25926.full

Interplay between Chromatin and Trans-actingFactors Regulating the Hoxd4 Promoter duringNeural Differentiation*

Received for publication, March 17, 2006, and in revised form, May 8, 2006 Published, JBC Papers in Press, June 6, 2006, DOI 10.1074/jbc.M602555200

Laila Kobrossy, Mojgan Rastegar1, and Mark Featherstone2

From the McGill Cancer Centre, McGill University, Montreal, Quebec H3G 1Y6 Canada

Correct patterning of the antero-posterior axis of the embry-onic trunk is dependent on spatiotemporally restricted Hoxgene expression. In this study, we identified components of theHoxd4 P1 promoter directing expression in neurally differenti-ating retinoic acid-treated P19 cells. We mapped three nucleo-somes that are subsequently remodeled into an open chromatinstate upon retinoic acid-induced Hoxd4 transcription. Thesenucleosomes spanned the Hoxd4 transcriptional start site inaddition to a GC-rich positive regulatory element located 3� tothe initiation site. We further identified two major cis-actingregulatory elements. An autoregulatory element was shown torecruit HOXD4 and its cofactor PBX1 and to positively regulateHoxd4 expression in differentiating P19 cells. Conversely, thePolycomb group (PcG) protein Ying-Yang 1 (YY1) binds to aninternucleosomal linker and represses Hoxd4 transcriptionbefore and during transcriptional activation. Sequential chro-matin immunoprecipitation studies revealed that the PcG pro-tein MEL18 was co-recruited with YY1 only in undifferentiatedP19 cells, suggesting a role for MEL18 in silencing Hoxd4 tran-scription in undifferentiated P19 cells. This study links for thefirst time local chromatin remodeling events that takeplacedur-ing transcriptional activation with the dynamics of transcrip-tion factor association andDNAaccessibility at aHox regulatoryregion.

Hox gene transcriptional activation marks the onset of anintricate series of events leading to proper embryonic pat-terning in all animals. The products ofHox genes, homeodo-main-containing HOX transcription factors, are essential inspecifying antero-posterior positional identity, hindbraindevelopment, limb formation, and numerous additionalmorphogenetic and organogenetic events (1, 2). Given theircrucial role in embryonic development, the genes encodingHOX proteins are highly conserved throughout the animal

kingdom, and their expression is tightly regulated (3). Inmammals, 39 Hox genes are organized into four clusters,each located on a different chromosome (1). Comparison ofthe clusters reveals 13 possible gene positions, althoughnone of the clusters retains a full complement of 13 genes.Hox genes occupying the same positions are termed paral-ogs, sharing high sequence identity and functional redun-dancy. One can assign a 3� and a 5� end to a cluster since allgenes are transcribed in the same direction. A unique featureof Hox gene clusters is a process termed “colinearity,” corre-lating both the timing of transcriptional activation and theanterior expression borders with the position of a particularHox gene along a cluster (4). Therefore, genes located more3� are expressed earlier and have a more anterior expressionborder than genes located more 5� along the cluster. Thisobservation and several other studies have led to the hypoth-esis that a sequential opening of chromatin, starting at the 3�end of a cluster and moving successively 5�, leads to therelease of silencing, first at the 3� end, and sequentiallyallowing the expression of more 5� genes with increasingtime (5, 6). Numerous studies have now established that it isthe strictly defined anterior expression border that mostdetermines HOX activity, and shifting this border eitheranteriorly or posteriorly leads to embryonic malformationsand homeotic transformations (7). A full understanding ofHox function therefore requires an explication of the mech-anisms governing spatiotemporally restricted expression.The regulation of Hox gene transcription is accomplished

through a set of enhancers located a few kilobases upstream ordownstream of the gene, although some enhancers have beenshown to be located hundreds of kilobases away (8). Transcrip-tion factors involved inHox gene regulation include HOX pro-teins themselves, acting with PBX,3 MEIS, and PREP cofactors(9, 10), Kreisler (11), KROX20 (12), and Sox-Oct family mem-bers (13). One of the most important factors regulating Hoxgene expression is retinoic acid (RA) (14). Functional retinoicacid-response elements (RAREs) have been identified forHoxa1, Hoxb1, Hoxa4, Hoxb4, Hoxd4, Hoxb5, Hoxb6, and

* This work was supported by Grant 49498 from the Canadian Institutes ofHealth Research (to M. F.). The costs of publication of this article weredefrayed in part by the payment of page charges. This article must there-fore be hereby marked “advertisement” in accordance with 18 U.S.C. Sec-tion 1734 solely to indicate this fact.

1 M. R. was supported by a CIHR Cancer Consortium Training Grant FellowshipAward from the McGill Cancer Centre, and by a Conrad F. Harrington Fel-lowship Award from the Faculty of Medicine, McGill University.

2 A Chercheur-National of the Fonds de la Recherche en Sante du Quebec. Towhom correspondence should be addressed: Mark Featherstone, McGillCancer Centre, McGill University, 3655 Promenade Sir William Osler, Mon-treal, QC, Canada H3G 1Y6. Tel.: 514-398-8937; Fax: 514-398-6769; E-mail:[email protected].

3 The abbreviations used are: PBX, pre-B cell transformation-related; RA, reti-noic acid; ARE, autoregulatory element; mARE, mouse ARE; RARE, RA-re-sponse element; PcG, Polycomb group; ChIP, chromatin immunoprecipi-tation; MNase, micrococcal nuclease; LM-PCR, ligation-mediated PCR;�-MEM, �-minimum essential medium; RT, reverse transcription; siRNA,small interfering RNA; EMSA, electrophoretic mobility shift assay; EXD,Extradenticle; CREB, cAMP-response element-binding protein; CBP, CREB-binding protein; MEIS, myeloid ecotropic viral integration site; TFIID, tran-scription factor II D; SWI/SNF, switch/sucrose non-fermenting.

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 281, NO. 36, pp. 25926 –25939, September 8, 2006© 2006 by The American Society for Biochemistry and Molecular Biology, Inc. Printed in the U.S.A.

25926 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 281 • NUMBER 36 • SEPTEMBER 8, 2006

at SY

RA

CU

SE

UN

IV, on January 15, 2013

ww

w.jbc.org

Dow

nloaded from

http://www.jbc.org/

Hoxb8 (15–23). In vivo mutations of the Hoxa1 and Hoxb1RAREs result in hindbrain patterning defects and cranial nervemalformations similar to those observed in the Hoxa1 andHoxb1 full knockouts, emphasizing a key role for retinoids incontrolling Hox gene expression during embryogenesis (17).Transgenic studies in mouse embryos where RARE sequenceswere mutated resulted in posteriorizedHox gene expression inboth the somitic mesoderm and the developing hindbrain (24,25). This, in addition to tissue culture studies (26), suggests apositive role for RA in activating Hox gene transcription.Hoxd4 is an ortholog of the Drosophila Hox gene Deformed

(Dfd). Hoxd4 expression begins at embryonic day 8.25 and hasan anterior border of expression between rhombomeres six andseven (r6/7) in the developing hindbrain and between somitesfour and five in the mesoderm (27). A 5� mesodermal enhancercontaining an RARE and anARE has been described earlier andshown to be functional in P19 cells (26, 28). A 3� neuralenhancer containing a DR5 type RARE is crucial for initiationand maintenance of Hoxd4 expression in the central nervoussystem (22, 29, 30). Two proximal promoters, an upstream pro-moter (P2) and a downstream promoter (P1), have been iden-tified (31). Transcripts originating from P1 have a more ante-rior border of expression in the central nervous system (r6/7)and are further anteriorized in response to RA, suggesting thatP1 is more responsive to signals originating from the 3� neuralenhancer (31).Similar to Hoxb4 (32), interactions between the Hoxd4 3�

neural enhancer and its proximal promoter P1 are important ininitiation of Hoxd4 gene expression in the hindbrain of trans-genic embryos (25) and in neurally differentiating P19 embry-onal carcinoma cells (33). We have correlated this enhancer-promoter interaction with chromatin changes occurring uponHoxd4 gene activation in response to RA in neurally differenti-ating P19 cells and in the central nervous system of developingmouse embryos. Chromatin opening occurred first at the 3�neural enhancer followed by the intervening sequences, culmi-nating at the proximal promoter P1 (33). These studies alsoestablished P19 cells as a valid system for studying Hoxd4enhancer-promoter function.In this study, we further characterized the Hoxd4 P1 pro-

moter. Because of the importance of enhancer-promoter com-munication and chromatin modifications that culminate at P1during neuralHoxd4 gene expression, we investigated the rolesof nucleosome position, chromatin remodeling, and cis-regula-tory elements involved in initiating Hoxd4 gene expression inneurally differentiating P19 cells. We show that nucleosomesare positioned at P1 and are remodeled in response to RA. Fur-thermore, we show that an ARE and a YY1 binding site regulatecorrectHoxd4 expression in P19 cells in response toRA, as doesa GC-rich motif that is essential for core promoter activity. Wealso show that YY1 continues to exert a repressive effect onHoxd4 transcription in P19 cells even after gene activation. Thisrepression is correlated with MEL18 recruitment to the YY1binding site at P1 during Hoxd4 transcriptional silencing.Finally, we discuss our results in the light of the evidence thatlinks chromatin to Hox gene regulation and YY1-mediatedrepression to PcG-mediated silencing of Hoxd4.

