JBC Connect with Cosmo for Collagen Detection

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Right arrow Citation Map
Services
Right arrow Email this article to a friend
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Right arrow reprints & permissions
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by Campoy, F. J.
Right arrow Articles by Bird, A.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by Campoy, F. J.
Right arrow Articles by Bird, A.
Social Bookmarking
Add to CiteULike   Add to Complore   Add to Connotea   Add to Del.icio.us   Add to Digg   Add to Reddit   Add to Technorati  
What's this?

Volume 270, Number 44, Issue of November 3, 1995 pp. 26473-26481
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.
Binding of Histone H1 to DNA Is Indifferent to Methylation at CpG Sequences (*)

(Received for publication, July 12, 1995)

Francisco J. Campoy ,  Richard R. Meehan ,  Stewart McKay ,  Julie Nixon ,  Adrian Bird (§)

From the Institute of Cell and Molecular Biology, University of Edinburgh, Kings Buildings, Edinburgh EH9 3JR, United Kingdom

ABSTRACT
INTRODUCTION
EXPERIMENTAL PROCEDURES
RESULTS
DISCUSSION
FOOTNOTES
ACKNOWLEDGEMENTS
REFERENCES

ABSTRACT

The possibility that histone H1 binds preferentially to DNA containing 5-methylcytosine in the dinucleotide CpG is appealing, as it could help to explain the repressive effects of methylation on gene activity. In this study, the affinity of purified H1 for methylated and non-methylated DNA sequences has been tested using both naked DNA and chromatin. Based on a variety of assays (bandshifts, filter-binding assays, Southwestern blots, and nuclease sensitivity assays), we conclude that H1 has no significant preference for binding to naked methylated DNA. Similarly, H1 showed the same affinities for methylated and non-methylated DNA when assembled into chromatin in a Xenopus oocyte extract. Thus potential cooperative interaction of H1 with polynucleosomal complexes is not enhanced by the presence of DNA methylation.

INTRODUCTION

The major DNA modification in vertebrates is methylation at the 5 position of cytosine in the dinucleotide CpG. About 60-90% of genomic CpGs contain 5-methyl cytosine (m^5C). Transfection assays with methylated genes have shown that CpG methylation often causes repression of transcription(1, 2, 3, 4) . The severity of repression depends on a number of parameters, particularly the location of methylation relative to the promoter (1, 5; but see (6) ), the local density of methyl-CpG pairs, and the strength of the promoter under test(7) . In general, high density methylation strongly inhibits transcription, whereas low density methylation can only inhibit weak promoters.

Several lines of evidence suggest that proteins that bind preferentially to methylated DNA are important components of the repression mechanism. Indeed, binding of the methyl-CpG-binding protein (MeCP) (^1)MeCP1 (8) shows the same dependence on local density of methylation as does transcriptional repression(4, 7) . Methylation-associated transcriptional repression may also arise by other routes. CpG methylation is known to prevent binding of some transcription factors, and this is likely to contribute to repression in some cases(9, 10) . Another potential cause of methylation-mediated repression is direct alteration of chromatin structure due to the presence of m^5C. Involvement of chromatin in mediating the effects of DNA methylation is suggested by the experiments of Buschhausen et al.(3, 11) . In addition, Keshet et al. found that methylated DNA is preferentially assembled into nuclease-resistant chromatin after transfection(12) . While these results could be explained by interaction between known methyl-CpG-binding proteins and chromatin, it is also possible that an ubiquitous component of chromatin, for example histones, interacts differentially with methylated DNA leading to an altered higher order structure that is incompatible with gene expression.

Attempts to detect an effect of DNA methylation on nucleosomal cores have not been successful(13, 14) . The linker histone H1, however, is an attractive candidate for mediator of the effects of methylation on chromatin. H1 plays a role in chromatin condensation(15, 16) . It is depleted in CpG islands, which are non-methylated and include the promoters of many genes, and in other potentially active genes, but is concentrated in inert chromatin(16, 17, 18) , and inhibits transcription in vitro(19, 20, 21) . If H1 were able to recognize methylated DNA, it could in theory contribute to the effects of methylation on chromatin structure and gene expression.

Studies using antibodies against m^5C have suggested a link between DNA methylation and histone H1 by showing that the modified base is preferentially localized in H1-containing nucleosomes(22) . Whether H1 has a higher affinity for methylated DNA is, however, an unresolved issue. Higurashi and Cole (23) did not detect any significant difference in the affinity of H1 for methylated or non-methylated DNA. Nevertheless, they suggest that the DNA complexed with histone H1 adopts a distinct conformation when it is methylated, since the methylated MspI sites became protected in the complexes. On the other hand, Levine et al.(24) reported that histone H1 bound preferentially to methylated DNA and claimed that H1 suppresses transcription from methylated templates in vitro. Jost and Hofsteenge (25) also implicated H1 as a methylated DNA-binding protein. They obtained a protein fraction from chicken liver enriched in histone H1 that showed binding to methylated CpG in a specific sequence context. The activity, known as MDBP-2, was originally identified as a 40-kDa protein that was very tissue specific in distribution (26) but was subsequently attributed to a dimer of H1 (25) . Finally, Johnson et al.(27) have concluded that H1 preferentially inhibits transcription from a methylated template, and they propose that methylated sites on the DNA serve as foci for long range chromatin condensation mediated by H1.

In this study, we have reinvestigated the interaction of purified chicken erythrocyte histone H1 (free of the erythrocyte variant H5) and rat kidney H1 with methylated and non-methylated DNA and chromatin using several assays. The results of comparative gel retardations, Southwestern blots, filter binding assays, and nuclease-protection assays indicate that histone H1 shows little or no preference for binding to naked methylated DNA. The apparent selective protection by H1 of methylated MspI sites can be explained by a characteristic of the endonuclease rather than the DNA itself. Most significantly, chromatin containing methylated DNA and chromatin containing non-methylated DNA have identical affinities for H1. We conclude that H1 is unlikely to be a primary mediator of the biological effects of DNA methylation.

EXPERIMENTAL PROCEDURES

Materials and DNA Probes

The restriction enzymes MspI, HaeIII, SacI, EcoRI, and SspI were from Boehringer Mannheim, and TaqI was from New England Biolabs. The methyltransferases SssI and HhaI, also from New England Biolabs, were used according to the manufacturer's instructions. Other enzymes employed were collagenase and creatine phosphokinase (from Sigma), micrococcal nuclease (Worthington), and protein kinase C (Promega).

Oligonucleotide DNA probes used in this study are shown in Table 1. The vitellogenin gene probes were synthesized according to the sequence used by Jost and Hofsteenge (25) corresponding to nucleotide positions -2 to +34 from the avian vitellogenin gene and contained one CpG. Poly(GAC) and poly(GAM) are ligated polymers of a synthetic 42-bp sequence containing 12 CGA repeats, which in poly(GAM) have the cytosine substituted for m^5C(28) . For the methylated vitellogenin and poly(GAM) oligonucleotides, the m^5C in the CpG sequences was introduced during synthesis. The Sat probe corresponds to a repeat unit of mouse satellite DNA cloned as a 234-bp insert of the pSat plasmids described elsewhere(28) . The pAdomal plasmid (29) contains a fragment of adenovirus 2 major late promoter cloned into pAT153. The plasmid contains 421 CpGs in 5785 bp total length or one CpG every 14 bp on average. The Sat probe and the pAdomal plasmid were methylated using SssI methyltransferase (CpG methylase) as described in Boyes and Bird (4) . The 135-bp CG11 probe was methylated in its 20 GCGC sites with HhaI methylase, or at 27 sites using CpG methylase(8) . For the interaction of H1 with chromatin, the plasmids pUC18 (2.7 kb) and pHsr11.9 (14.6 kb) were used. The plasmid pHsr11.9 consists of an EcoRI fragment from human rDNA cloned into pUC9 and was a gift from R. Anand.

