Inhibitors of DNA methylation and histone deacetylation ... - CiteSeerX

Uncorrected Version. Published on September 25, 2006 as DOI:10.1189/jlb.0106005

Inhibitors of DNA methylation and histone deacetylation independently relieve AML1/ETO-mediated lysozyme repression Rainer Claus,* Manfred Fliegauf,*,† Michael Stock,*,‡ Jesu´s Duque,* Mateusz Kolanczyk,*,§ and Michael Lu¨bbert*,1 *Department of Medicine, Division Hematology/Oncology and †Department of Pediatrics and Adolescent Medicine, University of Freiburg Medical Center, Freiburg, Germany; ‡Nikolaus Fiebiger Center of Molecular Medicine, Department of Experimental Medicine I, University Erlangen-Nu¨rnberg, Erlangen, Germany; and §Max-PlanckInstitute for Molecular Genetics, Berlin, Germany

Abstract: The human lysozyme (LZM) gene is highly methylated in LZM-nonexpressor immature myeloid and in nonmyeloid cells and unmethylated only in LZM-expressing cells. Extended methylation analyses of the CpG-poor 5ⴕ flanking region and of the exon 4 CpG island (both containing Alu elements) of the LZM gene were now performed. Marked demethylation was noted after treatment of AML1/ETO-positive Kasumi-1 cells with the DNA methyltransferase (DNMT) inhibitor 5-aza2’-deoxycytidine (5-azaCdR), not associated with cellular differentiation. LZM mRNA in Kasumi-1, but not in several AML1/ETO-negative myeloid cell lines, was specifically and independently up-regulated upon treatment with 5-azaCdR and, to a lesser extent, with the histone deacetylase (HDAC) inhibitor trichostatin A (TSA). Increased chromatin accessibility within the 5ⴕ LZM gene was observed concomitantly with 5-azaCdR-induced demethylation. In contrast, TSA treatment had no effect on chromatin accessibility, but, as shown by chromatin immunoprecipitation, resulted in increased acetylation of histones H3 and H4. Repression of LZM transcription is mediated by conditional AML1/ETO expression in an inducible cell line model (U-937) and reversed by siRNA "knockdown" of AML1/ETO in Kasumi-1 cells (Dunne et al., Oncogene 25: 2006). Antagonization of LZM repression following conditional expression of AML1/ETO was achieved by TSA. In conclusion, we demonstrate complex interactions between DNA methylation and histone modifications in mediating LZM repression, which implicate AML1/ ETO as one component involved in local chromatin remodeling. Interestingly, inhibitors of DNMTs and HDACs independently relieve repression of this CpG-poor gene in AML1/ETO-positive cells. J. Leukoc. Biol. 80: 000 – 000; 2006. Key Words: DNA methylation 䡠 5-aza-2’-deoxycytidine 䡠 chromatin modification 䡠 epigenetic silencing 䡠 acute myeloid leukemia 䡠 chromosomal translocation (8;21) 䡠 Alu 0741-5400/06/0080-0001 © Society for Leukocyte Biology

INTRODUCTION Methylation of cytosine residues (mostly at CpG islands, CpG dinucleotides clustered in regulatory regions of tissue- and development-specific genes) represents a key mechanism of postmitotic, epigenetically fixed but dynamic gene inactivation in normal cells [1, 2]. In hematopoietic development, this had first been demonstrated for erythropoiesis [3], later also for lymphoid [4] and myeloid differentiation [5, 6]. In recent years, aberrant methylation also of bona fide tumor suppressor and growth-regulatory genes has been recognized as the most frequent alteration both in hematologic neoplasms and solid tumors [7]. For instance, in acute leukemias, two genes encoding inhibitors of cyclin-dependent kinases 4 and 6, i.e., p15/ INK4B and p16/INK4A, are equally inactivated by DNA methylation as by genetic events [8]. Transcription of these genes is associated with 5⬘ demethylation and “open” chromatin in the presence of a methylated exonic CpG island (thus being permissive for transcription) [9]. Cytosine methylation is inversely correlated with gene transcription in most CpG-rich promoters. This state is functionally linked to a "closed," inactive chromatin structure, which, in turn, is mediated by local chromatin remodeling [10]. These different levels of epigenetic control of transcription are maintained by the physical interaction of different classes of proteins: MeCP (methylated CpG binding protein) 2 recruits histone deacetylases (HDACs) to stretches of densely methylated cytosine [11]; the Mi-2 protein also links DNA methylation, chromatin remodeling, and histone deacetylation [12]. Lysozyme (LZM, EC 3.2.1.17) is a secretory protein of 14.6 kDa mainly expressed in glandular epithelial cells and in the myelomonocytic lineage, representing 2.5% of the total cellular protein in mature granulocytes and macrophages [13, 14]. LZM is an enzyme that acts in nonspecific humoral immunity by disintegrating bacterial cell walls. LZM is located on chromosome 12 and is composed of four exons and three introns

1

Correspondence: Department of Medicine, Division Hematology/Oncology, University of Freiburg Medical Center, Hugstetter Str. 55, Freiburg D-79106, Germany. E-mail: [email protected] Received January 5, 2006; revised August 11, 2006; accepted August 15, 2006; doi: 10.1189/jlb.0106005.

Journal of Leukocyte Biology Volume 80, December 2006

Copyright 2006 by The Society for Leukocyte Biology.

1

covering 5856 bp [15]. Functional promoter activity is attributed to the 3.5 kb 5⬘ region [16]. DNA methylation of LZM and its murine ortholog/homolog is tightly regulated during (normal) myeloid cell differentiation [17, 18]. However, LZM expression is variable in primary AML blasts [19, 20], and control of its regulation at this early differentiation step of neoplastic cells is only poorly understood. We have previously demonstrated methylation changes within the LZM during normal myeloid cell differentiation toward monocytes, macrophages, and granulocytes, and in primary AML cells and cell lines [18, 20, 21]. Although several CpG sites were completely methylated in normal nonmyeloid cells, demethylation of a single CpG site within the 5⬘ LZM region was tightly coupled to its expression in differentiating myeloid precursor cells, with a complete demethylation in mature phagocytic blood cells (highest expression of LZM). In acute leukemias, a monocytic phenotype (AML of FAB subtypes M4, M5) is associated with LZM expression and partial gene demethylation, whereas in AML with partial myeloblastic maturation (FAB M2), less than 50% of the leukemias exhibited detectable levels of LZM mRNA [20]. We now extended these methylation studies of the human LZM gene to other myeloid cell lines and to a broad range of CpG sites using bisulfite-based approaches. To functionally address the role of epigenetic LZM gene repression, we investigated whether epigenetic states and epigenetically mediated transcriptional control were reversible by inhibitors of DNA methylation (5-aza-2'-deoxycytidine, 5-azaCdR) and histone deacetylation (trichostatin A, TSA). Finally, we provide evidence that the AML1/ETO leukemia-specific fusion protein is functionally involved in repressing the LZM gene.

MATERIALS AND METHODS Cells and tissue culture Peripheral blood lymphocytes, granulocytes, and monocytes were isolated from heparinized blood of healthy volunteers, as described [20]. All cell lines were obtained from DSMZ (German Collection of Microorganisms and Cell Cultures, Braunschweig, Germany). Kasumi-1 (established from a patient with AML-M2 carrying the t(8;21)positive cell line) was cultured in RPMI 1640 (GIBCO, Karlsruhe, Germany) supplemented with 15% (first few days after thawing 20%) heat-inactivated FBS. HL-60. KG-1, NB-4, and U-937 were grown in RPMI 1640 with 10% FBS. Medium additionally contained 1 mM sodium pyruvate and 100 U/ml penicillin-streptomycin. Cells were incubated at 37°C in a humidified atmosphere containing 5% CO2. Cell growth and viability were determined by trypan blue exclusion (final concentration 0,08%). Cell morphology was examined in cytospins using Pappenheim staining. Transfections of Kasumi-1 cells (kindly provided by PD Dr. Olaf Heidenreich, Tu¨bingen, Germany) were performed by electroporation (with siRNAs against AML1/ETO), as described previously [22].

