Regulatory Factor for X-box Family Proteins DifferentiallyInteract with Histone Deacetylases to Repress Collagen �2(I)Gene (COL1A2) Expression*

Received for publication, October 31, 2005, and in revised form, January 31, 2006 Published, JBC Papers in Press, February 6, 2006, DOI 10.1074/jbc.M511724200

Yong Xu‡, Pritam K. Sengupta‡§, Edward Seto¶, and Barbara D. Smith‡§1

From the ‡Department of Biochemistry, Boston University School of Medicine, Boston, Massachusetts 02118, §Veterans AffairsBoston Healthcare System, Boston, Massachusetts 02118, and ¶H. Lee Moffitt Cancer Center and Research Institute,Tampa, Florida 33612

Our studies indicate that the regulatory factor forX-box (RFX) fam-ily proteins repress collagen �2(I) gene (COL1A2) expression (Xu, Y.,Wang, L., Buttice, G., Sengupta, P. K., and Smith, B. D. (2003) J. Biol.Chem. 278, 49134–49144; Xu, Y., Wang, L., Buttice, G., Sengupta,P.K., andSmith, B.D. (2004) J. Biol. Chem.279, 41319–41332). In thisstudy,weexaminedthemechanism(s)underlyingtherepressionofcol-lagen gene by RFX proteins. Two members of the RFX family, RFX1and RFX5, associate with distinct sets of co-repressors on the collagentranscription start site in vitro. RFX5specifically interactswithhistonedeacetylase 2 (HDAC2) and themammalian transcriptional repressor(mSin3B), whereas RFX1 preferably interacts with HDAC1 andmSin3A. HDAC2 cooperates with RFX5 to down-regulate collagenpromoter activity, whereas HDAC1 enhances inhibition of collagenpromoter activity by RFX1. Interferon-� promotes the recruitment ofRFX5/HDAC2/mSin3B to the collagen transcription start site butdecreases theoccupancybyRFX1/mSin3Aasmanifestedbychromatinimmunoprecipitation assay. RFX1 binds to the methylated collagensequence with much higher affinity than unmethylated sequence,recruiting more HDAC1 and mSin3A. The DNA methyltransferaseinhibitor 5-aza-2�-deoxycytidine, which inhibits DNA methylation,reduces RFX1/HDAC1 binding to the collagen transcription start sitein chromatin immunoprecipitation assays. Finally, both RFX1 andRFX5 are acetylated in vivo. Trichostatin A stimulates the acetylationof RFX proteins and activates the collagen promoter activity. Collec-tively, our data strongly indicate two separate pathways for RFX pro-teins to repress collagen gene expression as follows: one for RFX5/HDAC2 in interferon-�-mediated repression, and the other forRFX1/HDAC1 inmethylation-mediated collagen silencing.

Collagen, currently consisting of more than 27 members, is a largefamily of extracellular matrix proteins that play vital structural andphysiological roles maintaining the integrity and contributing to home-ostasis of the human body (3). Because of their diverse structures anddistributions as well as complex interactions with other components ofthe extracellular matrix, expression and regulation of collagen proteinsare extremely complicated, yet critical processes, which occur at multi-ple levels as follows: transcriptional, post-transcriptional, translational,

and post-translational. Type I collagen, composed of two �1(I)(COL1A1) chains and one �2(I) (COL1A2) chain, is the most abun-dantly expressed member of the collagen genes, thereby playing a sig-nificant role in maintaining homeostasis. The transcription of thesegenes is important during development and repair of injury.Transcription of eukaryotic genes is controlled by coordination of

large complexes composed of activators/co-activators or repressors/co-repressors, which function to alter histone-DNA structure in highlyordered chromatin. Histones are subjected to a number of post-trans-lational modifications such as phosphorylation (4–7), acetylation(8–10), andmethylation (11–14). Thesemodifications, termed “histonecode,” greatly impact the chromatin structure, which in turn leads topermissive or unfavorable access of transcription factors to a particularpromoter causing the differential transcriptional activity of that pro-moter (12, 15). For example, a high level of acetylation on certain lysineson the tail of histones H3 andH4 is often associated with loose chroma-tin structure, favorable binding of activators/co-activators, and highrates of transcription, whereas a low level of acetylation is usuallyaccompanied by compact chromatin structure, denial of access of acti-vators/co-activators to promoters, and low rate of transcription (16–19). It is not only the intensity of acetylation/deacetylation but also thespecific sites becoming acetylated or deacetylated that dictate the over-all transcriptional outcome.Acetylation of histones is a dynamic and reversible process that is

catalyzed by two groups of antagonistic enzymes called histone acetyl-transferase and histone deacetylase (HDAC).2 So far, 18 differentHDACshave been identified and categorized into three groups based ontheir homology to three yeast HDACs. Class I HDACs, containingHDAC1, -2, -3, and -8, are homologous to yeast yRPD3 and areexpressed inmost tissues (20–25). Class II HDACs, which share homol-ogy with yeast yHDA1, consist of HDAC4, -5, -6, -7, -9, and -10 andshow tissue-specific expression patterns (26–30). HDAC11 has homol-ogy to both class I and II. Both class I and class II HDACs are sensitive tosuch inhibitors as trichostatin A (TSA) and sodium butyrate. Class IIIHDACs are mammalian homologs to yeast SIR2 protein that have dis-tinct catalytic domains compared with both class I and II HDACs andare dependent on NAD for enzymatic activity (31, 32). Class III HDACsare insensitive to TSA treatment; instead, their activity can be inhibitedby nicotinamide. Class I HDACs are invariably found inmultimolecularcomplexes containing other repressor/co-repressor proteins. For exam-

* This work was supported in part by Veterans Affairs merit review project, NHLBI GrantsR01-HL68094 and P01-HL013262-31 (to B. D. S.) from the National Institutes of Health,and American Heart Association Post-doctoral Fellowship Grant 0525981T (to Y. X.).The costs of publication of this article were defrayed in part by the payment of pagecharges. This article must therefore be hereby marked “advertisement” in accordancewith 18 U.S.C. Section 1734 solely to indicate this fact.

1 To whom correspondence should be addressed: Dept. of Biochemistry, Boston Univer-sity School of Medicine, 715 Albany St., Boston, MA 02118. Tel.: 617-638-4159; Fax:617-638-5339; E-mail: smith@biochem.bumc.bu.edu.

2 The abbreviations used are: HDAC, histone deacetylase; RFX, regulatory factor forX-box; ChIP, chromatin immunoprecipitation; aza-dC,5-aza-2�-deoxycytidine; TSA,trichostatin A; IFN-�, interferon-�; PMSF, phenylmethylsulfonyl fluoride; ANOVA,analysis of variance; MHC, major histocompatibility; DMEM, Dulbecco’s modifiedEagle’s medium; FBS, fetal bovine serum; siHDAC, silencing HDACs; CIITA, class IItransactivator.

THE JOURNAL OF BIOLOGICAL CHEMISTRY VOL. 281, NO. 14, pp. 9260 –9270, April 7, 2006Printed in the U.S.A.