Chick histone H1 was a generous gift from Jean Thomas (Cambridge). This histone H1 had been extracted from chicken erythrocyte nuclei by incubation with 0.65 M NaCl and purified by ion-exchange chromatography as described(30) . To check its purity, H1 was analyzed by discontinuous SDS-polyacrylamide gel electrophoresis. The H1 preparation was free from histone H5 and core histones. The rat histone H1 employed in the chromatin experiments was extracted from rat kidney nuclei that had been depleted of nonhistone proteins by low salt extraction (0.25 M NaCl), and subsequently extracted with 2 M NaCl, 5 M urea, 10 mM phosphate buffer, pH 7. This extract was fractionated on a Mono-S column and yielded pure H1, as evidenced by SDS-polyacrylamide gel electrophoresis.

Bandshift Assays

In all band shift assays, equal amounts of DNA were used for the methylated and non-methylated probes. In some cases there were slight differences in the specific activities of the probes. For the band shift assays with the satellite probes, increasing amounts of histone H1 (0-100 ng) were mixed with methylated and unmethylated satellite DNA probes (about 0.2 ng of each) in 20 mM HEPES buffer, pH 7.9, containing 50 mM NaCl, 3 mM MgCl 1 mM EDTA, 0.1% Triton X-100, 10 mM beta-mercaptoethanol, 4% glycerol, 0.7 µg/ml Escherichia coli DNA as competitor, bromphenol blue, and xylene cyanol. After 100 min on ice, samples were run in a 1.5% agarose gel with 0.5 times TBE buffer. Finally, the gel was dried onto DE81 paper and exposed. In the gel retardation assays with CG11, the HhaI-methylated or mock methylated probes (0.4 ng) were mixed with 0-80 ng of sonicated E. coli DNA as competitor and with a fixed amount (50 ng) of chick H1, in the HEPES buffer (see above). After incubation on ice, samples were resolved in 5% polyacrylamide gels in 0.5 times TBE buffer (TBE, 0.9 M Tris borate, 0.02 M EDTA, pH 8.0), and the gels were dried and exposed.

For the gel mobility assays using the vitellogenin promoter, increasing amounts of chick H1 (5-50 ng) were mixed with 1-3 ng of end-labeled probes, methylated or unmethylated, together with 100 ng of E. coli competitor DNA, and incubated for 30 min on ice. The incubation medium consisted of 50 mM HEPES, pH 7.5, 50-100 mM KCl, 1 mM EDTA, 1 mM dithiothreitol, and 10% glycerol. The probe and conditions corresponded to those used for MBDP-2 by Jost and Hofsteenge(25) . The samples were loaded onto 5% polyacrylamide gels prepared in 0.5 times TBE and electrophoresed at 120 V and 4 °C in 0.5 times TBE for 3-4 h. Finally, the gels were dried and exposed. In other cases, a fixed amount of H1 was incubated with the probes, together with various amounts of competitor DNA. In these assays, 25 ng of chick H1 were incubated with the labeled vitellogenin probes and 0-100 ng of competitor DNA, in 10 mM Tris-HCl buffer, pH 7.5, containing 1 mM EDTA, 10 mM beta-mercaptoethanol, 4% glycerol, and 0.1% Triton X-100 for 30 min on ice. Samples were run in 5% polyacrylamide gels as described above.

Filter Binding Assays

To check if the affinity of H1 for DNA is influenced by DNA methylation, we performed competition assays. A fixed amount (20 ng) of end-labeled linear pAdomal was mixed with various amounts (0-60 ng) of cold competitor plasmid, methylated or non-methylated. Histone H1 (10 ng) was added and DNAbullethistone complexes allowed to form in 10 mM Tris, pH 8, 35 mM NaCl, 7.5% glycerol, 0.2 mM DTT, 0.5 mg/ml BSA at room temperature for at least 2 h. After filtering the samples through nitrocellulose membranes (BA83, 0.2 µm, Schleicher and Schuell), the filters were washed with 1 ml of incubation buffer without BSA, dried, and scintillation counted. DNA complexed with protein was retained by the membranes. Samples lacking H1 were used to quantify free DNA retained nonspecifically by the filters. This nonspecific binding was less than 10% of input counts/minute and was subtracted from the sample values. In the absence of competitor, more than 50% of template DNA was retained in the filters.

Southwestern Blots

For slot-Southwestern blots with poly(GAC) and poly(GAM) probes, different amounts of chick histone H1 diluted in TBS were filtered through a nitrocellulose membrane (0.2 µm pore), using a Bio-Rad manifold. The amounts of H1 applied were 125, 250, 500, and 1000 ng/slot, in a total volume of 200 µl. The protein in the membrane was denatured by incubating 5 min in 6 M guanidine-HCl in binding buffer (20 mM HEPES, 40 mM KCl, 3 mM MgCl(2), 10 mM beta-mercaptoethanol, pH 7.9) and renatured by incubation in five successive 2-fold dilutions of guanidine-HCl in the same buffer for 5 min each. All the incubations were done at room temperature. The membrane was rinsed in binding buffer, blocked with 4% nonfat instant dried milk in binding buffer for at least 1 h, and rinsed several times in the buffer, before being cut in strips. Before adding the labeled probes, the strips were preincubated 20 min in binding buffer supplemented with 0.1% Triton X-100 and 2 µg/ml single-stranded E. coli DNA, plus different concentrations of sonicated double-stranded E. coli DNA (0, 10, or 20 µg/ml) as competitor. The labeled probes poly(GAM) or poly(GAC) were then added and the strips incubated 1 h. The membranes were washed three times for 5 min in binding buffer containing 0.01% Triton X-100, dried, and exposed to x-ray films.

For the Southwestern assays with the vitellogenin probes, up to 1 µg of chick H1/lane was applied to discontinuous polyacrylamide gels. As a control, a fusion protein containing glutathione transferase plus the first 392 amino acids of the methyl-CpG-binding protein MeCP2 was also loaded (see (31) ). After electrophoresis, gels were soaked 30 min in transfer buffer (25 mM Tris, 190 mM glycine) to wash off the SDS. The histone was transferred to nitrocellulose membranes using a Bio-Rad transfer device. The proteins in the filters were denatured in 6 M guanidine hydrochloride in binding buffer (20 mM HEPES, 40 mM KCl, 3 mM MgCl for 5 min, and renatured by rinsing the filters in five successive 2-fold dilutions of guanidine-HCl in binding buffer, 5 min each rinse. The filters were blocked for 30 min with 4% instant non-fat dried milk in binding buffer and washed in the buffer. The filters were incubated 10 min in binding buffer with 0.1% Triton X-100 and 2.5 µg/ml of competitor DNA. Then, about 500,000 counts/min of the end-labeled vitellogenin probes were added and the filters incubated 1 h at room temperature with gentle shaking. After washing the filters with 0.01% Triton X-100 in binding buffer, four times for 5 min, these were dried and exposed. A Molecular Dynamics PhosphorImager was employed to quantify the signals.