Drug treatment Cells were treated with increasing concentrations (10-1000 nM) of 5-aza-2'deoxycytidine (Sigma, Taufkirchen, Germany). 5-aza-2'-deoxycytidine (5-azaCdR, 5-aza) was dissolved in PBS (GIBCO). The compound was added in sequential pulses of 24 h duration. Treatment with TSA (Sigma) was initially performed with concentrations ranging from 10-5,000 nM. TSA was dissolved in ethanol. For subsequent experiments, concentrations of 200 nM were used for 72 h. All trans-retinoic acid (ATRA, Sigma) was dissolved in ethanol to a

2


final concentration of 1 ␮M. MS-275 (kindly provided by Schering AG, Berlin, Germany) was used at a final concentration of 100 nM. Combination experiments of both drugs were performed using different schedules of 72 h (3 ⫻ 24 h) 5-azaCdR pre-treatment (25, 50, 100 nM), complemented or followed by additional 36-72 h TSA treatment (50, 100, 200 nM).

Methylation-sensitive restriction analysis by PCR Genomic DNA was isolated from 1–5⫻107 cells by standard procedures. DNA was digested for 12 h using either SmaI or MspI (both New England Biolabs, Frankfurt, Germany). For PCR, LZM 5⬘ flanking region-specific primers LZM1/LZM2 were designed (LZM1: GGA AAG GGG AGG CAA AGT ATT GTG, LZM2: CAT GTG CCT GTT GTC TTA ATG GGT G) flanking SmaI restriction site S1. Confirmation of SmaI restriction site S3 in exon 4 was performed by specific primers LZM3/LZM4 (LZM3: TCC GTC AGT ATG TTC AAG GTT G, LZM4: CTA ACA CGG TGA AAC CCT GTC). The total reaction volume contained ⬃200-ng template DNA. Samples were amplified according to standard PCR conditions (LZM 5⬘: 95°C for 5 min; 28 cycles 95°C for 30 s, 55°C for 30 s, 72°C for 1 min; 72°C for 10 min; LZM exon 4: 95°C for 5 min; 30 cycles 95°C for 30 s, 54°C for 30 s, 72°C for 1 min; 72°C for 10 min) using Taq polymerase (Perkin Elmer, Boston, MA).

DNA probes A BamHI/RsaI restriction fragment of the pHL2 cDNA clone (kindly provided by Dr. S. Gordon [23], Oxford, UK) was used for hybridizing procedures. This 700-bp fragment is complementary to exon 1 through the central, coding portion of exon 4 of the human lysozyme gene. The fragment of the genomic clone of the human CD34 gene, the "Sma-C" fragment of the human myeloperoxidase (MPO) cDNA, and the GAPDH cDNA fragment have been used (18).

Methylation analysis by Southern blot analysis Genomic DNA samples were cleaved with HindIII (New England Biolabs). After purification by ethanol precipitation, a second restriction digest was performed using either methylation-sensitive SmaI, HpaII (New England Biolabs) or methylation insensitive MspI or XmaI (New England Biolabs). Cleaved DNA was size-fractionated on 0.8% agarose gels and blotted on Hybond N⫹ nylon membranes (GE Healthcare, Freiburg, Germany). Filters were hybridized at 65°C for 16 h using probes labeled by the random prime method with [␣-32P] dCTP. Filters were rinsed at 65°C to a final stringency of 0.1⫻ SSC and autoradiographed to X-ray films at – 80°C using intensifying screens (Kodak, Stuttgart, Germany).

RNA isolation and Northern blot analysis Total cellular RNA was extracted as described [24] and resolved on denaturing 1% formaldehyde-agarose gels. 28S and 18S rRNA were used as reference standards to verify the relative amount, integrity, and molecular size of the RNA. RNA was transferred to Hybond N⫹ nylon membranes (GE Healthcare). Hybridization was carried out as described above. Signals were quantified by densitometric calculation procedures using the image analysis software “ImageJ” (Wayne Rasband, NIH, [email protected]).

Bisulfite sequencing Bisulfite sequencing was carried out as described previously [25]. Two micrograms of HindIII-digested genomic DNA were denaturated in a volume of 20 ␮l by incubating for 20 min at 95°C and adding 5 ␮l 3 M NaOH. After addition of freshly prepared bisulfite-hydroquinone (12 ml 0.1 M hydroquinone, 208 ml 3.6 M sodium bisulfite) the solution was incubated for 16 h at 55°C in darkness. DNA was purified using QIAquick PCR Purification Kit according to manufacturer's advice (Qiagen, Milden, Germany). Remaining traces of sodium bisulfite were removed by treatment with 0.27 M NaOH and ethanol precipitation. CpG-rich stretches in the LZM 5⬘-region (LZM5: AAT ATG GAA AGG GGA GGT AAA GTA, LZM6: CTT TTC CAC AAA ATA AAC ATA ACA CA) and in exon 4 (LZM3: TCC GTC AGT ATG TTC AAG GTT G, LZM4: CTA ACA CGG TGA AAC CCT GTC) were amplified by PCR. PCR products were purified by gel extraction using QIAquick Gel Extraction Kit (Qiagen) and cloned into pCR4-TOPO (Invitrogen, Karlsruhe, Germany). Multiple clones

http://www.jleukbio.org

were sequenced using Abi Prism (Big)Dye Terminator Cycle Sequencing Ready Reaction Kit (Applied Biosystems, Foster City, CA).

MspI protection assay For isolation of nuclei, at least 5 ⫻ 107 cells were incubated in hypotonic buffer (RSB, 10 mM Tris pH 7.4, 3 mM MgCl2, 10 mM KCl) and 0.3% NP-40 at 4°C for 5 min, as described previously [5]. Nuclei were rinsed in cold 1⫻ RSB and cleaved by MspI (New England Biolabs) in increasing activities (ranging from 0 to 800 units) for 45 min at 37°C. Digestion was terminated by proteinase K digestion. Fragmented genomic DNA was purified as described. DNA samples were digested with HindIII (New England Biolabs) and analyzed by Southern blot analysis as described above.

Chromatin immunoprecipitation Chromatin immunoprecipitation (ChIP) was essentially performed as described previously (Upstate Biotech, Lake Placid, NY; www.upstatebiotech.com/support/protocols/chips.html) using Anti acetyl-H3 (#06-599) and Anti acetyl-H4 (#06-866) antibodies (Upstate Biotech) and normal rabbit serum (Santa Cruz Biotechnology, Heidelberg, Germany) as a negative control. A shortened washing procedure was used; after antibody binding protein A-sepharose solution was pelleted by centrifugation and washed 5 times using the sequence of buffers listed below. One milliliter of buffer was used for 1 min per wash: 1st min: 0.1% SDS; 1% Triton X-100; 2 mM EDTA; 20 mM Tris-HCl, pH 8.1; 150 mM NaCl; 2nd min: 0.1% SDS; 1% Triton X-100; 2 mM EDTA; 20 mM Tris-HCl, pH 8.1; 500 mM NaCl; and 3rd to 5th min: TE, pH 8.0. CpG-rich areas in the LZM 5⬘-region and exon 4 were amplified from immunoprecipitated fragmented genomic DNA by quantitative PCR using the primer pairs (LZM1/LZM2 and LZM3/LZM4) and analyzed by gel electrophoresis.