9260 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 281 • NUMBER 14 • APRIL 7, 2006

by guest on March 22, 2020

http://ww

w.jbc.org/

Dow

nloaded from

ple, the NuRD complex is composed of HDAC1, HDAC2, and methylgroup binding protein (33, 34). Another commonly found HDAC-con-taining complex is the Sin3 complex that consists of HDAC1 andHDAC2 in addition to Sin3-associated polypeptides SAP18 and SAP30(35–37).Themajor focus of research at this laboratory has been toward under-

standing the transcriptional events occurring at the transcription startsite of type I collagen genes. Earlier investigations have led to the dis-covery of a binding site for the regulatory factor for X-box (RFX) at thetranscription start site of both �1(I) and �2(I) genes, COL1A1 andCOL1A2 (38, 39). Binding of two members of RFX proteins, RFX1 andRFX5, to the start sites of the type I collagen genes represses theirexpression. RFX1 is able to form dimers with itself as well as with twoother RFX members, RFX2 and RFX3, with which RFX1 shares signifi-cant homology, including the dimerization domain that mediates thecomplex formation. On the other hand, RFX5, lacking the dimerizationdomain, is less homologous to other family members and forms a tri-meric complex with two other proteins, RFXB and RFXAP (1). RFX5 isalso responsible for recruiting class II transactivator (CIITA), themasterregulator for major histocompatibility II (MHC II) expression, to thecollagen transcription start site during IFN-� response (2). Althoughboth RFX1 and RFX5 down-regulate collagen expression, their distinctassociation with other transcription factors suggests that they areinvolved in different physiological and pathophysiological events lead-ing to the repression of collagen synthesis. Our results presented heredemonstrate that RFX1 and RFX5 differentially interact with class IHDACs, underlying the different pathways when repressing collagensynthesis.

MATERIALS AND METHODS

Cell Culture Maintenance and Treatment Protocols—Human lungfibroblasts, IMR-90, (IMR, NJ), human kidney cells 293FT (Invitrogen),and human fibrosarcoma cells HT1080 (ATCC, Manassas, VA) weremaintained in Dulbecco’s modified Eagle’s medium (DMEM) (Invitro-gen) supplemented with 10% fetal bovine serum (FBS, Hyclone) and 1%penicillin G/streptomycin (Sigma).In several studies, IMR-90 cells were treated with IFN-� and/or TSA

(Sigma). IMR-90 fibroblasts were plated in p150 tissue culture dishes at4� 106 cells/dish andmaintained in DMEMwith 10% FBS for 16–24 h.Cells were pretreated in DMEM with 0.4% FBS for 16 h prior to IFN-�treatment (100 units/ml in 0.4%DMEM for 0, 8, 16, or 24 h) and/or TSAtreatment (0.5–2 �M for 24 h where applicable). IFN-� and TSA wasadded together.In other studies, HT1080 cells were plated at a density of 5� 105 cells

per p35 tissue culture dish and treated with 5-aza-2�-deoxycytidine(aza-dC) (500 nM and 1 �M) for 3 days in DMEM adding fresh aza-dCeach day. In some studies HT1080 cells were treated with TSA (300 nM)for 24 h.

DNA Affinity Pull-down Assay—The collagen sequence (COL1A2�25/�30, GenBankTM accession number AF004877) with a HindIIIoverhang was synthesized as complementary strands and annealed asdescribed previously (40). Double-stranded collagen DNA was biotin-labeled by incubating with Klenow fragment (New England Biolabs,Beverly,MA) and biotin-14-dATP (Invitrogen) supplemented with reg-ular dCTP, dTTP, and dGTP at room temperature for 30 min. Thereaction mixture was phenol/chloroform-extracted and alcohol-pre-cipitated to remove unincorporated biotin.Nuclear protein extracts from IMR-90 cells were obtained as

described previously using 450mM sodiumchloride (1, 41). The strepta-vidin beads (Promega, Madison, WI) were washed three times with

ice-cold phosphate-buffered saline supplemented with 1 mM PMSF.Nuclear proteins (100–200�g) were pre-cleared by incubating with thewashed beads for 30 min at 4 °C on a shaking platform as describedpreviously (2). Pre-cleared nuclear proteins were prepared by capturingthe beads on a magnetic stand and removing the supernatant. Thesupernatant was then incubated with biotin-labeled collagen DNAprobe (�25/�30) for 1 h at room temperature in binding buffer (60mM

NaCl, 20 mM HEPES, pH 7.9, 0.1 mM EDTA, 4% glycerol, 2 mM dithio-threitol) supplemented with bovine serum albumin, poly(dI-dC), andsonicated salmon sperm DNA to remove nonspecific binding. DNA-protein complex formed was captured by the magnetic beads andwashed extensively with binding buffer supplemented with 0.01% Tri-ton X and 100 mM KCl. The bound proteins were eluted with 1� elec-trophoresis sample buffer by incubating at 90 °C for 10 min and ana-lyzed by SDS-polyacrylamide gels.

Plasmids, Transfections, and Luciferase Assays—The COL1A2-lucif-erase construct (pH20) (42) contains sequences from�221 to�54 bp ofmouse COL1A2 promoter fused to the luciferase reporter gene. Full-length class I HDAC expression constructs (HDAC1, HDAC2, andHDAC3) in pcDNA3 and silencing HDACs (43) were kindly providedby Dr. Edward Seto. Full-length FLAG-RFX5 were kindly provided byDr. Jenny Ting. RFX1 cDNA was excised from pHISB-RFX-1 plasmid,respectively, and inserted into the pIRES-hrGFP-2� construct (Strat-agene) that has green fluorescent protein-coding sequence. RFX1 plas-mid was digested with NotI and XhoI to produce the 5-kb vector alongwith 3-kb fragment. The RFX1 fragment was inserted into the EcoRI/XhoI site of the pIRES-hrGFP-2� plasmid. This bicistronic plasmid canencode the protein along with the green fluorescent protein whenexpressed in mammalian cell lines.Cells were plated at the density of 3 � 105 cells/well in 6-well tissue

culturedishes (for IMR-90cells) or5�106 cells perp100 tissueculturedish(for 293FT cells). Transfections were performed with Lipofectamine 2000reagent (Invitrogen) according to the manufacturer’s protocol. Cells wereharvested48hpost-transfection, and luciferaseassayswereperformedwitha luciferase reporter assay system (Promega).

Immunoprecipitations—To investigate whether factors interact invivo, co-immunoprecipitations were performed. Whole cell lysates(IMR-90 or 293FTwith transfected constructs as indicatedwhere appli-cable) were obtained by resuspending cell pellets in RIPA buffer (50mM

Tris, pH 7.4, 150 mMNaCl, 1% Triton X-100) with freshly added prote-ase inhibitor (Roche Applied Science) and PMSF (100 �g/ml RIPA).Anti-RFX5 (194, Rockland) or anti-RFX1 (I-19; Santa Cruz Biotechnol-ogy) antibody was added to and incubated with IMR-90 cell lysate over-night before being absorbed by protein A/G PLUS-agarose beads (SantaCruz Biotechnology). Precipitated immune complex was released byboiling with 1� SDS electrophoresis sample buffer. Alternatively,FLAG-conjugated beads (M2; Sigma) were added to and incubatedwith293FT cell lysate overnight. Precipitated immune complex was elutedwith 3� FLAG peptide (Sigma).

Westerns—Proteinswere separated by 8 or 10%polyacrylamide gel elec-trophoresis with pre-stained markers (Bio-Rad) for estimating molecularweight andefficiencyof transfer toblots. Proteinswere transferred tonitro-cellulose membranes (Bio-Rad) in a Mini-Trans-Blot Cell (Bio-Rad). Themembranes were blocked with 5% milk powder in Tris-buffered salinebuffer (TBST: 0.05%Tween20, 150mMNaCl, 100mMTris-HCl, pH7.4) at4 °C overnight and incubated for 3 h to monoclonal anti-FLAG (1:1000)(Sigma), polyclonal anti-RFX5 (194, 1:1000) (Rockland), polyclonal anti-mSin3A (K-20, 1:200) (SantaCruzBiotechnology), polyclonal anti-mSin3B(AK-12, 1:200) (Santa Cruz Biotechnology), anti-HDAC1 (H-51, 1:200)(Santa Cruz Biotechnology), polyclonal anti-HDAC2 (C-19, 1:200) (Santa

RFX Proteins Interact with HDACs

APRIL 7, 2006 • VOLUME 281 • NUMBER 14 JOURNAL OF BIOLOGICAL CHEMISTRY 9261

by guest on March 22, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Cruz Biotechnology), polyclonal anti-HDAC3 (H-99, 1:200) (Santa CruzBiotechnology), polyclonal N-acetyl-lysine (1:2000) (Cell Signaling), andpolyclonal anti-RFX1 (I-19, 1:100) (Santa Cruz Biotechnology) antibodiesat room temperature. After threewasheswithTBST, themembraneswereincubated with appropriate secondary antibodies, either anti-goat IgG(Sigma), anti-mouse IgG, or anti-rabbit IgG (AmershamBiosciences) con-jugated to horseradish peroxidase, for another 1 h at room temperature.Then protein blots were visualized using ECL reagent (PerkinElmer LifeSciences) on a Kodak image station (PerkinElmer Life Sciences).