Endonuclease Protection Assays

In these assays, control and methylated DNA were incubated with increasing amounts of H1, digested with various endonucleases, and the restriction fragments analyzed in agarose gels. The circular plasmid pAdomal was left untreated or methylated with SssI as described above. Methylated and unmethylated plasmids were made linear with SacI and then treated with proteinase K and purified by phenol-chloroform extraction and ethanol precipitation. Linear plasmid (100 ng), either methylated or unmethylated, was mixed with increasing amounts of chick H1 to give H1/DNA ratios of 0, 0.50, 0.75, 1, 2, and 4 µg of H1/µg of DNA in 200 mM NaCl, 10 mM Tris-HCl buffer, pH 7.5, and incubated on ice for 150 min to allow the formation of complexes. Alternatively, DNA and histone were mixed in buffer containing 600 mM NaCl and the mixture taken to 200 mM NaCl by stepwise dilution; the results were identical and are not shown. Digestion was started by adding two volumes of a buffer containing the restriction enzyme (MspI, HaeIII, or TaqI). The concentration of endonuclease was 100 units/µg of DNA for MspI and HaeIII, but 1000 units/µg for TaqI, since this enzyme shows optimal activity at 65 °C, and at 37 °C retains only about 10% of its activity. The reaction was carried out in 100 mM NaCl, 10 mM MgCl(2), 1 mM dithiothreitol, 10 mM Tris-HCl buffer, pH 7.5, at 37 °C for 75 min, and then stopped by adding EDTA to 5 mM. Samples were resolved in 1.5% agarose gels containing ethidium bromide, together with the undigested linear plasmids, and DNA size markers (123-bp DNA ladder). DNA was transferred onto a charged nylon membrane (Hybond N). Membranes were hybridized overnight at 68 °C in 0.5 M phosphate buffer, 7% SDS, 1 mM EDTA, pH 7.2, using the linear unmethylated pAdomal labeled by the random priming method as probe. Finally, membranes were washed at 68 °C, 2 times 10 min with 2 times SSC, 0.1% SDS, and 2 times 10 min with 0.1 times SSC, 0.1% SDS before exposure to x-ray film.

MspI Activity on Methylated and Unmethylated Free DNA

To compare the rates of cleavage of methylated and unmethylated DNA by the endonuclease MspI, the methylated and unmodified pAdomal plasmids, linearized with EcoRI, were treated with the enzyme in the absence of histone. Thus 0.5 µg of plasmid were digested with 0, 0.2, 1, 5, 10, and 100 units of MspI/µg of DNA, in 90 µl of 100 mM NaCl, 10 mM MgCl(2), 1 mM dithiothreitol, 10 mM Tris-HCl buffer, pH 7.5, at 37 °C for 75 min. The reactions were stopped by addition of EDTA to 5 mM and SDS to 0.5% and fragments analyzed as above.

H1 Footprinting of MspI Sites

Methylated and unmethylated pAdomal plasmids were linearized with EcoRI and end-labeled using [P]dATP and the Klenow fragment of DNA polymerase. The plasmids were cut with SspI to obtain the 5.6-kb linear fragments with only one labeled end, along with a 191-bp fragment. In the footprinting assays, 100 ng of methylated or unmethylated plasmid were mixed with histone H1 (0-4 µg/µg DNA) in 600 mM NaCl, 10 mM Tris-HCl buffer, pH 7.5, and incubated on ice. All samples contained the same counts/minute of end-labeled plasmid. To allow the formation of complexes, the samples were taken to 200 mM NaCl in a stepwise manner, by adding buffer without salt each 15 min, to decrease the concentration of NaCl by 50 mM. The samples were then treated with MspI (500 units/µg), and the restriction fragments were purified and resolved in 1.5% agarose gels.

Interaction of H1 with Minichromosomes

The two chosen plasmids were pUC18 (2.7 kb) and pHsr11.9 (14.6 kb). The latter plasmid consists of an 11.9-kb EcoRI fragment from human rDNA inserted into pUC9. Both plasmids were methylated with SssI methylase (or mock treated in the same conditions without enzyme). For the experiments involving assembly of the DNA into chromatin, we employed labeled rat histone H1. The histone was radiolabeled with protein kinase C and [-P]ATP following the method indicated by the manufacturers (Promega).

The DNA was assembled into chromatin by the method described by Rodriguez-Campos et al.(32) which uses a high speed supernatant from Xenopus oocytes as source of the histones and factors required. Xenopus oocytes are deficient in histone H1; therefore, minichromosomes packaged with this supernatant alone contain core histones but no histone H1(33) . The chromatin assembly reactions were carried out in 20 mM HEPES, pH 7.0, 5 mM KCl, 1 mM MgCl(2), 1 mM EGTA, 10 mM beta-glycerophosphate, 10% glycerol, 0.5 mM DTT (extraction buffer), supplemented with 3 mM ATP, 40 mM creatine phosphate, 2 µg/ml creatine phosphokinase, 60% oocyte extract, the desired amount of labeled H1, and DNA. The DNA consisted of equal amounts (0.75 µg/ml) of each of the two plasmids, pUC18 and pHsr11.9, only one of which was methylated in a given sample. In any experiment, two different concentrations of labeled H1 (0.2-1 µg H1/µg DNA) were used for each pair of plasmids. A mixture with all the components except the DNA was prepared and added to the DNAs to start the reactions, which proceed overnight at 27 °C.

After chromatin assembly, part of each sample was loaded onto sucrose gradients, and part was digested with microccocal nuclease. Minichromosomes assembled on each plasmid were separated by loading 500 µl of chromatin samples onto 15-30% sucrose gradients made up in 20 mM HEPES, pH 7, 100 mM NaCl, 0.2 mM EGTA, followed by centrifugation in a SW40 rotor, at 40,000 revolutions/min and 4 °C, for 2 h 45 min. Fractions of about 730 µl were collected from the bottom of the tubes, and their radioactivity determined by Cherenkov counting. DNA and proteins were purified from these fractions. For analysis of the DNA, aliquots of gradient fractions were digested with 0.2 mg/ml proteinase K in 0.5% SDS for several hours and ethanol precipitated by standard methods. The DNA was resolved in 1% agarose gels and transferred to nylon membranes by alkaline transfer. For protein analysis the fractions corresponding to each minichromosome were pooled and the proteins precipitated with 20% trichloroacetic acid in the presence of 0.05% Triton X-100 and 10 µg/ml protamine sulfate as carrier. After incubation 20 min on ice, samples were spun in a microfuge. The pellets were washed with 20% trichloroacetic acid and HCl-acetone, dried, and dissolved in loading mixture for their resolution in Triton-acid-urea gels (TAU gels). This loading mixture included 1 M acetic acid, 8 M urea, 0.1% Triton X-100, glycerol, and methyl green. TAU gels contained 1 M acetic acid, 8 M urea, and 0.5% Triton X-100, with 10.5% acrylamide in the separating gels. The running buffer was 0.1 M glycine, 1 M acetic acid. After electrophoresis, the gels were Coomassie stained, dried, and exposed.

To check that the DNA was organized into chromatin, after assembly part of each sample was digested with 100 units/ml of Micrococcus nuclease in the presence of 3 mM CaCl(2), at room temperature. Aliquots were taken at 2, 10, and 30 min, and the digestion stopped with 15 mM EGTA, 0.5% SDS on ice. These samples, plus an undigested control and an equivalent aliquot taken before assembly, were treated with 0.1 mg/ml RNase A, at 37 °C for at least 2 h, and with 0.5 mg/ml proteinase K in 0.5% SDS for several hours. The DNA was precipitated with ethanol and resolved in 1.5% agarose gels and transferred to nylon membranes for Southern analysis. The membranes were hybridized with two different probes. The EcoRI insert of pHsr11.9 was a probe specific for the pHsr plasmid. With linear pUC18 as a probe, the signal was much higher for the pUC18 plasmid, but pHsr was also detected due to homology of its vector portion with the probe.

