Biochimica et Biophysica Acta, 1090(1991)38-42 © 1991ElsevierSciencePublishers B.V.0167-4781/91/$03.50 ADONIS 016747819100191G
38
BBAEXP92279
Histones and D N A methylation in mammalian chromatin. Differential inhibition by histone H 1 Paola Caiafa ', Anna Reale 2, Paola Allegra ~, Matilde Rispoli ~, Maria D'Erme ~ and Roberto Strom 2 Departments of I Biochemical Sciences and of 2 Human Biopathology, Unirersity of Rome "La Sapienza" and C.N.R. Centre for Molecular Biology, Rome (Italy)
(Received 27 November1090) Keywords: Histone;DNA methylation;Chromatin Histones (from calf thymus or from human placenta), if renatured in the presence of EDTA, caused a severe inhibition of in vitro methylation of doubie-stranded DNA (from M/crococcus /uteus) by human placenta DNA methyltransferase. The absence of EDTA during the histone renaturation procedure abolished - at least in the 'physiological' range of the histones/DNA ratio - the inhibition. The H t component was responsible for this inhibition, no effect being exerted by the other histones. H n preparations were more effective if renatured in the presence of EDTA - 90% inhibition being reached at a 0.3:1 ( w / w ) H I / D N A ratio. It seems likely that the requirement for the presence of EDTA during the renaturation process is correlated to its ability to induce a fairly stable ordered conformation of the bistones, although this effect could also be shown with the 'inactive' H , . , H ~ and H 3 components, and was instead less evident with histone H n. The restriction to histone H t of the ability to inhibit enzymic DNA methylation may account for the lower methylation levels present in the internucleosomal DNA of mammalian chromatin.
Introduction DNA methylation in eukaryotic cells appears to play an important role in the modulation of gene expression [1-6]. The complexity of chromatin structure [7], whose organization in differently ordered levels depends essentially on the interactions between DNA and specific proteins, suggests anyhow a role of these interactions in regulating the activity of DNA methyltransferase and in determining the methylation levels of the corresponding DNA regions. Davis and coworkers [8] have indeed reported that the activity of the mouse ascites cell enzyme is almost completely inhibited by addition of a crude histone preparation in a 1:1 ratio with respect to DNA. In addition, Kautiainen and Jones [9], by studying the methyl-accepting activity of nuclei extracted by increas-
Abbreviations:CD, circular dichroism;SDS,sodium dodecylsulfate. Correspondence: P. Caiafa, Dipartimento di Scienze Biochimiche, Universitit di Roma, 'La Sapienza', Piazzale Aldo Moro, 5, 00185 Rome, Italy.
ing salt concentrations, have shown, on the other hand, that native chromatin can be transformed into an almost 20-fold better methyl-acceptor by extraction with 2 M NaCI. The present study was performed with the aim of ascertaining whether this inhibitory potency could be correlated to the structural role played by the different histones and whether relatively slight changes in the experimental conditions were capable of influencing it. Materials and Methods Histones
Commercial histone preparations (from calf thymus) were purchased from Boehringer-Mannheim, F.R.G. 'Total' human placenta histones were obtained by extracting whole placenta chromatin (prepared according to Riekwood and Birnie [10]) with 0.2 M HzSO 4 [11], followed by extensive dialyses and precipitation by 9 vol. of cold acetone. 'Core' histones were prepared by performing the acid extraction on chromatin which had been pelleted after dissociation in 0.6 M NaCI; they were completely deprived of H t by extracting them with HCIO4 according to Johns [12]. All histone prepa-
39 rations, as well as their subsequent manipulations, were carried out in the presence of 1 mM phenylmethylsulfonyl fluoride (from Fluka, Buchs, Switzerland), in order to inhibit chrorratin-bound proteinases. Monodimensional polyacrylt !de gel electrophoresis in 1% (w/v) sodium dodecyl sulfate (SDS) according to Laemmli [13] was used t o verify the composition and purity of the various preparations. Quantitative determination of proteins was carried out by a commercial adaptation (Bio-Rad, Richmond, CA, U.S.A.) of Bradford's procedure [14], verified by analysis of their acid hydrolysate (6 M HCI at l l 0 ° C for 24 h) on a LKB 4400 amino acid analyser.