RESULTS Progressive demethylation at a single cytosine located within the LZM promoter region during myeloid differentiation Normal myeloid precursor cells expressing LZM mRNA exhibit incomplete demethylation at a SmaI site termed "S1" (Fig. 1A) within the 5⬘ flanking region of this cytokine-inducible gene [18]. However, variable methylation of this site is found in myeloid leukemia cells [20]. This analysis was now extended to several other myeloid cell lines including Kasumi-1 and NB-4 with specific balanced chromosomal translocations resulting in the chimeric transcription factors AML1/ETO and PML/retinoic acid receptor (RAR)-␣, respectively. As shown in Fig. 1B, restriction analysis of the human LZM gene in four myeloid leukemic cell lines arrested at the late myeloblastic (HL-60, Kasumi-1), promyelocytic (NB-4), and monoblastic differentiation stage (U-937), respectively, revealed highly similar, incomplete demethylation at this SmaI site, which we previously mapped to the upstream region ⬃700 bp 5⬘ of the transcriptional start site. To confirm and extend mapping of this differentially methylated restriction site, first a semiquantitative PCR-based restriction assay was performed on genomic DNA from Kasumi-1, HL-60, U-937 cells, and normal monocytes. In case of complete digestion at the restriction site of interest (either because of use of a methylation-insensitive isoschizomer enzyme or because a methylation-sensitive enzyme digests to completion) an amplicon will be absent. Amplicons containing the SmaI sites within the 5⬘ flanking region (“S1”) and exon 4 (“S3”) of the LZM gene (Fig. 1A), respectively, could not be

Fig. 1. Myeloid leukemic cell lines with partial maturation exhibit incomplete demethylation at a single CpG dinucleotide within the LZM 5⬘ flanking region. (A) Map of the human LZM on gene (exons indicated by black boxes; arrow, transcriptional start site) showing the distribution and ratio of CpG dinucleotides vs. GpC dinucleotides. (B) DNA (10 ␮g) from early myeloblastic KG-1 cells, late myeloblastic HL-60 and Kasumi-1 cells, promyelocytic NB-4, monoblastic U-937 and normal peripheral blood monocytes was digested with HindIII (Hi), followed by restriction either with methylation-sensitive SmaI (S), or methylation-insensitive MspI (M) or XmaI (X) isoschizomers. DNA was size-fractionated by agarose gel electrophoresis, blotted and DNA hybridized to a human LZM cDNA probe as described in Materials and Methods. Identical results were obtained with MspI and XmaI because the CCGG sequences studied are embedded in CCCGGG sites. (C) DNA from Kasumi-1, HL-60, U-937 cells and peripheral blood monocytes was restricted first with HindIII, then with either MspI or SmaI. Following restriction, the DNA was reprecipitated and genomic PCR was performed with primers LZM1 and LZM2 to generate an amplicon of 498 bp containing the SmaI site 700 bp upstream of the transcription start site (amplicon a-a, top) and with primers LZM3 and LZM4 for a 522 bp amplicon a-b containing the SmaI site S3 within exon 4 of the human LZM gene (bottom). PCR were performed according to standard procedures. PCR product was size-fractionated on agarose gels and bands visualized by ethidium-bromide staining of the gels. Because of the sensitivity of the PCR assay in detecting the minor fraction of methylated alleles in these myeloid cells, the band indicating incomplete digestion at S1 was overrepresented in some experiments.

Claus et al. Regulation of lysozyme gene expression

3

Fig. 2. 5-aza-2'deoxycytidine (5-azaCdR) treatment of Kasumi-1 but not HL-60 or U-937 cells results in LZM gene demethylation and derepression of LZM transcription. (A) Treatment of Kasumi-1 with 5-azaCdR results in LZM gene demethylation. DNA from Kasumi-1 cells, either untreated or treated with 200 nM 5-azaCdR for 72 h was extracted, restricted first with HindIII then either with SmaI (SI), MspI (M), or HpaII (H). DNA was size-fractionated on agarose gel and hybridized to the LZM cDNA fragment, as described in Materials and Methods. Autoradiography was performed for 60 h. Note identical restriction patterns obtained with SmaI and HpaII. (B) Bisulfite sequencing was performed on DNA from Kasumi-1 cells, either untreated or treated with 200 nM 5-azaCdR for 72 h. Amplicons a-a and a-b were generated from sodium bisufite-treated DNA of Kasumi-1 cells, as described in Materials and Methods. PCR products were subcloned, and single alleles were analyzed for methylation status of each CpG by automated sequencing. Open circles indicate unmethylated CpGs, solid circles denote methylated CpGs. (C) RNA was extracted from the 3 myeloid cell lines, either untreated or after treatment with 5-azaCdR at 200 nM for the indicated time periods. Cells were pulsed (see arrow) every 24 h with fresh medium and 5-aza-CdR. Northern blot analysis (using a human LZM cDNA fragment) was performed followed by sequential hybridization with DNA probes for GAPDH, MPO and CD34 as indicated. LZM expression levels are represented as relative values compared with GAPDH, as determined by computer-based densitometry as described in Materials and Methods. (D) In a time-course experiment, Kasumi-1 cells were treated with 5-azaCdR, as described in (B) for the indicated time periods, followed by parallel restriction analysis with Hind III and SmaI and Northern blot analysis for LZM expression.

generated following methylation-insensitive HindIII/MspI double digests in all DNA samples (Fig. 1C). After double-digest with HindIII and methylation-sensitive SmaI, a faint band in Kasumi-1 cells, which is nearly absent in monocytes, indicated incomplete demethylation in Kasumi-1 cells and almost complete demethylation in monocytes, respectively, at the SmaI site S1. By the same assay, complete demethylation in monocytes but methylation in Kasumi-1 cells was noted at SmaI site S3 (part of a CpG island within LZM exon 4). Similar analyses with U-937 and HL-60 cells revealed a slightly higher degree of methylation at the S1 site whereas S3 methylation was comparable to Kasumi-1. The 5⬘ flanking region displays a CpG density at the expected, low frequency (ca. 1.5 CpGs per 100 bp), but the central exon 4 region fulfills criteria of a CpG island (5 CpGs per 100 bp; Fig. 1A, bottom).

A DNA methyltransferase inhibitor mediates derepression of the LZM gene in Kasumi-1 cells To determine the effect of DNA methyltransferase inhibition on CpG methylation in the LZM gene, we first compared 4


responses to a demethylating agent in different myeloid lines with similar methylation status at CpG site S1. In Kasumi-1, 5-azaCdR treatment was associated with partial demethylation at the two SmaI sites located in the CpG-rich regions in intron 3 (“S2”) and exon 4 (S3) of the LZM gene, respectively (Fig. 2A). This effect was not detectable at comparable levels in HL-60 and U-937 (data not shown). Regional extension of methylation analyses by bisulfite sequencing (Fig. 2B) and display of individual alleles showed that in Kasumi-1 cells the cytosine within S1 was the only strongly demethylated site, whereas the remaining CpGs were almost completely methylated. The exonic CpG island was completely methylated (Fig. 2B, top). However, 5-azaCdR treatment caused marked demethylation with spreading in both regions (bottom). Among the three cell lines, LZM mRNA expression varied widely and was highest in U-937 but hardly detectable in Kasumi-1 (Fig. 2C, left). We hypothesized that treatment with 5-azaCdR might increase LZM expression by demethylation of regulatory regions of the gene. All three cell lines were treated http://www.jleukbio.org

Fig. 3. Derepression of LZM transcription in Kasumi-1 attained by trichostatin A (TSA) treatment is not associated with DNA demethylation. (A) Kasumi-1 cells were treated with 200 nM TSA for the time periods indicated. Total RNA was extracted and subjected to Northern blot analysis, followed by hybridizations with cDNA probe for human LZM. (B) TSA treatment is not associated with changes in DNA methylation. DNA was extracted from Kasumi-1 cells that were treated with TSA for 72 h at a 200 nM concentration. Restriction analysis was performed with double digests of HindIII (Hi) with SmaI (S), HpaII (H), or MspI (M). As negative control (untreated cells), the left panel from Fig. 2B is shown. Note identical restriction patterns obtained with SmaI and HpaII. (C) No synergistic effects on LZM derepression by combined treatment with 5-azaCdR and TSA. Left: Kasumi-1 cells were treated with either 100 nM TSA for 72 h or with 100 nM 5-azaCdR (DAC) for 72 h. Combination of both drugs was performed as 72 h pretreatment with 5-azaCdR followed by 72 h TSA treatment in a schedule with 24 h overlapping treatment. Right: RNA was extracted and Northern blot analysis was performed as described above. While up-regulation of LZM by DAC is apparent, the addition of subliminal doses of TSA does not result in a synergistic upregulation.

