MOLECULAR AND CELLULAR BIOLOGY, Oct. 2008, p. 5912–5923 Vol. 28, No. 190270-7306/08/$08.00�0 doi:10.1128/MCB.00467-08Copyright © 2008, American Society for Microbiology. All Rights Reserved.

MBD3, a Component of the NuRD Complex, Facilitates ChromatinAlteration and Deposition of Epigenetic Marks�

Lluis Morey,1† Carmen Brenner,2 Francesco Fazi,3 Raffaella Villa,1 Arantxa Gutierrez,1 Marcus Buschbeck,1Clara Nervi,3 Saverio Minucci,4 Francois Fuks,2 and Luciano Di Croce1,5*

Centre de Regulacio Genomica, Universitat Pompeu Fabra, Barcelona, Spain1; Free University of Brussels, Brussels, Belgium2;San Raffaele Bio-Medical Park Foundation, Rome, Italy3; European Institute of Oncology, Milan, Italy4; and

ICREA and Centre de Regulacio Genomica, Barcelona, Spain5

Received 20 March 2008/Returned for modification 19 April 2008/Accepted 13 July 2008

In plants, as in mammals, mutations in SNF2-like DNA helicases/ATPases were shown to affect not onlychromatin structure but also global methylation patterns, suggesting a potential functional link betweenchromatin structure and epigenetic marks. The SNF2-like ATPase containing nucleosome remodeling anddeacetylase corepressor complex (NuRD) is involved in gene transcriptional repression and chromatin remod-eling. We have previously shown that the leukemogenic protein PML-RARa represses target genes throughrecruitment of DNA methytransferases and Polycomb complex. Here, we demonstrate a direct role of the NuRDcomplex in aberrant gene repression and transmission of epigenetic repressive marks in acute promyelocyticleukemia (APL). We show that PML-RARa binds and recruits NuRD to target genes, including to thetumor-suppressor gene RAR�2. In turn, the NuRD complex facilitates Polycomb binding and histone meth-ylation at lysine 27. Retinoic acid treatment, which is often used for patients at the early phase of the disease,reduced the promoter occupancy of the NuRD complex. Knockdown of the NuRD complex in leukemic cells notonly prevented histone deacetylation and chromatin compaction but also impaired DNA and histone methyl-ation, as well as stable silencing, thus favoring cellular differentiation. These results unveil an important rolefor NuRD in the establishment of altered epigenetic marks in APL, demonstrating an essential link betweenchromatin structure and epigenetics in leukemogenesis that could be exploited for therapeutic intervention.

The product of the chimeric gene generated by the 15;17chromosome translocation, PML-RARa, is a well-character-ized oncogenic transcription factor found in the majority ofhuman acute promyelocytic leukemias (APL) (8). Ectopic ex-pression of the fusion protein in hematopoietic precursor cellsblocks differentiation and promotes leukemia development(26). The oncogenic potential of PML-RARa is based on theaberrant silencing of genes, including several tumor suppressorgenes. PML-RARa, like the wild-type form of retinoic acidreceptor (RAR), represses transcription of target genesthrough binding to so-called RA responsive elements (RARE)and subsequent recruitment of corepressors such as histonedeacetylases (15, 22). In contrast to the wild-type RARa, thefusion protein PML-RARa is rendered insensitive to physio-logical concentrations of RA that would usually trigger tran-scriptional activation and therefore functions as a constitutiveand potent transcriptional repressor of RARE-containing pro-moters. However, pharmacological doses of RA (1 �M), whichare used for patients at the early phase of the disease, can leadto partial derepression of PML-RARa target genes (27). Wehave demonstrated that the transcriptional repression of PML-RARa target genes is further reinforced by recruitment of Poly-comb complex (32) and DNA methytransferases (DNMTs) (9,31). Once established, the PML-RARa-induced epigenetic mod-

ifications and chromatin changes are stable and maintainedthroughout cell divisions.

Recently, an interesting functional link between chromatinstructure and DNA methylation has been proposed (1): inplants, as in mammals, mutations in SNF2-like DNA helicases/ATPases were shown to affect not only chromatin structure butalso global methylation patterns.

In the past few years, several DNA helicase/ATPase-con-taining complexes have been characterized that facilitate orrepress transcription by utilizing the energy of ATP to alter thehistone-DNA interface within the nucleosome structure (35).Among these, the nucleosome remodeling and deacetylasecorepressor complex (NuRD) contains at least seven polypep-tides, including histone deacetylase 1 (HDAC1) and HDAC2,H4 interacting proteins (RbAp46/48), methyl-binding protein3 (MBD3), MTA-family members (MTA1 to MTA3) (12), andan SNF2-like chromatin-remodeling ATPase (Mi-2/CHD4) (7,29, 34, 37, 39). Genetic and molecular data suggest that MBD3is important for the integrity and stability of the NuRD com-plex (16), and it is implicated in the regulation of mouse em-bryonic stem cell pluripotency and self-renewal (19). The re-cruitment of the NuRD complex to DNA can occur throughinteraction with MBD2 (11) or with several sequence-specificDNA-binding proteins (12, 21, 23, 28).

We demonstrate here that NuRD plays an important role inthe hematopoietic differentiation block induced by PML-RARa expression. We show that PML-RARa binds and re-cruits NuRD to target genes, which in turn leads to chromatincompaction. Furthermore, binding of NuRD at target genesallows recruitment of the Polycomb repressive complex 2

* Corresponding author. Mailing address: Centre de RegulacioGenomica, Universitat Pompeu Fabra, Barcelona, Spain. Phone: 34-9331-60132. Fax: 34-93 3160 099. E-mail: luciano.dicroce@crg.es.

† Present address: Centre for Epigenetics and BRIC, University ofCopenhagen, Ole Maaløes Vej 5, 2200 Copenhagen, Denmark.

� Published ahead of print on 21 July 2008.

5912

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/m

cb o

n 12

Oct

ober

202

1 by

5.1

94.2

14.5

2.

(PRC2) and DNMT3a, with consequent promoter silencing.Retinoic acid (RA) treatment, which is often used for patientsat the early phase of the disease, reduced the occupancy of theNuRD complex at target genes. Together, these results unveila novel function for the NuRD complex in leukemogenesis,and establish a link between NuRD activity and epigeneticalterations in cancer.

MATERIALS AND METHODS

Plasmids. The FlaghMBD3a vector was kindly provided by F. Ishikawa (KyotoUniversity), pGEX-5X-3 mMBD3a and pGEX-5X-3 mMBD3b, antibodiesagainst MBD3 were kindly provided by D. Reinberg (University of Medicine andDentistry of New Jersey), pCLneo-mHDAC1-Myc and pME18s-Flag-mHDAC2were kindly provided by C. Seiser (University of Vienna), and pGEX-MTA1/2/3and pcDNA-Flag-MTA1/2/3 were kindly provided by W. M. Yang (NationalCheng Hsing University).

