Histone H4-Lysine 20 Monomethylation Is Increased in Promoter and Coding Regions of Active Genes and Correlates with Hyperacetylation*

Methylation and acetylation of position-specific lysine residues in the N-terminal tail of histones H3 and H4 play an important role in regulating chromatin structure and function. In the case of H3-Lys4, H3-Lys9, H3-Lys27, and H4-Lys20, the degree of methyla-tion was variable from the mono- to the di- or trimethylated state, each of which was presumed to be involved in the organization of chromatin and the activation or repression of genes. Here we inves-tigated the interplay between histone H4-Lys20 mono- and trim-ethylation and H4 acetylation at induced (β-major/β-minor glo-bin), repressed (c-myc), and silent (embryonic β-globin) genes during in vitro differentiation of mouse erythroleukemia cells. By using chromatin immunoprecipitation, we found that the β-majorand β-minor promoter and the β-globin coding regions as well as the promoter and the transcribed exon 2 regions of the highly expressed c-myc gene were hyperacetylated and monomethylated at H4-Lys20. Although activation of the β-globin gene resulted in an increase in hyperacetylated, monomethylated H4, down-regulation of the c-myc gene did not cause a decrease in hyperacetylated, monomethylated H4-Lys20, thus showing a stable pattern of histone modifications. Immunofluorescence microscopy studies revealed that monomethylated H4-Lys20 mainly overlaps with RNA pol II-stained euchromatic regions, thus indicating an association with transcriptionally engaged chromatin. Our chromatin immunopre-cipitation results demonstrated that in contrast to trimethylated H4-Lys20, which was found to inversely correlate with H4 hyper-acetylation, H4-Lys20 monomethylation is compatible with histone H4 hyperacetylation and correlates with the transcriptionally active or competent chromatin state.

Methylation and acetylation of position-specific lysine residues in the N-terminal tail of histones H3 and H4 play an important role in regulating chromatin structure and function. In the case of H3-Lys 4 , H3-Lys 9 , H3-Lys 27 , and H4-Lys 20 , the degree of methylation was variable from the mono-to the di-or trimethylated state, each of which was presumed to be involved in the organization of chromatin and the activation or repression of genes. Here we investigated the interplay between histone H4-Lys 20 mono-and trimethylation and H4 acetylation at induced (␤-major/␤-minor globin), repressed (c-myc), and silent (embryonic ␤-globin) genes during in vitro differentiation of mouse erythroleukemia cells. By using chromatin immunoprecipitation, we found that the ␤-major and ␤-minor promoter and the ␤-globin coding regions as well as the promoter and the transcribed exon 2 regions of the highly expressed c-myc gene were hyperacetylated and monomethylated at H4-Lys 20 . Although activation of the ␤-globin gene resulted in an increase in hyperacetylated, monomethylated H4, down-regulation of the c-myc gene did not cause a decrease in hyperacetylated, monomethylated H4-Lys 20 , thus showing a stable pattern of histone modifications. Immunofluorescence microscopy studies revealed that monomethylated H4-Lys 20 mainly overlaps with RNA pol IIstained euchromatic regions, thus indicating an association with transcriptionally engaged chromatin. Our chromatin immunoprecipitation results demonstrated that in contrast to trimethylated H4-Lys 20 , which was found to inversely correlate with H4 hyperacetylation, H4-Lys 20 monomethylation is compatible with histone H4 hyperacetylation and correlates with the transcriptionally active or competent chromatin state.
It has been proposed that distinct post-translational histone modifications act sequentially or in combination to form a "histone code" within chromatin (1). Acetylation and methylation of specific histone lysine residues can serve as a mark of either euchromatin or silent heterochromatin. Although methylation of H3-Lys 4 , H3-Lys 36 , and H3-Lys 79 has been linked to transcriptional activation and protection of euchromatin, H3-Lys 9 , H3-Lys 27 , and H4-Lys 20 methylation is generally thought to be associated with gene repression and heterochromatin formation (2)(3)(4). In this regard it must be noted that histone lysine residues can be mono-, di-, or trimethylated (5), thus extending the coding potential of a methylatable lysine position. Previous studies, however, focused only on detection of H3 (for review see Refs. 3 and 4) or H4 (6 -8) lysine methylation regardless of the methylation status. Recently, it was shown that a distinction between di-and trimethylation of various lysines of histone H3 is important for processes of transcriptional regulation or gene silencing (9 -11). Moreover, studies that focused on the in vivo distribution of mono-, di-, and trimethylated H3-Lys 9 and H3-Lys 27 demonstrate that mono-and dimethylated H3-Lys 9 and H3-Lys 27 are specifically localized to silent domains within euchromatin, whereas trimethylated H3-Lys 9 and monomethylated H3-Lys 27 were enriched at pericentric heterochromatin (12,13). In contrast to findings suggesting a role for H4-Lys 20 methylation in regulating gene expression, a recently published study demonstrates that H4-Lys 20 methylation, in particular trimethylation, plays no apparent role in gene regulation or heterochromatin function but is involved in DNA damage response in fission yeast (14).