RESULTS

Affinity of Histone H1 for Methylated and Unmethylated DNA

In our initial experiments, we used histone H1 extracted from chick erythrocyte nuclei with high salt and purified by ion-exchange chromatography (a gift from J. Thomas). Analysis by denaturing polyacrylamide gel electrophoresis showed only two bands, which correspond to the different H1 variants, and confirmed that the preparation contained neither histone H5 nor core histones (see (30) ). We first carried out bandshift assays to compare the affinity of H1 for a 269-bp mouse satellite probe with eight CpGs (see Table 1) that were either methylated or unmethylated. The labeled probes were incubated with increasing amounts of H1 from 0 to 100 ng, and the mobility of complexes was tested on agarose gels (Fig. 1A). Amounts of H1 above 30 ng gave histonebulletDNA complexes that migrated more slowly in the gel than the probe alone. The results show that under these conditions complexes are soluble at each H1/DNA ratio. Maximal retardation was obtained at 60 ng of H1, with further increases of histone concentration producing no higher effect. At each concentration, the migration of methylated and unmethylated probes was identical. Similar results were obtained with a probe that contained a higher density of methyl-CpGs (27 in 135 bp) and when H1 and DNA were mixed in 600 mM NaCl buffer and the complexes formed by lowering the salt concentration to 200 mM in a stepwise manner (data not shown). In other gel retardation experiments, the amounts of probe and histone H1 were kept constant, while various levels of E. coli DNA were included as competitor (Fig. 1B). The probe used was CG11, a 135-bp oligonucleotide containing 20 HhaI-methylatable sites. With low amounts of competitor DNA (up to 30 ng), histone H1 caused a retardation of both non-methylated and HhaI-methylated probes, this effect decreasing with higher amounts of competitor. The migration of both probes matched, indicating that the affinity of H1 for methylated and non-methylated CG11 was the same. The migration of the two probes was also indistinguishable when the assays were performed in buffers with very low ionic strength (1 mM sodium phosphate, pH 7.4, 0.2 mM EDTA, 4% glycerol, with or without 30 mM NaCl; gels not shown). Hence these band shift assays failed to show any effect of CpG methylation on histone H1 binding.

Figure 1: Gel retardation assays of complexes between histone H1 and methylated or unmethylated DNA sequences. A, assay with satellite DNA probes and different amounts of histone. Various amounts of chick H1, as indicated in nanograms above the lanes, were incubated with labeled satellite DNA probes, either unmethylated (NM lanes), or extensively methylated in its eight methylatable CpG groups (M lanes). The samples were resolved in a 1.5% agarose gel with 0.5 times TBE buffer and the gel dried and exposed. The two first and last lanes contained the free probes without histone. B, band shift of H1 and CG11 probes in the presence of increasing amounts of competitor DNA. In this case a fixed amount of H1 (50 ng) was incubated with the 135-bp probe CG11, either unmethylated (NM lanes) or HhaI-methylated (M lanes) in the presence of various amounts of E. coli DNA as indicated above the lanes. Samples were electrophoresed in 5% polyacrylamide gels with 0.5 times TBE buffer. F refers to free probes, without histone.


We next compared the affinity of H1 for methylated and non-methylated DNA by competition filter binding assays. A fixed amount of H1 (10 ng) was incubated with P-labeled non-methylated plasmid DNA (20 ng) in the presence of increasing amounts of unlabeled competitor DNA. Competitor was either non-methylated plasmid DNA or plasmid that had been methylated at all CpGs. The mixtures were passed through nitrocellulose filters to retain proteinbulletDNA complexes, and the amount of labeled DNA on the filters was measured in a scintillation counter. Each competitor decreased the amount of labeled complex to a similar extent, though the reduction was marginally greater with methylated DNA than with non-methylated DNA (Fig. 2). The results of this and other similar experiments indicated that methylated DNA was about 1.2-times (20%) more effective than non-methylated DNA as a competitor.

Figure 2: Filter binding assays comparing methylated and non-methylated DNAs as competitors for complex formation with H1. Complexes between H1 (10 ng) and non-methylated labeled pAdomal (20 ng) were formed in the presence of various concentrations (0-60 ng) of unlabeled methylated or non-methylated competitor plasmid. The samples were filtered through nitrocellulose to retain only the DNA bound to H1. Points show percentage of labeled DNA complexed with the histone, taking the ``no competitor'' level as 100%. Filled circles, non-methylated competitor; open circles, methylated competitor. Short bars indicate two filters for the same concentration.


A third assay for the affinity of histone H1 toward methylated and unmethylated DNA was based on slot-Southwesterns. Unlike the previous assays, where binding occurred in solution, Southwestern blots measured the affinity of single immobilized molecules of H1 for DNA. Various amounts of H1 protein were filtered onto a nitrocellulose membrane and incubated with P-labeled methylated or unmethylated DNA. The probes were poly(GAM) and poly(GAC), where M denotes m^5C (see Table 1). Poly(GAM) is a polymer of sequences containing 12 methyl-CpGs(28) . The results (Fig. 3) show that in the absence of competitor dsDNA, H1 bound to the same extent with both methylated (column 1) and unmethylated (column 4) probes, the signal decreasing with the amount of protein in a non-linear fashion. Non-linearity may indicate that the probe requires binding to several H1 molecules to be retained in the filter. Competitor E. coli DNA also bound to H1 and was able to displace most of the probe from the histone, but its effect was the same on poly(GAM) and poly(GAC) (compare columns 2 and 3 with 5 and 6). This indicates that the strength of the binding is equal irrespective of the methylation status of the probe.

Figure 3: Slot-Southwestern blot of histone H1 probed with poly(GAM) and poly(GAC). Increasing amounts of chick H1 (125, 250, 500, and 1000 ng/slot, from top to bottom) were filtered through a nitrocellulose membrane. After renaturation and blocking, strips were incubated for 1 h with labeled poly(GAM), which is methylated (columns 1-3), or poly(GAC), which is unmethylated (columns 4-6). Non-methylated E. coli DNA was included as competitor for strips 2 and 3 and 5 and 6. Strips 1 and 4 did not include double-stranded competitor DNA. After washing, filters were dried and exposed to an x-ray film.


Chicken Erythrocyte H1 Does Not Resemble MDBP-2

Pawlak et al.(26) have described MDBP-2, a protein from chick with a high affinity for methylated DNA sequences corresponding to the promoter of the vitellogenin gene, but a lower affinity for the same sequences when non-methylated. Later, MDBP-2 was reported to share sequences with histone H1 and to be recognized by anti-H1 antibodies (25) . The enriched protein was able to bind to methylated CpG in a variety of sequence contexts. Strikingly, it was reported that total liver H1 extracted with 0.2 M H(2)SO(4) also showed preferential binding to methylated DNA. We studied the effect of methylation on binding of our purified H1 preparation to the vitellogenin promoter (see Table 1). Using 50 ng of H1 and 100 ng of competitor DNA in conditions analogous to those employed by Jost and Hofsteenge(25) , we were unable to obtain complexes with the labeled promoter sequence as detected by bandshift assays (data not shown). By reducing the amount of competitor (using 25 ng of H1), complexes became apparent (Fig. 4A). There were no detectable differences in mobility between methylated or non-methylated complexes. When the levels of competitor were increased to 100 ng, no complexes were observed with either probe (compare lanes 9-12 in Fig. 4A).

Figure 4: Gel mobility and Southwestern assays for effect of methylation on binding of H1 to the vitellogenin promoter. A, band shift assay with different proportions of competitor DNA. 25 ng of chick H1 was incubated with the vitellogenin promoter probe (see ``Experimental Procedures''), along with 0-100 ng of competitor DNA. Samples were resolved in 5% polyacrylamide gels. M, methylated probe; NM, unmethylated probe. Amounts of competitor E. coli DNA are shown above the lanes. Lanes 11 and 12 show the free probes without histone, ran in the same gel. B, Southwestern assay. Aliquots of chick H1 were run in polyacrylamide gels, transferred onto nitrocellulose membranes, and incubated with end-labeled vitellogenin promoter probe in the presence of 2.5 µg/ml of competitor DNA. Lanes 1-3, filter hybridized with methylated probe; lanes 4-6, unmethylated probe. Lanes 2 and 5, 300 ng of H1; lanes 3 and 6, 1 µg of H1. Control lanes 1 and 4 correspond to a fusion protein containing the first 392 amino acids of MeCP2 which binds to DNA methylated at CpG(31) .