Chromatin Immunoprecipitation (ChIP)—Chromatin in control andIFN-� treated cells were cross-linkedwith 1% formaldehyde for 8min atroom temperature, sequentially washed with phosphate-bufferedsaline, Solution I (10 mMHEPES, pH 7.5, 10 mM EDTA, 0.5 mM EGTA,0.75% Triton X-100), and Solution II (10 mM HEPES, pH 7.5, 200 mM

NaCl, 1 mM EDTA, 0.5 mM EGTA). Cells were incubated in lysis buffer(150 mM NaCl, 25 mM Tris, pH 7.5, 1% Triton X-100, 0.1% SDS, 0.5%deoxycholate) supplemented with protease inhibitor tablet (RocheApplied Science) and PMSF. DNAwas fragmented into�500-bp piecesusing a Branson 250 sonicator. Aliquots of lysates containing 200 �g ofprotein were used for each immunoprecipitation reaction with anti-RFX5 (194; Rockland), anti-RFX1 (I-19; Santa Cruz Biotechnology),anti-mSin3A (K-20; Santa Cruz Biotechnology), anti-mSin3B (AK-12;Santa Cruz Biotechnology), anti-HDAC1 (H-51; Santa Cruz Biotech-nology), anti-HDAC2 (C-19; Santa Cruz Biotechnology), and anti-HDAC3 (H-99; Santa Cruz Biotechnology) antibodies followed byadsorption to protein A/G PLUS-agarose beads (Santa Cruz Biotech-nology). Precipitated DNA-protein complexes were washed sequen-tially with RIPA buffer (50 mM Tris, pH 8.0, 150 mM NaCl, 0.1% SDS,0.5% deoxycholate, 1% Nonidet P-40, 1 mM EDTA), high salt buffer (50mM Tris, pH 8.0, 500 mM NaCl, 0.1% SDS, 0.5% deoxycholate, 1% Non-idet P-40, 1 mM EDTA), LiCl buffer (50 mM Tris, pH 8.0, 250 mM LiCl,0.1% SDS, 0.5% deoxycholate, 1% Nonidet P-40, 1 mM EDTA), and TEbuffer (10 mM Tris, 1 mM EDTA pH 8.0), respectively. DNA-proteincross-link was reversed by heating the samples to 65 °C overnight. Pro-teins were digested with proteinase K (Sigma), and DNA was phenol/chloroform-extracted and precipitated by 100% ethanol. Dried DNAwas dissolved in 50�l of deionized distilledwater, and 10�l was used for

each real time PCR. The primers surrounding the collagen start site forreal time PCR have been described previously (1).

RNA Isolation and Real Time PCR—Cells were harvested, and RNAwas extracted using an RNeasy RNA isolation kit (Qiagen, Valencia,CA) according to the manufacturer’s protocol. Reverse transcriptasereactions were performed using a SuperScript first-strand synthesis sys-tem (Invitrogen) according to the manufacturer’s protocol. Real timePCRs were performed on a ABI Prism 7700 sequence detection PCRmachine (Applied Biosystems, Foster City, CA) according to manufac-turer’s protocol. The oligonucleotide forward and reverse PCR primersand fluorescent probes are described in Table 1.

RESULTS

RFX1 and RFX5 Differentially Interact with Class I HDACs onCOL1A2 Transcription Start Site—Previously, we reported that twomembers of the RFX family, RFX1 and RFX5, bind to the transcriptionstart site of the COL1A2 gene and repress its expression (39). DuringIFN-� treatment, when RFX5 occupies the COL1A2 transcription startsite, there is a coordinate decrease in acetylation of histones (1). Becausetranscriptional repression is usually associated with the recruitment ofHDACs, we examined whether RFX1 and/or RFX5 is responsible forrecruiting the HDACs to the collagen site. To this end, DNA affinitypull-down experiments were performed, as described previously (2),with nuclear proteins extracted from human IMR-90 cells and a biotin-labeled double-stranded DNA probe that spans from�25 to�30 of thecollagen promoter containing the RFX-binding site. Different DNA oli-gonucleotides were also used as competitors for RFX binding to test thespecificity of the interactions. Streptavidin-conjugated magnetic beadswere used to sequester the biotinylated probe with bound nuclearproteins.After extensive washing, bound proteins were eluted with SDS elec-

trophoresis buffer, separated on SDS-10% PAGE, and transferred tomembranes for Western analysis using specific antibodies for RFX andHDAC family members. No proteins were present in the eluates with-out a DNA probe (Fig. 1, A, lane 2, and B, lane 3). In the presence of theDNA probe, RFX1 is detected in the eluates along with HDAC1 (Fig. 1,A, lane 3, and B, lane 4). When a methylated sequence, which acts as a

FIGURE 1. RFX1 interacts with HDAC1 and RFX5interacts with HDAC2 at COL1A2 transcriptionstart site in vitro. A, DNA affinity pull-down assayswere performed with IMR-90 nuclear extract asdescribed under “Materials and Methods.” Eluateswere separated by 10% SDS gel and Westerns per-formed with anti-RFX1, HDAC1, RFX5, HDAC2, orHDAC3 antibody as indicated. The original lysate(10%) was separated in lane 1. B, IFN-� increasesthe interaction between RFX5 and co-repressormolecules on COL1A2 promoter in vitro. DNA affin-ity pull-down assays were performed with IMR-90nuclear extract treated with (�) or without (�)IFN-� (100 units/ml) as described under “Materialsand Methods.” Eluates were separated by 10% SDSgel and Westerns performed with anti-RFX1,HDAC1, RFX5, HDAC2, mSin3B, or HDAC3 anti-body as indicated. The original lysate (10%) wasseparated in lane 1 (�IFN-�) and lane 2 (�IFN-�).

RFX Proteins Interact with HDACs

9262 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 281 • NUMBER 14 • APRIL 7, 2006

by guest on March 22, 2020

http://ww

w.jbc.org/

Dow

nloaded from

competitor specifically eliminating the binding of RFX1 (mpBR), isadded as demonstrated previously (40), HDAC1 binding is also lost (Fig.1,A, lane 4, and B, lane 5). This indicates that RFX1 selectively interactswith HDAC1 on the collagen start site.In contrast, RFX5 is able to bind toDNA in the absence of RFX1 along

with HDAC2 (Fig. 1, A, lane 4, and B, lane 5), suggesting a selectiveinteraction between RFX5 and HDAC2. The MHC II X-box sequenceabolishes the binding of RFX1 (Fig. 1A, lane 5) and RFX5 (Fig. 1B, lane6), as well as the two HDACs, whereas neither Sp1 nor Ets consensussequence alters the binding of RFX1, HDAC1, RFX5, orHDAC2, imply-ing that the interactions are specific. A third class I HDAC, HDAC3, isnot detectable under any of these circumstances, probably because it isnot recruited to the collagen transcription start site through RFXproteins.