The first studies on methylation of H4-Lys 20 with antibodies to dimethylated H4-Lys 20 reported that dimethylated H4-Lys 20 acts in antagonizing H4-Lys 16 acetylation (7,8) and does not correlate with gene activity (6). Trimethylated H4-Lys 20 was found to be a marker of constitutive heterochromatin in murine interphase and metaphase cells (15) enriched at pericentric heterochromatin (16). Furthermore, it was shown that trimethylated H3-Lys 9 is required for induction of H4-Lys 20 trimethylation and that trimethylation of histone H3-Lys 9 and H4-Lys 20 functions as a repressive mark in gene-silencing mechanisms (5). ChIP 2 experiments demonstrated that the human histone deacetylase SirT1induced deacetylation of H4-Lys 16 is accompanied by H1 enrichment and the spreading of trimethylated H3-Lys 9 and monomethylated H4-Lys 20 at the promoter region of a repressed reporter system (17). Histone H3-Lys 27 trimethylation and H4-Lys 20 monomethylation were shown to be associated with Xist expression in ES cells and seem to mark the initiation of X inactivation. However, both H3-Lys 27 trimethylation and H4-Lys 20 monomethylation are maintained in the absence of transcriptional repression (18). Further investigations indicating a role of methylated H4-Lys 20 in gene activation were performed by Beisel et al. (19), who found that the Drosophila epigenetic activator ASH-1, a histone methyltransferase, activates transcription by dimethylation of H3-Lys 4 , H3-Lys 9 , and H4-Lys 20 at the promoter of target genes. Significant differences in subnuclear localization of the mono-and trimethyl versions of histone H4-Lys 20 were recently observed during mouse embryogenesis (20). H4 monomethyl Lys 20 was shown to be elevated in proliferating cells; in contrast, histone H4 trimethyl Lys 20 became enriched in differentiating cells during the mouse developmental process. Most recently, we investigated the changes in Lys 20 methylation states of various acetylated H4 histones as well as the interplay between the various H4-Lys 20 methylation states and acetylation during in vitro differentiation of mouse erythroleukemia cells (16). We found that trimethylated H4-Lys 20 histones increase during the differentiation process of mouse erythroleukemia (MEL) cells and that these hypermethylated H4 histones completely preclude histone H4 tri-and tetraacetylation.
With these observations in mind, we decided to investigate the methylation states of H4-Lys 20 in correlation with acetylation at induced (␤ major/minor globin) or repressed (c-myc) genes or at silent genes like embryonic ␤-globin in the MEL cell system. The ␤-globin locus, containing the erythroid-specific and developmentally regulated ␤-globin genes, is a particularly informative system for investigating the structure/function of histone modification patterns. The present study indicates that monomethylated H4-Lys 20 is not a principal feature of repressed gene regions. On the contrary, after induction of expression of the adult ␤-globin gene, we found increased monomethylated H4-Lys 20 paralleled by hyperacetylation of H4 at the ␤-major and ␤-minor promoter and ␤-globin transcribed region. Increased levels of monomethyl H4-Lys 20 and hyperacetylated H4 at the promoter (exon 1) and the transcribed exon 2 region were also found at the highly expressed c-myc gene. Most interestingly, monomethylation and hyperacetylation of the H4 histone also persist at low expression levels of c-myc, indicating that neither hypoacetylation nor decreased monomethylation of H4 is a prerequisite for c-myc gene down-regulation.
To confirm the findings obtained with ChIP, we investigated the H4 acetylation/methylation pattern using hydrophilic interaction liquid chromatography. These experiments revealed that monomethylation of H4-Lys 20 excludes neither acetylation of H4-Lys 16 nor hyperacetylation of the respective H4 histone molecule. These experiments therefore support our ChIP results showing a negative correlation between the patterns of trimethylated H4-Lys 20 and H4 hyperacetylation but a positive correlation between the patterns of monomethylated H4-Lys 20 and H4 hyperacetylation at distinct gene regions of the c-myc and ␤-globin genes. Immunofluorescence microscopy studies showed that trimethyl H4-Lys 20 is enriched mainly within DAPI-dense regions, which almost completely overlap with HP1␤-stained heterochromatin largely excluded, however, from active chromatin (RNApol II) regions. In contrast, monomethyl H4-Lys 20 mainly overlaps with RNApol II-stained euchromatic regions and is largely excluded from HP1␤-stained heterochromatin, indicating an association with transcriptionally engaged chromatin. To summarize our results, we conclude that hyperacetylated and monomethylated H4-Lys 20 may be important in maintaining the transcriptionally active or competent chromatin state.