MATERIALS AND METHODS

Tissue Culture and Transfections—P19 mouse embryonalcarcinoma cells were cultured in�-minimumessentialmedium(�-MEM) supplemented with 10% fetal bovine serum. For neu-ral differentiation, P19 cells were plated as low density mono-layers at 105 cells/ml�-MEMsupplementedwith 0.3�MRA for48 h. Transient transfection was performed using Lipo-fectamine 2000 reagent (Invitrogen). P19 cells were platedeither in the presence or in the absence of RA followed by trans-fection of different constructs in an antibiotic-free �-MEM.Twenty-four hours later, the medium was replaced with�-MEM containing antibiotics, and RA was added to the cellsundergoing differentiation. Transfected cells were harvested24 h later by scraping in ice-cold phosphate-buffered saline andresuspended in 100 �l of lysis buffer (10% Triton X-100, 1 M

K2HPO4, 1 M KH2PO4, 1 M dithiothreitol) for 5 min.Luciferase Vectors and Assays—Luciferase reporter con-

structswere designed in a pXP2 promoter-less background (34)containing the luciferase gene coupled to region CL of theHoxd4 3� neural enhancer (pXP2CL, see Fig. 2B, construct 1)(22). pXP2CLwas prepared by cloning EcoRV/BamHI-digestedCL fragments (in TOPO-II background) into SmaI/BamHI-di-gested pXP2 followed by sequencing. All Hoxd4 P1 deletionfragmentswere amplified by PCRusing P1-specific primers andPfx platinum (Invitrogen) as the heat-stable DNA polymerase.pSNlacZpA (29) was used as a template, and PCR productswere subcloned into TOPO-II (Invitrogen) and sequenced.BamHI-XhoI promoter fragments were cloned into BglII/XhoIsequentially digested pXP2CL. For luciferase assays, 30 �l ofcell extract was incubated in 100�l of luciferin solution (10mM

luciferin, 1 M Tris pH 7.8) and 12.5 �l of assay buffer (50 mM

ATP, 1 MMgCl2, 1 M Tris pH 7.8). Luciferase activity wasmeas-ured using a Lumat LB 9507 luminometer (EG&G Berthold).For measuring transfection efficiency, Rous sarcoma virus�-galactosidase plasmid was co-transfected with the luciferasereporter constructs, and �-galactosidase assays were per-formed as described (33). Final luciferase values were reportedas relative luciferase activity per �-galactosidase unit.Site-directed Mutagenesis—For site-directed mutagenesis of

the YY1 binding site, primers carrying the mutated sequences(Table 1) were used in two separate PCRs (each containingeither the 5� or the 3� primer containing the mutation), and theamplified products were gel-purified, combined, and used as atemplate for another nested PCR. The products of the final PCRwere also gel-purified, subcloned into TOPO-II, and sequencedto verify mutations. This was followed by BamHI/XhoI diges-tion and cloning into pXP2CL. For mutating the AREsequences, an NruI-BamHI fragment was released frompSXm34/35 already harboring the mutated ARE (28), blunt-ended, and cloned into construct 1 (see Fig. 2B).Micrococcal Nuclease (MNase) Digestion and Genomic DNA

Purification—Nuclei were prepared according to Carey andSmale (35). Undifferentiated or neurally differentiated P19 cellswere resuspended in Nonidet P-40 lysis buffer for 5 min, pel-leted, and resuspended in MNase digestion buffer. This wasfollowed by adding either 1 units or 5 units of MNase (RocheApplied Science) for 5min. The reactionwas stopped, and sam-

Chromatin and Trans-acting Factors Regulate Hoxd4 Transcription

SEPTEMBER 8, 2006 • VOLUME 281 • NUMBER 36 JOURNAL OF BIOLOGICAL CHEMISTRY 25927

at SY

RA

CU

SE

UN


ww

w.jbc.org

Dow

nloaded from

http://www.jbc.org/

ples were treated with 25 �g/�l proteinase K (Roche AppliedScience) overnight at 37 °C. For naked DNA controls, MNasedigestion was carried out following DNA purification of chro-matin extracted from P19 cells. Following phenol/chloroformextraction, samples were treated with 10 �g/�l RNase A for 2 hfollowed by another round of phenol/chloroform extraction.Finally, DNAwas ethanol-precipitated and resuspended in 100�l of H2O.One�g of DNAwas used for ligation-mediated PCR(LM-PCR, see below).Restriction Enzyme Accessibility—Following nuclei purifica-

tion, samples were resuspended in restriction enzyme digestionbuffer (35) followed by incubation with different concentra-tions ofKpnI andPvuII (NewEnglandBiolabs) for either 10minor 20 min at 37 °C. This was followed by proteinase K digestionand DNA purification. After determining the DNA concentra-tion, 1 �g of DNA was used for in vitro EcoRI (New EnglandBiolabs) digestion. Twenty-five percent of the digestion reac-tion was used for LM-PCR.LM-PCR—LM-PCR was performed as described previously

(36), with some modifications. Following DNA purification,digested DNA was ligated to a double-stranded unidirectionallinker, which provided a common 5� sequence for annealing toa PCR primer. This was followed by PCR amplification using a3� gene-specific primer and a 5� primer complementary to theunidirectional linker (Table 1). Finally, a second nested PCRwas performed using a radioactively labeled 3� gene-specificprimer, allowing us to detect specific MNase cleavage productsby autoradiography. Because the PCR templates consist ofDNA fragments linked to a 25-bp linker, the actual size of the

gene-specific product is 25 bp less than the size detected on thegel. All PCR reactions were carried out with the heat-stableDNA polymerase Pfu (Fermentas). First-strand synthesis reac-tions were performed for restriction enzyme-treated DNA butnot forMNase-treatedDNA. Instead,MNase-treatedDNAwasphosphorylated using polynucleotide kinase (Fermentas) fol-lowed directly by ligation with the unidirectional linker (Table1). The amplification PCR consisted of 18 (restriction enzyme)or 23 (MNase) PCR cycles with an extension time starting at 5min plus 15 s for each additional cycle. The labeling PCR con-sisted of three cycles for all the in vivo DNA treatments. Gene-specific primers complementary to Hoxd4 P1 were used, andtheir exact positions are given in Table 1. Real-time PCR usinggapdh-specific oligonucleotides was performed to ensure equalloading.ChIP Assays and Real-time PCR—ChIP experiments were

performed as described (33). Oligonucleotides used for real-time PCR are listed in Table 1. For sequential ChIP, a secondimmunoprecipitation procedurewas performed using chroma-tin samples consisting of antibody-protein-DNA complexesthat were eluted from the agarose beads of the first ChIP. Five�g of purified antibodies was used for each ChIP experiment.Anti-PBX1, -MEL18, and -YY1 antibodies were purchasedfrom Santa Cruz Biotechnology. The anti-HOXD4 rabbit poly-clonal antibody was described previously (33).Electromobility Shift Assays (EMSAs) and Supershifts—Nu-

clear extracts and EMSAs were performed as described (37).Five �g of polyclonal antibody was added to the binding reac-tion in supershift experiments. The primers used are listed in

TABLE 1Oligonucleotides used in this studyOligo, oligonucleotide; Ctrl, control; mYY1, mouse YY1.