with 5-azaCdR at low-cytotoxic concentrations, resulting in at least 50% growth inhibition (with ⱖ 80% cell viability) after 6 days of treatment (data not shown). Interestingly, LZM expression was unchanged in HL-60 and U-937 but increased ⬃10fold in Kasumi-1, with maximal levels attained between 48 and 60 h (Fig. 2C). To ask whether this was a consequence of partial induction of myeloid differentiation, blots were rehybridized with probes for the differentiation marker genes MPO and CD34, which, however, remained unchanged upon 5-azaCdR treatment. No changes in cell morphology or surface marker expression indicative of differentiation were obtained (data not shown). The temporal relationship between 5-azaCdR-induced demethylation and transcriptional activation in Kasumi-1 cells was examined next. Applying three pulses of 5-azaCdR every 24 h, a faint 5.0-kb band, indicating demethylation first appeared after 24 h, and became more predominant in Southern blot analysis after 72 h (Fig. 2D). Up-regulation of LZM expression was first observed after 24 to 48 h, with the maximum at 72 h, indicating that initial demethylation of SmaI sites S1 and S2 at least paralleled and possibly preceded gene derepression.

Derepression of the LZM gene in AML1/ETOpositive cells can also be attained by treatment with a histone deacetylase inhibitor DNA methylation as one major mechanism of epigenetic gene control is closely linked to local chromatin modifications such as histone acetylation/deacetylation. Direct interactions of methylated DNA with regulatory proteins (e.g., methylated DNA binding proteins, MBPs), and recruitment of silencing complexes containing corepressors with histone deacetylase activity (HDACs) are mechanisms of this functional connection. Thus, we hypothesized that inhibition of histone deacetylases might also derepress the human LZM gene. TSA at a 200-nM concentration had only limited cytotoxic effects on Kasumi-1 cells after a 3-day treatment (data not shown). The treatment resulted in an almost sevenfold up-regulation of LZM mRNA expression after 72 h (Fig. 3A). Hybridizations with the cDNAs for the differentiation markers MPO and CD34 (data not shown) showed no effect upon their expression, indicating that LZM up-regulation by TSA was unlikely to be a sequela of differentiation. Morphologically, cells showed increased cytoplasmic vacuolization, but neither nuclear segClaus et al. Regulation of lysozyme gene expression

5

mentation or other changes indicative of a differentiating effect, and no changes in surface marker expression analyzed by flow cytometry could be observed (data not shown). Treatment of U-937 and HL-60 with the same TSA concentration either did not affect LZM expression or resulted in a slight decrease in mRNA levels (data not shown). Up-regulation of LZM by TSA in Kasumi-1 was independent of DNA demethylation, as shown by an unaltered restriction pattern obtained with SmaI and HpaII (Fig. 3B), and as confirmed by methylation-sensitive restriction/PCR (data not shown). Because a synergistic effect of combined treatment with 5-azaCdR and TSA has been described for selected CpG-rich genes [26], we treated Kasumi-1 cells with different concentrations and schedules of TSA and 5-azaCdR. However, for subliminal doses of TSA (data shown for 100 nM, 72 h) following different dose levels of 5-azaCdR for 72 h (data shown for 100 nM), in 24 h overlapping schedule, no synergistic effect was detectable [Fig. 3C]. This was also apparent when rehybridizing with a probe for the gene encoding NIMA (data not shown), which is repressed by AML1/ETO induction [27]. Also, the combination of lower DAC concentrations (25 and 50 nM, 0-72 h) with modified exposure time to TSA at different dose levels (50, 100, and 200 nM, 72-144 h) did not induce a synergistic effect (Fig. 3C and data not shown).

Local chromatin configuration changes of the LZM gene during expression Local chromatin configuration is determined in part by the distribution, density, and methylation of CpG dinucleotides [28, 29, 53]. To address whether the distinct methylation profile noted in myeloid cells of different maturation degrees was associated with variably “active“ chromatin surrounding these differently methylated regions, MspI protection assay was performed. Chromatin accessibility patterns of the MspI sites targeted were relatively similar in HL-60, U-937 (all with ⬍20% methylation at site S1) as compared by densitometric quantification and distinct to those of peripheral blood lymphocytes (PBL), KG-1 early myeloblasts, and peripheral blood granulocytes (PMN) (Fig. 4A, bottom). Specifically, no 6.9-kb band indicative of accessibility to the nuclease at site S1 appeared even at the highest enzyme concentrations in KG-1 cells, PBL, and PMN. In contrast, chromatin surrounding this sequence was readily accessible to the nuclease in HL-60 and U-937. In Kasumi-1 cells exhibiting a similar degree of demethylation at S1, chromatin accessibility was observed to a lesser degree, as displayed by faint bands at 6.9 kb. Upon treatment of Kasumi-1 cells with 5-azaCdR for 72 h, appearance of additional bands and weakening of the top band (11.3 kb) indicated an increase in MspI accessibility with increasing demethylation. Interestingly, chromatin opening was also indicated by appearance of the 5.0 kb band, which is generated by SmaI restriction of “naked” DNA from 5-azaCdR-treated Kasumi-1 cells (see Fig. 2A). In contrast, no change in local chromatin configuration was observed in TSA-treated cells. Since the activation state of chromatin is also altered by changes of histone acetylation following histone deacetylase inhibition, chromatin immunoprecipitation was performed on both regions showing differential DNA methylation during 5-azaCdR-induced derepression (Fig. 4B). Treatment of Ka6


sumi-1 cells with TSA resulted in a substantial increase in the amount of acetylated histone H4 and H3 in the 5⬘ flanking region, and, to a lesser degree, within the CpG-rich region in exon 4 of the LZM gene. After 48 h, acetylation was decreased and was undetectable at 72 h for both histone fractions.

AML1/ETO-mediated repression of LZM transcription is relieved by trichostatin A G1 arrest and apoptosis are induced upon conditional expression of AML1/ETO in U-937 cells, as shown by others [30] as well as ourselves [27]. In addition, in our system of ecdysonemediated conditional expression, LZM transcription is repressed following AML1/ETO induction [Ref. 31 and Fig. 5] and the 5⬘ gene region containing 5 perfect AML1 binding sites (Fig. 5A). Very recent further evidence for LZM repression through the AML1/ETO stems from microarray results obtained with siRNA-mediated AML1/ETO knock-down, which results in LZM derepression, validated by quantitative PCR [32]. As shown in Fig. 5B by Northern blot analysis, it was confirmed that in Kasumi-1 cells, LZM mRNA is derepressed following transfection of the cells with siRNAs directed against AML1/ ETO [22]. Of note, derepression was more effective after a second transfection. AML1/ETO has been shown to recruit histone deacetylases to target genes of AML1, thereby mediating transcriptional repression via histone deacetylation [33]. We therefore asked whether inhibition of histone deacetylation by TSA would result in decreased LZM suppression. Several independent clones of stably transfected U-937 were treated for different time intervals with Ponasterone (Pon) A, a synthetic derivative of ecdysone, in the absence or presence of TSA. Treatment with PonA alone resulted in repression of LZM gene expression in several independent clones, but in wild-type U-937 cells, it did not affect LZM expression (Fig. 5C, left). This repressive effect of AML1/ETO could be inhibited by simultaneous treatment with TSA for 48 h (Fig. 5C, right); densitometric quantification of LZM expression (equalized for GAPDH expression) showed that while some increase of LZM expression occurred with TSA alone, the combined treatment with PonA and TSA completely mitigated the decrease associated with AML1/ETO induction. This was not accompanied by a significant reduction in cell viability usually observed after AML1/ETO induction [31].