and the DNA was precipitated at 0 ° C with 10% (w/v, final concentration) trichloroacetic acid. The pellets, washed again with trichloroacetic acid, were resuspended in 0.5 ml of 0.5 M NaOH and heated to 60 ° C for 20 rain to remove, by alkaline hydrolysis, any contaminating RNA. After cooling, DNA was precipitated again at 0 o C with 2 ml of 15% trichloroacetic acid and collected on a glass-fiber paper ( G F / C , Whatman), repeatedly washed with 5% trichloroacetic acid and then with 95% ethanol. The radioactivity was then measured in a Beckman LS-6800 liquid scintillation spectrometer. Results
Histone renaturation Prior to physico-chemical measurements and to enzymatic assays, histones were usually renatured by progressive dialyses as described by Hentzen and Bckhor [15], starting from 2 M NaC! + 5 M urea in 10 mM Tris-HCI buffer (pH 7.5), the final buffer being 10 mM Tris-HCi (pH 7.8); as it will be described under Results, the presence or absence of 5 mM EDTA (or of EGTA) throughout the dialysis steps was found to be of importance in determining the quality of the 'renatured' preparations. Circular dichroism Circular dichroism (CD) spectra were recorded with a Jasco J-500 A spectropolarimeter equipped with a DP-500 N data processor, using 2 mm light-path cells, at a temperature of 15 ( + 2 ) °C. The molar ellipticity, 0, was expressed as mdegrees cm 2 d m o l - t , referred to amino acid residues concentration. Histone concentration was usually around 10 -5 M. DNA methyltransferase assay D N A methyltransferase (EC 2.1.1.37) was purified from human placenta nuclei according to Carotti and coworkers [16], and usually assayed according to Caiafa and coworkers [17] in a 50 mM Tris-HC! buffer (pH 7.8) containing 10% (v/v) glycerol, 5 mM EDTA, 0.5 mM dithiothreitol, 10 / t C i / m l S-adenosyl-L-[methyl3H]methionine (New England Nuclear; specific activity 70-80 Ci/mmol) and 3 0 / t g / m l of Micrococcus luteus dsDNA. Specific proteins to be tested for their effect on D N A methylation were also present, when required, up to a 1 : 1 ratio (w/w) with respect to DNA. One enzyme unit was defined as that catalyzing the incorporation of 1 pmol of methyl groups into Micrococcus luteus dsDNA at 3 7 ° C under standard assay conditions. Assay mixtures were incubated for 30 min at 37 °C, the reaction being stopped by heating to 60 °C for 20 mitt in the presence of 1% (w/v) SDS and 0.2 p g / m l proteinase K. After cooling on ice, 300 / t g / m l of salmon sperm DNA were added to serve as carrier,
'Total' histones As shown in Fig. 1, the ability of 'total' histoncs from calf thymus (the same being true for human placenta histones) to affect in vitro enzymic D N A methylation was markedly different according to the presence or absence of EDTA during the renaturation steps, a marked inhibition being reached at a 1:1 histones-to-DNA ratio only in the former case. Using the same preparations of histones, renaturation in the presence or absence of 5 mM Na-EDTA was also found to modify the circular dichroism characteristics of the peptide backbone, even though most EDTA had been removed by 2 h dialysis to allow C'D measurements, which were actually performed in 5
,30[: •
-
¢0
"q--
"-4
\
70
'X
Z
~
'
'
/o
P r o t e i n / DNA rfltto (w: w )
Fig. 1. Effect, on the in vitro activity of human placenta D N A
methyitransferase,of total histones(from calf thymus) renatured by progressive dialysisat decreasing urea and NaCt cocgentratiousin the presence ( • . . . . . . • ) or absence (e-- -.--e) of 5 mM EDTA. Each point represents the mean r=ult of at least five different experiments in triplicate, + S.D.
40 I
[
r
~
;
mM Tris-HCI buffer (pH 7.8). As shown in Fig. 2, the overall amount of right-handed folding was indeed comparable, if Na-EDTA was present during renaturation, to that obtained upon addition of 1% ( w / v ) SDS, while it was quite negligible if EDTA had been omitted. Na-EGTA possessed the same renaturing capability as Na-EDTA; at the same concentration, Mg-EGTA was instead quite uneffective, while Ca-EGTA exhibited only a minor effect.
i
1O~ - r -
=
'l , i f I
~-//,
Single histones
,? 9
p
×
f
®
i i
/ ~..,/
!
-20 '
200
' 230 ~,~velengt h (rim)
'
'
260
Fig. 2. Circular dichroism spectra of total histones (from calf thymus) renatured by progressive dialysis at decreasing urea and NaCI concentrations in the presence ( . . . . . . ) or absence ( . . . . . ) of 5 mM EDTA. Since EDTA, if present, decreased the signal-to-noise ratio, all samples were subjected to 3 h dialysis against 5 mM Tris-HCI (pH 7.8) and adjusted to a final concentration of 0.1 mg protein/ml (usually by a 1:2 dilution) prior to actual recording of the CD spectra. The CD spectrum of total histones in I % sodium dodecylsulfate is also shown (full line).
Among the various histones, only H, was capable of inhibiting in vitro enzymic D N A methylation, 90% inhibition being reached at a histone-to-DNA ratio of approx. 0.3 when 5 mM EDTA had been present during renaturation, and at a ratio slightly above 0.6 when EDTA had been omitted (Fig. 3, upper left panel). None of the other histones, H2a, H2b and H 3, was able to affect significantly the D N A methylation process, up to a 1 : 1 ratio (Fig. 3, other panels). As for H 4, its considerable tendency to aggregation [18] may have interfered with the experimental finding of an apparent ineffectiveness. In terms of overall secondary structure, renaturation
H1 H1
H20 "
I"
H2Q
130
--a
100 " " l
T
,o.\ ,,
/
I\
%%
"
4O t
%. c.
8
lo i
i
i
i
i
i
i
H2b 130
H2b I
H3
H3
...-o
;.,k.
1C
i
t-
&..,,/.
o ,,lk
F
oz
-2C
×
× ®
i-
C
~ J.-.,L._ : - - a - . . . . .