LM-PCR 5�-3�Linker oligo LM-PCR 1 GCGGTGACCCGGGAGATCTGAATTCLinker oligo LM-PCR 2 GAATTCAGATCO-115 CTGAATGTCTGCTCTGGGTAGGACCCGAGGO-69 AGACAACGTTACAACCTCGGGTCCO-195 CCGAGCCTACCTGCACCATCTCTGAAAGCC

ChIP 5�-3�Hoxd4 ARE forward TACTCTTCTGTGCTGCTGTCHoxd4 ARE reverse TGCTTCTGCTGCTGCTATGHoxd4 P1 YY1 binding site forward GAACTCATTGCTGTAGCGAGHoxd4 P1 YY1 binding site reverse CAGAGCAGACATTCAGGCgapdh NAD binding domain forward AACGACCCCTTCATTGACgapdh NAD binding domain reverse TCCACGACATACTCAGCAC

EMSA 5�-3�YY1 consensus sense GGGGATCAGGGTCTCCATTTTGAAGCGGGATCTCCCYY1 consensus antisense GGGAGATCCCGCTTCAAAATGGAGACCCTGATCCCCHoxd4 P1 YY1 binding site sense CCTACCTGCACCATCTCTGAAAGCCAGGHoxd4 P1 YY1 binding site antisense CCTGGCTTTCAGAGATGGTGCAGGTAGGOligo A (1–33) ATGGTCGATGCAAAAACTTCATATATCTCCGACOligo B (34–66) ATGGCCAGAGACTGAGGCGCGGAGAGTACTGGCOligo C (1–66) ATGGTCGATGCAAAAACTTCATATATCTCCGACATG

GCCAGAGACTGAGGCGCGGAGAGTACTGGC

RT-PCR 5�-3�Hoxd4 homeobox forward CTACACCAGACAGCAAGTCCHoxd4 homeobox reverse CTATAAGGTCGTCAGGTCCG

Site-directed mutagenesis 5�-3�mYY1 forward ACAACGAGAGTAAAGCCAGGCTTGGTGAGTTCmYY1 reverse GCCTGGCTTTACTCTCGTTGTCAGGTAGGCTCGGTG

CATTTCTAA

siRNA 5�-3�YY1 siRNA sense AAGAUGAUGCUCCAAGAACdTdTYY1 siRNA antisense GUUCUUGGAGCAUCAUCUUdTdTCtrl siRNA sense CCUCGUCGUAGAACCUCCAdTdTCtrl siRNA antisense UGGAGGUUCUACGACGAGGdTdT



at SY

RA

CU

SE

UN


ww

w.jbc.org

Dow

nloaded from

http://www.jbc.org/

Table 1. Anti-Sp1 and anti-Sp3 antibodies were kind gifts of Dr.Christopher Mueller (Queen’s University).Whole Cell Extracts, Nuclear and Cytoplasmic Extracts, and

Immunoblotting—Whole cell extracts from P19 cells andimmunoblotting were performed as described (33). Nuclearand cytoplasmic fractions were also prepared as described (38).RNA Extraction and RT-PCR—RNA extraction and RT-PCR

were performed as described (33). For PCR, primers specific fortheHoxd4 homeobox (located in the second coding exon) wereused to assay for Hoxd4-specific transcripts, in addition togapdh-specific primers (spanning the coding region) that wereused as a control (Table 1).RNA Interference—For silencing yy1 gene expression, siRNA

oligonucleotides based on yy1 cDNA sequences were used asdescribed previously (39) (accession number NM_009537).The YY1 siRNA sequence is identical in the human,mouse, andXenopus cDNAs. Control siRNA oligonucleotides weredesigned as purine to pyrimidine (and vice versa) mutations ofthe YY1 siRNA primers (Table 1). For YY1 knockdown inundifferentiated P19 cells, both control and YY1 siRNA prim-ers having BglII- and HindIII-compatible ends were annealedand cloned into BglII/HindIII-digested pSUPER (OligoEngine,Seattle, WA), creating pSUPER-CTRL and pSUPER-YY1,respectively. Plasmids were transfected into undifferentiatedP19 cells using Lipofectamine 2000, and knockdown wasachieved 24 h later. For P19 cells differentiated with RA, thesame cDNA sequences were used for designing the siRNAprimers, this time delivered to the cells as ready-to-use2�-deprotected double-stranded siRNA duplex. The pSUPER-retro system was not used in these experiments because P19cells had to be pretreated with RA before transfection, optimaltiming of YY1 knockdown did not coincide with initiation ofHoxd4 transcription, and therefore, the effects of YY1depletiononHoxd4 gene expression could not bemonitored. Instead, P19cells were treated with RA for 24 h followed by transient trans-fection with YY1 siRNA oligonucleotides (see Fig. 6, YY1siRNA) or control oligonucleotides (see Fig. 6,Ctrl siRNA). Thecells were subjected to total protein and RNA extraction 24 hlater. Both primer sets were purchased from DharmaconResearch Inc. and transfected in P19 cells at a concentration of200 nM using Lipofectamine 2000. Cells were harvested 48 hlater.

RESULTS

Nucleosome Positioning and Remodeling at the Hoxd4 P1Promoter—In our previous study, we showed that Hoxd4 tran-scription is initiated from the correct start sites within 48 hfollowing RA treatment of P19 cells (33). RA responsivenesswas dependent on the Hoxd4 3� neural enhancer, which con-tains an RARE. Moreover, modifications characteristic of tran-scriptionally active chromatin (histone H3 acetylation andmethylation) were correlated with transcriptional activation,starting at the 3� end of the gene and concluding more 5� at P1.This conversion was accompanied by recruitment of RNA po-lymerase II to the Hoxd4 locus, a process that did not occur inthe absence of RA and not before those chromatin modifica-tions took place. These results suggested that the Hoxd4 P1promoter is not a nucleosome-free region and raised the

question of whether nucleosome positioning plays an activerole in mediating the repression and/or activation of Hoxd4transcription.To investigate nucleosome positioning at P1, we performed

high resolution nucleosome mapping using MNase digestioncoupled with LM-PCR analysis (40). A unique property ofMNase is its ability to create double-stranded nicks in internu-cleosomal regions of partially digested chromatin. To comparenucleosome positions before and after Hoxd4 transcriptionalactivation, we used nuclei extracted from P19 cells eitheruntreated or treated with RA for 48 h and thus coinciding withinitiation of Hoxd4 transcription. Two sets of nested primerswere designed to confirm nucleosome positions at P1, one setamplifying in the 3� direction relative to Hoxd4 transcriptionand the other amplifying in the 5� direction (Fig. 1A). Bothprimers were located upstream of the transcriptional start site(�1) at positions �115 (O-115) and �69 (O-69). Primer posi-tionswere chosen based on results from low resolutionMNase-coupled Southern blotting experiments that gave a rough esti-mation of nucleosome positions (data not shown). Two majorcleavage products could be detected using radiolabeled O-115and nuclei from untreated cells but not the naked DNA control(Fig. 1B, left panel). These sites thus define an internucleosomalregion spanning nucleotides �25 to �10, respectively. Thisindicated that a nucleosome (N2) is positioned at P1 starting at�25 and extending more 5�. Moreover, the lack of significantcleavage products 3� to nucleotide �10 suggested that an addi-tional nucleosome (N3) is positioned at P1 with a 5� border atposition �10 (Fig. 1B, left panel).We then compared cleavage products of RA-treated versus

untreated nuclei. Therewas a significant increase in intensity ofPCR products in RA-treated samples following cleavage at twodifferent MNase concentrations (1 and 5 units) (Fig. 1B, leftpanel). However, the position of the two major cleavage prod-ucts was unchanged, indicating that nucleosome sliding did nottake place following RA treatment (Fig. 1B, left panel), suggest-ing a more relaxed state of chromatin, making DNAmore sus-ceptible to enzymatic digestion.To confirm the position of N2, we designed an additional set

of PCR primers for amplification in the 5� direction (Fig. 1A).Two major cleavage products were detected with O-69 corre-sponding to gene-specific products starting at position�69 andextending until positions�171 and�183, respectively (Fig. 1B,right panel). These products were not detected using nakedDNA. Importantly, these results confirm the position of N2obtained with O-115 (Fig. 1A), fixing a nucleosome unit lengthof 146 bp with borders at positions �25 and �171. This alsosuggests that a third nucleosome (N1) is positioned upstreamofN2 with a 3� border at position �183. PCR products from sam-ples treated with RA were significantly more intense whencompared with untreated samples, suggesting that the latterwere protected by the presence of more compacted chromatin.To ensure that equal amounts of starting material were used,we performed real-time PCR using gapdh-specific oligonucleo-tides, which indicated that comparable DNA levels were pres-ent in the different samples (Fig. 1C). Therefore, these resultssuggest that upon RA treatment, chromatin relaxes at P1, is notaccompanied by nucleosome sliding, and leads to transcrip-



at SY

RA

CU

SE

UN


ww

w.jbc.org

Dow

nloaded from

http://www.jbc.org/



at SY

RA

CU

SE

UN


ww

w.jbc.org

Dow

nloaded from

http://www.jbc.org/

tional initiation at Hoxd4 P1. We note that the relative paucityof MNase I cut sites in naked DNA may suggest that ourapproach has not definitively proven that nucleosomes arepositioned at the Hoxd4 P1 promoter. However, several majorsensitive sites on naked DNA are indeed masked in chromatinpreparations, and the unit nucleosomedistance of 146 bp defin-ing the span of N2 is unlikely to occur by chance.To further investigate chromatin remodeling at P1, we

performed restriction enzyme accessibility assays focusingon N2 (Fig. 1D). To do this, KpnI and PvuII digests werecarried out using chromatin extracted from P19 cells eitheruntreated or treated with RA for 48 h followed by LM-PCRusing the gene-specific primers (O-115 and O-195 coupledwith KpnI and PvuII digests, respectively). Both KpnI andPvuII restriction sites were hypersensitive to increasing con-centrations of enzyme in samples treated with RA asrevealed by the increased intensity of cleavage products (Fig.1D, compare KpnI lanes 1 and 2, 3 and 4, and 5 and 6 andPvuII lanes 7 and 8, 9 and 10, and 11 and 12). These resultsshow that chromatin remodeling at P1 coincides with tran-scriptional initiation, corroborating the role of chromatindecondensation in activating Hoxd4 gene expression.Mapping of Cis-acting Regulatory Elements at P1—Nucleo-

some positioning and modification are coordinated with theplacement of cis-regulatory elements to control transcriptionalstatus. To determine the DNA sequences necessary for correctHoxd4 expression from P1 relative to nucleosome positioning,we constructed luciferase reporters driven by the Hoxd4 P1promoter (�800 to �140) and containing 540 bp of the RA-responsiveHoxd4 neural enhancer region (pXP2CL-Hoxd4P1,construct 3) (Fig. 2, A and B). 5� to 3� sequential deletion con-structs were used in transient transfection assays using bothundifferentiating and neurally differentiating P19 cells (treatedwith RA for 48 h). Control experiments showed that neither the3� neural enhancer nor the P1 promoter sequences alone coulddirect significant reporter activity (Fig. 2B, constructs 1 and 2).Therefore, transcriptional initiation can only be attributed tothe action of the neural enhancer on the P1 promoter.Maximum RA-responsiveness was achieved using construct