HA mMBD3b was subcloned from pGEX-5X-3 mMBD3b to pcDNA3 bydigestion with EcoRV and XhoI enzymes.

Cell lines and transfection. NB4 cells and U937-PR9 were cultured at 37°Cand 5% of CO2 in RPMI medium supplemented with 10% of fetal bovine serum.HEK 293T were cultured at 37°C and 5% of CO2 in Dulbecco modified Eaglemedium supplemented with 10% of fetal bovine serum and then transfected byusing the calcium phosphate coprecipitation method.

Immunoprecipitation and chromatin immunoprecipitation. For immunopre-cipitations, antibodies were coupled to protein A-Sepharose beads. Cell extractswere prepared in lysis buffer (50 mM HEPES [pH 7.5], 150 mM NaCl, 1 mMEDTA, 2.5 mM EGTA, 0.1% Tween 20, 1 mM phenylmethylsulfonyl fluoride,0.4 U of aprotinin and leupeptin/ml) and incubated with beads overnight at 4°C.Beads were washed three times with lysis buffer complemented with additional150 mM NaCl and 0.2% NP-40. Bound proteins were eluted with 2� Laemmlisample buffer and subjected to sodium dodecyl sulfate-polyacrylamide gel elec-trophoresis. Input lanes show 10% of the lysates used for precipitation. Theextracts were analyzed by Western blotting with antibodies specific for PML(PGM3), RARa (sc551), Mi2 (sc11378), MBD3 (sc9402), and MTA2 (sc28731)(Santa Cruz); Flag (Sigma); and HDAC1, HDAC2, and tubulin (Abcam).

For chromatin immunoprecipitation (ChIP), NB4 or U937-PR9 cells werecross-linked with 1% formaldehyde (Sigma) at 37°C for 10 min. Cells were rinsedtwice with ice-cold phosphate-buffered saline and collected. ChIPs were per-formed and analyzed as described previously (4) with antibodies specific foreither HDAC1 (Abcam), RNA polymerase II (Convence), p300, CBP, acetylatedH3, H3K27me3 (Upstate), MTA2, PML, HDAC2, and BRG1 (Santa Cruz). Theimmunoprecipitated DNA was quantified by real-time quantitative PCR (RocheLightCycler). The sequences of the PCR primers are available upon request.

Bisulfite genomic sequencing and DNA methyltransferase assay. Bisulfitegenomic sequencing was performed as described previously (32). Immunopre-cipitates were assayed for methyltransferase activity in a 100-�l reaction con-taining a 33-bp hemimethylated oligonucleotide substrate (500 ng), S-adenosyl-L-[methyl-3H]methionine (2 �l; 77 Ci/mmol), Tris-HCl (50 mM, pH 7.5), EDTA(5 mM), 50% glycerol, dithiothreitol (5 mM), and protease inhibitors. Afterincubation at 37°C for 1 h, unincorporated nuclides were removed by usingBiospin chromatography columns (Bio-Rad), and the incorporation of radioac-tivity was determined by liquid scintillation counting.

Immunoprecipitation in the presence of ethidium bromide. A total of 8 � 107

of cells was resuspended in 2 ml of lysis buffer (50 mM HEPES [pH 7.5], 150 mMNaCl, 1 mM EDTA, 2.5 mM EGTA, 0.1% Tween 20) with proteinase inhibitorsand then sonicated (Branson sonicator) at 10% output for 15 s. Then, 100 �g ofethidium bromide/ml was added to 1 ml of sonicated material, followed byincubation 30 min at 4°C. After 15 min of centrifugation at 14,000 rpm, thesupernatant was collected and used for PML-RAR immunoprecipitation over-night. Immunoprecipitated material was then washed three times with lysis buffersupplemented with 300 mM NaCl, 100 �g of ethidium bromide/ml, and protein-ase inhibitors.

RNA interference and retroviral infection. The target sequence used to silenceMBD3 was inserted as a short hairpin into the pRetroSuper (pRS) retroviralvector according to the manufacturer’s recommendations (OligoEngine).pLKO.1-Puro lentiviral vectors with the short hairpin target sequence to silenceMTA2 and Mi-2 were purchased from Sigma. pRS-Suz12 vector was kindlyprovided by K. Helin.

Portions (8 �g) of pRS-shMBD3, pRS-shSuz12, and pRS-scramble vectorswere transfected with 4 �g of the pVSVG plasmid by using the calcium phos-

phate precipitation method into HEK-293T/GP2 cells. After 2 and 3 days, su-pernatants containing the retroviruses were collected. Human U937-PR9 or NB4leukemic cells were spin infected for 90 min (3,200 rpm) in the presence ofprotamine sulfate (5 �g/ml) and viruses containing the shMBD3, shSuz12, orshScramble. After 36 h, infected cells were selected with puromycin (2 �g/ml forU937-PR9 or 1 �g/ml for NB4) for at least 72 h.

Next, 8-�g portions of pLKO.1-Puro-MTA2, pLKO.1-Puro-Mi-2, andpLKO.1-Puro-Random vectors were transfected with 4 �g of the pVSVG plas-mid and 5 �g of the p8.91 plasmid by using the calcium phosphate precipitationmethod into HEK-293T/GP2 cells. After 2 and 3 days, supernatants containinglentiviruses were collected. U937-PR9 cells were spin infected for 90 min (3,200rpm) in the presence of Polybrene (10 �g/ml), and viruses containing shMTA2,shMi-2, or shScramble. After 36 h, infected cells were selected with puromycin (2�g/ml for U937) for at least 72 h.

For rescue of phenotype experiments, MBD3 from the pcDNA-Flag-MBD3plasmid was mutated in bp 10 of the MBD3 short hairpin binding site by usinga QuikChange site-directed mutagenesis kit (Stratagene) according to the man-ufacturer’s recommendations. Mutated MBD3 (MBD3-rescue) was cloned intothe retroviral pMSCV2.2-green fluorescent protein (GFP) plasmid.

Then, 8 �g of pMSCV2.2-MBD3-T369A-GFP and pMSCV2.2-GFP vectorswas transfected with 4 �g of the pVSVG plasmid by the calcium phosphateprecipitation method into HEK-293T/GP2 cells. After 2 and 3 days, supernatantscontaining retroviruses were collected. U937-PR9 RNAi-scramble and U937-PR9 RNAi-MBD3 cells were spin infected for 90 min (3,200 rpm) in the pres-ence of protamine sulfate (5 �g/ml) and the virus containing the MBD3-rescue-GFP. After 36 h, the cells were GFP sorted.