MATERIALS AND METHODS
Cell Culture-MEL cells (line F4N) were grown in Dulbecco's minimum Eagle's medium containing 2ϫ nonessential amino acids, 1ϫ penicillin/streptomycin, and 10% fetal calf serum. Cells were cultured at initial cell density of 5 ϫ 10 4 /ml at 37°C and 5% CO 2 . Differentiation was induced by the addition of 2% Me 2 SO for 96 h or 1.75 mM sodium butyrate for 72 h. The percentage of benzidine-positive cells was determined as described by Orkin et al. (21).
Formaldehyde Cross-linking and Sonication-1 ϫ 10 8 cells were fixed with 0.4% formaldehyde in minimum Eagle's medium for 10 min at room temperature with gentle agitation to generate protein-DNA cross-links. To quench the reaction, glycine (125 mM final) was added. Cells were collected by centrifugation at 600 ϫ g for 10 min at 4°C and then washed two times in phosphate-buffered saline. After nuclei were prepared in cell lysis buffer at 4°C (10 mM Tris/HCl, pH 8.0, 10 mM NaCl, 0.2% Triton X-100, 1 mM PMSF, and complete protease inhibitor mixture (Roche Applied Science) as indicated in the product description), nuclei were lysed by incubation in nuclei lysis buffer (50 mM Tris/HCl, pH 8.1, 10 mM EDTA, 1% SDS, 1 mM PMSF, and complete protease inhibitor mixture) for 10 min on ice. The lysate was sonicated at 4°C with 10 pulses of 30 s each at 40% of maximum power with a Bandelin electronic UW 2070 (70 watts) sonicator with a 2-mm tip to generate DNA fragments from ϳ200 to 1000 bp with a maximum at ϳ500 bp. After centrifugation (10,000 ϫ g for 10 min at 4°C), the supernatant was used as soluble chromatin for chromatin immunoprecipitation assay.
Chromatin Immunoprecipitation Assay-ChIPs were performed as described by Forsberg et al. (22) with minor modifications. Briefly, the supernatant of soluble chromatin corresponding to ϳ4 ϫ 10 6 cells (ϳ3 A 260 units of DNA) were used for chromatin immunoprecipitations. One volume of soluble chromatin was diluted with 9 volumes of ChIP dilution buffer (17 mM Tris/HCl, pH 8.1, 1.2 mM EDTA, 167 mM NaCl, 1.1% Triton X-100, 0.01% SDS, 1 mM PMSF, and complete protease inhibitor mixture) and precleared with 100 l of protein A-agarose (50% slurry in ChIP dilution buffer) for 2 h at 4°C with gentle rocking. After centrifugation at 400 ϫ g for 3 min at 4°C, an aliquot (ϳ150 l) of precleared chromatin was removed and used in the subsequent PCR as input. The remainder of the chromatin (ϳ800 l) was incubated overnight at 4°C with gentle rotating with 10 g of the respective antibody or with an isotype-matched unspecific antibody as a negative control. Immune complexes were collected by incubation with protein A-agarose (50% slurry in ChIP dilution buffer) for 2 h at 4°C with gentle rotating. Protein A-agarose pellets were washed twice with 1 ml of low salt immune complex (IC) wash buffer (20 mM Tris/HCl, pH 8.1, 2 mM EDTA, 150 mM NaCl, 0.1% SDS, 1% Triton X-100), once with high salt IC wash buffer (20 mM Tris/HCl, pH 8.1, 2 mM EDTA, 500 mM NaCl, 0.1% SDS, 1% Triton X-100), once with LiCl wash buffer (10 mM Tris/ HCl, pH 8.1, 1 mM EDTA, 0.25 M LiCl, 1% Nonidet P-40, 1% deoxycholate), and twice with TE buffer (10 mM Tris/HCl, pH 8.0, 1 mM EDTA). Immune complexes were eluted twice as the bound fraction with 250 l of elution buffer (0.1 M NaHCO 3 , 1% SDS). NaCl (0.2 M final) was added to the bound fraction, and the input chromatin and crosslinks were reversed by incubation at 65°C overnight. Incubation with RNase A (0.05 mg/ml) at 37°C for 30 min was followed by digestion with proteinase K (50 g/ml) at 45°C for 2 h. DNA was purified by two extractions with phenol/chloroform and precipitated with ethanol. Purified DNA was resuspended in 50 l of water and quantified using picogreen fluorescence (Molecular Probes).