3 containing the full 940 bp of P1 sequences. Sequential dele-tion of P1 sequences identified two key regulatory units. P1region �800 to �580 possessed strong activator function sincedeletion of these sequences decreased reporter gene expressionby 10-fold (Fig. 2B, construct 4). A second regulatory elementspanned P1 at positions �195 to �115. Deletion of thesesequences resulted in increased reporter gene expression, sug-gesting the presence of significant repressor elements (Fig. 2B,

constructs 5 and 6). Further deletions of all P1 sequencesupstream of the transcriptional start site (�1) did not impairpromoter activity (Fig. 2B, construct 7 ). Deletion of sequences�66 to �140 only slightly decreased reporter activity, indicat-ing that crucial proximal promoter elements lie between �1 to�66 (Fig. 2B, construct 8). The presence of 3� promoter ele-ments has been shown for several proximal promoters includ-ing Hoxb4, Hoxa4, and TAFII55 (41).The Hoxd4 ARE Is a Positive Regulator of P1—The positive

regulatory region between�800 and�580 has previously beenshown to harbor an ARE (Fig. 3A) (28). Two TAAT/ATTAmotifs that bind HOXD4 in vitro are functional components ofthe ARE.Mutation of both sequences results in decreased tran-scriptional activity of reporter constructs driven by promotersP1 and P2 (28). However, those studies were conducted inundifferentiated P19 cells together with co-transfectedHOXD4 expression vectors, and therefore, they did not addressthe direct role of the ARE on enhancer-dependent transcrip-tion at P1. AlthoughHOXD4 antibodies were able to supershiftcomplexes formed between nuclear extracts and primers con-

FIGURE 1. Nucleosome positioning and remodeling at Hoxd4 P1. A, results from MNase digestion experiments coupled to LM-PCR using radiolabeledprimers O-115 and O-69 are shown schematically. The positions of digestion products are shown by the horizontal lines along with the size in bp. The resultingcoordinates of N1, N2, and N3 are relative to the P1 transcriptional start site (�1). B, undifferentiated P19 cells (�) or cells treated with RA (�) were digested witheither 1 unit (lanes 1 and 2) or 5 units (lanes 3 and 4) of MNase for 5 min, and naked DNA controls (N) were digested with 0.25 units of MNase for 5 min. Arrowsdenote major protections obtained by LM-PCR using primer O-115 (left panel ) and O-69 (right panel ) followed by gel electrophoresis. Sizes were determinedby running sequencing reactions from a known source on the same gel. C, loading controls for samples used in A and B. Real-time PCR was performed usinggapdh-specific oligonucleotides, and PCR was terminated during log phase. Quantitative representation of the total amount of DNA is shown (left) as well asthe ethidium bromide-stained gel of the PCR products. M, monolayer cultures not treated with RA. D, upper panel, schematic showing the positions ofnucleosomes N, N2, and N3, restriction sites of KpnI and PvuII, and the positions of primers O-115 and O-195. Lower panels, chromatin extracts were partiallydigested in vivo with different concentrations and durations of either KpnI (left) or PvuII (right). Following DNA purification, samples were fully digested withEcoRI for normalization. The gels show the cleavage products of KpnI and PvuII digests of chromatin samples extracted from untreated (�) or RA-treated (�)P19 cells. LM-PCR was performed following restriction enzyme digests using radiolabeled O-115 (for KpnI) or O-195 (for PvuII) followed by gel electrophoresisand autoradiography.

FIGURE 2. P1 deletion constructs and luciferase reporter assays. A, sche-matic drawing of pXP2CL-Hoxd4P1 firefly luciferase reporter bearing the 540bp Hoxd4 neural enhancer and Hoxd4 P1 sequences between positions �800and �140 relative to the transcriptional start site (�1). Equals construct 3below. B, luciferase assay results for different promoter deletion constructsreported as relative luciferase activity over �-galactosidase (�-gal ) units.Untreated (�RA, black bars) or RA-treated (�RA, gray bars) P19 cells wereco-transfected with deletion constructs together with Rous sarcoma virus-�-galactosidase vector as an internal control. All experiments were repeatedthree times, and error bars represent standard deviation of three independentexperiments. NE, Hoxd4 neural enhancer.



at SY

RA

CU

SE

UN


ww

w.jbc.org

Dow

nloaded from

http://www.jbc.org/

taining the ARE sequence in EMSA (28), a direct role of theHoxd4ARE in vivowas not addressed. To address the role of theARE in regulating transcriptional initiation at Hoxd4 P1, bothTAAT/ATTAmotifs of the ARE were mutated simultaneouslyin construct 3 (mARE). Mutating both AREs dramaticallydecreased reporter activity in the presence of RA (Fig. 3B), sup-porting a key role for the ARE in regulating transcriptional ini-tiation from P1.To assess HOXD4 binding to the ARE in vivo, we performed

ChIP experiments on the endogenous Hoxd4 locus using ananti-HOXD4 antibody and primers spanning the ARE. Therewas significant recruitment of HOXD4 to the Hoxd4 ARE butnot to the gapdh control locus followingRA treatment (Fig. 3C).TheHoxd4ARE does not harbor typical sites for cooperative

binding of HOXD4 and PBX. However, Extradenticle (EXD),the Drosophila ortholog of PBX1, has been shown to modifyDFD binding to the Dfd ARE despite a similar absence of EXDbinding sites (42). We therefore tested whether PBX1 is boundto theHoxd4ARE in P19 cells by performing ChIP experimentsusing anti-PBX1 antibodies. Our results showed that PBX1wassignificantly and specifically recruited to the ARE following RAtreatment (Fig. 3D), implying possible tethering of PBX to theARE via protein-protein interactions with HOXD4. ChIPexperiments using antibodies against another HOX partner,MEIS1, revealed no significant binding (Fig. 3E).YY1 Represses Transcription from P1—A negative regulatory

element was mapped between positions �195 and �115 (Fig.2B). We scanned this sequence for possible known transcrip-tion factor consensus binding sites. A CCAT core plus flankingsequences at position�182, located precisely in the short linkerseparating nucleosomes N1 and N2, bore high similarity to theconsensus binding site for transcription factor YY1 (Fig. 4A).YY1 was of special interest given its role inHoxb4 gene expres-sion, with binding sites at both the Hoxb4 promoter and theintronic enhancer. To investigate whether YY1 bindsHoxd4 P1at this region, we performed EMSAs using P1 probes spanningthe CCAT-containing sequence (d4-YY1) (Fig. 4B) and nuclearextracts obtained from P19 cells either untreated or treatedwith RA for 48 h. Although the expression level of YY1 does notchange following RA treatment (data not shown), we usedextracts from RA-treated and untreated P19 cells to monitorchanges in proteinmodification or protein-protein interactionsthat might influence DNA binding. As shown in Fig. 4B, a spe-cific protein-DNAcomplex could be detected using the d4-YY1

HOXD4 binding sites. B, site-directed mutation of both TAAT/ATTA motifs in apXP2CL-Hoxd4P1 gives rise to plasmid mARE. The mutated underlined nucle-otides are mutated into the sequence above shown in gray. The graph showsresults of luciferase reporter assays of RA-treated P19 cells transfected eitherwith pXP2CL-Hoxd4P1 (construct 3) or with mARE. The experiments wererepeated at least twice, and error bars represent S.E. C, ChIP experiments wereperformed with either polyclonal anti-HOXD4 antibodies or no antibody (NoAb) as a negative control using chromatin extracts of untreated (black bars) orRA-treated. Results are presented as percentage of input. Real-time PCR wasperformed using oligonucleotides specific for the Hoxd4 ARE or gapdh as acontrol (Table 1) and shows HOXD4 recruitment to the ARE following RAtreatment. Experiments were performed at least twice, and error bars repre-sent S.E. D, ChIP experiments using PBX-1 polyclonal antibodies show recruit-ment of PBX1 to the HOXD4 ARE in RA-treated cells. E, results from ChIP exper-iments using MEIS-1 polyclonal antibodies, showing that MEIS-1 is notrecruited to the Hoxd4 ARE in RA-treated and untreated P19 cells.

