rotor at 36000 r e v . / m i n for 16 h at 4°C in Beckman L2-65B ultracentrifuge. Results Formation of a nucleosome like complex between the octamer with 140-160 bp double-stranded D N A or 140-160 n ucleotides single-stranded DNA Fig. 2a, b, c, d shows the sedimentation profiles obtained when complexes of the histone octamer with a 140-160 bp long double-stranded D N A (at histone to D N A ratios of 0.4, 0.5, 0.7 and 0.9) are spun through a 5-20% sucrose gradient. Part of the D N A sediments as a peak at 5 S (as determined by velocity sedimentation in an analytical ultracentrifuge); in addition, there is a second peak of D N A at 11 S, in a position coincident with that of the nucleosome. The two bands of the gradients were analyzed for their histone composition by SDS-polyacrylamide gel electrophoresis: the 11 S fraction
A260
2O ~5 o
1.0-
v 10 ~
0.5-
i
1'o
3O Fraction n u m b e r
Fig. 2. 5-20% sucrose gradient sedimentation patterns of 140-160 bp double-stranded D N A complexes with histone octamer. Histone: D N A ratio ( I ) 0.4 g / g ; (A) 0.5 g / g ; (O) 0.7 g / g ; ( O ) 0.9 g / g . The arrow marks the position where nucleosome sediments.
239
A260 ~.~e~some
- 20
-10 ~ -5
0.5c
B
10
20 Froction number
30
Fig. 3. 5-20% sucrose gradient sedimentation patterns of 140-160 nucleotide single-stranded DNA complexes with histone octamer. Histone : DNA ratio (11) 0.4 g/g; (A) 0.5 g/g; (e) 0.7 g/g. The arrow marks the position where nucleosome sediments.
contains all four histones in equimolar amounts, whereas histones are absent in the 5 S fraction (not shown). Fig. 3a, b, c shows the sedimentation profiles obtained when complexes of the histones octamer with a 140-160 nucleotide long single-stranded D N A (at histone to D N A ratios of 0.4, 0.5 and 0.7) are spun through a 5-20% sucrose gradient. Experiments at higher histone-DNA ratios cannot be performed since the solubility of these complexes is much lower than that of the doublestranded D N A complexes. Also in this case part of the D N A sediments as a band at 4 S and there is also a second band at approx. 10 S. In addition, a fraction corresponding to 13 S (as determined in the analytical ultracentrifuge) appears at the bottom of the gradient as an edge. Using the procedure previously adopted by Palter et al. [4], we carried out, on every band, some characterizations: histone composition, histone to D N A ratio, electron microscopic visualization. In agreement with Palter et al. [4], the l0 S band results in a nucleosome like complex, in which the four histones are present in equimolar
.
,o
"-
.L,-.
Fig. 4. Electron microscopic images of nucleosome and DNAhistone complexes rotary-shadowed with platinum. (A) mononucleosomes isolated after micrococcal nuclease digestion of chromatin; (B) 140-160 nucleotides single-stranded DNA histone-complex (10 S); (C) 140-160 nucleotides single-stranded DNA-histone complex (13 S). Final magnification x 17300.
amounts (data not shown), with a p r o t e i n / D N A ratio of 1.2 very similar to the h i s t o n e / D N A ratio (1.3) found for the nucleosome and for the double-stranded 140-160 bp DNA-histone complex. The morphology, also, is very similar to that of the nucleosome as shown in Fig. 4B.
240
A26o
A260
0.5-
-20
1.Q-
-30
-~5 o
-20
2 -~o =~
~
03
¢ -10
5
O° -18o_37 o 0.5
110
~ 20 Frsction number
i 30
Fig. 5. 10-30% sucrose gradient sedimentation patterns of 140-160 nucleotides single-stranded D N A complexes with histone. H i s t o n e : D N A ratio is 0.5 g / g . The arrow marks the position where nucleosome sediments.
Sv 10
20
30
FrQction number
The 13 S broad band corresponding to the edge of the gradient shows characteristics very similar to those of the 11 S band; also in this case the h i s t o n e / D N A ratio is 1.2. When this fraction was visualized by electron microscopy (Fig. 4C) many aggregate particles, very similar to associate nucleosomes, are observed. Trying to further characterize this band we spun complexes of singlestranded DNA on a 10-30% sucrose gradient. Fig. 5 shows that, alongside the 10 S band, a continuous sequence of poorly resolved bands is present; these bands probably represent a continuous
Fig. 6. Electron microscopic images of poly(dT)-histone octamer complexes. Rotary-shadowed with platinum. Final magnification x 17300.