Differentiation assays. The NBT assay was performed using a commerciallyavailable nitroblue tetrazolium (NBT; Sigma). A 0.2-ml portion of cell suspen-sion at a density of 2 � 105 cells in RPMI–5% fetal bovine serum was mixed with0.2 ml of filtered 0.2% NBT solution and 3 �l of TPA 1 �M, followed byincubation for 30 min at 37°C. Subsequently, cytocentrifuge slides were prepared(200 rpm, 4 min). NBT-positive cells were determined by scoring 500 cells undera light microscope. Slides were also stained 1 min with modified Wright-Giemsastain (Sigma), rinsed in phosphate-buffered saline, stained 10 min in modifiedGiemsa stain (Sigma) 1:20 with water (Sigma), rinsed with water, and subjectedto morphological examination under a light microscopy.

RNA purification and reverse transcription-PCR analysis. RNA from U937-PR9 MBD3 RNA interference cells (RNAi MBD3) or from U937-PR9 controlcells (RNAi control) after no treatment, after RA treatment (1 nM, 36 h), andafter RA treatment (1 nM, 12 h) with subsequent Zn induction (100 mM Zn,24 h) was extracted by using a Qiagen RNeasy minikit, retrotranscripted (avianmyeloblastosis virus; Roche), and assayed for the expression of RAR�2 by usingquantitative real-time PCR (Roche LightCycler). The sequences of the PCRprimers are available upon request.

RESULTS

NuRD knockdown in leukemic cells. It has been proposedthat chromatin remodeling machineries might contribute tothe establishment and/or maintenance of epigenetic marks. Weand others have previously demonstrated that PML-RARaimposes an altered pattern of DNA methylation (9) and lysine27 methylation (32) at its target genes. We hypothesized thatNuRD could be implicated in PML-RARa-induced gene si-lencing since (i) one of the HDACs enzymes present in thecomplex has been demonstrated to associate with the PMLmoiety of PML-RARa (36); (ii) NuRD complex is mainlyinvolved in gene repression, while other chromatin remodelingcomplexes (e.g., SWI/SNF) are either exclusively involved ingene activation or function in both activation and repression;and (iii) NuRD has been implicated in hormone receptor-mediated gene silencing (37). We thus decided to investigatewhether the NuRD complex would be involved in the mainte-nance of the cancerous phenotype in fully established patient-derived NB4 leukemic cells. It has been previously shown thatreduction of MBD3 protein levels leads to the disassembly ofthe NuRD complex concomitantly with decreased intracellularlevels of some of other subunits (16, 19). Therefore, in order to

VOL. 28, 2008 NuRD-DEPENDENT EPIGENETIC MARK DEPOSITION 5913

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/m

cb o

n 12

Oct

ober

202

1 by

5.1

94.2

14.5

2.

FIG. 1. NuRD knockdown in APL facilitates differentiation. (A) Western blots analysis of mock and MBD3 interference NB4 cells. Equalamounts of cell extracts from mock (RNAi-Scr) and RNAi cells (RNAi-MBD3) were blotted with the indicated antibodies. (B) Differentiation of

5914 MOREY ET AL. MOL. CELL. BIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/m

cb o

n 12

Oct

ober

202

1 by

5.1

94.2

14.5

2.

ablate NuRD activity, we generated a stable MBD3 knock-down in NB4 cells (RNAi MBD3), using a retroviral vector-based shRNA approach, with strongly reduced expression ofMBD3 (more than 70%, Fig. 1A). Under this experimentalcondition, the levels of MTA2 proteins were also reduced,while the levels of HDAC enzymes were not affected (Fig. 1Aand data not shown). Importantly, coimmunoprecipitationanalysis confirmed that the NuRD complex was not formed, asalso previously described (16, 19; data not shown). Next, weassessed differentiation based on morphological and functionalcriteria. NB4 RNAi control and MBD3 RNAi cells werestained with Wright-Giemsa (Fig. 1B, left panel). As shown,MBD3 RNAi cells displayed a more mature morphology, asobserved by several features: (i) they contained a reducednucleus/cytoplasm ratio, (ii) the dark blue-gray cytoplasm be-came lighter and often contained granules, and (iii) they fre-quently displayed an irregular nucleus shape. Furthermore, theNBT reduction test, which measures functional differentiation,was performed using standard methodology. The percentageof cells containing intracellular-reduced black formazan depos-its was determined. While we observed only a modest amountof spontaneous differentiated cells in control samples, we ob-served a significantly increased number of NBT-positive cellsin RNAi MBD3 NB4 (Fig. 1B, right panel). Together, theseresults show that MBD3 protein is required to fully maintainthe undifferentiated and leukemic phenotype of NB4 cells.

Epigenetic changes in NuRD knockdown cells. We havepreviously demonstrated that the leukemic potential of PML-RARa relies—at least in part—on its ability to induce epige-netic alterations at target genes (17), which include DNAmethylation (9) and trimethylation of lysine 27 of histone H3(H3K27me3) (32), which is catalyzed by the Polycomb repres-sive complex 2 (PRC2). Thus, based on the results describedabove, we hypothesized that MBD3 knockdown could affectDNA and histone methylation at PML-RARa-repressed genes.Among several PML-RARa target genes, RARb2 has been char-acterized in great detail (reference 10 and reference therein). Thegenomic bisulfite analysis confirmed that in control cells theRARb2 promoter is hypermethylated to a similar extent as pre-viously reported, while in cells lacking functional NuRD complexthe levels of CpG methylation were significantly reduced (Fig.1C). In parallel with the previous analysis, we next performedChIP experiments for the H3K27me3. A three- to fourfold re-duction in H3K27me3 was observed in MBD3 knockdown cellscompared to the mock control (RNAi control) (Fig. 1D). Simi-larly, promoter methylation was significantly reduced in RNAiMTA2 (Fig. 1E and data not shown), while an increased numberof NBT-positive cells were observed in RNAi MTA2 NB4 com-pared to control cells (Fig. 1F).

As a further control for the specificity of our results, we

generated a new stable MBD3 knockdown in acute myeloidleukemia (AML) HL60 cells (Fig. 1G), which are derived froman AML patient and are negative for PML-RAR� but positivefor wild-type RAR�. We thus assessed differentiation based onmorphological and functional criteria. No differences betweenHL60 control RNAi and MBD3 RNAi cells were observed fordifferentiation-associated antigens, morphology, or NBT re-duction (Fig. 1H and data not shown). We then analyzed bybisulfite genomic sequencing the methylation status of theRAR�2 promoter in HL60 cells with or without MBD3. Thelevel of methylated CpGs in control HL60 cells is similar tothat observed in NB4 cells; however, knockdown of MBD3does not cause any decrease of DNA methylation in HL60 cells(Fig. 1I), in contrast to the 70% reduction observed in NB4MBD3 RNAi (Fig. 1G).

These results revealed an unexpected role of the NuRDcomplex in the metabolism of epigenetic marks in leukemiccells and prompted us to further investigate not only whetherNuRD is present at target promoters but also whether it par-ticipates in initiating epigenetic alterations in PML-RARa-expressing cells.