DISCUSSION Myeloid-specific expression of developmentally regulated genes by epigenetic mechanisms via promoter demethylation has been demonstrated for the myeloperoxidase (MPO) gene [5], the c-FMS gene [6], the SERPINB5 promoter [2], and other genes. Similarly, chromatin remodeling during myeloid differentiation was found for the MPO gene [5], the c-FMS gene [34], and at the ELA2 locus [35]. The human lysozyme gene also provides an excellent model for development-specific gene methylation. Previously, we had mapped a single CpG dinucleotide discriminating between myeloid LZM expressor cells of intermediate and late maturation stages (⬍20% methylation) and LZM nonexpressor early myeloid and nonhematopoietic http://www.jleukbio.org

Fig. 4. Chromatin accessibility to MspI within the 5⬘ LZM gene differs in myeloid cells of different maturation and in Kasumi-1 cells is altered by 5-azaCdR but not TSA (A) For MspI accessibility assay, viable nuclei were isolated by hypotonic lysis, as described in Materials and Methods from the following cells: peripheral blood lymphocytes (PBL), KG-1 early myeloblast cells, U-937 monoblasts, HL-60, and Kasumi-1 acute myeloblast cells peripheral blood granulocytes (PMN), which were untreated or treated with 5-azaCdR (200 nM, 72 h) or TSA (200 nM, 72 h), respectively. Nuclei were incubated with increasing concentrations of MspI (50 – 800 U for 30 – 60 min at 37°C), as described in Materials and Methods. DNA was then extracted, digested with HindIII, and subjected to Southern blot analysis using a LZM cDNA probe. Lanes marked M indicate naked DNA from the respective cell type digested with MspI. (B) Acetylation of histones H3 and H4 within the LZM gene is increased by TSA in Kasumi-1 cells. Chromatin immunoprecipitation with antibody against acetylated histone H4 and H3 was performed with nuclei from untreated Kasumi-1 cells and cells treated for the indicated time periods with TSA (200 nM), as described in Materials and Methods. The amount of DNA bound to the acetylated histones was quantified by performing 30 cycles of genomic PCR with primers LZM 1 and LZM 2, resulting in amplicon a-a (5⬘ upstream region of the LZM gene, top) or primers LZM 3 and LZM 4 (resulting in amplicon a-b, LZM exon 4, bottom). Agarose gel electrophoresis was performed, followed by visualization of bands under UV transillumination. Protein input quantification was carried out for indicated time points using Bio-Rad protein quantification kit.


7

Fig. 5. Conditional AML1/ETO expression in U-937 cells results in repression of LZM transcription. (A) Gene map of the human LZM gene locus, with perfect AML1/RUNX1 binding sites (according to the consensus sequence TGT/CGGT) indicated. Arrows indicate size and orientation of Alu repeats. (B) AML1/ETO knock-down by siRNA in Kasumi-1 cells results in derepression of LZM mRNA. Briefly, RNA was extracted after single and repeat transfections by electroporation [22] and analyzed by Northern blot analysis, as described above. Minus sign, mock transfection; siAGF1, active antisense RNA against AML1/ETO; siAGF6, mismatch control. (C) 9/14/18 cells (U-937 cells with conditional AML1/ETO expression) were treated with ponasterone (5 ␮M) for 48 h (to induce AML1/ETO) in the absence or presence of TSA (200 nM), Northern blot analysis was performed for LZM and GAPDH. Numeric values represent the relative intensity of the different LZM bands equalized for GAPDH expression.

cells (⬎80% methylation) to ⬃700 bp upstream of the transcriptional start site [21]. Localization of this site (situated in a relatively CpG-poor area of the lysozyme promoter region) was now both confirmed and validated on an extended range of myeloid cells by Southern blot analysis and by methylationsensitive restriction/PCR, and by methylation analysis extended to neighboring cytosines by bisulfite sequencing. For a second CpG dinucleotide, shown in our previous study to be demethylated only in terminally differentiated normal phagocytic cells (highest LZM expression), and located within a CpG-rich region in exon 4, this was also confirmed by both techniques. This CpG island, which, similar to the MRD1, Keratin 18, ␣-1 globin, and other genes [36 – 40] is part of an Alu sequence, is fully methylated in different myeloid cell lines of intermediate maturation stage.

Selective LZM gene derepression by two distinct epigenetic modifications By low-cytotoxic concentrations of the demethylating agent 5-azaCdR, demethylation both in the 5⬘ flanking region and the exonic CpG island was achieved in the Kasumi-1 cell line. Demethylation, which is considered a passive mechanism dependent on cell division [41], could be detected already after 24-48 h. In contrast, two different myeloid cell lines, HL-60 and U-937, exhibiting a similar methylation pattern, did not show comparable demethylation upon 5-azaCdR treatment (data not shown). 8


Of note, Kasumi-1 cells (which by other differentiation markers are very similar to HL-60 cells) had very low LZM mRNA levels, and an at least 10-fold up-regulation of LZM mRNA by DNA methyltransferase inhibitor treatment was observed in this cell line but not in HL-60 and U-937. Differentiation markers demonstrated that this derepression of the LZM gene by demethylation was not associated with cellular differentiation. Time course experiments indicated demethylation at least concomitantly with increased mRNA expression, compatible with the assumption that up-regulation of transcription is an effect of demethylation within cis regulatory regions of the LZM gene. A recent study of the IL-2 gene activation in T-cells implies an additional, early and active enzymatic demethylation mechanism [42]. Effects of 5-azaCdR on histone methylation, which precede those on cytosine methylation by at least 48 h, have been described [43]. This activity may also be operative in our model. Low CpG density within the LZM promoter region despite its transcriptional repression in Kasumi-1 cells suggests other repressive mechanisms. Methylcytosine-binding proteins such as MeCP2 promote chromatin condensation, leading to local and even distant (⬎500 bp) repression of transcription by interacting with corepressor complexes that contain histone deacetylase activity [11, 44] and, as more recently demonstrated, histone methyltransferase activity [45]. Functional interactions between methylated DNA and a repressive chromatin conformation mediated by MBD2/3 have also been elucihttp://www.jleukbio.org

dated in Drosophila [46]. Moreover, direct interactions between DNMTs and histone deacetylases were recently described [47– 49]. We therefore treated Kasumi-1 cells with TSA, a powerful and specific inhibitor of HDACs. An almost sevenfold upregulation of LZM mRNA levels was observed in Kasumi-1 but not HL-60 or U-937 cells, which was less striking and delayed compared with 5-azaCdR mediated transcriptional activation. TSA treated Kasumi-1 cells did not show demethylation of DNA, indicating that DNA methylation and histone deacetylation are simultaneously and independently operative in this model, and both contribute to LZM gene regulation in Kasumi-1. Because a synergism between both repressive mechanisms has been described for other genes [26, 50 –52], Kasumi-1 cells were sequentially treated with 5-azaCdR and TSA at various doses and schedules. However, no synergistic effects could be observed.