The behavior of H1 with the vitellogenin probes was also studied in Southwestern assays (Fig. 4B). Different amounts of chick H1 were run in polyacrylamide gels and transferred onto nitrocellulose filters. As a control, recombinant MeCP2, a protein which shows a strong preference for binding to methylated DNA, was also included (31) . Equal counts/minute of end-labeled vitellogenin probes in either the methylated (left) or unmethylated (right) state were incubated with identical filters in the presence of competitor E. coli DNA. The two probes gave the same signal with H1 (compare lanes 3 and 6, for 1 µg of H1), indicating that the affinity of H1 for methylated and unmethylated DNA is the same. MeCP2, on the other hand, showed a much higher binding to the methylated oligonucleotide (see lanes 1 and 4).

Protection by Histone H1 against Endonuclease Digestion

The restriction endonuclease MspI recognizes the sequence CCGG and is able to cut even if the internal cytosine is methylated(34) . Higurashi and Cole (23) reported that DNA methylated at that internal cytosine was resistant to MspI digestion at a subset of cleavage sites when in complexes with histone H1. We carried out similar endonuclease protection assays using the plasmid pAdomal which contains 421 CpGs in a total length of 5785 bp (one CpG every 14 bp on average). The plasmid was methylated with SssI methyltransferase (CpG methylase), which methylates all CpGs, and thereby mimics the mammalian methyltransferase(35) . Methylated and unmethylated plasmids were then linearized with SacI and mixed with varying amounts of chicken H1 under conditions suitable for formation of H1bulletDNA complexes (see ``Experimental Procedures''). Complexes were challenged with MspI, and restriction fragments were resolved in 1.5% agarose gels (Fig. 5). Without histone, the methylated (lane 2) and unmethylated (lane 10) DNAs were both digested to small fragments. The protective effect of H1 was first seen on the methylated plasmid at a ratio of 0.75-1 µg H1/µg DNA. The methylated DNA was only partially digested showing several large fragments, whereas unmethylated DNA was almost completely digested (compare lanes 4 and 5 with 12 and 13). At 2 µg H1/µg DNA, protection of methylated DNA was almost complete (lane 6), but unmodified DNA still showed some degree of digestion (lane 14). Thus histone H1 protects methylated DNA better than unmethylated DNA against cleavage by MspI, in agreement with Higurashi and Cole(23) .

Figure 5: Digestion of histone H1bulletDNA complexes with restriction endonucleases. Methylated (left) or unmethylated (right) pAdomal plasmids, linearized with SacI, were mixed with chick histone H1 at different H1:DNA ratios (0-4 µg/µg DNA), as indicated on the top of the lanes. H1bulletDNA complexes were formed for 150 min on ice and then digested with MspI (100 units/µg DNA, upper panel), HaeIII (100 units/µg, center panel), or TaqI (1000 nits/µg, lower panel). The purified restriction fragments were resolved in 1.5% agarose gels, transferred onto a nylon membrane, and hybridized using labeled linear plasmid as probe. The films also show the linear plasmids untreated with restriction enzymes (uncut, lanes 1 and 9) and a 123-bp DNA ladder as size markers (lanes 8).


To test whether methylated DNA complexes were protected against all restriction enzymes, we challenged them with HaeIII, which cuts the sequence GGCC. The pattern of restriction with HaeIII was the same for methylated and unmethylated plasmids (Fig. 5, central panel) indicating that methylated DNA complexes were not generally protected against nucleases. The site for HaeIII does not contain the methylated sequence CpG. We therefore tested the sensitivity of the complexes to TaqI, which cuts the sequence TCGA and is not blocked by CpG methylation(36) . Surprisingly, the restriction patterns for methylated and unmethylated pAdomal complexes were once more identical (Fig. 5, lower panel). At a ratio of 0.5 µg H1/µg DNA cleavage was complete; at 0.75-1 µg H1/µg DNA cleavage was partial; and at 2 µg H1/µg DNA cleavage of both methylated and unmethylated DNAs was blocked. Thus CpG methylation appears to be irrelevant to the TaqI sensitivity of H1bulletDNA complexes, implying that H1 does not dramatically alter the conformation of methylated CpGs.

The above results suggest that the protection of MspI sites is not due to generalized inaccessibility of methylated CpGs in the complexes, but to magnification of a pre-existing sensitivity of MspI to the presence of m^5CpG within its recognition site. According to this view, the preferential protection of methylated MspI sites in the complexes is due to MspI itself and does not reflect conformational inaccessibility of methylated DNA. In support of this, we found that MspI digests naked DNA more slowly when it is methylated (Fig. 6), in agreement with Butkus et al.(37) . For instance, an MspI concentration of 10 units/µg of DNA was enough to digest completely the unmethylated plasmid, while the methylated DNA was only partially cleaved at this enzyme concentration and required 100 units/µg to be completely cut. Thus MspI, while able to cut methylated DNA, is inhibited about 2-fold by CpG methylation, and this may contribute to the inhibition of cleavage seen in complexes between methylated DNA and H1.

Figure 6: MspI activity on methylated and unmethylated naked DNA. Methylated and unmethylated pAdomal plasmids, linearized with EcoRI, were cleaved with increasing amounts of MspI in the absence of histone, in 100 mM NaCl, 10 mM MgCl(2), 1 mM dithiothreitol, 10 mM Tris-HCl buffer, pH 7.5, at 37 °C for 75 min. The restriction fragments were purified and analyzed on a 1.5% agarose gel, transferred onto a membrane, and hybridized with labeled plasmid. Lanes 2-7 correspond to the methylated plasmid, and lanes 9-14 to unmethylated DNA. The concentration of MspI increases from left to right as shown above the lanes (in units/µg DNA). Lanes 1, 8, and 15 show DNA size markers (123-bp ladder).


Higurashi and Cole (23) reported that H1 protects a subset of methylated MspI sites from cleavage. Our nuclease protection data did not show evidence for selective protection, as all sites were resistant to MspI at high H1 concentrations. To investigate this question more fully, we carried out H1 footprinting assays of methylated and unmethylated DNA. The pAdomal plasmids were labeled at one terminus, complexed with increasing amounts of H1, and treated with a constant amount of MspI. Increasing the amount of H1 conferred increasing resistance to MspI (Fig. 7), and this effect was stronger for methylated DNA as expected from the data in Fig. 6. However, when the bands resulting from partial digestion in the methylated and unmethylated lanes were compared, they were found to be the same (Fig. 7). Thus no site seemed to be specifically protected in the complexes with H1.

Figure 7: H1 footprinting of MspI sites. Methylated and unmethylated pAdomal plasmids were made linear with EcoRI and end-labeled. The plasmids were mixed with histone H1 in various ratios (as indicated in µg H1/µg DNA above lanes), and the complexes digested with MspI. The fragments were resolved in 1.5% agarose gels, and the gels were dried and exposed. Lanes 1-7, methylated plasmid; lanes 9-15, unmethylated DNA; lane 8, DNA size markers (123-bp ladder). Lanes 1 and 9 show the end-labeled probes, untreated with MspI. Each band corresponds to one MspI site.