FIGURE 3. Hoxd4 ARE is crucial for P1 activity and recruits HOXD4 andPBX1 in RA-treated P19 cells. A, conservation between mouse and zebrafishof DNA sequences spanning the Hoxd4 ARE including two TAAT/ATTA



at SY

RA

CU

SE

UN


ww

w.jbc.org

Dow

nloaded from

http://www.jbc.org/

probe, and complex formation wasspecifically inhibited either by coldprobe or by a YY1 consensus site. Anonspecific competitor did notinhibit complex formation (Fig. 4B).The presence of YY1 in the shiftedcomplex was assessed with anti-YY1 antibodies. As shown in Fig.4C, the protein-DNA complex wassignificantly inhibited by anti-YY1antibodies but not with severalother nonspecific antibodies (Fig.4C), further supporting the pres-ence of YY1 in the shifted complex.The relevance of YY1 to P1 func-

tionwas investigated bymutation ofits binding site in Construct 4 (Fig.5A, mYY1). Mutating the YY1 siteled to an increase in reporter geneexpression (Fig. 5B) that was com-parable with that seen followingdeletion of sequences�195 to�115(Fig. 2B, constructs 5 and 6). Toassess whether this activity could becorrelated with the presence of YY1at this site in vivo, we performed

FIGURE 4. Site-specific YY1 binding to Hoxd4 P1. A, sequence comparison between Hoxd4 P1 repressorysequences and the YY1 binding site consensus sequence. The location of the YY1 binding site is shown relativeto nucleosomes N1 and N2. B, EMSA using the Hoxd4 YY1 binding site as the labeled probe and nuclear extracts(NE) from untreated (�) or RA-treated (�) P19 cells. Competition with either 10� or 100� concentrations ofcold probe or cold YY1 consensus oligonucleotides (cons.) inhibit complex formation, whereas the same con-centrations of a nonspecific oligonucleotides (nonsp.) do not. C, anti-YY1 antibodies but not other unrelatedantibodies (anti-RAR�, anti-PBX1, and anti-TFIID) inhibit complex formation between nuclear extracts andHoxd4 P1 YY1 binding site.

FIGURE 5. YY1 represses reporter construct activity and binds to Hoxd4 P1 in vivo. A, schematic showing mutation of Hoxd4 P1 YY1 binding site (mYY1) inthe construct 4 background (Fig. 2). B, luciferase reporter assays showing increased activation of the reporter construct containing the mutated Hoxd4 YY1binding site when compared with the wild-type construct. Values are the averages of two independent experiments with error bars representing S.E. C, ChIPexperiments with anti-YY1 antibodies using chromatin extracts from untreated (black bars) or RA-treated (gray bars) P19 cells. Mock ChIP experiments using noantibodies (No Ab) were performed as controls. Real-time PCR was performed using primers specific for the region of the P1 YY1 binding site or gapdh as acontrol, and values are presented as percent of input material. IP, immunoprecipitation.



at SY

RA

CU

SE

UN


ww

w.jbc.org

Dow

nloaded from

http://www.jbc.org/

ChIP experiments using chromatin extracts from RA-treatedand untreated P19 cells and immunoprecipitating with anti-YY1 antibodies. Significant levels of YY1 were present at theendogenous P1 YY1 binding site in neurally differentiating RA-treatedP19 cells aswell as in untreated cells (Fig. 5C), indicatingthat YY1 binds to P1 before and after transcriptional activation.This observation suggests a role in modulating Hoxd4 expres-sion even after transcription has been initiated.

We reasoned that YY1 may main-tain Hoxd4 in a repressed state inundifferentiated P19 cells. Followingneural differentiation, YY1 couldeither switch to an activator, as it doesat the IFNB gene (43), or persist as arepressor, thereby balancing stimula-tory signals to achieve appropriateamounts of Hoxd4 transcripts. Toexamine the functional consequencesof YY1 binding, we used siRNA toknock down yy1 expression. P19 cellswere transiently transfected with yy1siRNA and control expression vec-tors, and total protein and RNAwereextracted 48 h later. There was a sig-nificant decrease in YY1 proteinexpression following yy1 knockdownin two independent experiments (Fig.6A, upper panel), which was not thecase in untransfected cells or cellstransfected with the control vector.Actin levels were equivalent in allsamples.Hoxd4 transcript levelsweredetermined by RT-PCR (Fig. 6A,lower panel). yy1 knockdown, but notcontrol transfections, markedly stim-ulated endogenousHoxd4 expression(Fig. 6A, lower panel). These resultswere reproducible and specific forHoxd4, as gapdh levels remainunchanged. These data strongly sug-gest that YY1 represses Hoxd4 tran-scription and that even partial loss ofYY1 relieves this silencing despite theabsence of RA.To investigate whether YY1 acts

as an activator or a repressor fol-lowing Hoxd4 induction, we per-formed YY1 knockdown by directtransfection of siRNA double-stranded oligonucleotides (39). Asshown in Fig. 6B, YY1 levels weresignificantly reduced followingtransfection with YY1 siRNA oli-gonucleotides, whereas they re-mained constant in untransfectedcells and cells transfected with con-trol siRNA oligonucleotides. Actinlevels remained the same in all sam-

ples tested, confirming that YY1 knockdown was specific.Interestingly, RT-PCR results showed that following YY1knockdown,Hoxd4-specific transcript levels were significantlyincreased in both experiments, whereas gapdh RNA levelsremained constant, suggesting that YY1 continues to repressHoxd4 expression even after Hoxd4 induction.

Onemechanism of YY1-mediated transcriptional repressionis through recruitment of PcG proteins (44, 45). In situ hybrid-

FIGURE 6. Knockdown of yy1 increases Hoxd4 transcription in P19 cells. A, upper panel, knockdown of yy1 inundifferentiated P19 cells. Western blots (WB) using anti-YY1 antibodies representing two independent yy1knockdown experiments (sets 1 and 2) using whole cell extracts of untransfected (�) or transfected P19 cellswith either pSUPER-YY1 (YY1 siRNA) or pSUPER-CTRL (ctrl siRNA) vectors. Membranes were reprobed withanti-actin antibodies as a loading control. Lower panels, RT was performed on total RNA extracted from P19 cellsused in the upper panel. PCR was performed using primers specific for Hoxd4 homeobox region (Hoxd4) before(�) and after RT or gapdh coding region (gapdh) as a control. B, YY1 knockdown in RA-treated P19 cells. Upperand lower panels are as for panel A. Results show specific increases in Hoxd4 transcripts following YY1 knock-down. C, ChIP with MEL18 antibodies or no antibody (No Ab) as a negative control using chromatin extractsfrom untreated (black bars) or RA-treated (gray bars) P19 cells. Real-time PCR was performed using primersspanning the Hoxd4 YY1 binding site (Hoxd4 P1) or corresponding to a gapdh control. IP, immunoprecipitation.D, sequential ChIP performed first with anti-YY1 antibodies followed by a second ChIP with anti-MEL18 anti-bodies. This was followed by real-time PCR using either Hoxd4 P1- or gapdh-specific oligonucleotides. E, West-ern blot analysis showing the expression of MEL18 in nuclear (N) and cytoplasmic (C ) fractions of P19 cellseither treated (�) or untreated (�) with RA. TFIID expression was analyzed in both fractions to ensure efficientextraction of nuclear versus cytoplasmic fractions. HOXD4 protein expression was performed to confirm tran-scriptional activation following RA treatment.



at SY

RA

CU

SE

UN


ww

w.jbc.org

Dow

nloaded from

http://www.jbc.org/

ization experiments have shown that Hoxd4 transcription isaffected inmouse embryos null for the PcGgeneMel18 (46, 47).We speculated that repression of Hoxd4 by YY1 may be medi-ated via a MEL18-dependent mechanism. We therefore con-ducted ChIP experiments to assess MEL18 recruitment to P1before and after Hoxd4 transcriptional activation. MEL18 wasindeed significantly recruited to Hoxd4 P1 in undifferentiatedP19 cells but was lost following RA treatment (Fig. 6C). Toconfirm that this losswas not due to a global decrease inMEL18protein expression or to its relocalization to the cytoplasm, weseparated nuclear and cytoplasmic fractions from both mono-layer and RA-treated P19 cells. This was followed by immuno-blot analysis usingMEL18-specific antibodies (Fig. 6E). Expres-sion of MEL18 was evident in both nuclear and cytoplasmicfractions and was not significantly altered in either extract fol-lowing RA treatment (Fig. 6E). TFIIDwas almost entirely local-ized to the nuclear fractions, indicating efficient separation ofnuclear and cytoplasmic components (Fig. 6E). HOXD4expressionwas detectable only in RA-treated cells as confirmedby immunoblots using HOXD4-specific antibodies (Fig. 6E).These results confirm that the reduced association ofMEL18 atP1 in P19 cells following RA treatment is indeed due todecreased recruitment of MEL18 to the Hoxd4 locus, asopposed to a global decrease in protein expression or subcellu-lar relocalization.To determine whether YY1 andMEL18 are present together

at P1, extracts immunoprecipitated with �-YY1 antibodieswere subjected to a second round of ChIP using MEL18 anti-bodies. As shown in Fig. 6D, YY1 and MEL18 were boundtogether to P1 only in undifferentiated P19 cells (Fig. 6D), sug-gesting that MEL18 is released from a P1-bound YY1 complexfollowing RA treatment and Hoxd4 transcriptional activation.This further implies that the repressive effects of YY1 followingHoxd4 induction must be mediated through a MEL18-inde-pendent mechanism.A Positive Regulatory Element Resides Downstream of the