IFN-� Enhances the Interaction between RFX5 and Co-repressors—Ear-lier studies indicated that IFN-� increased RFX5 protein levels, nuclearlocalization, and occupation on the collagen start site in IMR-90 cells(1). Because there was an interaction between RFX5 and HDACs, RFX5may recruit more HDACs and IFN-� could enhance the associationbetween RFX5 and HDACs at the collagen transcription start site.Increased HDACs could inhibit histone acetylation and ultimatelyrepress collagen transcription. To test this hypothesis, DNA affinitypull-down experiments were performed first with nuclear proteinsextracted from IMR-90 cells treated with or without IFN-� (100 units)for 24 h.Both RFX1 and HDAC1 bind to the collagen sequence with similar

affinity either with or without IFN-� treatment (Fig. 1B, comparelane 4 and lane 10). With the addition of methylated mpBR, bindingof RFX1 as well as HDAC1 is eliminated, whereas RFX5, HDAC2, andmSin3B associate with the probe (Fig. 1B, lane 5). Binding of RFX5,HDAC2, and mSin3B are all enhanced with IFN-� treatment (Fig. 1B,compare lane 5 to lane 11). The specificity of the interactions wasmain-tained in the presence of IFN-� (Fig. 1B, lanes 12–14), suggesting thatmore co-repressors are recruited to the collagen transcription start siteby RFX5.

Interactions betweenRFXProteins andHDACsAreDNA-independent—Next, co-immunoprecipitation experimentswere performedbetweenRFXproteins and HDACs to determine whether interactions depend on DNA.The first immunoprecipitation experiments were performed usingepitope-tagged RFX5 and HDAC-expressed proteins in 293FT cellsbecause they are fast-growing human kidney cells that stably express thelarge T antigen of SV40 allowing increased amounts of overexpressed pro-tein. FLAG-conjugated beads were used to precipitate protein extractsfrom 293FT cells transfected with either FLAG-tagged RFX5 or HDAC2.Eluates were separated by SDS-polyacrylamide gels and examined forendogenous proteins that might co-precipitate with FLAG-tagged pro-teins.As depicted in Fig. 2A, FLAG-RFX5co-precipitateswith endogenousHDAC2 (left panel) and vice versa (right panel), indicative of a reciprocalinteraction. In both cases, mSin3B is present in the immune complex.Next, nuclear proteins from IMR-90 cells treated with or without

IFN-� were precipitated using an anti-RFX5 or anti-HDAC2 antibodyto examine endogenous interactions. Eluates were separated by SDS-polyacrylamide gels and examined for HDAC2 (Fig. 2A, left panel) orRFX5 and HDAC2 (right panel). HDAC2 is immunoprecipitated withthe RFX5 antibody but not with a pre-immune IgG (Fig. 2B, right panel).More HDAC2 is co-immunoprecipitated with RFX5 antibody fromIFN-� treated cells. HDAC2 antibody also co-immunoprecipitatedincreased amounts of RFX5 after IFN-� treatment (Fig. 2B, right panel)suggesting increased interactions between these proteins during IFN-�treatment.Finally, to examine endogenous interactions, RFX1 was immunopre-

cipitated from IMR-90 nuclear extracts with an anti-RFX1 antibody.Proteins in the supernatant and precipitate were separated by SDS-polyacrylamide gels and examined for endogenous proteins that remainin the supernatant or co-precipitate. A considerable fraction of HDAC1co-precipitates with RFX1 (Fig. 2C). In contrast to RFX5, RFX1 did notco-precipitate with HDAC2 (Fig. 2C). These data suggest that RFX pro-teins probably associate with HDAC before recruiting them to the col-lagen transcription start site.

FIGURE 2. RFX proteins interact with HDACs independent of DNA. A, RFX5, HDAC2, and mSin3B interact with each other reciprocally. 293FT cell extracts (500 �g) with transfectedFLAG-RFX5 (left panel) or FLAG-HDAC2 (right panel) were precipitated using FLAG-conjugated beads (40 �l) (M2) or pre-immune IgG-conjugated beads (40 �l) (IgG) as described under“Materials and Methods.” Eluates were separated by 10% SDS gel and Westerns performed with anti-FLAG, RFX5, HDAC2, or mSin3B antibody as indicated. 10% of the original lysate was alsoloaded as input. B, RFX5 interacts with HDAC2 in IMR-90 cells. Whole cell extracts (500 �g) prepared from IMR-90 cells with (�) or without (�) IFN-� treatment were precipitated using ananti-RFX5 antibody (5�g) (left panel), an anti-HDAC2 antibody (5�g) (right panel), or anti-pre-immune IgG (5�g) as indicated. Eluates were separated by 10% SDS gel and Westerns performedwith an anti-HDAC2 antibody (left panel) or both anti-HDAC and anti-RFX5 antibody. The original lysate proteins (10%) was also loaded and labeled as input. IP, immunoprecipitation. C, RFX1interacts with HDAC1 but not HDAC2 in IMR-90 cells. Nuclear extracts (300 �g) prepared from IMR-90 cells were precipitated using an anti-RFX1 antibody (5 �g). Eluates were separated by10% SDS gel and Westerns performed with anti-HDAC1 or HDAC2 antibody as indicated. 10% of the original lysate was also loaded as input. Sup, supernatant.

RFX Proteins Interact with HDACs

APRIL 7, 2006 • VOLUME 281 • NUMBER 14 JOURNAL OF BIOLOGICAL CHEMISTRY 9263

by guest on March 22, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Class I HDACs Display Distinct Functions Regulating COL1A2 Pro-moter Activity—Because there were interactions between RFX proteinsand HDACs, we examined whether these interactions bear any func-tional significance. Transient transfections were performed in IMR-90cells with different HDAC expression constructs with or without RFXexpression plasmids. As shown in Fig. 3, both HDAC1 and HDAC2repress the collagen promoter activity, but HDAC3 does not signifi-cantly alterCOL1A2 promoter activity, which is in accordance with ourbinding data (Fig. 1). RFX5 represses better with HDAC2 (Fig. 3A).Similarly, RFX1 represses better with HDAC1 (Fig. 3B).Next, siHDACs (43) were transfected with or without RFX expres-

sion constructs (Fig. 3, C and D). Silencing either HDAC1 or HDAC2,but not HDAC3, activated collagen promoter activity. Surprisingly,without these HDAC enzymes, RFX5 also further activated the collagenpromoter (Fig. 3C). RFX1 remained a repressor with siHDAC2 andsiHDAC3 but lost its ability to repress collagen promoter withoutHDAC1. This further demonstrates the different functional interac-tions between the RFX family members.

RFX5/HDAC2/mSin3B Are Involved in IFN-�-mediated CollagenTranscriptional Repression—Because there is differential associationbetweenRFXproteins andHDACs, both physically and functionally, wehypothesized that RFX5/HDAC2 and RFX1/HDAC1might be involvedin different mechanisms responsible for collagen repression. Our pre-vious reports suggest that RFX5 complexmight be responsible formedi-ating IFN-� repression of collagen transcription (1, 2). Because certain

co-repressor proteins were associated with RFX5 on the collagen startsite in vitro in DNA affinity pull-down assays (Fig. 1) and the associationcould be increased by IFN-� (Fig. 1B), it was hypothesized that theserepressorsmight be recruited to the collagen site by RFX5 during IFN-�response in vivo to repress collagen expression. Chromatin immuno-precipitation (ChIP) assays were performed with anti-RFX5, RFX1,HDAC1, HDAC2, HDAC3, mSin3A as well as mSin3B antibodies inIMR-90 cells treated with IFN-� for 0, 8, 16, or 24 h.