70 40 i t
-2O
10
"
2OO
'
d~
'
Protein
1'o / DNA
23O
26O 2OO
23O
260
WovelenCjt h (nm)
'
o'~
'
;o
ratio ( w w)
Fig. 3. Effect, on the in vitro activity of haman placenta D N A methyltransferase, of histones Hi, H ~ , H~o and H 3 (from calf thymus) renatured by progressive dialys/s at decreasing urea and NaCI concentrations in the presence ( a . . . . . . a ) or absence (e . . . . . e ) of 5 mM EDTA. Each point represents the mean result of at least five different experiments in triplicate, 4- S.D.
Fig. 4. Circular dichroism spectra of histoncs He, H ~ , Hzb and H 3 (from calf thymus) renatured by progressive dialysis at decreasing urea and NaCI concentrations in the presence ( . . . . . . ) or absence ( . . . . . ) of 5 mM EDTA. Histonc H I was also renatured by progressive dialysis in the presence of 30 mM EDTA ( . . . . . . ). After renaturation, all samples were processed as indicated for total histones in the legend of Fig. 2. The spectra of the histones in !% sodium dodecylsulfate are also shown (full lines).
41 tained if 30 mM Na-EDTA (or Na-EGTA) was used (Fig. 4, upper left panel). The other histones exhibited instead a fairly ordered conformation upon addition of 5 mM Na-EDTA (Fig. 4, other panels).
i
L_S7 .......
'Core' histones and reconstitution experiments &
~
7c
.< z 121
40
10
I
I O5
Protein I D N A
I cargo
I '10
(w:w)
Fig. 5. Effect, on the in vitro activity of human placenta DNA methyltransferase, of Ht-depleted 'core' histoncs (from human placenta) renatured by progressive dialysis at decreasing urea and NaCI concentrations in the presence ( • . . . . . . • )c.r absence (e . . . . . e) of 5 mM EDTA. Each point represents the mean result of at least 5 different experiments in tripli:ate, ~ S.D.
Preparations of placental 'core' histones, obtained from Hi-depleted chromatin (and in which the electrophoretic analysis had demonstrated the presence of H ~ , Ha,, Ha and H 4 in equal amounts, as well as the absence of H~) exerted practically no effect on in vitro DNA metbylation (Fig. 5), independently of the presence or absence of 5 mM EDTA during the renaturation procedure. The presence of EDTA during renaturation resulted in an ordered overall conformation (Fig. 6). Addition of H l to such 'core' histone preparations in a 1:8 molar ratio with respect to the sum of the other histones restored the inhibitory effect on DNA methylation present in preparations of total histones, provided that renaturatio~ was performed in the presence of EDTA. Quite surprisingly, no inhibition was exerted, in preliminary experiments (data not shown), by Hi + H2b mixtures in the same 1:8 molar ratio. Discussion
in the presence of 5 mM Na-EDTA could induc~, as far as H t histone was concerned, only ,: l;mited amount of right-handed folding able to persist after removal of this muitivalent anion - a quite high level being at10[- !
l
,
i
~
l
li~" l
~
x
./Z"
g
LV
x
(D
-2G
Wavetength [nm)
Fig. 6. Circular dichroism spectra of H t-depleted "core' histories (from humanplacenta)renatured by progressivedialysisat decreasing urea and NaCI concentrations in the presence (. . . . . . ) or absence (. . . . . ) of 5 mM EDTA. After renaturation,the samples were processed as indicatedfor total historiesin the legendof Fig. 2. The spectra of HI-depleted "core" histones in 1% sodium dodecyl sulfate are alsoshown{fullline).
The restriction, to a single histone species, of the ability to suppress in vitro DNA methylation is consistent with previous reports that nucleosomal core DNA contains more methylcytosine than the internucleosomai linker DNA [19-22]. In our experiments, the presence of EDTA during the renaturation process played a critical role, presumably by inducing, through local compensation of clusters of positive charges, fairly ordered protein conformation(s) which remained to some extent even after EDTA removal. Among the various histones, H, was indeed the one which was most reluctant to the conformational effect of EDTA - not too great a surprise if we recall how rich it is in cationic side chains. In the absence of other histones, even a random conformation of H, was capable of exerting some inhibition of DNA methylation; addition of EDTA during the renaturation process resulted anyhow in a potentiation of this inhibitory effect. Since it was verified that addition of H I caused no appreciable modification of the CD spectrum of DNA in the 250300 nm region (experiments not shown), the inh~itory effect of H~ cannot be ascribed to a non-specific precipitation of DNA by a cationic protein. In total histones (whether prepared as such or obtained by reconstitution of H, plus core histories) the inhibition was instead critically dependent on the EDTA-induced formation of an ordered structure. A possible role of homeotypic a n d / o r heterotypic bonds [7] in regulating the inhibitory potency of histone H i is s,~ggested by the