Transcriptional Start Site—Finally, we attempted to identifythe regulatory element responsible for core promoter activitylocated between positions �1 and �66. Two Luc reporter con-structs were designed carrying sequences�1 to�33 (construct9) or �34 to �66 (construct 10) and their activities when com-pared with construct 8 (Fig. 2B) in RA-treated P19 cells.Although construct 10 had comparable reporter activity withconstruct 8, that of construct 9 was significantly diminished(Fig. 7A). Consistent with activity mediated by the downstreamregion, oligonucleotides spanning 1–66 and 34–66 formed asingle major complex with nuclear extracts in EMSA, contraryto primer 1–33, which showed no significant complex forma-tion (Fig. 7B). Specificity of complex formation was then testedby EMSAusing primer 34–66 as a probe accompanied by com-petition reactions using an increased amount of cold probe (Fig.7C). Interestingly, the specific binding of primer 34–66 to P19nuclear extracts was accompanied by increased binding toextracts from RA-treated P19 cells (compare � with �).Sequences �34 to �66 contain a GC-rich sequence compara-ble with an Sp1/Sp3 consensus binding site. Given that anincrease in complex formation was observed when nuclearextracts fromRA-treated P19 cellswere used and that Sp1DNA

binding ability is increased in the presence of RA (48, 49), weperformed supershift experiments to investigate Sp1 and Sp3 aspossible binding candidates. However, incubation with eitherSp1 or Sp3 antibodies did not inhibit or supershift complexformation (data not shown), suggesting that a transcription fac-tor other than Sp1 or Sp3 binds this region.

DISCUSSION

The role of chromatin in regulating Hox gene transcrip-tional initiation has recently come under the spotlightthrough studies showing sequential 3� to 5� chromatindecondensation of the HoxB cluster during mouse embry-onic development (5, 50) and regulation of HoxD geneexpression through chromatin remodeling during limbdevelopment (51) and neural differentiation (33). In thisstudy, we showed that nucleosomes are positioned at theHoxd4 P1 promoter and are remodeled following RA treat-ment of P19 cells. This remodeling leads to chromatin

FIGURE 7. A positive promoter element maps downstream of the Hoxd4transcriptional start site. A, luciferase reporter constructs and their activitiesin transfected RA-treated P19 cells. Construct 8 (1– 66) is the same as thatpresented in Fig. 2B. Results are the averages of two independent experi-ments, and error bars represent S.E. Rel Luc, relative luciferase. B, EMSAs usinglabeled probe A (containing P1 sequences �1 to �33), probe B (�34 to �66),or probe C (�1 to �66) (Table 1). Nuclear extracts (NE) from untreated (�) orRA-treated (�) P19 cells were used. The arrow points to a specific DNA-pro-tein complex. C, competition experiments using probe B (�34 to �66).



at SY

RA

CU

SE

UN


ww

w.jbc.org

Dow

nloaded from

http://www.jbc.org/

decondensation and coincides with the initiation of Hoxd4gene transcription. We also identified three major cis-actingregulatory elements impinging upon P1: an ARE and a GC-rich element acting as positive regulators and a YY1 bindingsite that represses P1 transcription before and during neuraldifferentiation in P19 cells.Chromatin remodeling can result in translational reposi-

tioning of a nucleosome, such as is achieved by nucleosomesliding, and subsequent exposure of a transcription factorbinding site, as is the case for the IFN-� promoter (52). Forpromoters such as IL-12 p40, however, covalent histonemodifications may lead to chromatin decondensation with-out altering nucleosome position (40). In this study, weshowed that three nucleosomes are positioned at P1 inundifferentiated P19 cells where Hoxd4 transcription isrepressed. N3 was shown to span the transcriptional startsite (�1) and the positive GC-rich cis-element (�34 to �66),implying a functional relevance for N3 positioning. We havealready shown that histone H3 is not acetylated at lysine 9 inthe region of P1 in undifferentiated P19 cells, confirming astate of closed transcriptionally inactive chromatin (33).Upon RA treatment and P19 cell differentiation down theneural pathway, chromatin relaxation takes place at P1 asshown by MNase- and restriction enzyme-coupled LM-PCRexperiments (Fig. 1) but is not accompanied by nucleosomerepositioning. This chromatin relaxation might be a directconsequence of the CBP recruitment and histone hyper-acetylation known to take place during Hoxd4 transcrip-tional activation (33).SWI/SNF and ISWI are two major ATP-dependent chro-

matin remodeling complexes. Although SWI/SNF is able toincrease nucleosomal accessibility to DNase and restrictionenzymes in the absence of translational movement, ISWIcannot. This feature of SWI/SNF-mediated remodeling hasbeen proposed to result from changes to DNA topology, theconformation of the histone octamer, or both. In one sce-nario, the energy of ATP would be used by SWI/SNF to gen-erate a transient topological intermediate that would col-lapse into a stable and conformationally altered state (53).The differential ability of SWI/SNF to induce such chroma-tin changes in the absence of translational movement hasbeen ascribed to the ATPase domain of the SWI/SNF-spe-cific motor protein BRG1 (54). Together, these results sug-gest that the Hoxd4 promoter is a target of SWI/SNF but notISWI.Maintenance of a repressed or an activated Hox transcrip-

tional status has long been attributed to members of the PcGand Trithorax protein families (55, 56), which execute theirfunctions through chromatin modification. In this study, wedescribe a repressive role for YY1 in regulating Hoxd4expression. YY1, a homolog of Drosophila PcG group pro-tein Pleiohomeotic, is a multifunctional zinc finger-contain-ing transcription factor that acts either as a transcriptionalactivator or as a repressor (57). YY1 mutant embryos dieshortly following implantation, and heterozygous embryosdevelop neurulation defects (58). Loss of function studiesperformed in Xenopus report antero-posterior patterningdefects, reduced head structures, and abnormalities in mid-

brain-hindbrain boundaries consistent with a role for YY1 incentral nervous system development and the control of Hoxgene expression. Interestingly, YY1 has been shown to acti-vateHoxb7 gene expression in tissue culture (59) and to bindHoxb4 promoter and CR1 enhancer region as part of an over-lapping NFY/YY1 site (60). This site seems to be conservedfor Hoxb4 and might also be present in the Hoxc8 earlyenhancer (60). However, the YY1 binding site described inthis study does not resemble the Hoxb4 binding site since itdoes not seem to contain an overlapping NFY binding motifand is not evolutionarily conserved with zebrafish, unlikeother more downstream P1 sequences (25), suggesting anovel role for YY1 in fine-tuning Hoxd4 expression levels inmammalian embryonic development.YY1 binds specifically to the �183 site in vitro and to a

short region spanning this site in vivo, and mutation of thissequence in reporter constructs relieves the repressive effecton transcriptional initiation from P1. Interestingly, this YY1binding site is located within the internucleosomal regionseparating N1 and N2. YY1 binding to this site might not besignificantly altered following nucleosome remodeling at P1after RA treatment since nucleosome sliding does not takeplace. Alternatively, prior YY1 binding may impose bothnucleosome positioning and a subsequent closed chromatinstate. Regardless, our results show that this site is accessibleto YY1 even in the repressed state where positioned nucleo-somes are condensed and Hoxd4 is not transcribed.Our RNA interference results indicate that YY1 represses