PCRs were performed in 1ϫ buffer (16 mM (NH 4 ) 2 SO 4 , 67 mM Tris/ HCl, pH 8.8, 1.5 mM MgCl 2 , 0.01% Tween 20), 0.25 mM dNTPs, 2% Me 2 SO, 2 units of BioTherm Taq-DNA polymerase, 0.25 M of each primer set, 5 ng (10 ng) of DNA from bound fraction, control, or input. We used a touchdown PCR protocol starting at 95°C for 3 min (10 min in the case of hot start DNA polymerase), followed by 5 cycles of 95°C for 30 s, 55°C for 1 min, and 72°C for 1 min; then 5 cycles of 95°C for 30 s, 53°C for 1 min, and 72°C for 1 min; and finally 30 cycles of 95°C for 30 s, 50°C for 1 min, and 72°C for 1 min. PCR products were collected after 20, 25, 30, 35, and 40 cycles and separated on 2% agarose gels. Gels were scanned using the Typhoon gel and blot imager (Amersham Biosciences), and trace or volume intensity was measured using Quantity One software (Bio-Rad). Curves from input and bound were generated for each primer pair, and the enrichment or depletion of the bound fraction as compared with the input was calculated using the linear range of the two curves.
Western Blotting-Histones were resolved in SDS-loading buffer, fractionated by SDS-PAGE, and transferred to a nitrocellulose membrane (Hybond ECL, Amersham Biosciences). The membrane was probed with antibodies using standard techniques and detected by enhanced chemiluminescence (ECL, Amersham Biosciences).
Immunofluorescence-Mouse erythroleukemia cells were grown in suspension culture, collected by centrifugation, and washed with phosphate-buffered saline. An amount of 10 5 cells was cytospun onto slides. Cells were permeabilized using Triton X-100 (1% for 2 min and 0.1% for 10 min) in KCM buffer (120 mM KCl, 20 mM NaCl, 0.5 mM EDTA, 10 mM Tris/HCl, pH 7.5), blocked for 30 min with 2.5% bovine serum albumin/KCM, and sequentially incubated with the primary and secondary antibodies (in 1% bovine serum albumin/KCM). Cells were fixed for 5 min with 4% paraformaldehyde in KCM, stained with DAPI for 1 min, and mounted in ProLong antifade mounting medium (Molecular Probes). For double labeling with rabbit polyclonal to tetraacetylated H4 as first primary antibody and either rabbit polyclonal to H4-mono-me-Lys 20 or rabbit polyclonal to H4-tri-me-Lys 20 as second primary antibody, we used Cy3-conjugated Fab fragment goat anti-rabbit IgG as first secondary antibody to achieve effective blocking of the first primary antibody (tetraacetylated H4). Staining was visualized using a ϫ100 objective on a Zeiss Axioplan 2 microscope and SPOT camera (Diagnostic Instruments). Images were captured using MetaVue software (Universal Imaging Corp.), processed by deconvolution analysis using AutoDeblur Deconvolution software (AutoQuant Imaging, Inc.), and analyzed using Adobe Photoshop 8.0.
Hydrophilic Interaction Liquid Chromatography-The histone fraction H4 (ϳ120 g) was isolated by reversed phase-high performance liquid chromatography as described (23) and further separated on a SynChropak CM300 column (250 ϫ 4.6 mm inner diameter; 6.5-m particle size; 30-nm pore size; Agilent Technologies, Vienna, Austria) at 30°C and at a constant flow of 1.0 ml/min using a multistep gradient starting at solvent A/solvent B (100:0) (solvent A: 70% acetonitrile, 0.015 M triethanolamine/H 3 PO 4 , pH 3.0; solvent B: 65% acetonitrile, 0.015 M triethanolamine/H 3 PO 4 , pH 3.0, and 0.68 M NaClO 4 ). The concentration of solvent B was increased from 0 to 10% B in 2 min and from 10 to 40% in 30 min and was then maintained at 40% for 10 min. The modified histone H4 isoforms obtained by HILIC were identified by ESI-MS as described (23). 20 and Trimethyl H4-Lys 20 during Interphase and at Metaphase in MEL Cells-To localize modified histones, we used immunofluorescence with monospecific antibodies. The distribution of mono-and trimethylated H4-Lys 20 during MEL cell interphase was investigated by dual immunofluorescence staining (Fig.  1). We found that trimethylated H4-Lys 20 is enriched mainly in DAPIdense regions with almost complete overlapping with HP1␤-stained heterochromatin (Fig. 1A), which is largely excluded from active chromatin (RNApol II) (Fig. 1B) regions. In contrast, monomethylated H4-Lys 20 mainly overlaps with RNApol II-stained euchromatic regions (Fig. 1D) and not with DAPI-dense regions (Fig. 1C), thus indicating an association with transcriptionally engaged chromatin. A comparable distribution pattern of both monomethylated (data not shown) and trimethylated H4-Lys 20 was also seen in differentiated MEL cells (16).