RA reduces NuRD occupancy at PML-RARa target genes. Inorder to explore the possibility of NuRD occupancy at PML-RARa target genes, we performed ChIP experiments in NB4cells. Several NuRD proteins were found associated at theRARE region of the endogenous RARb2 promoter (Fig. 2A)in untreated cells, while a pharmacological dose of RA (1 �M)caused a reduction of the NuRD complex, with a correspond-ing increase of histone acetyltransferases p300 and CBP and ofhistone H3 acetylation levels (Fig. 2A). Similarly, the promoteroccupancy of Brahma-related gene 1 (BRG1), a subunit of theSWI/SNF chromatin remodeling complex, as well as polymer-ase II, was increased upon RA treatment, resulting in pro-moter reactivation (Fig. 2B). Importantly, similar analysis wasperformed at the pS2 promoter, a well-characterized estrogen-induced gene that is insensitive to RA stimulation (25, 30).None of the NuRD subunit was found associated at the pS2promoter (data not shown). To strengthen our observations,we expanded our analysis to other PML-RARa target genes(24). Among these, we characterized the hPSCD4 and hNFE2promoters based on the presence of several RARE elementswithin the regulative promoter regions. ChIP experiments con-firmed that, similar to RARb2, promoters of these other targetgenes showed a similar reduction of MTA2 and concomitantlyan increase of H3ac upon RA administration (Fig. 2C).

Targeting NuRD complex at PML-RARa target genes. Wehypothesized that PML-RARa recruits the NuRD complex toits target genes. We performed ChIP experiments in the he-matopoietic precursor cells U937-PR9. In contrast to NB4,which constitutively express PML-RARa, U937-PR9 cells con-

mock and RNAi-MBD3 NB4 cells into granulocytes. Cells were stained with Wright-Giemsa and analyzed for the morphology under the lightmicroscopy. Differentiation was evaluated by NBT reduction assay. (C) NuRD knockdown affects the DNA methylation levels at the RAR�2 gene.DNA extracted from mock (RNAi-Scr) and RNAi-MBD3 NB4 cells was used for bisulfite genomic sequencing. The methylation status of eachCpG dinucleotide in each sequenced cloned is depicted by a red square if the position was methylated or a white square if it was not. (D) NuRDknockdown affects H3 methylation at lysine 27 (H3K27me3). ChIP assays were performed in mock (RNAi-Scr) and RNAi-MBD3 NB4 cells asdescribed in Fig. 2A. The promoter of RAR�2 was amplified by real-time PCR. Errors bars indicate the standard deviation obtained from threeindependent experiments. (E and F) Western blots and differentiation (NBT) analysis of mock and MTA2 interference NB4 cells. (G and I)Western blots, differentiation (NBT), and DNA methylation analysis of mock and MBD3 interference HL60 cells.

VOL. 28, 2008 NuRD-DEPENDENT EPIGENETIC MARK DEPOSITION 5915

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/m

cb o

n 12

Oct

ober

202

1 by

5.1

94.2

14.5

2.

FIG. 2. Binding of PML-RAR� and NuRD complex to the endogenous RAR�2 promoter is RA sensitive. (A) In NB4 cells, the NuRD complex isbound to the RAR�2 promoter and it is release after 8 h of RA (1 �M). NB4 cells were subjected to ChIP analysis as indicated in the figure. The RAR�2promoter was amplified by real-time PCR. Errors bars indicate the standard deviation obtained from three independent experiments. (B) EndogenousRAR�2 expression is upregulated after 8 h of RA (1 �M) treatment. Semiquantitative RT-PCR of cDNA from 1 �g of RNA. Tubulin expression levelswere used as a PCR control. (C) The NuRD complex is released from NFE2 and PSCD4 promoters after pharmacological RA treatment. A ChIP assaywas performed as in panel A. Errors bars indicate the standard deviation obtained from three independent experiments. (D) PML-RAR� recruits theNuRD complex to the RAR�2 promoter. U937-PR9 cells, treated sequentially with RA (1 nM) to activate endogenous RARs and then with 100 �MZn (12 h) to induce PML-RAR� expression, were subjected to ChIP analysis, as indicated in the figure. The RAR�2 promoter was amplified by real-timePCR. Errors bars indicate the standard deviation obtained from three independent experiments.

5916

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/m

cb o

n 12

Oct

ober

202

1 by

5.1

94.2

14.5

2.

tain a zinc-inducible promoter controlling PML-RARa expres-sion, making them an ideal cellular system for studying theearly epigenetic events associated with the oncogene expres-sion. PML-RARa binding to RARb2 promoter was restrictedto the promoter region harboring the RARE (Fig. 2D) (31).Upon expression of PML-RARa, HDAC2 and MTA2 wereenriched at endogenous target genes, while the occupancy ofhistone acetyltransferase enzyme p300 and histone H3 acety-lation was reduced (Fig. 2D and 5B).

Increased occupancy of NuRD complex at target genescould be due to its direct targeting. We thus investigatedwhether PML-RARa interacts with the NuRD remodelingcomplex. We found that both MBD3a and MBD3b isoforms

associated with the oncoprotein when co-overexpressed in293T cells (Fig. 3A). We then extended our analysis to othercomponents of the NuRD complex. PML-RARa interactedwith MTA2 and to a lesser extent with MTA1 and MTA3. Finally,both HDAC1 and HDAC2 associated with PML-RARa (datanot shown).

To investigate whether PML-RARa associates with physio-logical levels of the NuRD complex, we performed coimmu-noprecipitation experiments using lysates from the two best-characterized APL model systems: patient-derived NB4 andU937-PR9 cells. We consistently found that in the same coim-munoprecipitated material, PML-RARa associated with all ofthe NuRD components, including MTA2, Mi-2, and HDAC1/2

FIG. 3. Endogenous interaction between PML-RAR� and NuRD subunits in leukemic cells. (A) 293T cells were transfected as indicated inthe figure, and extracts were immunoprecipitated with anti-PML-RAR� or anti-HA antibodies. Western blots of input lysate and immunopre-cipitates were analyzed by using antisera against RAR� or HA. (B) Interaction of NuRD subunits and PML-RAR� in U937-PR9 cells uponPML-RAR� induction (100 �M Zn for 14 h). Cell extracts from U937-PR9 cells were immunoprecipitated using PGM3, anti-MTA2, anti-HDAC1,or anti-HDAC2 antibodies. Immunocomplexes were detected by Western blotting with antibodies as indicated in the figure. (C) Interaction ofNuRD subunits and PML-RAR� in the human APL-derived NB4 leukemic cells. Cell extracts of NB4 cells, untreated or after RA (1 �M)treatment for 45 min, were immunoprecipitated with PGM3 antibody, and immunocomplexes were detected by Western blots as in panel B.(D) Interaction of NuRD and PML-RAR� is DNA independent. Cell extracts of NB4 cells, untreated or incubated in the presence of ethidiumbromide (100 �g/ml) for 30 min, were immunoprecipitated using PGM3 or control antibodies, and immunocomplexes were detected by Westernblots as in panel B.