Hoxd4 gene expression in undifferentiated P19 cells whereHoxd4 is known to be silenced (33). The fly homolog of YY1,Pleiohomeotic, binds Polycomb-response elements andrecruits two types of PcG complexes, one of which containsthe fly homolog of MEL18, posterior sex combs (PSC) (61).Recently, data from Srinivasan and Atchison (44) demon-strate that YY1 binds to Polycomb-response elements inDro-sophila and subsequently recruits other PcG proteins toDNA. Similarly, our ChIP results are consistent with recruit-ment of MEL18 to the P1 YY1 site in undifferentiated P19cells (Fig. 6) and with results from MEL18 mutant mouseembryos that exhibit ectopic Hoxd4 gene expression (47).This suggests a possible mechanism by which YY1 bindingrecruits a MEL18-containing complex that maintains silenc-ing at the Hoxd4 locus. This is supported by the location ofthe YY1 binding site in an internucleosomal region, allowingYY1 to bind the repressed Hoxd4 gene and then recruitMEL18 so as to maintain silenced chromatin at P1. Ourattempts to verify whether MEL18 recruitment to P1 isdecreased following YY1 knockdown using ChIP did notreveal significant results (data not shown), possibly due tothe incomplete knockdown of YY1 in our experiments.Alternatively, MEL18 binding may be independent of YY1 orbecome so once recruited.YY1 remains bound to P1 following Hoxd4 activation in

neurally differentiating P19 cells, and its knockdown resultsin further increases in Hoxd4 transcription. However,MEL18 is no longer present at the active P1 promoter, sug-gesting that the release of MEL18, but not YY1, is requiredfor gene induction. Our results further demonstrate a role



at SY

RA

CU

SE

UN


ww

w.jbc.org

Dow

nloaded from

http://www.jbc.org/

for YY1 in dampening activated transcription atHoxd4. Whywould repressive YY1 function be required at transcription-ally active loci? Repressive histone deacetylase and chroma-tin remodeling functions are required for limiting cyclicalrounds of transcriptional initiation at the estrogen-receptortarget gene pS2 (62). Similarly, YY1 may recruit such activi-ties to curtail RA-induced Hoxd4 expression. Regardless ofthe inhibitory mechanism, YY1 binding to P1 attenuates butdoes not abolish transcription, and therefore, could fine-tune the amount ofHoxd4 transcript and protein levels avail-able under specific conditions. A recent study in Drosophilasuggests that the repressive effect of the HOX protein Ultra-bithorax (UBX) on Distal-less (Dll) transcription in the limbis highly concentration-dependent (63). Therefore, YY1might be required to fine-tune the appropriate amount ofHOXD4 required to repress or activate downstream targetgenes. Alternatively, the relief of repression seen inRA-treated cells following YY1 knockdown may be an indi-rect result of altered regulation at other loci.

TheHoxd4 ARE was first charac-terized in non-differentiating P19cells and in the absence of the neuralenhancer (28). In this study, weinvestigated the role of the ARE inregulating P1 in concert with theenhancer during neural differentia-tion of P19 cells. We found that theARE plays a crucial role in initiatingtranscription from P1 only after RAtreatment, indicating that AREactivity was also dependent on theHoxd4 3� neural enhancer. More-over, we showed binding of HOXD4to theARE in vivo byChIP, confirm-ing previous results obtained inEMSA (28). Interestingly, we alsodetected significant PBX1 recruit-ment to the Hoxd4 ARE inRA-treated P19 cells, although theARE in question does not clearlyharbor a HOX-PBX binding site(TGATTNAT). Previous work inDrosophila embryos suggests thatalthough DFD can interact withsimple DNA binding sites in theabsence of EXD, protein-proteintethering of EXD is required torelease the intrinsic DFD transacti-vation function, which is otherwisemasked by the homeodomain (42).HOXD4might similarly tether PBXgiven the high degree of functionalconservation among the Dfd andHoxd4 AREs and among HOX pro-teins and their cofactors (64).Finally, we alsomapped a positive

regulatory element located 3� to thetranscriptional start site (Fig. 2B).

This element is spanned by nucleosome N3 (Fig. 1C) and istherefore more accessible to transcription factor binding afterRA treatment. We tried to identify the transcription factorbinding to the GC-rich motif at the core promoter down-stream to the transcriptional start sites. Although Sp1 hasbeen shown to regulate the expression of several Hox genes(59, 65, 66), we conclude that it does not act through theGC-rich motif at P1 (data not shown). Other proteins ofinterest include AP-2, which has been shown to regulateHoxa2 (67), and USF-1 (59, 68), which binds Hoxb4 andHoxb7. On the other hand, this GC-rich motif may be func-tionally similar to other previously described core promoterelements, such as the downstream promoter element, whichhave been shown to participate in initiating transcriptionfrom numerous proximal promoters (69). Further studiesare needed to confirm these suggestions.In conclusion, we have shown that multiple levels of con-

trol at P1 influence the regulation ofHoxd4 gene expression.An integration of these results with those of our earlier study

FIGURE 8. A model of events at the Hoxd4 neural enhancer and P1 promoter leading to transcriptionalactivation. The diagram depicts the Hoxd4 P1 promoter with the three positioned nucleosomes mapped inthis study, an internucleosomal YY1�MEL18 complex, the upstream ARE, the transcriptional start site (arrow),and the 3� Hoxd4 neural enhancer plus bound retinoid X receptor� retinoic acid receptor (RXR�RAR) het-erodimer. Nucleosomes are present at the Hoxd4 enhancer but have not been mapped (dotted circles), and therelative position of the RXR�RAR heterodimer is arbitrary. Results from the current study and from Rastegar et al.(33) suggest the following order of events: (1) RA is bound by retinoid receptors (RXR�RAR) present at thepreviously characterized RARE in the Hoxd4 neural enhancer (22, 25), leading to (2) recruitment of coactivatorssuch as CBP and (3) methylation and acetylation of histone H3 on lysine 4 and lysine 9, respectively, andN-terminal acetylation of histone H4. These modifications are hallmarks of transcriptionally active chromatinand correlate with an “open” or accessible state. Recruitment of additional transcription factors and coactiva-tors to the enhancer, perhaps in conjunction with a 5� spreading of histone modifications (33), leads to (4)release of MEL18 from a complex with YY1 at the P1 promoter. HOXD4 subsequently binds the upstream AREand tethers PBX to this region via protein-protein interactions, whereas a positively acting GC-rich elementrecruits an unknown factor (GC) just downstream of the transcriptional start site (5). Recruitment of coactiva-tors such as CBP (6) leads to covalent modification of histones H3 and H4 (7) and relaxation of DNA without achange in nucleosome position. RNA polymerase II (Pol II) arrives at the enhancer (8) before the promoter (33),placing Hoxd4 among a growing list of genes in which the enhancer may deliver (9) polymerase to the pro-moter (70). RNA polymerase II, in cooperation with the basal transcriptional machinery, now engages thesterically accessible promoter region, and transcription is initiated. The continued presence of YY1 may berequired to limit successive rounds of reinitiation. Note that although this chronology is experimentally vali-dated at several points, the sequence is partly speculative.



at SY

RA

CU

SE

UN


ww

w.jbc.org

Dow

nloaded from

http://www.jbc.org/

(33) suggests a model for the concerted action of factorrecruitment and chromatin modification or remodeling atthe Hoxd4 enhancer and P1 promoter (Fig. 8). Further studywill explore the mechanistic links between the enhancer andpromoter that jointly specify the fine temporal and spatialpatterns of Hoxd4 expression in the developing embryo.

Acknowledgments—We thank C.Mueller for anti-Sp1 and -Sp3 anti-bodies, and S. Smale and members of the Featherstone laboratory forhelpful discussions.

REFERENCES1. Krumlauf, R. (1994) Cell 78, 191–2012. Kmita, M., and Duboule, D. (2003) Science 301, 331–3333. Manzanares, M., Wada, H., Itasaki, N., Trainor, P. A., Krumlauf, R., and

Holland, P. W. (2000) Nature 408, 854–8574. Duboule, D. (1998) Curr. Opin. Genet. Dev. 8, 514–5185. Duboule, D., and Deschamps, J. (2004) Dev. Cell 6, 738–7406. Roelen, B. A., de Graaff, W., Forlani, S., and Deschamps, J. (2002) Mech.

Dev. 119, 81–907. Rijli, F.M.,Mark,M., Lakkaraju, S., Dierich, A., Dolle, P., and Chambon, P.