Complex interplay between DNA methylation and chromatin configuration within the human LZM gene locus To further investigate the dynamic interaction between DNA methylation and chromatin remodeling, MspI protection studies were performed, revealing a contribution of chromatin closing to LZM gene repression: untreated Kasumi-1 exhibited less chromatin opening around the S1 upstream CpG site than HL-60 and U-937 myeloid cells, whereas LZM nonexpressing KG-1 early myeloblastic cells and peripheral blood lymphocytes (both with an almost completely methylated LZM gene) did not disclose accessible chromatin at this site. In Kasumi-1 cells, chromatin accessibility to MspI increased after treatment with 5-azaCdR but not with TSA. These results are compatible with a model of regional CpG methylation, resulting in a closed chromatin configuration that is much less accessible to Msp I or other nucleases [29, 53], by attracting MeCP2 and HDACs. In addition, inaccessible chromatin in both myeloid and nonmyeloid LZM nonexpressing cells (including normal lymphocytes) confirms that restriction of LZM expression to the myeloid lineage only at intermediate and late stages of maturation is mediated by partial demethylation and chromatin opening. An apparent exception to this model, peripheral blood granulocytes (which are completely demethylated at all CpG sites examined) also exhibit a pattern of closed chromatin. This may be attributable to the known high degree of overall chromatin condensation in these cells with low transcriptional activity [54]. Peripheral blood monocytes would probably provide a much better cell type to further test this hypothesis. Because TSA treatment of Kasumi-1 cells changed neither local chromatin configuration (at least as assessed by Msp I protection) nor DNA methylation, chromatin immunoprecipitation of histones H3 and H4 was performed, revealing increased histone acetylation within both the 5⬘ flanking region and the exonic CpG island of the LZM gene, with highest levels attained already after 24 h, that is, before the highest increase of mRNA expression. In Fig. 6, a model of the interplay of methylation and chromatin accessibility is presented: LZM nonexpressor early myeloid KG-1 cells exhibit methylation and inaccessible chromatin of the gene locus, whereas in LZM expressing more mature myeloid cell lines Kasumi-1 and HL60, partial demethylation and an "open" chromatin configura-

Fig. 6. Model of the interactions of epigenetic mechanisms regulating LZM expression in myeloid cell lines. Lollipops represent CpG dinucleotides within SmaI sites evaluated for their methylation status within the 5⬘ region (S1), intron 3 (S2), and the exon 4 CpG island (S3). Detailed regional methylation analyses are only shown for the first two lines (Kasumi-1 -/⫹ 5-azaCdR), bisulfite sequencing was not done for the others. Different sizes of arrows indicate differential local chromatin accessibility. LZM expression is detected by Northern blot analysis. GAPDH and for KG-1 ethidium bromide (*) was used for loading control. Comparable results have been obtained by Nguyen et al. for p14/ARF and p16/INK4a promoters in bladder cancer cell lines [53].

tion within the 5⬘ flanking region can be detected, which in Kasumi-1 are both enhanced by a demethylating agent.

AML1/ETO fusion protein represses the human LZM gene Several leukemia-specific DNA-binding proteins have recently been implicated in chromatin remodeling and DNA methylation. PML/RAR-␣ recruits both HDAC activity and a DNA methyltransferase when binding to its target sequence in the RAR ␤ promoter [55], thereby methylating and thus silencing expression of this gene. The MLL gene specifically targets unmethylated, but not methylated, CpG-rich sequences [56]. Recently, the Myc transcription factor, a mediator of cell growth and differentiation, was shown to act as a transcriptional repressor not only by interfering with transcriptional activators but also by recruiting DNA methyltransferase activity via DNMT3a [57]. The functional link between DNA methylation and transcriptional activity of the LZM gene in AML1/ETO expressor cells differed strikingly from AML1/ETO-negative cells in our studies. Either further demethylation (leading to chromatin opening) or acetylation of histones was necessary to up-regulate the gene in Kasumi-1 and to achieve expression levels comparable to HL-60 cells (AML1/ETO-negative). We have previously hypothesized [31] that AML1/ETO has a role in LZM gene regulation in Kasumi-1 cells, as five perfect AML1 binding sites are present in the LZM 5⬘ flanking region (Fig. 5A). Indeed, we could demonstrate that LZM transcription is repressed by AML1/ETO in a U-937 cell line model with inducible expression of AML1/ETO [27, 31]. Very recently, the group of Heidenreich identified LZM as one of only 11 genes derepressed in Kasumi-1 cells by using both oligonucleotide and cDNA microarrays after AML1/ETO mRNA expression was repressed by siRNAs [32]. This constitutes further proof that Claus et al. Regulation of lysozyme gene expression

9

the human LZM gene is a (direct or indirect) target of AML1/ ETO. AML1/ETO is responsible for the specific phenotype of this leukemia type, by targeting multiple genes, several of which are transcriptionally repressed, such as the C/EBP-␣ transcription factor [58] and the p14 cell cycle inhibitor [59]. Leukemia cells expressing AML1/ETO exhibit a distinct cell morphology, arrested differentiation, increased apoptosis, and distinct gene expression patterns [60]. These features are thus most likely determined by the ability of this DNA binding fusion protein to activate or repress (direct and indirect) target genes [30, 61, 62]. AML1/ETO is involved in both local chromatin remodeling and DNA methylation changes by recruiting histone deacetylases [33] and DNMT 1 [63] to its target genes. However, it is unclear whether it also interacts with proteins maintaining DNA methylation. Furthermore, it is not known whether the ability of AML1/ETO to bind to its specific consensus sites is modified by the DNA methylation status of its target. Indirect evidence that AML1/ETO-dependent repression of LZM transcription is mediated at least partially by deacetylation of histones is provided by the ability of TSA to completely abolish this repression (Fig. 5C). Four AML1 binding sites are clustered within an 800-bp region upstream of the transcription start site and neighboring a differentially methylated region. Fine-mapping of this region, for example, by promoter/reporter studies, in vivo footprinting or ChIP are necessary to further prove a direct interaction of AML1/ETO protein with this upstream region of the LZM gene. Specifically, because one AML1 consensus binding site in the 5⬘ region contains a CpG dinucleotide, the methylation status of this cytosine may influence AML1/ETO binding [64]. However, more experimental evidence is needed to determine which (if any) of the 5 putative AML1 binding sites— or which combination of binding sites—is functionally involved in this repression. In conclusion, we show that DNA methylation at distinct regions of the human LZM gene is functionally linked to its transcriptional repression, which in AML1/ETO positive Kasumi-1 is relieved by pharmacological hypomethylation and, independently, by histone deacetylation. The complex interplay between DNA methylation and local chromatin structure of this CpG-poor gene provides a model for combinations of both derepressive principles (inhibitors of DNMTs and HDACs), which warrant further preclinical and clinical investigation.

ACKNOWLEDGMENTS We wish to thank Mahmoud Abdelkarim for excellent technical assistance, Gang Chen, Maike Buchner and Melanie Feuerstein for help with bisulfite sequencing, Dr. Olaf Heidenreich (Tu¨bingen) for reagents and helpful suggestions, Dr. Milton Werner (Rockefeller University, New York, NY) for critical reading of the manuscript, and Dr. Florian Otto for continued insightful discussions. This work was supported by Deutsche Jose´ Carreras Leuka¨mie-Stiftung e.V., grants R 00/14 and 06/42f. Jesu´s Duque is supported by DAAD-La Caixa Grant A/05/29785, ref. 314. 10