Affinity of H1 for Chromatin Containing Methylated DNA

Histone H1 is found naturally in nucleosomal chromatin, and therefore assays of its interaction with naked DNA may be inadequate. To address the effect of methylation on H1-DNA interaction in a situation more similar to that in vivo, we assembled plasmids into chromatin in a Xenopus oocyte extract in the presence of added rat histone H1(32) . To control the experiment internally, two plasmids of greatly differing sizes were incubated together with the extract; one plasmid in the methylated form at all CpGs and the other non-methylated. The experiment was done both ways round, with either the large plasmid or the small one being methylated. After assembly, the different size of the plasmids allowed the separation of the two sorts of minichromosomes by centrifugation in sucrose gradients. The use of kinase-labeled H1 made it possible to follow the added H1 and compare its incorporation into the minichromosomes.

To test if both plasmids were assembled into chromatin, the samples incubated with the extract were treated with nuclease from Micrococcus, and the resultant DNA fragments resolved in agarose gels and visualized by Southern analysis (Fig. 8A). Two different probes were used for this analysis: the insert of pHsr, as a probe specific for this plasmid (Fig. 8A), and linear pUC18, which gives a much stronger signal for pUC18 than for pHsr (data not shown). The results showed that both plasmids were efficiently assembled into chromatin whether methylated or not, as limited digestion with the nuclease gave an extensive ladder of fragments with a spacing of about 160 base pairs. The nuclease fragments gave the same pattern with both probes; no differences in degree of digestion or in spacing between the two plasmids were observed. This result demonstrated that both plasmids were properly organized into long stretches of regularly spaced nucleosomes. Thus, any cooperative binding of H1 to adjacent nucleosomes (perhaps enhanced by methylation) should occur in this system.

Figure 8: The incorporation of histone H1 into polynucleosomal chromatin is not affected by DNA methylation. Samples containing equal amounts of the plasmids pHsr11.9 (14.6 kb) and pUC18 (2.7 kb) were organized into chromatin using a Xenopus oocyte extract. One of the plasmids in each sample had been methylated in the CpGs with SssI methylase, the other was unmodified. The desired amount of radiolabeled rat histone H1 (0.2-1 µg H1/µg DNA) was mixed with the oocyte extract and supplemented with an ATP regenerating system, before its addition to the DNA. The assembly reaction proceed overnight at 27 °C. Part of each sample was then digested with Micrococcus nuclease, and part was loaded onto sucrose gradients. A, Micrococcus nuclease pattern of assembled chromatin. Chromatin samples were digested with 100 units/ml of Micrococcus nucleus for 0-30 min as indicated, and the reactions stopped with 15 mM EGTA, 0.5% SLS. The purified DNA was resolved in 1.5% agarose gels and transferred to membranes for Southern blot analysis. The representative patterns shown were obtained using the EcoRI insert of pHsr as probe. B, indicates samples before chromatin assembly, and M are DNA size markers. B, separation of the minichromosomes in sucrose gradients. The minichromosomes containing the pHsr11.9 or pUC18 plasmids were separated by centrifugation through sucrose gradients. DNA was purified from aliquots of the gradient fractions, resolved in 1% agarose gels and detected by Southern analysis. The film corresponds to one of the samples (closed circles in C) probed with pUC, which hybridizes with both plasmids. Using the EcoRI insert of pHsr as probe, only the longer plasmid gave signal (not shown). The amount of each plasmid was the same, although as the probe hybridizes only with the ``vector'' portion of pHsr (2.7 kb out of 14.6), the intensity of the signal is lower for pHsr. M stands for DNA size markers. C, distribution of H1 in the sucrose gradients. The content of labeled H1 in the gradient fractions was determined by Cherenkov counting. For each gradient, only one of the plasmids was methylated. Thus, either pHsr was methylated and pUC was non-methylated (open circles) or vice versa (closed circles). These profiles were obtained for 0.5 µg of added H1/µg DNA. D, as in C, but for a double amount of added H1 (1 µg H1/µg DNA).


Sucrose gradient centrifugation separated the fast sedimenting pHsr minichromosomes (14.6 kb) from the smaller pUC minichromosomes (2.7 kb), as shown by probing a blot of the DNA from the fractions resolved by agarose gel electrophoresis with plasmid-specific DNA sequences (Fig. 8B). Using pUC18 as probe, both plasmids were revealed since this probe also hybridizes with the vector part of pHsr although the signal from this plasmid is weaker. When the 11.9-kb insert of pHsr was utilized as probe, only the bigger plasmid was seen (data not shown). Most clearly for pUC, a faster migrating band of supercoiled (form I) DNA and a slower band of open circle (form II) are seen. We think that a certain amount of nicking occurs during the purification of the DNA from the fractions, particularly for the longer pHsr, giving rise to open circular forms.

The distribution of added histone H1 along the gradients was monitored by Cherenkov counting of the fractions. It is apparent from Fig. 8C that the profile of radioactivity did not depend on which plasmid was methylated. Two peaks of radioactivity were observed, corresponding to those fractions containing either pHsr or pUC minichromosomes (as seen by the DNA analysis of Fig. 8B). Analysis of the proteins in each peak by trichloroacetic acid precipitation and electrophoresis in TAU gels confirmed that all the radioactivity was due to histone H1 (data not shown). It is clear that the profiles for both combinations of plasmids can be superimposed. For instance, the amount of histone H1 incorporated into pHsr minichromosomes (left peak) was the same whether that plasmid is methylated (open circles) or non-methylated (closed circles). The same is true for the pUC plasmid (in this case the open circles indicating the non-methylated form). This demonstrated that the binding of the linker histone to chromatin is unaffected by methylation. Although the level of radioactivity was higher for the slower peak, this was seen for both samples and therefore cannot be attributed to methylation. This difference probably arose from some non-incorporated H1 trailing from the top of the gradient.

A possible trivial explanation for the absence of a methylation effect on H1 binding is that the concentrations of H1 were sufficiently high to saturate both minichromosomes. This seemed unlikely as, although the amount of added H1 was 0.5 µg/µg of DNA, which could give about 2.5 H1 molecules/nucleosome, most H1 did not get incorporated into chromatin and remained at the top of the gradients (Fig. 8C and data not shown). We suspect that H1 is sequestered by binding to a component of the extract, perhaps RNA ( (40) and data not shown). Thus the amount of incorporated H1 was a minority of that added, and the proportion of H1 in the minichromosomes was only of about 1 molecule each 8-10 nucleosomes. To rule out the possibility that both minichromosomes were saturated with H1, the amount of added H1 was doubled (Fig. 8D). This had the effect of doubling the amount of H1 incorporated into both minichromosomes, thereby confirming that H1 incorporation was not at saturating levels. Once again the affinity of H1 for methylated chromatin was the same as for unmethylated chromatin, as had been observed for naked DNA.

Another possible explanation for the results is that methylation is either lost or gained from the plasmids during incubation with the extract. Previous studies have shown that this does not happen in intact Xenopus oocytes, from which these extracts are derived(38, 39) . To check this, we isolated DNA from assembled minichromosomes and treated with the methylation-sensitive enzyme HpaII. The methylated plasmid was still resistant, and the unmethylated plasmid was still sensitive to this enzyme (data not shown). Thus no change in methylation had occurred during chromatin assembly.

DISCUSSION

Indifference of H1 to CpG Methylation

The results reported here argue against a central role for histone H1 in mediating the effects of DNA methylation on chromatin structure and gene expression. A combination of assays failed to reveal any clear preference of H1 for binding to methylated DNA rather than non-methylated DNA. The results with assembled chromatin were particularly persuasive. In these experiments methylated and non-methylated plasmids were mixed prior to assembly in a Xenopus oocyte extract in the presence of added H1. Although the two plasmids were in competition for H1, no differential affinity due to methylation was detectable. This was not due to the presence of saturating amounts of the linker histone, as loading of minichromosomes was shown to be proportional to the concentration of added H1. The results show that nucleosomal arrays containing densely methylated DNA are not preferred sites of H1 binding. Considering all the evidence together, it now seems highly unlikely that the altered chromatin structure that has been attributed to methylated DNA is due to its association with core or linker histones.