VOL. 28, 2008 NuRD-DEPENDENT EPIGENETIC MARK DEPOSITION 5917

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/m

cb o

n 12

Oct

ober

202

1 by

5.1

94.2

14.5

2.

FIG. 4. The NuRD complex contributes to RARb2 promoter epigenetic silencing. (A) NuRD complex is necessary for the complete repressionof the RAR�2 gene induced by PML-RAR�. U937-PR9 cells were treated with physiological concentration of RA (1 nM). mRNA levels ofendogenous RAR�2 gene were measured by real-time PCR. The results were normalized against �-actin mRNA levels. (B) Western blots analysisof mock (RNAi-Scr) and MBD3 interference U937-PR9 cells (RNAi-MBD3). Equal amounts of cell extract from mock and RNAi-MBD3 cellswere blotted with the indicated antibodies. (C) NuRD facilitates DNMT3a binding at RAR�2 genes. ChIP assays were performed in mock

5918 MOREY ET AL. MOL. CELL. BIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/m

cb o

n 12

Oct

ober

202

1 by

5.1

94.2

14.5

2.

(Fig. 3B and C). The interaction resisted the presence ofethidium bromide in the precipitation reaction (Fig. 3D), aswell as a high-salt extraction method (data not shown), thusexcluding the possibility of a DNA-mediated protein associa-tion, suggesting a direct interaction between PML-RARa andthe NuRD complex. RA induces structural change in PML-RARa that dissociates all of the corepressors binding andincreases the affinity for the binding of coactivators (31). In-terestingly, the interaction between PML-RARa and NuRDwas dramatically reduced in the presence of pharmacologicaldoses of RA (Fig. 3C).

MBD3 is required for epigenetic silencing of PML-RARatarget promoters. We next wanted to investigate whetherMBD3 is required for PML-RARa-mediated gene silencing.First, we tested the ability of PML-RARa to downregulate itstarget gene RARb2 in U937-PR9 cells when stimulated withphysiological concentration of RA (1 nM) (Fig. 4A, lane 3versus lane 7). We then generated a stable MBD3 knockdownin U937-PR9 cells (RNAi MBD3). A reduction of �75% ofMBD3 protein was achieved under these conditions (Fig. 4B).Similarly to NB4 cells, in U937-PR9 cells the knockdown ofMBD3 led to a reduction of intracellular MTA2, but it did notaffect HDAC1/2 and DNMTs (Fig. 4B and data not shown). Incontrast to control cells, in RNAi MBD3 cells the RARb2promoter was not efficiently silenced by PML-RARa (Fig. 4A,lanes 3 and 4 versus lanes 7 and 8).

We and others have previously demonstrated that PML-RARa and other transcription factors can repress transcriptionby recruitment of DNMTs (e.g., STAT5, AML1-ETO, andMyc) (2, 38), leading to promoter hypermethylation. We nextanalyzed the occupancy of DNMTs at RARb2 promoter inboth RNAi control cells and in RNAi MBD3 cells. As shown inFig. 4C, the accumulation of DNMT3a, which was observedin the presence of PML-RARa, was significantly reduced inRNAi MBD3 cells. We did not observe any variation in DNMT1occupancy. Previously, DNMT3a has been reported to be able tobind MBD3 (5). In addition, we found that HDAC1 and theNuRD-specific subunit MTA2 bind to in vitro-translatedDNMT3a (data not shown). These results prompted us to inves-tigate the status of methylation of the RARb2 promoter in de-pendence of PML-RARa and NuRD. Thus, we performed bisul-fite genomic sequencing analysis in the same experimental settingas described above. As previously reported, the expression ofPML-RARa caused RARb2 promoter hypermethylation in wild-type or mock-infected leukemic cells (Fig. 4D and data notshown). Interestingly, in MBD3 knockdown cells the methylationof CpGs was reduced by 50%. Of note, the protein levels ofDNMT3a and DNMT1, as well as the global methyltransferaseactivity, were not altered by RNAi MBD3 (data not shown).

We next investigated whether Mi2, the ATPase subunit ofthe NuRD complex, is necessary for DNA methylation of

PML-RAR�-target genes. We thus generated a stable Mi2knockdown in U937-PR9 cells (RNAi-Mi2; Fig. 4E) and per-formed a bisulfite genomic sequencing of RAR�2 promoterusing the RNAi-Scr and RNAi-Mi2 cell lines. We observedthat in Mi2 knockdown cells, PML-RAR-induced DNA meth-ylation was reduced by more than 50% (Fig. 4D). Similarly,promoter methylation was significantly reduced in RNAiMTA2 (Fig. 4D and F), corroborating our conclusion that theNuRD complex and its associated enzymatic activities play animportant role in PML-RARa-induced DNA methylation.

Role of the NuRD complex in the cross talk between DNAand histone methylation. The results described above suggestthat the NuRD complex is required for PML-RARa-inducedde novo DNA methylation. We recently demonstrated that thepresence of DNA methylation and Polycomb are both neces-sary for maintenance of the epigenetic alterations in leukemiccells that constitutively express PML-RARa (32). We thusdecided to investigate how epigenetic silencing is initiated fol-lowing the expression of the oncoprotein PML-RARa. Forthis, we analyzed not only the kinetics of NuRD, DNMTs, andPolycomb recruitment but also their interdependence for theestablishment of the epigenetic repressive marks at promoters.

Time course ChIP analysis performed in U937-PR9 cellsindicated that the NuRD complex is efficiently loaded atRARb2 promoter after only 8 h of expression of PML-RARa,as measured by the presence of the diagnostic subunit MTA2(Fig. 5A). Similar analyses suggested that EZH2 and DNMT3are found significantly associated at the promoter region atlater time points, with DNMT3 being strongly recruited at48 h. Interestingly, in cells lacking active NuRD complex,PML-RARa-mediated EZH2 recruitment and H3K27me3were severely compromised (Fig. 5B), while the global H3acetylation level was higher.

The specificity of our observations was further investigatedby reintroducing a MBD3 silent mutant (3), which is insensitiveto the shRNAs (Fig. 5C), into MBD3 knockdown cells. To thisend, we generated stable cell lines wherein the MBD3 rescuevector (MBD3-rescue) or a control vector was infected intoRNAi-MBD3 cells. GFP-positive infected cells were thensorted, and MBD3 protein levels were analyzed by Westernblotting (Fig. 5D, left panel). Interestingly, the interaction be-tween MTA2 and HDAC1 was restored in MBD3-rescuedcells (Fig. 5D, right panel). In addition, aberrant H3K27 meth-ylation and gene silencing was also observed upon PML-RARaexpression (Fig. 5E and F). We conclude that the observedeffects of the shRNAi-MBD3 are due to specific effects on theintended target and cannot be explained by off-target effects.Similarly, knockdown of the Mi2 or MTA2 subunit of theNuRD complex (Fig. 4E and F) impaired deposition of theH3K27 methylation marks (Fig. 5G and H). We next exploredwhether knockdown of the Polycomb repressive complex 2

(RNAi-Scr) and RNAi-MBD3 U937-PR9 cells as described in Fig. 2A. The promoter of RAR�2 was amplified by real-time PCR. Errors barsindicate the standard deviation obtained from three independent experiments. (D) NuRD knockdown reduces DNA methylation of RAR�2 gene.DNA extracted from mock (RNAi-Scr), RNAi-MBD3, RNAi-Mi2, and RNAi-MTA2 U937-PR9 cells untreated or treated with 100 �M Zn for24 h was used for bisulfite genomic sequencing. The methylation status of each CpG dinucleotide in each sequenced cloned is depicted by a redsquare if the position was methylated or a white square if it was not. (E and F) Western blots analysis of mock, Mi2, and MTA2 interferenceU937-PR9 cells. Equal amounts of cell extracts from mock (RNAi-Scr) and RNAi cells were blotted with the indicated antibodies.