Distribution of Monomethyl H4-Lys
Mono-and Trimethyl H4-Lys 20 Histones Are Enriched after Induction of MEL Cell Differentiation-We next asked whether the levels of modified histones change during the MEL cell differentiation process using Me 2 SO. Histones from induced and uninduced MEL cells were prepared and run on SDS-PAGE. After blotting to nitrocellulose, specific antibodies to mono-, di-, and trimethylated H4-Lys 20 , acetylated H4-Lys 16 , and hyperacetylated H4 were used for immunological Western blot analysis. Although the level of dimethylated H4-Lys 20 did not change after treatment with Me 2 SO, the monomethylated H4-Lys 20 level increased about 1.5-fold and that of trimethylated H4-Lys 20 about 2-fold (Fig. 2). The level of acetylated H4-Lys 16 and the level of hyperacetylated histone H4 were clearly diminished in Me 2 SO-treated MEL cells as compared with untreated controls (Fig. 2). These results prompted us to analyze the relationship between the distribution of various epigenetic marks and gene activation or repression during MEL cell differentiation.
Because it is well known that during MEL cell differentiation the adult ␤-globin gene is largely expressed (24) and the c-myc gene is rapidly down-regulated (25)(26)(27), we used these gene loci to investigate histone H4 acetylation and methylation status at distinct sites. As determined by RT-PCR (Fig. 3), the increase in ␤-major globin transcript is about 60-fold and that in ␤-minor about 3-fold after 96 h of induction (Fig. 3,  A and B). The myc transcripts decreased by about 10 -15-fold after 96 h of induction, which is consistent with earlier reports (25-27) on the down-regulation of this gene, whereas the GAPDH transcript levels remained unchanged before and after induction (Fig. 3, C and D).
Examination of Acetyl H4-Lys 16 and Hyperacetyl H4 Pattern in Differentiated and Untreated MEL Cells at the ␤-Globin Gene Region Using ChIP-Using various antibodies specific to mono-or trimethylated H4 lysine 20 and acetylated H4 lysine 16 as well as hyperacetylated H4 in chromatin immunoprecipitation studies, we determined the extent of specific sequence enrichment in the mouse ␤-globin gene region. In order to determine the enrichment or depletion of immunoprecipitation for a specific antibody, the total amount of DNA in the bound fraction was measured with the pico green DNA quantification method, and an equal amount of input DNA (before immunoprecipitation) was run in parallel as a reference standard. Fig. 4 shows, as a representative panel of data, the specific PCR products generated after different cycle numbers for the ␤-globin gene exon 2 region using anti-tetraacetylated H4 antibody for immunoprecipitation. To calculate the relative abundance of each sequence, we used the difference in the number of PCR cycles for the bound and the input fraction needed to reach a fixed threshold value. The enrichment or depletion of a sequence in the bound fraction was calculated from the difference between the threshold cycle number for the bound and for the input fraction (28). By using this data analysis method, we were able to calculate a 13-fold increase in hyperacetylated H4 at the ␤-globin exon 2 region in the bound fraction as compared with the input fraction of induced MEL cells. Data from the analysis of MEL cells before and after induction of differentiation are given in Fig. 5, which shows the distribution of hyperacetylated H4 and the acetyl H4-Lys 16 (Fig. 5, A and B, respectively) as well as the monoand trimethylated H4-Lys 20 (Fig. 5, C and D, respectively) across the mouse ␤-globin locus. In uninduced MEL cells, hyperacetylated histone H4 was enriched to about 5-fold in the locus control region at hypersensitive sites 1-3 (HS1-3) as well as at the ␤-minor promoter and in the transcribed exon 2 and the exon 2-intron 2 regions of ␤-globin; in the HS6 region and the ␤-major globin promoter, enrichment was only 3-fold. Neither the inactive Ey-globin promoter nor the ␤H1 promoter was significantly enriched with hyperacetylated histone H4 in untreated or induced cells. The induction of MEL cells thus causing a more than 60-fold increase in the ␤-globin gene transcript (see Fig. 4A) resulted in 2-4-fold higher levels of hyperacetylated H4 histones in the HS1, -2, and -3, the ␤-major promoter, and the transcribed exon 2 region. No change in hyperacetylation, however, was found in HS6 or in Ey or ␤H1 promoter regions (Fig. 5A). Thus, for histone acetylation, our results largely agree with those of earlier studies of the mouse ␤-globin locus (22,24,29).