(1993) Cell 75, 1333–13498. Maconochie, M., Nonchev, S., Morrison, A., and Krumlauf, R. (1996)

Annu. Rev. Genet 30, 529–5569. Moens, C. B., and Selleri, L. (2006) Dev. Biol. 291, 193–20610. Featherstone, M. (2003) in Murine Homeobox Gene Control of Embry-

onic Patterning and Organogenesis (Lufkin, T., ed) pp. 1–42, ElsevierScience, Amsterdam

11. Manzanares, M., Trainor, P. A., Nonchev, S., Ariza-McNaughton, L.,Brodie, J., Gould, A., Marshall, H., Morrison, A., Kwan, C. T., Sham,M. H., Wilkinson, D. G., and Krumlauf, R. (1999) Dev. Biol. 211,220–237

12. Nonchev, S., Maconochie, M., Vesque, C., Aparicio, S., Ariza-McNaugh-ton, L., Manzanares, M., Maruthainar, K., Kuroiwa, A., Brenner, S., Char-nay, P., and Krumlauf, R. (1996) Proc. Natl. Acad. Sci. U. S. A. 93,9339–9345

13. Di Rocco, G., Gavalas, A., Popperl, H., Krumlauf, R., Mavilio, F., andZappavigna, V. (2001) J. Biol. Chem. 276, 20506–20515

14. Deschamps, J., and van Nes, J. (2005) Development (Camb.) 132,2931–2942

15. Gould, A., Itasaki, N., and Krumlauf, R. (1998) Neuron 21, 39–5116. Huang, D., Chen, S. W., and Gudas, L. J. (2002)Dev. Dyn. 223, 353–37017. Dupe, V., Davenne,M., Brocard, J., Dolle, P.,Mark,M., Dierich, A., Cham-

bon, P., and Rijli, F. M. (1997) Development (Camb.) 124, 399–41018. Langston, A. W., and Gudas, L. J. (1992)Mech. Dev. 38, 217–22719. Marshall, H., Studer, M., Popperl, H., Aparicio, S., Kuroiwa, A., Brenner,

S., and Krumlauf, R. (1994) Nature 370, 567–57120. Packer, A. I., Crotty, D. A., Elwell, V. A., and Wolgemuth, D. J. (1998)

Development (Camb.) 125, 1991–199821. Studer,M., Popperl, H.,Marshall, H., Kuroiwa,A., andKrumlauf, R. (1994)

Science 265, 1728–173222. Zhang, F., Nagy Kovacs, E., and Featherstone, M. S. (2000)Mech. Dev. 96,

79–8923. Oosterveen, T., Niederreither, K., Dolle, P., Chambon, P., Meijlink, F., and

Deschamps, J. (2003) EMBO J. 22, 262–26924. Marshall, H., Morrison, A., Studer, M., Popperl, H., and Krumlauf, R.

(1996) FASEB. J. 10, 969–97825. Nolte, C., Amores, A., Nagy Kovacs, E., Postlethwait, J., and Featherstone,

M. (2003)Mech. Dev. 120, 325–33526. Popperl, H., and Featherstone, M. S. (1993)Mol. Cell. Biol. 13, 257–26527. Featherstone,M. S., Baron, A., Gaunt, S. J., Mattei, M. G., and Duboule, D.

(1988) Proc. Natl. Acad. Sci. U. S. A. 85, 4760–476428. Popperl, H., and Featherstone, M. S. (1992) EMBO J. 11, 3673–368029. Zhang, F., Popperl, H., Morrison, A., Kovacs, E. N., Prideaux, V., Schwarz,

L., Krumlauf, R., Rossant, J., and Featherstone,M. S. (1997)Mech. Dev. 67,

49–5830. Morrison, A., Ariza-McNaughton, L., Gould, A., Featherstone, M., and

Krumlauf, R. (1997) Development (Camb.) 124, 3135–314631. Folberg, A., Kovacs, E. N., and Featherstone, M. S. (1997) J. Biol. Chem.

272, 29151–2915732. Brend, T., Gilthorpe, J., Summerbell, D., and Rigby, P. W. (2003)Develop-

ment (Camb.) 130, 2717–272833. Rastegar, M., Kobrossy, L., Kovacs, E. N., Rambaldi, I., and Featherstone,

M. (2004)Mol. Cell. Biol. 24, 8090–810334. Nordeen, S. K. (1988) BioTechniques 6, 454–45835. Carey, M., and Smale, S. T. (2000) Transcriptional Regulation in Eu-

karyotes: Concepts, Strategies, and Techniques, pp. 338–346, Cold SpringHarbor Laboratory Press, Cold Spring Harbor, NY

36. Garrity, P. A., and Wold, B. J. (1992) Proc. Natl. Acad. Sci. U. S. A. 89,1021–1025

37. Houle, M., Prinos, P., Iulianella, A., Bouchard, N., and Lohnes, D. (2000)Mol. Cell. Biol. 20, 6579–6586

38. Haller, K., Rambaldi, I., Daniels, E., and Featherstone, M. (2004) J. Biol.Chem. 279, 49384–49394

39. Kurisaki, K., Kurisaki, A., Valcourt, U., Terentiev, A. A., Pardali, K., TenDijke, P., Heldin, C. H., Ericsson, J., and Moustakas, A. (2003) Mol. Cell.Biol. 23, 4494–4510

40. Weinmann, A. S., Plevy, S. E., and Smale, S. T. (1999) Immunity 11,665–675

41. Gutman, A., Gilthorpe, J., and Rigby, P. W. (1994) Mol. Cell. Biol. 14,8143–8154

42. Li, X., Murre, C., and McGinnis, W. (1999) EMBO J. 18, 198–21143. Weill, L., Shestakova, E., and Bonnefoy, E. (2003) J. Virol. 77,

2903–291444. Srinivasan, L., and Atchison, M. L. (2004) Genes Dev. 18, 2596–260145. Atchison, L., Ghias, A., Wilkinson, F., Bonini, N., and Atchison, M. L.

(2003) EMBO J. 22, 1347–135846. Akasaka, T., Kanno, M., Balling, R., Mieza, M. A., Taniguchi, M., and

Koseki, H. (1996) Development (Camb.) 122, 1513–152247. Akasaka, T., van Lohuizen, M., van der Lugt, N., Mizutani-Koseki, Y.,

Kanno,M., Taniguchi,M., Vidal,M., Alkema,M., Berns, A., andKoseki,H.(2001) Development (Camb.) 128, 1587–1597

48. Akiyama, H., Tanaka, T., Maeno, T., Kanai, H., Kimura, Y., Kishi, S.,and Kurabayashi, M. (2002) Investig. Ophthalmol. Vis. Sci. 43,1367–1374

49. Horie, S., Ishii, H., Matsumoto, F., Kusano, M., Kizaki, K., Matsuda, J., andKazama, M. (2001) J. Biol. Chem. 276, 2440–2450

50. Chambeyron, S., Da Silva, N. R., Lawson, K. A., and Bickmore, W. A.(2005) Development (Camb.) 132, 2215–2223

51. Barna,M.,Merghoub, T., Costoya, J. A., Ruggero, D., Branford,M., Bergia,A., Samori, B., and Pandolfi, P. P. (2002) Dev. Cell 3, 499–510

52. Lomvardas, S., and Thanos, D. (2001) Cell 106, 685–69653. Narlikar, G. J., Fan, H. Y., and Kingston, R. E. (2002) Cell 108, 475–48754. Fan, H. Y., Trotter, K. W., Archer, T. K., and Kingston, R. E. (2005) Mol.

Cell 17, 805–81555. Gould, A. (1997) Curr. Opin. Genet. Dev. 7, 488–49456. Ross, J. M., and Zarkower, D. (2003) Dev. Cell 4, 891–90157. Thomas, M. J., and Seto, E. (1999) Gene (Amst.) 236, 197–20858. Donohoe, M. E., Zhang, X., McGinnis, L., Biggers, J., Li, E., and Shi, Y.

(1999)Mol. Cell. Biol. 19, 7237–724459. Meccia, E., Bottero, L., Felicetti, F., Peschle, C., Colombo, M. P., and Care,

A. (2003) Biochim. Biophys. Acta 1626, 1–960. Gilthorpe, J., Vandromme, M., Brend, T., Gutman, A., Summerbell, D.,

Totty, N., and Rigby, P. W. (2002) Development (Camb.) 129,3887–3899

61. Mohd-Sarip, A., Cleard, F., Mishra, R. K., Karch, F., and Verrijzer, C. P.(2005) Genes Dev. 19, 1755–1760

62. Metivier, R., Penot, G., Hubner, M. R., Reid, G., Brand, H., Kos, M., andGannon, F. (2003) Cell 115, 751–763

63. Tour, E., Hittinger, C. T., andMcGinnis, W. (2005)Development (Camb.)132, 5271–5281

64. Krumlauf, R. (1992) BioEssays 14, 245–25265. Galliot, B., Dolle, P., Vigneron, M., Featherstone, M. S., Baron, A., and



at SY

RA

CU

SE

UN


ww

w.jbc.org

Dow

nloaded from

http://www.jbc.org/

Duboule, D. (1989) Development (Camb.) 107, 343–35966. Kim, M. H., Cho, M., and Park, D. (1998) Somatic Cell Mol. Genet. 24,

371–37467. Maconochie, M., Krishnamurthy, R., Nonchev, S., Meier, P., Manzanares,

M., Mitchell, P. J., and Krumlauf, R. (1999) Development (Camb.) 126,1483–1494

68. Giannola, D. M., Shlomchik, W. D., Jegathesan, M., Liebowitz, D.,Abrams, C. S., Kadesch, T., Dancis, A., and Emerson, S. G. (2000) J. Exp.Med. 192, 1479–1490

69. Butler, J. E., and Kadonaga, J. T. (2002) Genes Dev. 16, 2583–259270. Szutorisz, H., Dillon, N., and Tora, L. (2005) Trends Biochem. Sci 30,

593–599



at SY

RA

CU

SE

UN


ww

w.jbc.org

Dow

nloaded from

http://www.jbc.org/

25926.full

Documents

Transcript of 25926.full