42 lower efficiency obtained in the reconstitution experiments. As fr,r the other histories, none of them, though capable of undergoing a clearcut transition to an ordered conformation upon addition of EDTA, was able to suppress DNA methylation. The inhibitory effect of H~ on DNA methylation and its well-known tendency to bind to the 'linker' regions of double-stranded DNA [23] could account for the relatively higher methylation levels of nucleosomai core DNA, which is, however, also characterized by a higher relative abundance of CpG sequences [5]. The hypomethylation of the linker DNA would also be reflected on the stability of the higher levels of chromatin organization; in this optics it is worth mentioning that histone H~ reportedly plays a fairly specific role in compacting the solenoid structure [24], and that it has a high affinity for the 'scaffold-associa .~d regions' of the chromatin DNA [25]. If should, however, be noticed that, in the cell cycle, since the peak of DNA methylation occurs, upon replication, prior to H~ binding, only the 'delayed' methylation can be expected to suffer from the inhibitory effect on this histone. As for a ~x~ssible relationship between the in vitro inhibitory effect of H, on the enzymatic reaction leading to the formation of 5-methylcytosine and the in vivo biological role attributed to DNA methylation [1-6], it seems unlikely that H,, being one of the major chromatin proteins, can play a direct role in regulating the activation or deactivation of single genes. Although H, has been reported to be absent from the 'active chromatin' regions, where are located the undermethylated, CpG-rich regulatory sequences of the housekeeping genes [26], these CpG-rich islands represents indeed less than 2% of the whole genome, and specific regulatory proteins [27], different from histone H~, must contribute, maybe together with other factor(s) [28], in keeping these islands in an unmethylated state. In our view, instead, a general hypomethylation of linker DNA, such as would derive from the inhibitory effect of H, on the DNA methylation process, can be of importance in ensuring the appropriate level of chromatin condensation. Ac"lamwledgements This work was supported by the Italian Ministry of Education (60%, Progetti di Ateneo) and took advan-
tage of the facilities made available by the Fondazione 'lstituto Pasteur - Fondazione Cenci Bolognetti'. References 1 Razin, A. and Riggs, A.D. (1980) Science 210, 604-610. 2 Naveh-Many, T. and Cedar, H. (1981) Proc. Natl. Acad. Sci. USA 78, 4246-4250. 3 Ehrlich, M. and Wang, R.Y.H. (1981) Science 212, 1350-1357. 4 Doerfler, W. (1983) Annu. Rev. Biochem. 52, 93-124. 5 Adams, R.L.P. and Burdon, R. (1985) in Molecular Biology of DNA Methylation. Springer Series in Molecular Biology (Rich, A., ed.). Springer-Verlag, New York. 6 Jones, PA. (1986)Cell 40. 485-486. 7 Van Holde, K.E. (1989) in Chromatin. Springer Series in Molecular Biology (Rich, A., ed.). Springer-Verlag, New York. 8 Davis, T., Rinaldi, A., Clark, L. and Adams, R.L.P. (1986) Biochim. Biophys. Acta 866, 233-241. 9 Kautiainen, T.L. and Jones, P.A. (1985) Biochemistry 24, ! 1931197. 10 Rick'wood, D. and Birnie, G.D. (1976) in Subnuclear Components (Birnie, G.D., ed.), pp. 129-186. Buttetworths. London. 11 Monahan, J.J. and Hall, R.H. (1973) Can. J. Biochem. 51, 70n720. 12 Johns, E.W. (1977) Methods Cell Biol. 16, p. 183-203. 13 Laemmli, U.K. (1979) Nature 227, 680-685. 14 Bradford, M.M. (1976) Anal. Biochem. 72, 248-254. 15 Hentzen, P.C. and Bekhor, l. (1985) in Progress in Nonhistone Protein Research (Bekhor, l.. ed.), VOl. !, Chapter 4, pp. 75-101. CRC Press, Boca Raton. 16 Carotti, D., Palini, F., Maslrantonio, S., Rispoli, M., Strom, R., Amato, A. Campagnari, F. and Whitehead, E.P. (1986) Biochim. Biophys. Acta 86, 135-143. 17 Caiafa, P., Mastrantonio, S.. Cacace, F., Attin~, M., Rispoli, M. and Strom, R. (1988) Biochim. Biophys. Acta 951, 191-200. 18 Smerdon, M.J. and lsemberg, 1. (1976) Biochemistry 15, 40464049. 19 Razin, A. and Cedar, H. (1977) Proc. Natl. Acad. bci. USA 74, 2725-2728. 20 Solage, A. and Cedar, H. 0978) Biochemistry 17, 2934-2938. 21 Adams, R.L.P., Davis, T., Fulton, J., Kirk, D., Oureshi, M. and Burdon, R.H. (1984)Curr. Top. Microbiol. Immunol. 108, 143156. 22 Calafa, P., Attin'~, M., Cacace, F.. Tomas~ni. A. and Strom, R. (1986) Biochim. Biophys. Acta 867, 195-200. 23 NolL M. and Kornberg, R.D. 0977) J. Mol. Biol. 109, 393-404. 24 Thoma, F., Koller, T. and Klug, A. (1979) J. Cell Biol. 83, 403-427. 25 Kas, E., Izaurralde, E. and Laemmli, U.K. (1989) J. MOt. Biol. 210, 587-599. 26 Tazi, J. and Bird, A. (1990) Cell 60, 909-920. 27 Bird, A.P. (1986) Nature 321, 20'.3-213. 28 Carotti, D., Palini, F., Lavia, P. and Strom, R. (1989) Nucleic Acids Res. 17, 9219-9229.