REFERENCES 1. Bird, A. P. (1986) CpG-rich islands and the function of DNA methylation. Nature 321, 209 –213. 2. Futscher, B. W., Oshiro, M. M., Wozniak, R. J., Holtan, N., Hanigan, C. L., Duan, H., Domann, F. E. (2002) Role for DNA methylation in the control of cell type specific maspin expression. Nat. Genet. 31, 175–179. 3. Van der Ploeg, L. H., Flavell, R. A. (1980) DNA methylation in the human gamma delta beta-globin locus in erythroid and nonerythroid tissues. Cell 19, 947–958. 4. Lichtenstein, M., Keini, G., Cedar, H., Bergman, Y. (1994) B cell-specific demethylation: a novel role for the intronic kappa chain enhancer sequence. Cell 76, 913–923. 5. Lu¨bbert, M., Miller, C. W., Koeffler, H. P. (1991) Changes of DNA methylation and chromatin structure in the human myeloperoxidase gene during myeloid differentiation. Blood 78, 345–356. 6. Felgner, J., Kreipe, H., Heidorn, K., Jaquet, K., Heuss, R., Zschunke, F., Radzun, H. J., Parwaresch, M. R. (1992) Lineage-specific methylation of the c-fms gene in blood cells and macrophages. Leukemia 6, 420 – 425. 7. Baylin, S. B., Herman, J. G., Graff, J. R., Vertino, P. M., Issa, J. P. (1998) Alterations in DNA methylation: a fundamental aspect of neoplasia. Adv. Cancer Res. 72, 141–196. 8. Drexler, H. G. (1998) Review of alterations of the cyclin-dependent kinase inhibitor INK4 family genes p15, p16, p18 and p19 in human leukemialymphoma cells. Leukemia 12, 845– 859. 9. Nguyen, C., Liang, G., Nguyen, T. T., Tsao-Wei, D., Groshen, S., Lu¨bbert, M., Zhou, J. H., Benedict, W. F., Jones, P. A. (2001) Susceptibility of nonpromoter CpG islands to de novo methylation in normal and neoplastic cells. J. Natl. Cancer Inst. 93, 1465–1472. 10. Bird, A. P., Wolffe, A. P. (1999) Methylation-induced repression– belts, braces, and chromatin. Cell 99, 451– 454. 11. Jones, P. L., Veenstra, G. J., Wade, P. A., Vermaak, D., Kass, S. U., Landsberger, N., Strouboulis, J., Wolffe, A. P. (1998) Methylated DNA and MeCP2 recruit histone deacetylase to repress transcription. Nat. Genet. 19, 187–191. 12. Wade, P. A., Gegonne, A., Jones, P. L., Ballestar, E., Aubry, F., Wolffe, A. P. (1999) Mi-2 complex couples DNA methylation to chromatin remodelling and histone deacetylation. Nat. Genet. 23, 62– 66. 13. Jolles, P., Schoentgen, F., Jolles, J., Dobson, D. E., Prager, E. M., Wilson, A. C. (1984) Stomach lysozymes of ruminants. II. Amino acid sequence of cow lysozyme 2 and immunological comparisons with other lysozymes. J. Biol. Chem. 259, 11617–11625. 14. Gordon, S., Todd, J., Cohn, Z. A. (1974) In vitro synthesis and secretion of lysozyme by mononuclear phagocytes. J. Exp. Med. 139, 1228 –1248. 15. Peters, C. W., Kruse, U., Pollwein, R., Grzeschik, K. H., Sippel, A. E. (1989) The human lysozyme gene. Sequence organization and chromosomal localization. Eur. J. Biochem. 182, 507–516. 16. Clarke, S., Greaves, D. R., Chung, L. P., Tree, P., Gordon, S. (1996) The human lysozyme promoter directs reporter gene expression to activated myelomonocytic cells in transgenic mice. Proc. Natl. Acad. Sci. USA 93, 1434 –1438. 17. Ammerpohl, O., Schmitz, A., Steinmiller, CL., Renkawitz, R. (1998) Repression of the mouse M-lysozyme gene involves both hindrance of enhancer factor binding to the methylated enhancer and histone deacetylation. Nucleic Acids Res 26, 5256-5260. 18. Lu¨bbert, M., Brugger, W., Mertelsmann, R., Kanz, L. (1996) Developmental regulation of myeloid gene expression and demethylation during ex vivo culture of peripheral blood progenitor cells. Blood 87, 447– 455. 19. Leculier, C., Couprie, N., Francina, A., Archimbaud, E., Adeleine, P., Treille, D., Denoyel, G., Fiere, D. (1992) Specific detection of monocytic lysozyme within normal and leukemic cells. Blood 79, 760 –764. 20. Lu¨bbert, M., Mertelsmann, R., Herrmann, F. (1997) Cytosine methylation changes during normal hematopoiesis and in acute myeloid leukemia. Leukemia 11, Suppl 1, S12–S18. 21. Lu¨bbert, M., Henschler, R., Kreutz, M., Andreesen, R., Mertelsmann, R., Herrmann, F. (1997) The human lysozyme gene undergoes stepwise demethylation during phagocyte maturation. Leukemia 11, 990 –997. 22. Heidenreich, O., Krauter, J., Riehle, H., Hadwiger, P., John, M., Heil, G., Vornlocher, H. P., Nordheim, A. (2003) AML1/MTG8 oncogene suppression by small interfering RNAs supports myeloid differentiation of t(8; 21)-positive leukemic cells. Blood 101, 3157–3163. 23. Chung, L. P., Keshav, S., Gordon, S. (1988) Cloning the human lysozyme cDNA: inverted Alu repeat in the mRNA and in situ hybridization for macrophages and Paneth cells. Proc. Natl. Acad. Sci. USA 85, 6227– 6231. 24. Chomczynski, P., Sacchi, N. (1987) Single-step method of RNA isolation by acid guanidinium thiocyanate-phenol-chloroform extraction. Anal. Biochem. 162, 156 –159.

http://www.jleukbio.org

25. Clark, S. J., Harrison, J., Paul, C. L., Frommer, M. (1994) High sensitivity mapping of methylated cytosines. Nucleic Acids Res. 22, 2990 –2997. 26. Cameron, E. E., Bachman, K. E., Myohanen, S., Herman, J. G., Baylin, S. B. (1999) Synergy of demethylation and histone deacetylase inhibition in the re-expression of genes silenced in cancer. Nat. Genet. 21, 103–107. 27. Fliegauf, M., Stock, M., Berg, T., Lu¨bbert, M. (2004) Williams-Beuren syndrome critical region-5/non-T-cell activation linker: a novel target gene of AML1/ETO. Oncogene 23, 9070 –9081. 28. Keshet, I., Lieman-Hurwitz, J., Cedar, H. (1986) DNA methylation affects the formation of active chromatin. Cell 44, 535–543. 29. Antequera, F., Boyes, J., Bird, A. (1990) High levels of de novo methylation and altered chromatin structure at CpG islands in cell lines. Cell 62, 503–514. 30. Burel, S. A., Harakawa, N., Zhou, L., Pabst, T., Tenen, D. G., Zhang, D. E. (2001) Dichotomy of AML1-ETO functions: growth arrest versus block of differentiation. Mol. Cell. Biol. 21, 5577–5590. 31. Fliegauf, M., Burger, J., Claus, R., Stock, M., Otto, F., Lu¨bbert, M. (2003) Inducible AML1/ETO expression in U-937 myeloid cells: a model to identify genes targeted by the leukemia-specific fusion protein. In: Acute Leukemias IX, Basic Research, Experimental Approaches and Novel Therapies. Hematology and Blood Transfusion. (Hiddemann, W., Bu¨chner, T., Ritter, J., Unterhalt, M., and Haferlach, T., eds), pp. 28-33, Springer Verlag, Berlin. 32. Dunne, J., Cullmann, C., Ritter, M., Soria, N.M., Drescher, B., Debernardi, S., Skoulakis, S., Hartmann, O., Krause, M., Krauter, J., Neubauer, A., Young, B.D., Heidenreich, O. (2006) siRNA-mediated AML1/MTG8 depletion affects differentiation and proliferation-associated gene expression in t(8;21)-positive cell lines and primary AML blasts. Oncogene. 33. Lutterbach, B., Westendorf, J. J., Linggi, B., Patten, A., Moniwa, M., Davie, J. R., Huynh, K. D., Bardwell, V. J., Lavinsky, R. M., Rosenfeld, M. G., et al. (1998) ETO, a target of t(8;21) in acute leukemia, interacts with the N-CoR and mSin3 corepressors. Mol. Cell. Biol. 18, 7176 –7184. 34. Follows, G. A., Tagoh, H., Richards, S. J., Melnik, S., Dickinson, H., de Wynter, E., Lefevre, P., Morgan, G. J., Bonifer, C. (2005) c-FMS chromatin structure and expression in normal and leukaemic myelopoiesis. Oncogene 24, 3643–3651. 35. Wong, E. T., Jenne, D. E., Zimmer, M., Porter, S. D., Gilks, C. B. (1999) Changes in chromatin organization at the neutrophil elastase locus associated with myeloid cell differentiation. Blood 94, 3730 –3736. 36. Fryxell, K. B., McGee, S. B., Simoneaux, D. K., Willman, C. L., Cornwell, M. M. (1999) Methylation analysis of the human multidrug resistance 1 gene in normal and leukemic hematopoietic cells. Leukemia 13, 910 – 917. 37. Thorey, I. S., Cecena, G., Reynolds, W., Oshima, R. G. (1993) Alu sequence involvement in transcriptional insulation of the keratin 18 gene in transgenic mice. Mol. Cell. Biol. 13, 6742– 6751. 38. Kochanek, S., Renz, D., Doerfler, W. (1995) Transcriptional silencing of human Alu sequences and inhibition of protein binding in the box B regulatory elements by 5⬘-CG-3⬘ methylation. FEBS Lett. 360, 115–120. 39. Zrihan-Licht, S., Weiss, M., Keydar, I., Wreschner, D. H. (1995) DNA methylation status of the MUC1 gene coding for a breast-cancer-associated protein. Int. J. Cancer 62, 245–251. 40. Liang, G., Chan, M. F., Tomigahara, Y., Tsai, Y. C., Gonzales, F. A., Li, E., Laird, P. W., Jones, P. A. (2002) Cooperativity between DNA methyltransferases in the maintenance methylation of repetitive elements. Mol. Cell. Biol. 22, 480 – 491. 41. Wolffe, A. P., Jones, P. L., Wade, P. A. (1999) DNA demethylation. Proc. Natl. Acad. Sci. USA 96, 5894 –5896. 42. Bruniquel, D., Schwartz, R. H. (2003) Selective, stable demethylation of the interleukin-2 gene enhances transcription by an active process. Nat. Immunol. 4, 235–240. 43. Nguyen, C. T., Weisenberger, D. J., Velicescu, M., Gonzales, F. A., Lin, J. C., Liang, G., Jones, P. A. (2002) Histone H3-lysine 9 methylation is associated with aberrant gene silencing in cancer cells and is rapidly reversed by 5-aza-2'-deoxycytidine. Cancer Res. 62, 6456 – 6461. 44. Kaludov, N. K., Wolffe, A. P. (2000) MeCP2 driven transcriptional repression in vitro: selectivity for methylated DNA, action at a distance and contacts with the basal transcription machinery. Nucleic Acids Res. 28, 1921–1928. 45. Fuks, F., Hurd, P. J., Wolf, D., Nan, X., Bird, A. P., Kouzarides, T. (2003) The methyl-CpG-binding protein MeCP2 links DNA methylation to histone methylation. J. Biol. Chem. 278, 4035– 4040.