We next determined the levels of acetylated H4-Lys 16 , which is known as the primary acetylation site in monoacetylated H4 histone   16 , and hyperacetylated (Hyperacet) H4 was determined using specific antibodies and Western blot analysis. Histones were prepared from induced or untreated cells run on SDS-PAGE and blotted to nitrocellulose membrane. Horseradish peroxidase-conjugated secondary antibodies and ECL were used for detection. Quantification was performed with Quantity One software (Bio-Rad). All samples were analyzed in triplicate. Standard deviations are indicated by error bars. NOVEMBER 18, 2005 • VOLUME 280 • NUMBER 46 molecules, using a specific antibody able to recognize acetylated H4-Lys 16 . The acetylation pattern of H4-Lys 16 across the ␤-globin region did not resemble the hyperacetylation pattern of H4 with the exception of the HS3 region, which showed a 3-fold increase after induction. This increase is comparable with the 4-fold increase in hyperacetylated H4 histone (Fig. 5B). In this context it must be mentioned that the anti-acetylated H4-Lys 16 antibody used, although mainly and specifically detecting acetylated lysine 16 of histone H4, is not able to distinguish whether or not other H4 sites are also acetylated. Likewise, we cannot rule out the possibility that the anti-tetraacetylated H4 antibody used, which reacts mainly with hyperacetylated H4 histones (di-, tri-, and tetraacetylated isoforms), also detects monoacetylated H4. Despite this possible unspecificity, the anti-tetraacetylated H4 antibody is used for detecting transcriptionally active chromatin regions (for example see Ref. 30).

Pattern of Hyperacetylation and Lys 20 Monomethylation of H4
Pattern of Monomethyl and Trimethyl H4-Lys 20 in the ␤-Globin Gene Region-To investigate further whether mono-and/or trimethylation of histone H4-Lys 20 could be related to activation or inactivation of the ␤-globin gene, we determined the methylation pattern of H4 histone lysine 20 (Fig. 5, C and D). Although in untreated control MEL cells the monomethylated H4-Lys 20 was uniformly low across the entire region with a slight increase only in the ␤-minor promoter region, the monomethylation peaks significantly increased by about 3-4-fold at the ␤-major promoter and in the transcribed exon 2 region of the globin gene after induction of differentiation (Fig. 5C). Distribution of trimethylated H4-Lys 20 , however, was quite different (Fig. 5D). Whereas the inactive Ey and H1␤ promoters showed an enrichment of trimethylated H4-Lys 20 , all other regions displayed normal or diminished trimethylation. In this context, the various H4-Lys 20 mono-and trimethylation patterns of HS6 and the other hypersensitive sites examined (HS1-3) are worth noting. In contrast to the modest H4-Lys 20 monomethylation of HS1-3 in uninduced and induced MEL cells, we observed an increase in H4-Lys 20 monomethylation of HS6 after MEL cell induction (Fig.  5C). On the other hand, a decrease in H4-Lys 20 trimethylation of HS6 similar to that found in HS1-3 after induction was not detected (Fig.  5D). Summarizing our results, we concluded that a pronounced negative correlation exists between patterns of H4-Lys 20 trimethylation and H4 hyperacetylation (compare Fig. 5, A and D), whereas a positive correlation exists between the patterns of H4-Lys 20 monomethylation and H4 hyperacetylation in distinct regions of the ␤-globin gene (compare  . A representative panel of data generated by semiquantitative PCR using a primer pair for the ␤-globin gene exon 2 region of induced MEL cells. An antibody to tetraacetylated H4 was used for immunoprecipitation. PCR was stopped after the given numbers of cycles and the samples were loaded on a 2% agarose gel. The gel was scanned using a Typhoon imager (Amersham Biosciences) and quantified with the Bio-Rad Quantity One software. Intensity is given in arbitrary units. A representative gel is shown below the curve of the bound and the input fractions. Enrichment of a sequence in the bound fraction was calculated from the difference between the threshold cycle number for the bound fraction (C b ) and the threshold cycle number for the input DNA (C i ). As control, we used an isotype-matched antibody for immunoprecipitation. M, DNA molecular weight marker.  (25)(26)(27). We used the c-myc gene to determine the consequences of down-regulation of a gene during the Me 2 SO-induced differentiation process on the acetylation and methylation status of histone H4. Both the myc exon 1 region, where the P1 and P2 myc promoters are situated (27), and the transcribed myc exon 2 region showed a considerable amount of hyperacetylated H4 before induction of differentiation (Fig. 6A). A decrease in histone H4 hyperacetylation was not found at the promoter nor in the  (Fig. 6A). In the promoter region only 3-fold enrichment of monomethylated H4-Lys 20 was detected, whereas the exon 2 region showed substantial H4-Lys 20 monomethylation (Fig. 6C). Similar to the result achieved with hyperacetylation, a decrease in monomethylated H4-Lys 20 was not found in exon 1 nor in exon 2 region of the c-myc gene after induction of differentiation (Fig. 6C). At neither of the myc exons analyzed was an enrichment of trimethylated H4-Lys 20 observed before or after induction (Fig. 6D).