Similar amounts of genomic DNA surrounding the COL1A2 tran-scription start site is precipitated by either RFX5 or RFX1 antibody (Fig.4A). During IFN-� treatment, however, binding of RFX5 to the collagensequence is greatly enhanced, but RFX1 binding is gradually and slightlydecreased, suggesting that RFX5, but not RFX1, is involved in IFN-�-mediated collagen repression.More DNAs are precipitated bymSin3A binding to the collagen start

site thanmSin3B before IFN-� is added (Fig. 4B, compare 0 hmSin3A to0 h mSin3B). Although binding of mSin3B increases with time duringIFN-� treatment,mSin3Aoccupancy is continuously decreased. By 24 hof IFN-� treatment, mSin3A binding is decreased to the level of mSin3Bbinding before IFN-� treatment, whereas binding of mSin3B is stimu-lated to the level of mSin3A binding at the beginning of IFN-� treat-ment. In other words, it appears as though IFN-� induces an exchangeof mSin3A for mSin3B on the collagen transcription start site.Meanwhile, when the occupancy of HDACs was examined, it was

discovered that HDAC2 was the predominant form of histone deacety-

FIGURE 3. RFX proteins functionally interactwith different HDACs on COL1A2 promoter.A, HDAC2 enhances repression of COL1A2 pro-moter by RFX5. A COL1A2 promoter construct(pH20, 0.5 �g) was co-transfected with differentclass I HDAC constructs (0.5 �g) in the presence(�) or absence (�) of RFX5 (0.5 �g), along withGFP (0.1 �g) for normalization, into IMR-90 cells induplicate wells as described under “Materials andMethods.” Average luciferase activities were nor-malized by both protein concentration and GFPfluorescence and were expressed as relative per-centage activity compared with the control groupin which an empty vector was transfected. Thisrepresentative experiment was repeated at leastthree times. B, HDAC1 enhances repression ofCOL1A2 promoter by RFX1. A COL1A2 promoterconstruct (pH20, 0.5 �g) was co-transfected withdifferent class I HDAC constructs (0.5 �g) in thepresence (�) or absence (�) of RFX5 (0.5 �g),along with GFP (0.1 �g) for normalization, intoIMR-90 cells in duplicate wells as described under“Materials and Methods.” Average luciferase activ-ities were normalized by both protein concentra-tion and GFP fluorescence and were expressed rel-ative to percentage activity compared with thecontrol group in which an empty vector was trans-fected. This representative experiment wasrepeated at least three times. C and D, silencingHDAC constructs activate COL1A2 promoter activ-ity. A COL1A2 promoter construct (pH20, 0.5 �g)was co-transfected with different class I siHDACconstructs (43) in the presence (�) or absence (�)of RFX5 (C, 0.5 �g) or RFX1 (D, 0.5 �g), along withGFP (0.1 �g) for normalization, into IMR-90 cells intriplicate wells as described under “Materials andMethods.” Average luciferase activities were nor-malized by both protein concentration and GFPfluorescence and were expressed as relative lucif-erase activity per �g of protein per ng of GFP. Thisrepresentative experiment was repeated at leasttwice.

RFX Proteins Interact with HDACs

9264 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 281 • NUMBER 14 • APRIL 7, 2006

by guest on March 22, 2020

http://ww

w.jbc.org/

Dow

nloaded from

lase present on the collagen transcription start site before IFN-� treat-ment. Binding of HDAC1 andHDAC3wasminimal compared with thebinding of HDAC2, although the amount of DNA precipitated by eitheranti-HDAC1 or anti-HDAC3 antibody was significantly above controllevels (Fig. 4C).

RFX1/HDAC1/mSin3A Are Involved in Methylation-mediatedCollagen Transcriptional Repression—Our recent studies indicatethat collagen expression is greatly decreased when the gene is meth-ylated (44) and that RFX1 binding to the collagen site is methylation-sensitive (38, 40), raising the possibility that RFX1 might be involvedin methylation-mediated collagen repression by recruiting co-re-pressors to the collagen transcription start site. To examine the valid-

ity of this hypothesis,DNAaffinity pull-downexperimentswereperformedusing either a regular biotinylated COL1A2DNA probe (�25 to �30)or the same probe that was methylated at the �7 site that enhancesthe binding of RFX1 to the collagen transcription start site (40). Asdepicted in Fig. 5A, binding of RFX1 was much more robust on themethylated probe than on the unmethylated probe as expected.Interestingly, both HDAC1 and mSin3A bind to the methylatedprobe more strongly just like RFX1, supporting the notion that dur-ing methylation of collagen DNA more RFX1 is able to bind to thetranscription start site and recruit more co-repressor complexes torepress collagen transcription. No RFX5 binding is detectable underthese conditions.

FIGURE 4. IFN-� induces a distinct set of repressors/co-repressors on COL1A2 transcription start site in vivo. A, IFN-� treatment leads to increased occupancy of RFX5 as well asdecreased occupancy of RFX1 on COL1A2 transcription start site. Chromatin immunoprecipitation assays were performed with IMR-90 cells treated with IFN-� (100 units/ml) for 0, 8,16, or 24 h as indicated using anti-RFX1 or anti-RFX5 antibodies as described under “Materials and Methods.” This experiment was repeated at least twice in duplicate wells. Arepresentative graph was shown. Data are expressed as picograms of DNA precipitated by the indicated antibody per ng of total genomic DNA. B, IFN-� treatment leads to increasedoccupancy of mSin3A as well as decreased occupancy of mSin3B on COL1A2 transcription start site. Chromatin immunoprecipitation assays were performed with IMR-90 cells treatedwith IFN-� (100units/ml) for 0, 8, 16, or 24 h as indicated using anti-mSin3A or anti-mSin3B antibodies as described under “Materials and Methods.” This experiment was repeatedthree times in duplicate wells and plotted as average � S.D. Data are expressed as picograms of DNA precipitated by indicated antibody per ng of total genomic DNA. One-wayANOVA was used to assess the statistical significance. **, p � 0.05; ***, p � 0.01. IFN-� treatment leads to increased occupancy of HDAC2, but not HDAC1 or HDAC3, on COL1A2transcription start site. Chromatin immunoprecipitation assays were performed with IMR-90 cells treated with IFN-� (100 units/ml) for 0, 8, 16, or 24 h as indicated using anti-HDAC1,HDAC2, or anti-HDAC3 antibodies as described under “Materials and Methods.” This experiment was repeated at least twice in duplicate wells. A representative graph is shown. Dataare expressed as picograms of DNA precipitated by the indicated antibody per ng of total genomic DNA.

RFX Proteins Interact with HDACs

APRIL 7, 2006 • VOLUME 281 • NUMBER 14 JOURNAL OF BIOLOGICAL CHEMISTRY 9265

by guest on March 22, 2020

http://ww

w.jbc.org/

Dow

nloaded from

The inhibitor, aza-dC, a compound that inhibits DNA methyltrans-ferases, dramatically (40-fold) increases collagen gene expression in afibrosarcoma cell line, HT1080, in which 50% of collagen genomic DNAis methylated (38, 44). Therefore, we examined whether the effect ofaza-dC is mediated through diminished binding of RFX1 as well co-repressors to the collagen transcription start site because of de-methy-lation. ChIP assays were performed with HT1080 cells treated withdifferent concentrations of aza-dC for 72 h using anti-RFX1,HDAC1, aswell as HDAC2 antibodies. RFX1 binding is decreased with aza-dCtreatment in a dose-response manner by up to 60% (Fig. 5B). Interest-ingly, HDAC1 binds to the partially methylated collagen transcriptionstart site in vivomuch more strongly than HDAC2 (Fig. 5C). Binding ofHDAC1 is also greatly decreased by aza-dC treatment similar to RFX1,further confirming that RFX1/HDAC1 might be involved in DNAmethylation-mediated collagen transcriptional repression.