Previous studies have given conflicting impressions about the affinity of H1 for methylated DNA. Southwestern assays of mouse and rat nuclear proteins failed to detect differences in the affinities of methylated and non-methylated probes for H1(28) . Higurashi and Cole (23) found slightly enhanced binding of H1 to non-methylated DNA but considered the magnitude of this preference too small to be significant. Our filter binding assays showed preferential binding to methylated DNA, but once again the effect was very small (1.2-fold). On the other hand, Levine et al.(24) found that methylated DNA could successfully disrupt a pre-existing H1bulletDNA complex, whereas non-methylated DNA could not. The failure of excess non-methylated DNA to remove H1 is in fact more surprising than the success of methylated DNA. This result would not be expected of soluble H1bulletDNA complexes, but would be expected if the preformed complexes had precipitated, or if H1 was in excess over DNA. Since complexes were assayed only by filter binding, the presence of precipitates may have gone undetected in the study. In view of the known tendency of H1bulletDNA complexes to precipitate, and the disagreement with present data and data of Higurashi and Cole(23) , it is important to test this possibility.

The activity of MDBP-2, a protein that shows preferential binding to a methylated site in the promoter of the vitellogenin gene, has been attributed to the histone H1 fraction of chicken liver(25) . Our results show that purified histone H1 from chicken erythrocytes does not bind preferentially to the vitellogenin promoter when it is methylated. Thus, we conclude that erythrocytes lack the MDBP2-H1 form that was detected by others.

H1 and Repression of Transcription

The studies by Levine et al.(24) and Johnson et al.(27) using naked DNA templates also showed that transcription of methylated genes were repressed preferentially by H1. If there is no preference of H1 for methylated DNA, then how can this result be explained? A possible explanation is that one or more transcription factors that bind to the promoters under test are directly affected by methylation. This would weaken the promoter and cause it to be inhibited more easily by a nonspecific repressor of transcription such as H1. In fact the adenovirus major late promoter does contain a binding site for the transcription factor MLTF, which is known to be blocked by methylation of CpG in its binding site(9) . This factor will have been displaced in the fully methylated constructs that were used in the experiments of Levine et al.(24) , and the promoter must therefore have been compromised before the addition of H1. It follows that the preferential inhibition of transcription from the methylated template seen in their study could result from the weakness of the methylated promoter, rather than discriminatory binding by H1. It is not known whether the factors responsible for driving transcription of the tRNA gene in the study by Johnson et al.(27) are directly affected by methylation.

Methyl Sensitivity of MspI

MspI has been extensively used in studies of DNA methylation, as it can cleave the methylated sequence Cm^5CGG whereas its isoschizomer HpaII cannot(34) . Once complexed with H1, however, Cm^5CGG sites become partially resistant to MspI. We found no evidence for general inaccessibility of methylated CpG sites in H1 complexes, as TaqI, which also recognizes CpG, was able to cleave methylated and unmethylated complexes equally. It seems unlikely that H1 binds to m^5CpG at CCGG but not at TCGA. Our results suggest that MspI resistance is due to the exquisite sensitivity of this endonuclease to perturbations in the structure of methylated sites. This may be because the detailed mechanism of cleavage is different on methylated versus unmethylated DNA. The finding that MspI is kinetically inhibited by the presence of methyl-CpG in naked DNA (our data and (37) ) is compatible with this idea. For example, the cleavage of methylated DNA may require a higher distortion of the helix, and this distortion could be hampered by H1 binding. In this case the reluctance of MspI to cleave complexed methylated DNA could be explained without the need to invoke differential interactions between H1 and methylated versus unmethylated DNA. In conclusion, it seems that MspI is not giving reliable information about the accessibility/conformation of methylated sites in H1bulletDNA complexes, as apparently identical complexes are cleaved differently depending on the methylation status of the DNA. Since MspI has been used to infer binding of proteins or conformational changes at methylated CpG in intact nuclei(6, 41, 42, 43) , it may be necessary to interpret these earlier studies with caution in the light of results presented here.

FOOTNOTES

*
This work was supported by the Imperial Cancer Research Fund, The Wellcome Trust, the Howard Hughes Medical Institute, and the Human Capital and Mobility Programme of the European Community. The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore by hereby marked ``advertisement'' in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

§
To whom correspondence should be addressed. Tel.: 44-131-650-5670; Fax: 44-131-650-5379.

(^1)
The abbreviations used are: MeCP, methyl-CpG binding protein; bp, base pair(s); MDBP, methylated DNA-binding protein; DTT, dithiothreitol; BSA, bovine serum albumin; kb, kilobase(s).

ACKNOWLEDGEMENTS

We are grateful to Dr. Jean Thomas for the gift of chick histone H1 and for critical discussion of our results. We also thank Frank Johnson for photography.