Histone H4-Lys 20 Monomethylation Is Compatible with Histone H4 Hyperacetylation-Investigations by Nishioka et al. (8) concerning the human methyltransferase PR-Set7 have shown that methylation of H4-Lys 20 and H4-Lys 16 acetylation are inversely correlated. Because acetylation of H4-Lys 16 in human cells is regarded as a mark of transcriptionally active chromatin (31), the authors suggested that methylated H4-Lys 20 maintains silent chromatin. The study by Nishioka et al. (8) did not discriminate between the individual methylation states. As PR-Set7 is known to be a monomethyltransferase, the incompatibility observed between H4-Lys 20 methylation and H4-Lys 16 acetylation obviously should be caused by H4-Lys 20 monomethylation. Most interestingly, the ChIP experiments in the present study revealed a positive correlation between patterns of monomethylated H4-Lys 20 and hyperacetylated H4 or acetylated H4-Lys 16 in the case of myc exons 1 and 2 (Fig. 6, A-C) and between monomethylated H4-Lys 20 and hyperacetylated H4 in the ␤-major promoter and the exon 2 region of the ␤-globin gene (Fig. 5, A and C). To confirm our ChIP results, we used immunofluorescence to examine the distribution of mono-or trimethylated H4-Lys 20 and hyperacetylated H4 during MEL cell interphase by dual immunofluorescence staining (Fig. 7). We found that trimethylated H4 Lys 20 excludes epitope recognition of the anti-tetraacetylated H4 antibody (Fig. 7A), whereas monomethylated H4 Lys 20 mainly overlaps with hyperacetylated H4 regions (Fig. 7B). To further support our findings with an additional method, we induced MEL cells with the deacetylase inhibitor butyrate for 72 h, and we investigated the H4 acetylation/ methylation pattern using a hydrophilic interaction liquid chromatographic procedure enabling the simultaneous separation of non-, mono-, di-, and trimethylated lysine 20 of non-, mono-, di-, tri-, and tetraacetylated H4 histones (Fig. 8). As shown, H4-Lys 20 mono-or dimethylation does not preclude acetylation of the neighboring H4-Lys 16 (Fig. 8, arrows), and tetraacetylated H4 exists in an H4-Lys 20 mono- (Fig.  8, arrowhead) and dimethylated form. A negative interplay, however, exists between H4-Lys 20 trimethylation and H4 hyperacetylation, which was previously demonstrated by Sarg et al. (16) and is also supported by the present report (Figs. 5-7).

DISCUSSION
The goal of this study was to investigate the relationship between acetylation/methylation marks in histone H4 and gene activation or repression in the course of in vitro differentiation of MEL cells. We performed chromatin immunoprecipitation assays to analyze genes largely expressed like ␤-major and ␤-minor globin genes, as well as quiescent genes like the embryonic Ey and ␤H1 genes, and the rapidly down-regulated but transcriptionally competent genes like the c-myc gene.
As to the function of histone H4-Lys 20 methylation, there is no clarity at present. Although in the past methylated H4-Lys 20 histone was largely believed to be associated with transcriptionally active rather than repressed genes (32,33), recently published reports indicate that methylated histone H4-Lys 20 is associated with silent chromatin (7,8). However, many of these reports disregard the fact that a specific lysine of histones can be methylated in vivo to form monomethyl-, dimethyl-, or trimethyl-lysine (34). Therefore, it is not possible to surmise the transcriptional state of a gene based on the fact that a given lysine in its histone is methylated or not. Instead, the precise methylation state of the lysine has to be used as an indicator of an active or repressed gene (9).
By using specific antibodies, this study demonstrates that mono-and trimethylated H4-Lys 20 histones are localized to distinct domains of chromatin (Fig. 1). Trimethylated H4-Lys 20 is mainly enriched in DAPIdense regions (Fig. 1, A and B, and Fig. 7A), reflecting the characteristic accumulation at pericentric heterochromatin (data not shown). Major functions of trimethylated H4-Lys 20 heterochromatin include the formation of a specialized chromatin structure around centromeres, which is vital to the structural integrity of the centromere, and the protection of genome stability by silencing transposable elements (35,36). However, outside the DAPI-dense regions, we found additional weak trimethylation of H4-Lys 20 (Fig. 1, A and B, and Fig. 7A). The 2-3-fold increase in trimethylated H4-Lys 20 at the embryonic quiescent Ey or ␤H1 promoter region (Fig. 5D) might reflect this type of H4-Lys 20 trimethylation. It must be mentioned that the repressed myc promoter and transcribed gene region did not show enrichment of trimethylated H4-Lys 20 in our ChIP experiments (Fig. 6), indicating that trimethylated H4-Lys 20 might play a role in repression of some genes but is not a generalized mark of gene repression.