VOL. 28, 2008 NuRD-DEPENDENT EPIGENETIC MARK DEPOSITION 5919

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/m

cb o

n 12

Oct

ober

202

1 by

5.1

94.2

14.5

2.

FIG. 5. DNA and histone methylation dependency in APL leukemic cells. (A) Kinetics of NuRD, EZH2, and Dnmt3a/b recruitment to RAR�2promoter. U937-PR9 cells, treated with 100 �M Zn for 8, 24, and 48 h to induce PML-RAR� expression, were subjected to ChIP analysis asindicated in the figure. The RAR�2 promoter was amplified by real-time PCR. Errors bars indicate the standard deviation obtained from threeindependent experiments. (B) NuRD complex is required for PML-RAR�-mediated Polycomb recruitment and H3K27me3 at RAR�2 promoter.ChIP assays were performed in mock (RNAi-Scr) and RNAi-MBD3 U937-PR9 cells, as described in Fig. 2A. The promoter of RAR�2 was

5920 MOREY ET AL. MOL. CELL. BIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/m

cb o

n 12

Oct

ober

202

1 by

5.1

94.2

14.5

2.

(PRC2) would also affect NuRD occupancy and gain of DNAmethylation at RARb2 promoter. We generated a stableSUZ12 knockdown U937-PR9 cell line (RNAi SUZ12). Areduction of �70% of SUZ12 protein was achieved under

these conditions compared to the mock knockdown cells(RNAi control) (Fig. 6A), and a corresponding decrease inbulk histone H3 tri-methyl K27 level was also observed. ChIPanalysis for MTA2 indicates that the PRC2 complex is not

amplified by real-time PCR. Errors bars indicate the standard deviation obtained from three independent experiments. (C) Schematic representation ofthe silent mutation in MBD3. (D) A silent mutant of MBD3 rescues interaction between NuRD components. Equal amounts of cell extract from mock,MBD3 interference (RNAi-MBD3), and MBD3-rescue U937-PR9 cells were blotted with the indicated antibodies (D, left panel) or were immunopre-cipitated using HDAC1 or control antibodies. The presence of MTA2 in the immunocomplexes was detected by Western blots (D, right panel). (E) Asilent mutant of MBD3 rescues from MBD3 knockdown phenotype. mRNA levels of endogenous RAR�2 gene were measured by real-time PCR. Theresults were normalized against �-actin mRNA levels. (F to H) ChIP analysis was performed with the antibodies indicated in the figure. The RAR�2promoter was amplified by real-time PCR. Errors bars indicate the standard deviation obtained from three independent experiments.

FIG. 6. NuRD occupancy at RAR�2 promoter is independent of PRC2 complex. (A) Western blot analysis of mock and Suz12 interferenceU937-PR9 cells. Equal amounts of cell extract from mock and RNAi-MBD3 cells were blotted with the indicated antibodies (left panel). ChIPanalysis was performed using antibodies indicated in the figure. The RAR�2 promoter was amplified by real-time PCR. Errors bars indicate thestandard deviation obtained from three independent experiments (right panel). (B) Polycomb knockdown reduces DNA methylation of theRAR�2 gene. DNA extracted from mock (RNAi-Scr) and RNAi-Suz12 U937-PR9 cells untreated or treated with 100 �M Zn for 24 h was usedfor bisulfite genomic sequencing. The methylation status of each CpG dinucleotide in each sequenced cloned is depicted by a red square if theposition was methylated or a white square if it was not.

VOL. 28, 2008 NuRD-DEPENDENT EPIGENETIC MARK DEPOSITION 5921

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/m

cb o

n 12

Oct

ober

202

1 by

5.1

94.2

14.5

2.

required for PML-RARa-mediated NuRD recruitment, al-though EZH2 binding and H3K27 trimethylation were foundto be reduced. Interestingly, in cells lacking PRC2 complex,DNA methylation fails to accumulate at the RARb2 promoterupon PML-RARa expression, while in RNAi control cells,PML-RARa caused promoter hypermethylation (Fig. 6B). To-gether, these experiments suggest that the NuRD complex isrequired for the establishment of DNA and histone methyl-ation at lysine 27, while PRC2 likely plays a role in sustainingactive DNA hypermethylation (Fig. 7).

DISCUSSION

The intimate link between NuRD and DNA methylation waspostulated since the identification of MBD2 and MBD3 withinthe MeCP1/NuRD complex. It was later demonstrated that,although MBD2 is not an integral subunit of NuRD, it asso-ciates and recruits the NuRD complex to methylated DNA.More recently, Feng and Zhang provided additional mecha-nistic insight on the MBD2-dependent binding of NuRD tomethylated nucleosomes (11). Our data now reveal a novelfunction of NuRD in this scenario. We found that the estab-lishment of DNA methylation patterns at PML-RARa targetgenes is altered in cells lacking the NuRD complex. Based on

this finding we speculate that the remodeling activity of NuRDmay facilitate access of DNMTs to chromatin template fordeposition of methyl groups at CpG sites. Alternatively, acontinuous binding of NuRD to silenced promoters could berequired for the maintenance (and perpetuation during cellcycle) of methylated CpGs. It would be interesting to investi-gate in the future whether DNA methylation levels of othergenes, which are not PML-RARa direct targets, are also af-fected in cells lacking NuRD complex and more in general howchromatin structure influences DNA methylation pattern.Consistent with this hypothesis, the chromatin remodelingSNF2-like protein DDM1 was shown to be essential for fullmethylation of the Arabidopsis thaliana genome (33). In mam-mals, mutations in mouse Lsh genes (6) and human ATRX (13,14), both of which encode relatives of the chromatin-remod-eling protein SNF2, have significant effects on global DNAmethylation patterns. Thus, both global and targeted DNAmethylation might require a dedicated chromatin remodelingmachinery. The aforementioned link between NuRD andMBDs, as well as the interaction between NuRD and DNMTsdocumented here, identify NuRD as an important player inDNA methylation metabolism in tumor cells and likely in nor-mal cells.