TSAHasDifferential Cell-specific Effects on Steady StatemRNALevels—BecauseHDACsareclearly involved in the repressionof collagen transcrip-tion, we postulated that inhibition of overall HDACactivitywould increasecollagen transcription. TSA is a general inhibitor for class I and II HDACsthat alters expression of many genes presumably through histone deacety-lation. First, HT1080 cells were treated with several doses of TSA (50, 100,500, and 1000 nM) for 24 h. There was no change in collagenmRNA levelsat low doses. At higher doses of TSA there was a 5- (500 nM) or 10-fold(1000 nM) increase in collagenmRNA (Fig. 6A).On the other hand, when IMR-90 cells were treated with TSA (500

nM), there was no significant increase in collagenmRNA levels (Fig. 7B).IFN-� repressed collagen mRNA levels by 50% in 24 h. TSA partiallyreversed the repression of collagen mRNA expression.ThemRNA steady state levels are a combination of transcription and

degradation or processing of mRNA. In order to examine early tran-

FIGURE 5. Methylation of the collagen gene(COL1A2) enhances the interaction betweenRFX1 and co-repressors. A, methylation enhancesthe binding of RFX1, HDAC1, as well as mSin3A toCOL1A2 sequence in vitro. DNA affinity pull-downassays were performed with IMR-90 nuclear extractusing either unmethylated probe (U) or methylatedprobe (M) as described under “Materials and Meth-ods.” Eluates were separated by 10% SDS gel andWesterns performed with anti-RFX1, HDAC1, RFX5,or mSin3A antibody as indicated. 10% of the originallysate was also loaded as input (I). B and C, Aza-dCtreatment of HT1080 cells leads to decreased occu-pancy of RFX1 and HDAC1 on COL1A2 transcriptionstart site. Chromatin immunoprecipitation assayswere performed with HT1080 cells treated withAza-dC at indicated concentrations for 72 consecu-tive hours using anti-RFX1 (B), anti-HDAC1, or anti-HDAC2 (C) antibodies as described under “Materialsand Methods.” This experiment was repeated at leasttwice and a representative graph was shown. Dataare expressed as picograms of DNA precipitated byindicated antibody per ng of total genomic DNA.

FIGURE 6. TSA stimulates COL1A2 mRNA in a cell-specific manner. A, HT1080 cells were untreated ortreated with 300 nM of TSA for 24 h before harvestingas described under “Methods and Materials.” TotalRNAs were prepared and transcribed to form cDNA,and real time PCRs were performed with the cDNAsamples using primers to detect COL1A2 mRNA. Band C, IMR-90 cells were untreated or treated withIFN-� (100 units/ml) and/or TSA (500 nM) for 24 hbefore harvesting as described under “Materials andMethods.” Total RNAs were prepared and tran-scribed to form cDNA, and real time PCRs were per-formed with the cDNA samples using primers todetect COL1A2 mRNA (B) or COL1A2 heterogeneousnuclear RNA (hnRNA) (C) (see Table 1 for primers).Each experiment was repeated at least three times induplicate wells. Data are expressed as relative RNAlevels compared with control levels, normalized to18 S RNA, and presented as average � S.D. A, pairedsample t test, two-tailed; B and C, one-way ANOVAwas performed to evaluate the statistical signifi-cance; ***, p � 0.01.

RFX Proteins Interact with HDACs

9266 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 281 • NUMBER 14 • APRIL 7, 2006

by guest on March 22, 2020

http://ww

w.jbc.org/

Dow

nloaded from

scription, primers within the second intron and second exon (Table 1)were used to measure heterogeneous RNA transcripts. Total RNA wasDNase-treated, and no reverse transcriptase was used as controls to besure that RNA, not DNA, was measured by these primers. TSA treat-ment increased transcription of heterogeneous collagen mRNA morethan steady state mRNA in the IMR-90 cells (Fig. 7C), suggesting thatTSA increases transcription and degradation or processing of mRNA.IFN-� repressed collagen heterogeneous RNA to the same extent assteady state mRNA.

TSA Stimulates Collagen Promoter Activity and Partially BlocksRepression by RFX Proteins—Even though TSA effects on collagenmRNAwere low, TSA activated collagen promoter-luciferase activity inIMR-90 cells (Fig. 8A). Overexpression of RFX1 and RFX5 repressedcollagen synthesis, and TSA treatment was able to partially block thisrepression (Fig. 8B). RFX5 seemed to be more sensitive to TSA thanRFX1 because TSA (500 nM) almost completely blocked repression byRFX5 overexpression.

RFX Proteins Are Acetylated by TSA Treatment—Although HDACsare active on histones, several acetylated transcription factors aredeacetylated byHDACs. To determinewhether RFX proteins are acety-lated, IMR-90 cells were treated with 1 �M TSA, and the acetylation ofeither RFX5 or RFX1 was examined by immunoprecipitation followedby Western blot with a specific anti-lysine antibody. Most intriguingly,

both RFX5 (Fig. 7C) and RFX1 (Fig. 7D) are acetylated in vivo, and theiracetylation is dramatically stimulated by TSA treatment within 24 h.

DISCUSSION

Previously, we demonstrated that there is a binding site for the RFXfamily of transcription factors within the first exon (�1 to �20) of theCOL1A2 gene (40). Two members of the RFX family, RFX5 and RFX1,bind to this site and repress collagen expression (39). RFX1homodimersand RFX1/RFX2 heterodimers bind to the collagen gene with higheraffinity especially when it is methylated on the coding strand (38–40,44, 45). In this study, it is clear that RFX1 interacts on the collagen startsite withHDAC1 byDNAaffinity precipitation (Fig. 1). This interactionof RFX1 andHDAC1with collagenDNA is increasedwhen theCpG sitewithin the RFX consensus sequence is methylated (Fig. 5A). In normalfibroblasts, there is a small but measurable amount of RFX1 bindingwithin chromatin as judged by ChIP assays (Fig. 4A). Collagen type Igenes aremethylated inDNA from certain cancer lines and in colorectaltumors (44). The collagen gene inHT1080 cells is 50%methylated at theCpG within the RFX1-binding site (44), and aza-dC increases collagenexpression as much as 40-fold in this fibrosarcoma cell line (44, 46).RFX1 occupies the site 2-foldmore inHT1080 cells than in IMR-90 cells(data not shown). In HCT116 cells that have higher methylation statusof the collagen gene, RFX1 occupancy is further increased (data not

FIGURE 7. TSA stimulates COL1A2 promoter activ-ity in a dose-response manner. A and B, COL1A2promoter construct (pH20, 0.5 �g) was co-trans-fected with or without RFX expression constructs(0.5 �g), along with GFP (0.1 �g) for normalization,into IMR-90 cells in duplicate wells as describedunder “Materials and Methods.” 24 h after transfec-tions, cells were treated with TSA at indicated con-centrations for additional 24 h before harvesting.Luciferase activities were normalized by both pro-tein concentration and GFP fluorescence and wereexpressed relative percentage activity comparedwith the control group in which an empty vector wastransfected. This experiment was repeated threetimes and data were shown as average � S.D. One-way ANOVA was used to evaluate the statistical sig-nificance. IP, immunoprecipitation; IB, immunoblot.**, p � 0.05 and ***, p � 0.01. C and D, TSA enhancesthe acetylation of both RFX5 and RFX1. Whole cellextracts prepared from IMR-90 cells with TSA (500nM) treatment for the indicated period of time wereprecipitated using an anti-RFX5 antibody or an anti-RFX1 as indicated. Eluates were separated by 10%SDS gel and Westerns performed with anti-acetyl-ly-sine, RFX5, or RFX1 antibody as indicated.