In contrast to H4-Lys 20 trimethylation, monomethylated H4-Lys 20 histones are excluded from DAPI-dense regions and mainly overlap with RNApol II-stained euchromatic regions (Fig. 1) and regions where histone H4 is hyperacetylated (Fig. 7A), indicating an association with transcriptionally engaged chromatin. Our ChIP experiments support this assumption, because we observed increased monomethylation of H4-Lys 20 at the promoters and transcribed gene regions of active genes (␤-major globin gene, c-myc gene). There is a pronounced positive correlation between patterns of hyperacetylated and monomethylated histone H4-Lys 20 , most obviously in the ␤-major promoter and the exon 2 region of the ␤-globin gene (Fig. 5, A and C) as well as in the c-myc exon1 (promoter region) and c-myc exon 2 (transcribed gene) (Fig. 6). Most unexpectedly, we also found high levels of hyperacetylated and monomethylated histone H4-Lys 20 in the c-myc exon 1 and exon 2 region after induction of MEL cells (96 h of Me 2 SO treatment), when c-myc expression is substantially diminished. The down-regulation of the c-myc gene in MEL cells induced to differentiate by Me 2 SO treatment is known to be the result of both a transcriptional inhibition consisting of a block of transcript elongation and a post-transcriptional regulation consisting of an increased rate of mRNA degradation (37). Because the expression of c-myc is known to decline in a polyphasic manner during the MEL cell differentiation process (25,26), we suppose that the c-myc gene is still engaged in the transcription initiation process, persisting in a transcriptionally competent condition and thus explaining the consistently high levels of hyperacetylated and monomethylated H4 histones. The finding that the RNA pol II complex is located at the c-myc P2 promoter (exon 1), even when c-myc is not expressed (27), might explain our results. Furthermore, a correlation between acetylation and active transcription is not imperative. For instance, the inactive ␤-major and ␤-minor globin  (Cy3, red). A, distribution of trimethylated H4-Lys 20 (green) and hyperacetylated H4 (red). B, distribution of monomethylated H4-Lys 20 (green) and hyperacetylated H4 (red). The far right column of each panel simultaneously shows the Cy3 (red) and FITC (green) patterns (merge), giving a yellow stain when the two antibodies bind in the same region. DNA was counterstained with DAPI to highlight the foci of heterochromatin. Scale bar, 5 m. The histone H4 fraction was analyzed on a SynChropak CM300 column (250 ϫ 4.6 mm) at 30°C and a constant flow of 1.0 ml/min using a two-step gradient (see "Materials and Methods"). Cells were grown for 72 h in the presence of 1.75 mM sodium butyrate. The isolated protein fractions obtained by HILIC were identified by ESI-MS (23) and designated ac0 -ac4 for the non-, mono-, di-, tri-, and tetraacetylated histone H4 forms and 0 -3 for the non-, mono-, di-, and trimethylated histone H4 forms. The tetraacetylated (Lys 5 , Lys 8 , Lys 12 , and Lys 16 ) monomethylated (Lys 20 ) histone H4 form is marked with an arrowhead. gene promoter (22,24) and the inactive transcribed globin gene region (Fig. 5A) of uninduced MEL cells or the transcriptionally inactive ␤-minor globin promoter in yolk sac (22) are all hyperacetylated. Therefore, we assume that hypoacetylation of H4 is not a prerequisite of c-myc gene down-regulation and that mono-and trimethylated H4-Lys 20 histones are not involved in repression of the c-myc gene. We suggest that persistent hyperacetylation in down-regulated c-myc gene regions is probably caused by the repression of a specific histone deacetylase during the MEL cell differentiation process. For several Drosophila genes, it was recently shown (38) that high levels of methylation (H3-Lys 27 and H3-Lys 9 ) at the promoter do not prevent strong transcription, suggesting that multiple-regulated interactions between methylated histone regions at promoters and transcribed genes are required to create a silenced locus. Beisel et al. (19) showed that the epigenetic activator ASH-1, a multicatalytic histone methyltransferase, activates transcription by methylation of H3-Lys 4 , H3-Lys 9 , and H4-Lys 20 at the promoter of target genes in Drosophila. This trivalent methylation pattern established by ASH-1 seems to prevent the interaction of repressors with ASH-1 target genes.
We showed recently (16) that histone H4 hyperacetylation (tri-and tetraacetylation) precludes Lys 20 trimethylation of H4. The present study supports this finding, because ChIP experiments (Figs. 5 and 6) and immunofluorescence studies (Fig. 7A) revealed a negative correlation between patterns of trimethylated H4-Lys 20 and H4 hyperacetylation. A positive correlation, however, was found between monomethylated H4-Lys 20 and H4 hyperacetylation, in particular in distinct gene regions of c-myc and ␤-globin genes. The fact that H4-Lys 20 monomethylation is absolutely compatible with hyperacetylated H4 was also confirmed by immunofluorescence studies (Fig. 7B) and HILIC analysis of H4 histones obtained after treating MEL cells with the deacetylase inhibitor butyrate (Fig. 8). Taken together, in contrast to H4-Lys 20 trimethylation, which is associated with chromatin compaction and repression of gene activity (16), H4-Lys 20 monomethylation correlates with the transcriptionally active or competent chromatin state.