38
BBAEXP92279
Histones and D N A methylation in mammalian chromatin. Differential inhibition by histone H 1 Paola Caiafa ', Anna Reale 2, Paola Allegra ~, Matilde Rispoli ~, Maria D'Erme ~ and Roberto Strom 2 Departments of I Biochemical Sciences and of 2 Human Biopathology, Unirersity of Rome "La Sapienza" and C.N.R. Centre for Molecular Biology, Rome (Italy)
(Received 27 November1090) Keywords: Histone;DNA methylation;Chromatin Histones (from calf thymus or from human placenta), if renatured in the presence of EDTA, caused a severe inhibition of in vitro methylation of doubie-stranded DNA (from M/crococcus /uteus) by human placenta DNA methyltransferase. The absence of EDTA during the histone renaturation procedure abolished - at least in the 'physiological' range of the histones/DNA ratio - the inhibition. The H t component was responsible for this inhibition, no effect being exerted by the other histones. H n preparations were more effective if renatured in the presence of EDTA - 90% inhibition being reached at a 0.3:1 ( w / w ) H I / D N A ratio. It seems likely that the requirement for the presence of EDTA during the renaturation process is correlated to its ability to induce a fairly stable ordered conformation of the bistones, although this effect could also be shown with the 'inactive' H , . , H ~ and H 3 components, and was instead less evident with histone H n. The restriction to histone H t of the ability to inhibit enzymic DNA methylation may account for the lower methylation levels present in the internucleosomal DNA of mammalian chromatin.
Introduction DNA methylation in eukaryotic cells appears to play an important role in the modulation of gene expression [1-6]. The complexity of chromatin structure [7], whose organization in differently ordered levels depends essentially on the interactions between DNA and specific proteins, suggests anyhow a role of these interactions in regulating the activity of DNA methyltransferase and in determining the methylation levels of the corresponding DNA regions. Davis and coworkers [8] have indeed reported that the activity of the mouse ascites cell enzyme is almost completely inhibited by addition of a crude histone preparation in a 1:1 ratio with respect to DNA. In addition, Kautiainen and Jones [9], by studying the methyl-accepting activity of nuclei extracted by increas-
Abbreviations:CD, circular dichroism;SDS,sodium dodecylsulfate. Correspondence: P. Caiafa, Dipartimento di Scienze Biochimiche, Universitit di Roma, 'La Sapienza', Piazzale Aldo Moro, 5, 00185 Rome, Italy.
ing salt concentrations, have shown, on the other hand, that native chromatin can be transformed into an almost 20-fold better methyl-acceptor by extraction with 2 M NaCI. The present study was performed with the aim of ascertaining whether this inhibitory potency could be correlated to the structural role played by the different histones and whether relatively slight changes in the experimental conditions were capable of influencing it. Materials and Methods Histones
Commercial histone preparations (from calf thymus) were purchased from Boehringer-Mannheim, F.R.G. 'Total' human placenta histones were obtained by extracting whole placenta chromatin (prepared according to Riekwood and Birnie [10]) with 0.2 M HzSO 4 [11], followed by extensive dialyses and precipitation by 9 vol. of cold acetone. 'Core' histones were prepared by performing the acid extraction on chromatin which had been pelleted after dissociation in 0.6 M NaCI; they were completely deprived of H t by extracting them with HCIO4 according to Johns [12]. All histone prepa-
39 rations, as well as their subsequent manipulations, were carried out in the presence of 1 mM phenylmethylsulfonyl fluoride (from Fluka, Buchs, Switzerland), in order to inhibit chrorratin-bound proteinases. Monodimensional polyacrylt !de gel electrophoresis in 1% (w/v) sodium dodecyl sulfate (SDS) according to Laemmli [13] was used t o verify the composition and purity of the various preparations. Quantitative determination of proteins was carried out by a commercial adaptation (Bio-Rad, Richmond, CA, U.S.A.) of Bradford's procedure [14], verified by analysis of their acid hydrolysate (6 M HCI at l l 0 ° C for 24 h) on a LKB 4400 amino acid analyser.