RFX Proteins Interact with HDACs

APRIL 7, 2006 • VOLUME 281 • NUMBER 14 JOURNAL OF BIOLOGICAL CHEMISTRY 9267

by guest on March 22, 2020

http://ww

w.jbc.org/

Dow

nloaded from

shown), suggesting that RFX1 does interactmorewith the collagen startsite when the gene is methylated. Aza-dC, which decreases methylationby inhibiting DNAmethyltransferases, also decreases the occupancy ofRFX1 as well as HDAC1 on the collagen start site (Fig. 5). These dataagain suggest that RFX1 and HDAC1 interact at the DNA within chro-matin. In fact, RFX1 also co-immunoprecipitates with HDAC1, but notHDAC2 or HDAC3, suggesting that these protein-protein interactionsoccur independent of RFX1-DNA interactions.On the other hand, RFX5 interacts with two other RFX proteins,

RFXB and RFXAP, as well as CIITA to repress collagen transcriptionduring IFN-� stimulation (1, 2). This study demonstrates that RFX5occupies the collagen gene, and RFX5 occupancy increases with IFN-�treatment as RFX1 occupancy decreases (Fig. 4A). HDAC2, but notHDAC1 or HDAC3, also increases on the collagen gene with time ofIFN-� treatment (Fig. 4B) suggesting that HDAC2 is responsible for thedeacetylation of histones surrounding the collagen start site observedearlier (1). Therefore, although both RFX1 and RFX5 are repressors forthe COL1A2 gene, the underlying mechanisms might differ through

different co-repressor interactions that could account for altered cellu-lar responses impacting collagen expression. Two collagen-relatedevents, namely IFN-�-regulated collagen repression during pulmo-nary fibrosis or inflammation and methylation-mediated collagendown-regulation in carcinogenesis, have led to an association ofthese two events with the RFX family. RFX1 might be responsible formethylation, whereas RFX5 may be responsible for IFN-�-mediatedcollagen down-regulation.CIITA is also recruited to the collagen transcription start site by

RFX5with time of IFN-� treatment (2). Earlier studies demonstrate thatCIITA interacts with HDAC1 and mSin3A to terminate MHC II acti-vation (47, 48). However, this investigation suggests that HDAC2 andSin3B occupy the collagen transcription start site during IFN-� treat-ment (Fig. 4B). Our preliminary results suggest that CIITA can complexwithHDAC2 and Sin3B (data not shown). Although our effort so far hasbeen concentrated on RFX5 and CIITA, recent evidence has been pub-lished that the other two members of the RFX5 complex, RFXB andRFXAP, are both capable of interacting with certain histone-modifying

TABLE 1Primers for mRNA and hnRNA real time PCR

Gene Amplicon location Exon SequencesCOL1A2mRNAa 7568–7650/8929–8965Forward primer Exon 5 5�-GCCCCCCAGGCAGAGA-3�TaqMan probe Exon 5/6 6FAM-CCTGGTCTCGGTGGGAACTTTGCTG-TAMRAb

Reverse primer Exon 6 5�-CCAACTCCTTTTCCATCATACTGA-3�

COL1A2 hnRNAa 2468–2552Forward primer Exon1 5�-CTTGCAGTAACCTTATGCCTAGCA-3�TaqMan probe Exon 1/Intron 1 6FAM-CATGCCAATGTAAGTGCCTTCAGCTTGTT-TAMRAb

Reverse primer Intron 1 5�-CCCATCTAACCTCTCTACCCAGTCT-3�a GenBankTM accession number AF004877.b The TaqMan probes contained a fluorescent dye (FAM) and a quencher (TAMRA).

FIGURE 8. Model for RFX family repression of collagen gene expression. A schematic representation illustrating two distinct pathways for collagen repression by RFX1 and RFX5.RFX1 is in a complex with HDAC1 and Sin3A. Methylation of the collagen gene at CpG �7 activates RFX1 interaction with the collagen gene bringing HDAC1 to the site. RFX5differentially interacts with HDAC2 and Sin3B. IFN-� treatment increases expression of RFX5 complex proteins and CIITA allowing increased interaction with HDAC2/Sin3B as well.Once the RFX5 complex is assembled, it interacts with CIITA which may also interact with HDACs. The drawing at the bottom represents some possible interactions showing proteinssuch as TBD, TIFF, and TIFF and NFY that can interact with CIITA. The collagen promoter cis-acting consensus sites are designated (TATA, TATA box, or TFIID-binding site, CCAAT,CBF/NF-Y-binding site (78).

RFX Proteins Interact with HDACs

9268 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 281 • NUMBER 14 • APRIL 7, 2006

by guest on March 22, 2020

http://ww

w.jbc.org/

Dow

nloaded from

factors (49, 50), suggesting a universal role for RFX5 proteins in tran-scriptional control. RFXAP, for example, complexes with Brahmarelated gene 1 (BRG1), an ATPase-dependent chromatin remodelingmolecule, to regulate MHC II expression (49). RFXANK is a bindingpartner for HDAC4, although the biological significance of this interac-tion remains unclear (50). Therefore, there might be extensive interac-tions between histone-modifying proteins and RFX-associated tran-scription factors. This may explain why silencing of any one of the classI HDACs still caused an increase in collagen promoter activity. Cer-tainly other classes of HDACs, such as HDAC4 discussed above, couldbe involved in collagen repression through the RFX complex.Co-repressor complexes have been isolated containing several

HDACswithHDAC1 andHDAC2 usually isolated together with Sin3A(36, 37). However, recently interactions of HDACswith individual tran-scription factors have been noted especially in the Sp1/kruppel zincfinger family (51–54). HDACs interact with Sp1 protein without activa-tion of the enzyme activity by blocking access of the Sp1 protein to theDNA (53). Collagen is an Sp1-activated promoter, and therefore, therecruitment of HDAC to the collagen genemay block assess to Sp1 sitesin the promoter. In this case, HDAC represses transcription withoutactive catalytic activity.TSA inhibits multiple histone deacetylases in class I and class II caus-

ing wide ranging changes in gene expression. The effects of TSA oncollagen gene expression are dependent on cell type. In our studies, TSAactivated gene expression in the cancer cell line HT1080. Gene expres-sion of collagen in human hepatoma cells is induced by TSA accompa-nied by changes in histone acetylation (55). However, one of the celllines, PLC/PRF, had no detectable collagen expression. Our investiga-tions on this cell line indicate that the collagen gene is very highlymeth-ylated, and therefore, TSA cannot activate a gene silenced by completemethylation. On the other hand, TSA can induce collagen gene expres-sion in a variety of cancer cell lines with partially methylated collagengenes such as breast cancer cell line, MCF7, or colorectal cancer cellline, HCT116.3

Our studies also indicate that in fibroblasts TSA produces no signif-icant changes in steady state mRNA levels. However, this seems to becaused by a combination of increased transcription with accompanyingincreases in degradation and/or processing. Others have demonstratedthat TSA suppresses collagen synthesis in dermal fibroblasts andhepatic stellate cells (56–58). Although steady state mRNA and proteinlevels of collagen decrease after TSA treatment, nuclear run-on studiesindicated that transcription of type I collagen actuallywas activated thussupporting our results with heterogeneous nuclear RNA (58). Thedecrease in collagen found by these investigators is most likely due toindirect mechanisms involving increased degradation or processing ofcollagen mRNA and protein because of multiple changes in geneexpression from TSA inhibition of HDACs.The actual increased acetylation of RFX1 and RFX5 proteins with