Fig. 7. 5-20% sucrose gradient sedimentation patterns of 140-160 bp double-stranded D N A complexes with histone at different temperatures.
aggregation process due to the poor solubility of the complex. Association equilibria as a function of the temperature for double- and single-stranded DNA were also studied. Figs. 7 and 8 report the profiles of 5-20% sucrose gradient of the reconstitutes at different temperatures. In both cases the association process appears almost independent from the temperature, although at 55°C the complex with the single-stranded DNA has such a low solubility that it precipitates immediately from the solution. Since renaturation of DNA in the presence of histones has been reported [6], analyses of S t nuclease were performed. The results reported in Table I rule out renaturation in our experimental conditions, at least in the range (10%) of experimental error. It is, moreover, possible to carry out complexation with the single-stranded polynucleotide poly(dT) and, although it is difficult to isolate the complex on sucrose gradient owing to its low solubility, we have found, also in this case, the formation of 12 S complex by analysis of the reconstitute at the analytical centrifuge (not shown). This complex can be visualized with elec-
241
tron microscopy (Fig. 6) and shows a morphology very similar to that of the nucleosome.
A260 Nucl.~,~osome
-2O
1.0"
-15
P
-10
-5
1'o
2'o
3'0
Froction number
Fig. 8. 5-20% sucrose gradient sedimentation patterns of 140-160 nucleotides single-stranded DNA complexes with histone at different temperatures.
X-ray diffraction analysis The complex between single-stranded DNA and the octamer of histones was characterized by lowangle X-ray diffraction analysis and compared with chromatin. The diffraction pattern from chromatin has long been described by a series of low-angle diffraction reflections of spacings of about 55, 38 and 26 ,~ [18]. A quite similar diffraction pattern characterizes reconstituted nucleohistones obtained from double-stranded DNA, and the four histones H2A, H2B, H3 and H4, as well as the complex between DNA and H3 and H4 fractions alone. It is now clear that such a typical pattern arises mainly from the overall shape of the nucleosome which resembles a flat cylindroid particle with a diameter of approx. 110 A and a height of 55 ,~, corresponding to the envelope of DNA supercoiled around the histone octamer core [13]. Since the electron microscopy visualization shows that particles, with a morphology very simi-
i
TABLE I MEASUREMENT OF NUCLEASE S I DIGESTION IN DOUBLE- AND SINGLE-STRANDED DNA-HISTONE COMPLEXES The digestion percentages are calculated assuming that the absorbance for a fully digested DNA is l A-times larger than initial absorption (as determined from the digestion of singlestranded DNA at different times). DNA-histone complexes
l(sl>
% digestion 0.1 M NaCI
140-160 nucleotides single-stranded DNA-histone complex (10 S) 50 140-160 nucleotides single-stranded DNA 100 140-160 bp double-stranded DNA 0 140-160 nucleotides single stranded DNA (4 S) 100 2. l03 nucleotides single-stranded DNA-histone complex (X-ray) 50 2.103 nucleotides single-stranded DNA 100
1.0 M NaCI
70 70 0 70
|
I
I
I
0
0.02
0.04
0.06
s ( A ) -1
70 70
Fig. 9. Densitometer profiles of X-ray diffraction patterns of chromatin (lower) and single-stranded DNA-histone complex (upper).
242 lar to the nucleosome, are formed in the interaction between single-stranded D N A and histones, the low-angle X-ray diffraction of single-stranded D N A and histone octamer complex was tried. While less defined than in reconstituted nucleosomes and in chromatin, the X-ray diffraction patterns reveals, as shown in Fig. 9, the presence of reflections at 55 and 38 ,~ (approx. the same spacings as chromatin). It is, in fact, conceivable that single-stranded D N A is able to wrap around the basic surface of the histone octamer following the best pathway of neutralization of phosphates because of its larger flexibility with respect to double-stranded DNA. In such cases, while a spheroidal shape of the complex is the most plausible to be expected, nervertheless, the scattering pattern would be very similar to that of the chromatin, providing the radial coordinates are of similar dimensions. Attempts to produce low-angle diffraction rings from core octamer prepared identically to the reconstituted samples, except for the addition of DNA, were unsuccessful. This result was expected taking into account that the octamer dissociates in the four histones, going from 2 to 0.3 M NaCI (unless DNA is present), as in a typical reconstitution experiment, and that it was free of DNA as indicated by A230J-,4260 ratio of approx. 20 (see Materials and Methods). Discussion
Specific interactions between single-stranded D N A and the histone octamer, as previously shown by Alberts [3], produce a 'nucleosome like particle'. The possibility that this complex, morphologically and hydrodynamically very similar to nucleosome, is an artifact due to DNA renaturation promoted by histones is ruled out by the analysis with S~ nuclease; no difference is, in fact, observed between the digestion pattern of single-stranded D N A purified on hydroxyapatite and the DNAoctamer complex, when the reaction is carried out in 1 M NaC1. A similar complex, is obtained using poly(dT) instead of DNA; since this polymer is unable to assume double helix structure in a very large range of physicochemical conditions [7], this result clearly confirms that a single-chain polynucleotide can wrap around the histone octamer,