and the DNA was precipitated at 0 ° C with 10% (w/v, final concentration) trichloroacetic acid. The pellets, washed again with trichloroacetic acid, were resuspended in 0.5 ml of 0.5 M NaOH and heated to 60 ° C for 20 rain to remove, by alkaline hydrolysis, any contaminating RNA. After cooling, DNA was precipitated again at 0 o C with 2 ml of 15% trichloroacetic acid and collected on a glass-fiber paper ( G F / C , Whatman), repeatedly washed with 5% trichloroacetic acid and then with 95% ethanol. The radioactivity was then measured in a Beckman LS-6800 liquid scintillation spectrometer. Results
Histone renaturation Prior to physico-chemical measurements and to enzymatic assays, histones were usually renatured by progressive dialyses as described by Hentzen and Bckhor [15], starting from 2 M NaC! + 5 M urea in 10 mM Tris-HCI buffer (pH 7.5), the final buffer being 10 mM Tris-HCi (pH 7.8); as it will be described under Results, the presence or absence of 5 mM EDTA (or of EGTA) throughout the dialysis steps was found to be of importance in determining the quality of the 'renatured' preparations. Circular dichroism Circular dichroism (CD) spectra were recorded with a Jasco J-500 A spectropolarimeter equipped with a DP-500 N data processor, using 2 mm light-path cells, at a temperature of 15 ( + 2 ) °C. The molar ellipticity, 0, was expressed as mdegrees cm 2 d m o l - t , referred to amino acid residues concentration. Histone concentration was usually around 10 -5 M. DNA methyltransferase assay D N A methyltransferase (EC 2.1.1.37) was purified from human placenta nuclei according to Carotti and coworkers [16], and usually assayed according to Caiafa and coworkers [17] in a 50 mM Tris-HC! buffer (pH 7.8) containing 10% (v/v) glycerol, 5 mM EDTA, 0.5 mM dithiothreitol, 10 / t C i / m l S-adenosyl-L-[methyl3H]methionine (New England Nuclear; specific activity 70-80 Ci/mmol) and 3 0 / t g / m l of Micrococcus luteus dsDNA. Specific proteins to be tested for their effect on D N A methylation were also present, when required, up to a 1 : 1 ratio (w/w) with respect to DNA. One enzyme unit was defined as that catalyzing the incorporation of 1 pmol of methyl groups into Micrococcus luteus dsDNA at 3 7 ° C under standard assay conditions. Assay mixtures were incubated for 30 min at 37 °C, the reaction being stopped by heating to 60 °C for 20 mitt in the presence of 1% (w/v) SDS and 0.2 p g / m l proteinase K. After cooling on ice, 300 / t g / m l of salmon sperm DNA were added to serve as carrier,
'Total' histones As shown in Fig. 1, the ability of 'total' histoncs from calf thymus (the same being true for human placenta histones) to affect in vitro enzymic D N A methylation was markedly different according to the presence or absence of EDTA during the renaturation steps, a marked inhibition being reached at a 1:1 histones-to-DNA ratio only in the former case. Using the same preparations of histones, renaturation in the presence or absence of 5 mM Na-EDTA was also found to modify the circular dichroism characteristics of the peptide backbone, even though most EDTA had been removed by 2 h dialysis to allow C'D measurements, which were actually performed in 5
,30[: •
-
¢0
"q--
"-4
\
70
'X
Z
~
'
'
/o
P r o t e i n / DNA rfltto (w: w )
Fig. 1. Effect, on the in vitro activity of human placenta D N A
methyitransferase,of total histones(from calf thymus) renatured by progressive dialysisat decreasing urea and NaCt cocgentratiousin the presence ( • . . . . . . • ) or absence (e-- -.--e) of 5 mM EDTA. Each point represents the mean r=ult of at least five different experiments in triplicate, + S.D.
40 I
[
r
~
;
mM Tris-HCI buffer (pH 7.8). As shown in Fig. 2, the overall amount of right-handed folding was indeed comparable, if Na-EDTA was present during renaturation, to that obtained upon addition of 1% ( w / v ) SDS, while it was quite negligible if EDTA had been omitted. Na-EGTA possessed the same renaturing capability as Na-EDTA; at the same concentration, Mg-EGTA was instead quite uneffective, while Ca-EGTA exhibited only a minor effect.
i
1O~ - r -
=
'l , i f I
~-//,
Single histones
,? 9
p
×
f
®
i i
/ ~..,/
!
-20 '
200
' 230 ~,~velengt h (rim)
'
'
260
Fig. 2. Circular dichroism spectra of total histones (from calf thymus) renatured by progressive dialysis at decreasing urea and NaCI concentrations in the presence ( . . . . . . ) or absence ( . . . . . ) of 5 mM EDTA. Since EDTA, if present, decreased the signal-to-noise ratio, all samples were subjected to 3 h dialysis against 5 mM Tris-HCI (pH 7.8) and adjusted to a final concentration of 0.1 mg protein/ml (usually by a 1:2 dilution) prior to actual recording of the CD spectra. The CD spectrum of total histones in I % sodium dodecylsulfate is also shown (full line).
Among the various histones, only H, was capable of inhibiting in vitro enzymic D N A methylation, 90% inhibition being reached at a histone-to-DNA ratio of approx. 0.3 when 5 mM EDTA had been present during renaturation, and at a ratio slightly above 0.6 when EDTA had been omitted (Fig. 3, upper left panel). None of the other histones, H2a, H2b and H 3, was able to affect significantly the D N A methylation process, up to a 1 : 1 ratio (Fig. 3, other panels). As for H 4, its considerable tendency to aggregation [18] may have interfered with the experimental finding of an apparent ineffectiveness. In terms of overall secondary structure, renaturation
H1 H1
H20 "
I"
H2Q
130
--a
100 " " l
T
,o.\ ,,
/
I\
%%
"
4O t
%. c.
8
lo i
i
i
i
i
i
i
H2b 130
H2b I
H3
H3
...-o
;.,k.
1C
i
t-
&..,,/.
o ,,lk
F
oz
-2C
×
× ®
i-
C
~ J.-.,L._ : - - a - . . . . .