We have previously shown that PML-RARa recruitsHDAC3 to its target genes (31). The interaction with NuRDnot only leads to chromatin remodeling and changes inDNA methylation but also causes a recruitment of two ad-ditional HDAC enzymes, which display differences in sub-strate specificity with respect to HDAC3 (18). This mightexpand the ability of PML-RARa to remove acetyl groupsfrom histone tails, thus preparing them for further modifi-cations. We have recently documented increased epigeneticmarks (such as H3K27me) upon PML-RARa expression(32). The data presented here suggest that NuRD recruit-ment at target genes is required not only for PML-RARa-induced DNA methylation but also for deposition of Poly-comb repressive marks. Interestingly, genetic analyses inDrosophila have demonstrated that the ATPase dMi-2 par-ticipates in Polycomb repression (20), thus anticipating adecade ago a role of NuRD in histone deacetylation andchromatin changes that in turn allow binding of Polycombprotein complexes. In leukemic cells, PML-RARa recruitsboth NuRD and Polycomb complexes to its target genes (32)and coordinates enzymatic activities to ensure a stable epi-genetic gene silencing. Our kinetic analysis of corepressoroccupancy suggests a precise sequence of events that occurat target genes, with NuRD-mediated chromatin remodelingbeing necessary for further deposition of epigenetic repres-sive marks. Moreover, knockdown of PRC2 complex preventseither de novo hypermethylation or its stability. We suggest thatPML-RARa represses gene transcription through several distinctmechanisms, including histone deacetylation, DNA methylation,histone modification, chromatin compaction, and heterochroma-tinization. The data presented here demonstrate that the NuRDcomplex plays a pivotal role in both the establishment and main-tenance of aberrant epigenetic silencing imposed by PML-RARaand will help in identifying potential molecular targets of inter-vention in cancer.

FIG. 7. Model of promoter repression in APL leukemia. The on-coprotein PML-RAR� binds to a well-defined DNA sequence andrecruits NuRD complex, which in turn allow occupancy of Polycombcomplex and DNMTs. The activity of these epigenetic modifier en-zymes leads to modifications of histone tails, DNA methylation, andtranscriptional silencing.

5922 MOREY ET AL. MOL. CELL. BIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/m

cb o

n 12

Oct

ober

202

1 by

5.1

94.2

14.5

2.

ACKNOWLEDGMENTS

We gratefully acknowledge D. Reinberg, M. Brackertz, C. Seiser,and W. M. Yang for providing vectors. We thank the members of DiCroce’s laboratory and V. Raker, G. Vicent, and R. Shiekhattar forhelpful discussions and for critically reading the manuscript.

This study was supported by grants from the Spanish-MEC(BFU2007-63059) and Consolider “Marato TV3” to L.D.C. M.B. issupported by a fellowship from the Deutsche Forschungsgesellschaft.

REFERENCES

1. Bourc’his, D., and T. H. Bestor. 2002. Helicase homologues maintain cyto-sine methylation in plants and mammals. Bioessays 24:297–299.

2. Brenner, C., R. Deplus, C. Didelot, D. Danovi, P. G. Pelicci, B. Amati, T.Kouzarides, Y. de Launoit, L. Di Croce, and F. Fuks. 2005. Myc repressestranscription through recruitment of DNA methyltransferase corepressor.EMBO J. 24:336–346.

3. Brummelkamp, T. R., R. Bernards, and R. Agami. 2002. A system for stableexpression of short interfering RNAs in mammalian cells. Science 296:550–553.

4. Buschbeck, M., I. Uribesalgo, A. Ledl, A. Gutierrez, S. Minucci, S. Muller,and L. Di Croce. 2007. PML4 induces differentiation by Myc destabilization.Oncogene 26:3415–3422.

5. Datta, J., S. Majumder, S. Bai, K. Ghoshal, H. Kutay, D. S. Smith, J. W.Crabb, and S. T. Jacob. 2005. Physical and functional interaction of DNAmethyltransferase 3A with Mbd3 and Brg1 in mouse lymphosarcoma cells.Cancer Res. 65:10891–10900.

6. Dennis, K., T. Fan, T. Geiman, Q. Yan, and K. Muegge. 2001. Lsh, a memberof the SNF2 family, is required for genome-wide methylation. Genes Dev.15:2940–2944.

7. Denslow, S. A., and P. A. Wade. 2007. The human Mi-2/NuRD complex andgene regulation. Oncogene 26:5433–5438.

8. Di Croce, L. 2005. Chromatin modifying activity of leukemia associatedfusion proteins. Hum. Mol. Genet. 14:R77–R84.

9. Di Croce, L., V. A. Raker, M. Corsaro, F. Fazi, M. Fanelli, M. Faretta, F.Fuks, F. Lo Coco, T. Kouzarides, C. Nervi, S. Minucci, and P. G. Pelicci.2002. Methyltransferase recruitment and DNA hypermethylation of targetpromoters by an oncogenic transcription factor. Science 295:1079–1082.

10. Fazi, F., G. Zardo, V. Gelmetti, L. Travaglini, A. Ciolfi, L. Di Croce, A. Rosa,I. Bozzoni, F. Grignani, F. Lo-Coco, P. G. Pelicci, and C. Nervi. 2007.Heterochromatic gene repression of the retinoic acid pathway in acute my-eloid leukemia. Blood 109:4432–4440.

11. Feng, Q., and Y. Zhang. 2001. The MeCP1 complex represses transcriptionthrough preferential binding, remodeling, and deacetylating methylated nu-cleosomes. Genes Dev. 15:827–832.

12. Fujita, N., D. L. Jaye, C. Geigerman, A. Akyildiz, M. R. Mooney, J. M. Boss,and P. A. Wade. 2004. MTA3 and the Mi-2/NuRD complex regulate cell fateduring B lymphocyte differentiation. Cell 119:75–86.

13. Gibbons, R. J., T. L. McDowell, S. Raman, D. M. O’Rourke, D. Garrick, H.Ayyub, and D. R. Higgs. 2000. Mutations in ATRX, encoding a SWI/SNF-like protein, cause diverse changes in the pattern of DNA methylation. Nat.Genet. 24:368–371.

14. Gibbons, R. J., D. J. Picketts, L. Villard, and D. R. Higgs. 1995. Mutationsin a putative global transcriptional regulator cause X-linked mental retarda-tion with alpha-thalassemia (ATR-X syndrome). Cell 80:837–845.

15. Grignani, F., S. De Matteis, C. Nervi, L. Tomassoni, V. Gelmetti, M. Cioce,M. Fanelli, M. Ruthardt, F. F. Ferrara, I. Zamir, C. Seiser, M. A. Lazar, S.Minucci, and P. G. Pelicci. 1998. Fusion proteins of the retinoic acid recep-tor-alpha recruit histone deacetylase in promyelocytic leukaemia. Nature391:815–818.