46. Marhold, J., Kramer, K., Kremmer, E., Lyko, F. (2004) The Drosophila MBD2/3 protein mediates interactions between the MI-2 chromatin complex and CpT/A-methylated DNA. Development 131, 6033– 6039. 47. Bachman, K. E., Rountree, M. R., Baylin, S. B. (2001) Dnmt3a and Dnmt3b are transcriptional repressors that exhibit unique localization properties to heterochromatin. J. Biol. Chem. 276, 32282–32287. 48. Fuks, F., Burgers, W. A., Godin, N., Kasai, M., Kouzarides, T. (2001) Dnmt3a binds deacetylases and is recruited by a sequence-specific repressor to silence transcription. EMBO J. 20, 2536 –2544. 49. Rountree, M. R., Bachman, K. E., Baylin, S. B. (2000) DNMT1 binds HDAC2 and a new co-repressor, DMAP1, to form a complex at replication foci. Nat. Genet. 25, 269 –277. 50. Ghoshal, K., Datta, J., Majumder, S., Bai, S., Dong, X., Parthun, M., Jacob, S. T. (2002) Inhibitors of histone deacetylase and DNA methyltransferase synergistically activate the methylated metallothionein I promoter by activating the transcription factor MTF-1 and forming an open chromatin structure. Mol. Cell. Biol. 22, 8302– 8319. 51. Klisovic, M. I., Maghraby, E. A., Parthun, M. R., Guimond, M., Sklenar, A. R., Whitman, S. P., Chan, K. K., Murphy, T., Anon, J., Archer, K. J., et al. (2003) Depsipeptide (FR 901228) promotes histone acetylation, gene transcription, apoptosis and its activity is enhanced by DNA methyltransferase inhibitors in AML1/ETO-positive leukemic cells. Leukemia 17, 350 –358. 52. Lu, Q., Ray, D., Gutsch, D., Richardson, B. (2002) Effect of DNA methylation and chromatin structure on ITGAL expression. Blood 99, 4503– 4508. 53. Nguyen, C. T., Gonzales, F. A., Jones, P. A. (2001) Altered chromatin structure associated with methylation-induced gene silencing in cancer cells: correlation of accessibility, methylation, MeCP2 binding and acetylation. Nucleic Acids Res. 29, 4598 – 4606. 54. Szabo, G., Jr., Damjanovich, S., Sumegi, J., Klein, G. (1987) Overall changes in chromatin sensitivity to DNase I during differentiation. Exp. Cell Res. 169, 158 –168. 55. Di Croce, L., Raker, V. A., Corsaro, M., Fazi, F., Fanelli, M., Faretta, M., Fuks, F., Lo Coco, F., Kouzarides, T., Nervi, C., et al. (2002) Methyltransferase recruitment and DNA hypermethylation of target promoters by an oncogenic transcription factor. Science 295, 1079 –1082. 56. Birke, M., Schreiner, S., Garcia-Cuellar, M. P., Mahr, K., Titgemeyer, F., Slany, R. K. (2002) The MT domain of the proto-oncoprotein MLL binds to CpG-containing DNA and discriminates against methylation. Nucleic Acids Res. 30, 958 –965. 57. Brenner, C., Deplus, R., Didelot, C., Loriot, A., Vire, E., De Smet, C., Gutierrez, A., Danovi, D., Bernard, D., Boon, T., et al. (2005) Myc represses transcription through recruitment of DNA methyltransferase corepressor. EMBO J. 24, 336 –346. 58. Pabst, T., Mueller, B. U., Harakawa, N., Schoch, C., Haferlach, T., Behre, G., Hiddemann, W., Zhang, D. E., Tenen, D. G. (2001) AML1-ETO downregulates the granulocytic differentiation factor C/EBPalpha in t(8; 21) myeloid leukemia. Nat. Med. 7, 444 – 451. 59. Linggi, B., Muller-Tidow, C., van de Locht, L., Hu, M., Nip, J., Serve, H., Berdel, W. E., van der Reijden, B., Quelle, D. E., Rowley, J. D., et al. (2002) The t(8;21) fusion protein, AML1 ETO, specifically represses the transcription of the p14(ARF) tumor suppressor in acute myeloid leukemia. Nat. Med. 8, 743–750. 60. Nucifora, G., Rowley, J. D. (1994) The AML1 and ETO genes in acute myeloid leukemia with a t(8;21). Leuk. Lymphoma 14, 353–362. 61. Mulloy, J. C., Cammenga, J., MacKenzie, K. L., Berguido, F. J., Moore, M. A., Nimer, S. D. (2002) The AML1-ETO fusion protein promotes the expansion of human hematopoietic stem cells. Blood 99, 15–23. 62. Lasa, A., Carnicer, M. J., Aventin, A., Estivill, C., Brunet, S., Sierra, J., Nomdedeu, J. F. (2004) MEIS 1 expression is downregulated through promoter hypermethylation in AML1-ETO acute myeloid leukemias. Leukemia 18, 1231–1237. 63. Liu, S., Shen, T., Huynh, L., Klisovic, M. I., Rush, L. J., Ford, J. L., Yu, J., Becknell, B., Li, Y., Liu, C., et al. (2005) Interplay of RUNX1/MTG8 and DNA methyltransferase 1 in acute myeloid leukemia. Cancer Res. 65, 1277–1284. 64. Tate, P. H., Bird, A. P. (1993) Effects of DNA methylation on DNAbinding proteins and gene expression. Curr. Opin. Genet. Dev. 3, 226 – 231.


11