70 40 i t
-2O
10
"
2OO
'
d~
'
Protein
1'o / DNA
23O
26O 2OO
23O
260
WovelenCjt h (nm)
'
o'~
'
;o
ratio ( w w)
Fig. 3. Effect, on the in vitro activity of haman placenta D N A methyltransferase, of histones Hi, H ~ , H~o and H 3 (from calf thymus) renatured by progressive dialys/s at decreasing urea and NaCI concentrations in the presence ( a . . . . . . a ) or absence (e . . . . . e ) of 5 mM EDTA. Each point represents the mean result of at least five different experiments in triplicate, 4- S.D.
Fig. 4. Circular dichroism spectra of histoncs He, H ~ , Hzb and H 3 (from calf thymus) renatured by progressive dialysis at decreasing urea and NaCI concentrations in the presence ( . . . . . . ) or absence ( . . . . . ) of 5 mM EDTA. Histonc H I was also renatured by progressive dialysis in the presence of 30 mM EDTA ( . . . . . . ). After renaturation, all samples were processed as indicated for total histones in the legend of Fig. 2. The spectra of the histones in !% sodium dodecylsulfate are also shown (full lines).
41 tained if 30 mM Na-EDTA (or Na-EGTA) was used (Fig. 4, upper left panel). The other histones exhibited instead a fairly ordered conformation upon addition of 5 mM Na-EDTA (Fig. 4, other panels).
i
L_S7 .......
'Core' histones and reconstitution experiments &
~
7c
.< z 121
40
10
I
I O5
Protein I D N A
I cargo
I '10
(w:w)
Fig. 5. Effect, on the in vitro activity of human placenta DNA methyltransferase, of Ht-depleted 'core' histoncs (from human placenta) renatured by progressive dialysis at decreasing urea and NaCI concentrations in the presence ( • . . . . . . • )c.r absence (e . . . . . e) of 5 mM EDTA. Each point represents the mean result of at least 5 different experiments in tripli:ate, ~ S.D.
Preparations of placental 'core' histones, obtained from Hi-depleted chromatin (and in which the electrophoretic analysis had demonstrated the presence of H ~ , Ha,, Ha and H 4 in equal amounts, as well as the absence of H~) exerted practically no effect on in vitro DNA metbylation (Fig. 5), independently of the presence or absence of 5 mM EDTA during the renaturation procedure. The presence of EDTA during renaturation resulted in an ordered overall conformation (Fig. 6). Addition of H l to such 'core' histone preparations in a 1:8 molar ratio with respect to the sum of the other histones restored the inhibitory effect on DNA methylation present in preparations of total histones, provided that renaturatio~ was performed in the presence of EDTA. Quite surprisingly, no inhibition was exerted, in preliminary experiments (data not shown), by Hi + H2b mixtures in the same 1:8 molar ratio. Discussion
in the presence of 5 mM Na-EDTA could induc~, as far as H t histone was concerned, only ,: l;mited amount of right-handed folding able to persist after removal of this muitivalent anion - a quite high level being at10[- !
l
,
i
~
l
li~" l
~
x
./Z"
g
LV
x
(D
-2G
Wavetength [nm)
Fig. 6. Circular dichroism spectra of H t-depleted "core' histories (from humanplacenta)renatured by progressivedialysisat decreasing urea and NaCI concentrations in the presence (. . . . . . ) or absence (. . . . . ) of 5 mM EDTA. After renaturation,the samples were processed as indicatedfor total historiesin the legendof Fig. 2. The spectra of HI-depleted "core" histones in 1% sodium dodecyl sulfate are alsoshown{fullline).
The restriction, to a single histone species, of the ability to suppress in vitro DNA methylation is consistent with previous reports that nucleosomal core DNA contains more methylcytosine than the internucleosomai linker DNA [19-22]. In our experiments, the presence of EDTA during the renaturation process played a critical role, presumably by inducing, through local compensation of clusters of positive charges, fairly ordered protein conformation(s) which remained to some extent even after EDTA removal. Among the various histones, H, was indeed the one which was most reluctant to the conformational effect of EDTA - not too great a surprise if we recall how rich it is in cationic side chains. In the absence of other histones, even a random conformation of H, was capable of exerting some inhibition of DNA methylation; addition of EDTA during the renaturation process resulted anyhow in a potentiation of this inhibitory effect. Since it was verified that addition of H I caused no appreciable modification of the CD spectrum of DNA in the 250300 nm region (experiments not shown), the inh~itory effect of H~ cannot be ascribed to a non-specific precipitation of DNA by a cationic protein. In total histones (whether prepared as such or obtained by reconstitution of H, plus core histories) the inhibition was instead critically dependent on the EDTA-induced formation of an ordered structure. A possible role of homeotypic a n d / o r heterotypic bonds [7] in regulating the inhibitory potency of histone H i is s,~ggested by the