thereby forming a particle with morphological and hydrodynamic properties very similar to those of the nucleosome. This finding is further supported by the low-angle X-ray diffraction pattern which suggests, in analogy with the chromatin pattern, the formation of a particle with geometric features similar to the nucleosome. In this case, however, the broaded maxima indicates that the complex with single-stranded DNA is more disordered. Although X-ray analysis is one of the few methods available to reveal tertiary structures, this method does not give quantitative estimations. This point is extremely important when considering recombination of histones with single-stranded DNA, since a percentage of DNA renaturation could be responsible for prominent features which are observed in the diffraction pattern. However, the data reported in Table I show that the renaturation of DNA, if present, is undetectable with the S~-nuclease assay, and is lower than 10%. These data provide evidence that the low-angle X-ray diffraction from core octamer singlestranded DNA complexes is not due to renaturated DNA since it is directly comparable to the diffraction from equal amounts of core octamer double-stranded DNA complexes exposed for the same lenght of time. X-ray diffraction of poly(dT)-octamer complex could have been a resolutory analysis, but it proved impossible because of the disordered precipitates it forms. From the data reported in Figs. 2 and 3, the percentage of double- or single-stranded DNA complexes with histone octamer versus increasing h i s t o n e / D N A input ratios, can be determined. The resulting values are reported in Fig. 10. In the case of the single-stranded DNA we have, obviously, considered as bound DNA the amount corresponding either to the 10 or 13 S band. The resulting trend is very similar in both cases. It can be observed that the difference between the octamer initially added and the complexed one is almost negligible, thus indicating a very strong binding that prevents a precise estimate of the association constants in both cases. This finding is qualitatively in agreement with the high value reported by Stein [20] for the nucleosome association constant, namely 10 H M-~. A high association constant explains the invariance of association equilibrium in the range of 0-55°C. This
243
Acknowledgements
100
/
@/ =o =
o_
/
/
/
/
/
/
/
We are grateful to Drs. M. Teresa Carri and Gioacchino Micheli for electron microscopy images, and to Mr. Arcangelo Di Francesco for skillful technical aid.
©
References
50
E
LL
//="
/ ?¢
/
/ ,
/
/
/
/
~
I
l
I
I
,
I
1.0
0.5 R added
Fig. 10. Percenta&e of 140-160 nucleotide (O) or 140-160 bp (e) DNA complexes with histone octamer evaluated by sedimentation analysis (Figs. 2 and 3); the areas of the 11 or 10 S bands were normalized, dividing for the sum of the respective area and the area of the free D N A band. The input ratios between the histone octamer and the DNA are given in molarity (R added).
finding predicts that the stability of the complex between histones and single-stranded D N A will be similar to the one of the nucleosome in physiological conditions. The similarity of the association process in the two cases could suggest that, as proposed by Weintraub et al. [2], in the active chromatin the histone octamer can bind alternatively the doubleor single-stranded D N A with low energy requirement, and this mechanism is certainly necessary for any model which postulates the maintenance of the nucleosome organization during replication and transcription.
1 McGee, J.D. and Felsenfeld, G. (1980) Annu. Rev. Biochem. 49. 1115-1156 2 Weintraub, H., Worcel, A. and Alberts, B. (1976) Cell 9, 409 - 417 3 Palter, K.B. and Alberts, B.M. (1979) J. Biol. Chem. 254, 11160-11169 4 Palter, K.B., Foe, V.E. and Alberts, B.M. (1979) Cell 18, 451-467 5 Mathis, D., Oudet, P. and Chambon, P. (1980) Prog. Nucleic Acid Res. Mol. Biol. 24, 1-55 6 Cox, M.M. and Lehman, I.R. (1981) Nucleic Acids Res. 9, 389-400 7 Zmuolzka, B., Bollum, F.J. and Shugar, D. (1969) J. Mol. Biol. 46, 169-183 8 Hewish, D.R. and Burgoyne, L.A. (1973) Biochem. Biophys. Res. Commun. 52, 504-510 9 Lutter, L.C. (1978) J. Mol. Biol. 124, 391-420 10 Thomas, J.O. and Butler, P.J. (1977) J. Mol. Biol. 116, 769- 781 11 Rhodes, D. (1979) Nucleic Acids Res. 6, 1805-1815 12 Chung, S.Y., Hill, W.E. and Doty, P. (1978) Proc. Natl. Acad. Sci. U.S.A. 75, 1680-1684 13 Thomas, J.O. and Kornberg, R.D. (1975) Proc. Natl. Acad. Sci. U.S.A. 72, 2626-2630 14 Tapiero, H., Monier, M.N., Shaol, D. and Hard, J. (1974) Nucleic Acids Res. 1,309-322 15 Thoma, F. and Koller, T. (1977) Cell 12, 101-107 16 Ohba, Y., Moremitsu, Y. and Watarai, A. (1979) Eur. J. Biochem. 100, 285-293 17 Case, S.T., Mongeon, L. and Baker, R.F. (1974) Biochim. Biophys. Acta 349, 1-12 18 Zubay, G., Wilkins, M.W.F. and Blout, E.N. (1962) J. Mol. Biol. 4, 69-72 19 Bosely, P.G., Bradbury, E.M., Butler-Browne, G.S., Carpenter, B.G. and Stephens, R.M. (1976) Eur. J. Biochem. 62, 21-31 20 Stein, A. (1979) J. Mol. Biol. 130, 103-134