REFERENCES

  1. Busslinger, M., Hurst, J., and Flavell, R. A. (1983) Cell 34, 197-206 [CrossRef][Medline] [Order article via Infotrieve]
  2. Keshet, I., Yisraeli, J., and Cedar, H. (1985) Proc. Natl. Acad. Sci. U. S. A. 82, 2560-2564 [Abstract/Free Full Text]
  3. Buschhausen, G., Wittig, B., Graessmann, M., and Graessmann, A. (1987) Proc. Natl. Acad. Sci. U. S. A. 84, 1177-1181 [Abstract/Free Full Text]
  4. Boyes, J., and Bird, A. (1991) Cell 64, 1123-1134 [CrossRef][Medline] [Order article via Infotrieve]
  5. Murray, E. J., and Grosveld, F. (1987) EMBO J. 6, 2329-2335 [Medline] [Order article via Infotrieve]
  6. Kass, S. U., Goddard, J. P., and Adams, R. L. P. (1993) Mol. Cell. Biol. 13, 7372-7379 [Abstract/Free Full Text]
  7. Boyes, J., and Bird, A. (1992) EMBO J. 11, 327-333 [Medline] [Order article via Infotrieve]
  8. Meehan, R. R., Lewis, J. D., McKay, S., Kleiner, E. L., and Bird, A. P. (1989) Cell 58, 499-507 [CrossRef][Medline] [Order article via Infotrieve]
  9. Watt, F., and Molloy, P. L. (1988) Genes & Dev. 2, 1136-1143
  10. Iguchi-Ariga, S. M. M., and Schaffner, W. (1989) Genes & Dev. 3, 612-619
  11. Buschhausen, G., Graessmann, M., and Graessmann, A. (1985) Nucleic Acids Res. 13, 5503-5513 [Abstract/Free Full Text]
  12. Keshet, I., Lieman-Hurwitz, J., and Cedar, H. (1986) Cell 44, 535-543 [CrossRef][Medline] [Order article via Infotrieve]
  13. Felsenfeld, G., Nickol, J., Behe, M., McGhee, J. D., and Jackson, D. (1982) Cold Spring Harbor Symp. Quant. Biol. 47, 577-584
  14. Drew, M. R., and McCall, M. J. (1987) J. Mol. Biol. 197, 485-511 [CrossRef][Medline] [Order article via Infotrieve]
  15. Thoma, F., Koller, T., and Klug, A. (1979) J. Cell Biol. 83, 403-427 [Abstract/Free Full Text]
  16. Kamakaka, R. T., and Thomas, J. O. (1990) EMBO J. 9, 3997-4006 [Medline] [Order article via Infotrieve]
  17. Tazi, J., and Bird, A. (1990) Cell 60, 909-920 [CrossRef][Medline] [Order article via Infotrieve]
  18. Weintraub, H. (1984) Cell 38, 17-27 [CrossRef][Medline] [Order article via Infotrieve]
  19. Wolffe, A. P. (1989) EMBO J. 8, 527-537 [Medline] [Order article via Infotrieve]
  20. Croston, G. E., Kerrigan, L. A., Lira, L. M., Marshak, D. R., and Kadonaga, J. T. (1991) Science 251, 643-649 [Abstract/Free Full Text]
  21. Laybourn, P. J., and Kadonaga, J. T. (1991) Science 254, 238-245 [Abstract/Free Full Text]
  22. Ball, D. J., Gross, D. S., and Garrard, W. T. (1983) Proc. Natl. Acad. Sci. U. S. A. 80, 5490-5494 [Abstract/Free Full Text]
  23. Higurashi, M., and Cole, R. D. (1991) J. Biol. Chem. 266, 8619-8625 [Abstract/Free Full Text]
  24. Levine, A., Yeivin, A., Ben-Asher, E., Aloni, Y., and Razin, A. (1993) J. Biol. Chem. 268, 21754-21759 [Abstract/Free Full Text]
  25. Jost, J.-P., and Hofsteenge, J. (1992) Proc. Natl. Acad. Sci. U. S. A. 89, 9499-9503 [Abstract/Free Full Text]
  26. Pawlak, A., Bryans, M., and Jost, J.-P. (1991) Nucleic Acids Res. 19, 1029-1034 [Abstract/Free Full Text]
  27. Johnson, C. A., Goddard, J. P., and Adams, R. L. P. (1995) Biochem. J. 305, 791-798
  28. Lewis, J. D., Meehan, R. R., Henzel, W. J., Maurer-Fogy, I., Jeppesen, P., Klein, F., and Bird, A. (1992) Cell 69, 905-914 [CrossRef][Medline] [Order article via Infotrieve]
  29. Heiermann, R., and Pongs, O. (1985) Nucleic Acids Res. 13, 2709-2730 [Abstract/Free Full Text]
  30. Clark, D. J., and Thomas, J. O. (1986) J. Mol. Biol. 187, 569-580 [CrossRef][Medline] [Order article via Infotrieve]
  31. Nan, X., Meehan, R. R., and Bird, A. (1993) Nucleic Acids Res. 21, 4886-4892 [Abstract/Free Full Text]
  32. Rodríguez-Campos, A., Shimamura, A., and Worcel, A. (1989) J. Mol. Biol. 209, 135-150 [CrossRef][Medline] [Order article via Infotrieve]
  33. Shimamura, A., Tremethick, D., and Worcel, A. (1988) Mol. Cell. Biol. 8, 4257-4269 [Abstract/Free Full Text]
  34. Waalwijk, C., and Flavell, R. A. (1978) Nucleic Acids Res. 5, 3231-3236 [Abstract/Free Full Text]
  35. Renbaum, P., Abrahamove, D., Fainsod, A., Wilson, G., Rottem, S., and Razin, A. (1990) Nucleic Acids Res. 18, 1145-1152 [Abstract/Free Full Text]
  36. McClelland, M., and Nelson, M. (1992) Nucleic Acids Res. 20, 2145-2157
  37. Butkus, V., Petrauskiene, L., Maneliene, Z., Klimasauskas, S., Laucys, V., and Janulaitis, S. (1987) Nucleic Acids Res. 15, 7091-7102 [Abstract/Free Full Text]
  38. Vardimon, L., Kressmann, A., Cedar, H., Maechler, M., and Doerfler, W. (1982) Proc. Natl. Acad. Sci. U. S. A. 79, 1073-1077 [Abstract/Free Full Text]
  39. Macleod, D., and Bird, A. P. (1983) Nature 306, 200-203 [CrossRef][Medline] [Order article via Infotrieve]
  40. Halmer, L., and Gruss, C. (1995) Nucleic Acids Res. 23, 773-778 [Abstract/Free Full Text]
  41. Antequera, F., MacLeod, D., and Bird, A. P. (1989) Cell 58, 509-517 [CrossRef][Medline] [Order article via Infotrieve]
  42. Levine, A., Cantoni, G. L., and Razin, A. (1991) Proc. Natl. Acad. Sci. U. S. A. 88, 6515-6518 [Abstract/Free Full Text]
  43. Hsieh, C.-L., and Lieber, M. R. (1992) EMBO J. 11, 315-325 [Medline] [Order article via Infotrieve]
©1995 by The American Society for Biochemistry and Molecular Biology, Inc.

Add to CiteULike CiteULike   Add to Complore Complore   Add to Connotea Connotea   Add to Del.icio.us Del.icio.us   Add to Digg Digg   Add to Reddit Reddit   Add to Technorati Technorati    What's this?


This article has been cited by other articles:


Home page
 J. Cell Biol.Home page
N. Gilbert, I. Thomson, S. Boyle, J. Allan, B. Ramsahoye, and W. A. Bickmore
DNA methylation affects nuclear organization, histone modifications, and linker histone binding but not chromatin compaction
J. Cell Biol., May 7, 2007; 177(3): 401 - 411.
[Abstract] [Full Text] [PDF]

Home page
 Eukaryot CellHome page
H. D. Folco, M. Freitag, A. Ramon, E. D. Temporini, M. E. Alvarez, I. Garcia, C. Scazzocchio, E. U. Selker, and A. L. Rosa
Histone H1 Is Required for Proper Regulation of Pyruvate Decarboxylase Gene Expression in Neurosporacrassa
Eukaryot. Cell, April 1, 2003; 2(2): 341 - 350.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
N. Steward, M. Ito, Y. Yamaguchi, N. Koizumi, and H. Sano
Periodic DNA Methylation in Maize Nucleosomes and Demethylation by Environmental Stress
J. Biol. Chem., September 27, 2002; 277(40): 37741 - 37746.
[Abstract] [Full Text] [PDF]

Home page
 FASEB J.Home page
J. ZLATANOVA, P. CAIAFA, and K. VAN HOLDE
Linker histone binding and displacement: versatile mechanism for transcriptional regulation
FASEB J, September 1, 2000; 14(12): 1697 - 1704.
[Abstract] [Full Text]

Home page
 J. Cell Sci.Home page
A Akhmanova, T Verkerk, A Langeveld, F Grosveld, and N Galjart
Characterisation of transcriptionally active and inactive chromatin domains in neurons
J. Cell Sci., January 12, 2000; 113(24): 4463 - 4474.
[Abstract] [PDF]

Home page
 Mol. Cell. Biol.Home page
J. L. Barra, L. Rhounim, J.-L. Rossignol, and G. Faugeron
Histone H1 Is Dispensable for Methylation-Associated Gene Silencing in Ascobolus immersus and Essential for Long Life Span
Mol. Cell. Biol., January 1, 2000; 20(1): 61 - 69.
[Abstract] [Full Text]

Home page
 J. Biol. Chem.Home page
R. Feil, M. D. Boyano, N. D. Allen, and G. Kelsey
Parental Chromosome-specific Chromatin Conformation in the Imprinted U2af1-rs1 Gene in the Mouse
J. Biol. Chem., August 15, 1997; 272(33): 20893 - 20900.
[Abstract] [Full Text] [PDF]

Home page
 J. Biol. Chem.Home page
J. S. Godde, S. U. Kass, M. C. Hirst, and A. P. Wolffe
Nucleosome Assembly on Methylated CGG Triplet Repeats in the Fragile X Mental Retardation Gene 1Promoter
J. Biol. Chem., October 4, 1996; 271(40): 24325 - 24328.
[Abstract] [Full Text] [PDF]