42 lower efficiency obtained in the reconstitution experiments. As fr,r the other histories, none of them, though capable of undergoing a clearcut transition to an ordered conformation upon addition of EDTA, was able to suppress DNA methylation. The inhibitory effect of H~ on DNA methylation and its well-known tendency to bind to the 'linker' regions of double-stranded DNA [23] could account for the relatively higher methylation levels of nucleosomai core DNA, which is, however, also characterized by a higher relative abundance of CpG sequences [5]. The hypomethylation of the linker DNA would also be reflected on the stability of the higher levels of chromatin organization; in this optics it is worth mentioning that histone H~ reportedly plays a fairly specific role in compacting the solenoid structure [24], and that it has a high affinity for the 'scaffold-associa .~d regions' of the chromatin DNA [25]. If should, however, be noticed that, in the cell cycle, since the peak of DNA methylation occurs, upon replication, prior to H~ binding, only the 'delayed' methylation can be expected to suffer from the inhibitory effect on this histone. As for a ~x~ssible relationship between the in vitro inhibitory effect of H, on the enzymatic reaction leading to the formation of 5-methylcytosine and the in vivo biological role attributed to DNA methylation [1-6], it seems unlikely that H,, being one of the major chromatin proteins, can play a direct role in regulating the activation or deactivation of single genes. Although H, has been reported to be absent from the 'active chromatin' regions, where are located the undermethylated, CpG-rich regulatory sequences of the housekeeping genes [26], these CpG-rich islands represents indeed less than 2% of the whole genome, and specific regulatory proteins [27], different from histone H~, must contribute, maybe together with other factor(s) [28], in keeping these islands in an unmethylated state. In our view, instead, a general hypomethylation of linker DNA, such as would derive from the inhibitory effect of H, on the DNA methylation process, can be of importance in ensuring the appropriate level of chromatin condensation. Ac"lamwledgements This work was supported by the Italian Ministry of Education (60%, Progetti di Ateneo) and took advan-
tage of the facilities made available by the Fondazione 'lstituto Pasteur - Fondazione Cenci Bolognetti'. References 1 Razin, A. and Riggs, A.D. (1980) Science 210, 604-610. 2 Naveh-Many, T. and Cedar, H. (1981) Proc. Natl. Acad. Sci. USA 78, 4246-4250. 3 Ehrlich, M. and Wang, R.Y.H. (1981) Science 212, 1350-1357. 4 Doerfler, W. (1983) Annu. Rev. Biochem. 52, 93-124. 5 Adams, R.L.P. and Burdon, R. (1985) in Molecular Biology of DNA Methylation. Springer Series in Molecular Biology (Rich, A., ed.). Springer-Verlag, New York. 6 Jones, PA. (1986)Cell 40. 485-486. 7 Van Holde, K.E. (1989) in Chromatin. Springer Series in Molecular Biology (Rich, A., ed.). Springer-Verlag, New York. 8 Davis, T., Rinaldi, A., Clark, L. and Adams, R.L.P. (1986) Biochim. Biophys. Acta 866, 233-241. 9 Kautiainen, T.L. and Jones, P.A. (1985) Biochemistry 24, ! 1931197. 10 Rick'wood, D. and Birnie, G.D. (1976) in Subnuclear Components (Birnie, G.D., ed.), pp. 129-186. Buttetworths. London. 11 Monahan, J.J. and Hall, R.H. (1973) Can. J. Biochem. 51, 70n720. 12 Johns, E.W. (1977) Methods Cell Biol. 16, p. 183-203. 13 Laemmli, U.K. (1979) Nature 227, 680-685. 14 Bradford, M.M. (1976) Anal. Biochem. 72, 248-254. 15 Hentzen, P.C. and Bekhor, l. (1985) in Progress in Nonhistone Protein Research (Bekhor, l.. ed.), VOl. !, Chapter 4, pp. 75-101. CRC Press, Boca Raton. 16 Carotti, D., Palini, F., Maslrantonio, S., Rispoli, M., Strom, R., Amato, A. Campagnari, F. and Whitehead, E.P. (1986) Biochim. Biophys. Acta 86, 135-143. 17 Caiafa, P., Mastrantonio, S.. Cacace, F., Attin~, M., Rispoli, M. and Strom, R. (1988) Biochim. Biophys. Acta 951, 191-200. 18 Smerdon, M.J. and lsemberg, 1. (1976) Biochemistry 15, 40464049. 19 Razin, A. and Cedar, H. (1977) Proc. Natl. Acad. bci. USA 74, 2725-2728. 20 Solage, A. and Cedar, H. 0978) Biochemistry 17, 2934-2938. 21 Adams, R.L.P., Davis, T., Fulton, J., Kirk, D., Oureshi, M. and Burdon, R.H. (1984)Curr. Top. Microbiol. Immunol. 108, 143156. 22 Calafa, P., Attin'~, M., Cacace, F.. Tomas~ni. A. and Strom, R. (1986) Biochim. Biophys. Acta 867, 195-200. 23 NolL M. and Kornberg, R.D. 0977) J. Mol. Biol. 109, 393-404. 24 Thoma, F., Koller, T. and Klug, A. (1979) J. Cell Biol. 83, 403-427. 25 Kas, E., Izaurralde, E. and Laemmli, U.K. (1989) J. MOt. Biol. 210, 587-599. 26 Tazi, J. and Bird, A. (1990) Cell 60, 909-920. 27 Bird, A.P. (1986) Nature 321, 20'.3-213. 28 Carotti, D., Palini, F., Lavia, P. and Strom, R. (1989) Nucleic Acids Res. 17, 9219-9229.