16. Hendrich, B., J. Guy, B. Ramsahoye, V. A. Wilson, and A. Bird. 2001. Closelyrelated proteins MBD2 and MBD3 play distinctive but interacting roles inmouse development. Genes Dev. 15:710–723.

17. Hormaeche, I., and J. D. Licht. 2007. Chromatin modulation by oncogenictranscription factors: new complexity, new therapeutic targets. Cancer Cell11:475–478.

18. Johnson, C. A., D. A. White, J. S. Lavender, L. P. O’Neill, and B. M. Turner.2002. Human class I histone deacetylase complexes show enhanced catalyticactivity in the presence of ATP and coimmunoprecipitate with the ATP-dependent chaperone protein Hsp70. J. Biol. Chem. 277:9590–9597.

19. Kaji, K., I. M. Caballero, R. MacLeod, J. Nichols, V. A. Wilson, and B.

Hendrich. 2006. The NuRD component Mbd3 is required for pluripotency ofembryonic stem cells. Nat. Cell Biol. 8:285–292.

20. Kehle, J., D. Beuchle, S. Treuheit, B. Christen, J. A. Kennison, M. Bienz, andJ. Muller. 1998. dMi-2, a hunchback-interacting protein that functions inpolycomb repression. Science 282:1897–1900.

21. Kim, J., S. Sif, B. Jones, A. Jackson, J. Koipally, E. Heller, S. Winandy, A.Viel, A. Sawyer, T. Ikeda, R. Kingston, and K. Georgopoulos. 1999. IkarosDNA-binding proteins direct formation of chromatin remodeling complexesin lymphocytes. Immunity 10:345–355.

22. Lin, R. J., L. Nagy, S. Inoue, W. Shao, W. H. Miller, Jr., and R. M. Evans.1998. Role of the histone deacetylase complex in acute promyelocytic leu-kaemia. Nature 391:811–814.

23. Luo, J., F. Su, D. Chen, A. Shiloh, and W. Gu. 2000. Deacetylation of p53modulates its effect on cell growth and apoptosis. Nature 408:377–381.

24. Meani, N., S. Minardi, S. Licciulli, V. Gelmetti, F. L. Coco, C. Nervi, P. G.Pelicci, H. Muller, and M. Alcalay. 2005. Molecular signature of retinoic acidtreatment in acute promyelocytic leukemia. Oncogene 24:3358–3368.

25. Metivier, R., G. Penot, M. R. Hubner, G. Reid, H. Brand, M. Kos, and F.Gannon. 2003. Estrogen receptor-alpha directs ordered, cyclical, and com-binatorial recruitment of cofactors on a natural target promoter. Cell 115:751–763.

26. Minucci, S., S. Monestiroli, S. Giavara, S. Ronzoni, F. Marchesi, A. Insinga,D. Diverio, P. Gasparini, M. Capillo, E. Colombo, C. Matteucci, F. Con-tegno, F. Lo-Coco, E. Scanziani, A. Gobbi, and P. G. Pelicci. 2002. PML-RAR induces promyelocytic leukemias with high efficiency following retro-viral gene transfer into purified murine hematopoietic progenitors. Blood100:2989–2995.

27. Pandolfi, P. P. 2001. Oncogenes and tumor suppressors in the molecularpathogenesis of acute promyelocytic leukemia. Hum. Mol. Genet. 10:769–775.

28. Schultz, D. C., J. R. Friedman, and F. J. Rauscher III. 2001. Targetinghistone deacetylase complexes via KRAB-zinc finger proteins: the PHD andbromodomains of KAP-1 form a cooperative unit that recruits a novelisoform of the Mi-2alpha subunit of NuRD. Genes Dev. 15:428–443.

29. Tong, J. K., C. A. Hassig, G. R. Schnitzler, R. E. Kingston, and S. L.Schreiber. 1998. Chromatin deacetylation by an ATP-dependent nucleo-some remodeling complex. Nature 395:917–921.

30. Tora, L., J. White, C. Brou, D. Tasset, N. Webster, E. Scheer, and P.Chambon. 1989. The human estrogen receptor has two independent non-acidic transcriptional activation functions. Cell 59:477–487.

31. Villa, R., L. Morey, V. A. Raker, M. Buschbeck, A. Gutierrez, F. De Santis,M. Corsaro, F. Varas, D. Bossi, S. Minucci, P. G. Pelicci, and L. Di Croce.2006. The methyl-CpG binding protein MBD1 is required for PML-RARafunction. Proc. Natl. Acad. Sci. USA 103:1400–1405.

32. Villa, R., D. Pasini, A. Gutierrez, L. Morey, M. Occhionorelli, E. Vire, J. F.Nomdedeu, T. Jenuwein, P. G. Pelicci, S. Minucci, F. Fuks, K. Helin, and L.Di Croce. 2007. Role of the Polycomb repressive complex 2 in acute promy-elocytic leukemia. Cancer Cell 11:513–525.

33. Vongs, A., T. Kakutani, R. A. Martienssen, and E. J. Richards. 1993. Ara-bidopsis thaliana DNA methylation mutants. Science 260:1926–1928.

34. Wade, P. A., P. L. Jones, D. Vermaak, and A. P. Wolffe. 1998. A multiplesubunit Mi-2 histone deacetylase from Xenopus laevis cofractionates with anassociated Snf2 superfamily ATPase. Curr. Biol. 8:843–846.

35. Workman, J. L., and R. E. Kingston. 1998. Alteration of nucleosome struc-ture as a mechanism of transcriptional regulation. Annu. Rev. Biochem.67:545–579.

36. Wu, W. S., S. Vallian, E. Seto, W. M. Yang, D. Edmondson, S. Roth, andK. S. Chang. 2001. The growth suppressor PML represses transcription byfunctionally and physically interacting with histone deacetylases. Mol. Cell.Biol. 21:2259–2268.

37. Xue, Y., J. Wong, G. T. Moreno, M. K. Young, J. Cote, and W. Wang. 1998.NURD, a novel complex with both ATP-dependent chromatin-remodelingand histone deacetylase activities. Mol. Cell 2:851–861.

38. Zhang, Q., H. Y. Wang, M. Marzec, P. N. Raghunath, T. Nagasawa, andM. A. Wasik. 2005. STAT3- and DNA methyltransferase 1-mediated epige-netic silencing of SHP-1 tyrosine phosphatase tumor suppressor gene inmalignant T lymphocytes. Proc. Natl. Acad. Sci. USA 102:6948–6953.

39. Zhang, Y., G. LeRoy, H. P. Seelig, W. S. Lane, and D. Reinberg. 1998. Thedermatomyositis-specific autoantigen Mi2 is a component of a complex con-taining histone deacetylase and nucleosome remodeling activities. Cell 95:279–289.

VOL. 28, 2008 NuRD-DEPENDENT EPIGENETIC MARK DEPOSITION 5923

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/m

cb o

n 12

Oct

ober

202

1 by

5.1

94.2

14.5

2.