TSA treatment suggests that the mechanism for repression of collagenthrough the RFX family may require active deacetylation activity ofHDAC. The role of acetylation of RFX family needs to be further ana-lyzed. Protein acetylation of transcription factors has been implicated inmultiple cellular processes, including differentiation, metabolism, pro-liferation, and survival against stress (59–62). NF-�B, c-Myc, Sp3, p53,FOXO,GATA, and SMAD7 have all been identified as targets for acety-lation (61, 63–66). For example, acetylation of SMAD7 is involved withstabilizing the protein (67). On the other hand, acetylation of p53 isrequired for efficient recruitment of co-activators to promoter regions

as well as activation of target genes (68–70). Transcriptional outputmediated by FOXO is controlled via a two-tiered mechanism of phos-phorylation and acetylation (71). FOXO and RFX family have similarwinged helix DNA binding domains (72) so they could be regulated in asimilar manner by acetylation. The acetylationmay alter the interactionof RFX1 with co-repressors or with DNA. The overall consequence ofacetylation is not well understood and seems to vary from one factor toanother and to depend on specific circumstances.The roles of HDACs in several diseases, especially in lung inflamma-

tion (73, 74) and in cancer (75), have been recently outlined. HDAC2specifically may decrease in chronic obstructive pulmonary diseasepatients causing corticosteroid resistance (74, 76). HDAC1 has beenassociated with certain types of breast cancer (77). These studies pointout the biological significance of acetylation and deacetylation and theneed to investigate the mechanism of action on individual genes asso-ciated with these diseases.In summary, the RFX family represses collagen transcription at the

transcription start site through either methylation-specific binding byRFX1–3 family members or through IFN-�-stimulated binding duringinflammation by the RFX5-CIITA complex (Fig. 8). In each case, a spe-cific co-repressor complex most likely contains RFX family members.RFX1 interacts best with HDAC1 and mSin3A, whereas RFX5/CIITAinteracts with HDAC2 and mSin3B on the collagen transcription startsite. During IFN-� treatment, RFX5 synthesis, translocation to thenucleus, and complex formation are increased (1). CIITA is expressedand recruited to the collagen start site with RFX5 (2). Most likely RFX5becomes deacetylated by HDAC2, which increases complex interac-tions and repressor activity of RFX5. If the collagen gene is methylated,RFX1 interacts with the collagen gene and represses transcriptionthrough HDAC1 deacetylation of chromatin. This study reveals thatdifferent members of the same family of transcription factors represscollagen transcription through similar but distinct mechanisms.

REFERENCES1. Xu, Y., Wang, L., Buttice, G., Sengupta, P. K., and Smith, B. D. (2003) J. Biol. Chem.

278, 49134–491442. Xu, Y., Wang, L., Buttice, G., Sengupta, P. K., and Smith, B. D. (2004) J. Biol. Chem.

279, 41319–413323. Myllyharju, J., and Kivirikko, K. I. (2004) Trends Genet. 20, 33–434. Hauser, C., Schuettengruber, B., Bartl, S., Lagger, G., and Seiser, C. (2002)Mol. Cell.

Biol. 22, 7820–78305. Biade, S., Stobbe, C. C., Boyd, J. T., and Chapman, J. D. (2001) Int. J. Radiat. Biol. 77,

1033–10426. Mishra, S. K., Mandal, M., Mazumdar, A., and Kumar, R. (2001) FEBS Lett. 507,

88–947. Loury, R., and Sassone-Corsi, P. (2003)Methods 31, 40–488. Tishchenko, L. I., Miul’berg, A. A., and Ashmarin, I. P. (1971) Biokhimiya 36,

595–6039. Krieger, D. E., Levine, R., Merrifield, R. B., Vidali, G., and Allfrey, V. G. (1974) J. Biol.

Chem. 249, 332–33410. Boffa, L. C., Vidali, G., Mann, R. S., and Allfrey, V. G. (1978) J. Biol. Chem. 253,

3364–336611. Hendzel, M. J., and Davie, J. R. (1991) Biochem. J. 273, 753–75812. Jenuwein, T., and Allis, C. D. (2001) Science 293, 1074–108013. Boggs, B. A., Cheung, P., Heard, E., Spector, D. L., Chinault, A. C., and Allis, C. D.

(2002) Nat. Genet. 30, 73–7614. Fournier, C., Goto, Y., Ballestar, E., Delaval, K., Hever, A. M., Esteller, M., and Feil, R.

(2002) EMBO J. 21, 6560–657015. Strahl, B. D., and Allis, C. D. (2000) Nature 403, 41–4516. Boffa, L. C., Walker, J., Chen, T. A., Sterner, R., Mariani, M. R., and Allfrey, V. G.

(1990) Eur. J. Biochem. 194, 811–82317. Kuo, M. H., and Allis, C. D. (1998) BioEssays 20, 615–62618. Walia, H., Chen, H. Y., Sun, J. M., Holth, L. T., and Davie, J. R. (1998) J. Biol. Chem.

273, 14516–1452219. Alland, L.,Muhle, R., Hou,H., Jr., Potes, J., Chin, L., Schreiber-Agus, N., andDePinho,

R. A. (1997) Nature 387, 49–5520. Laherty, C.D., Yang,W.M., Sun, J.M., Davie, J. R., Seto, E., andEisenman, R.N. (1997)3 Y. Xu, P. K. Sengupta, E. Seto, and B. D. Smith, unpublished results.

RFX Proteins Interact with HDACs

APRIL 7, 2006 • VOLUME 281 • NUMBER 14 JOURNAL OF BIOLOGICAL CHEMISTRY 9269

by guest on March 22, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Cell 89, 349–35621. Gao, L., Cueto, M. A., Asselbergs, F., and Atadja, P. (2002) J. Biol. Chem. 277,

25748–2575522. Dangond, F., Henriksson, M., Zardo, G., Caiafa, P., Ekstrom, T. J., and Gray, S. G.

(2001) Int. J. Oncol. 19, 773–77723. Kasten, M. M., Dorland, S., and Stillman, D. J. (1997)Mol. Cell. Biol. 17, 4852–485824. Buggy, J. J., Sideris,M. L.,Mak, P., Lorimer, D.D.,McIntosh, B., andClark, J.M. (2000)

Biochem. J. 350, 199–20525. Dangond, F., Hafler, D. A., Tong, J. K., Randall, J., Kojima, R., Utku, N., and Gullans,

S. R. (1998) Biochem. Biophys. Res. Commun. 242, 648–65226. Fischle, W., Emiliani, S., Hendzel, M. J., Nagase, T., Nomura, N., Voelter, W., and

Verdin, E. (1999) J. Biol. Chem. 274, 11713–1172027. Grozinger, C. M., Hassig, C. A., and Schreiber, S. L. (1999) Proc. Natl. Acad. Sci.

U. S. A. 96, 4868–487328. Fischer, D. D., Cai, R., Bhatia, U., Asselbergs, F. A., Song, C., Terry, R., Trogani, N.,

Widmer, R., Atadja, P., and Cohen, D. (2002) J. Biol. Chem. 277, 6656–666629. Guardiola, A. R., and Yao, T. P. (2002) J. Biol. Chem. 277, 3350–335630. Kao, H. Y., Lee, C. H., Komarov, A., Han, C. C., and Evans, R. M. (2002) J. Biol. Chem.

277, 187–19331. Kyrylenko, S., Kyrylenko,O., Suuronen, T., and Salminen, A. (2003)Cell.Mol. Life Sci.

60, 1990–199732. Thiagalingam, S., Cheng, K. H., Lee, H. J., Mineva, N., Thiagalingam, A., and Ponte,

J. F. (2003) Ann. N. Y. Acad. Sci. 983, 84–10033. Jones, P. L., Veenstra, G. J., Wade, P. A., Vermaak, D., Kass, S. U., Landsberger, N.,

Strouboulis, J., and Wolffe, A. P. (1998) Nat. Genet. 19, 187–19134. Zhang, Y., Ng, H. H., Erdjument-Bromage, H., Tempst, P., Bird, A., and Reinberg, D.

(1999) Genes Dev. 13, 1924–193535. Hassig, C. A., Tong, J. K., Fleischer, T. C., Owa, T., Grable, P. G., Ayer, D. E., and

Schreiber, S. L. (1998) Proc. Natl. Acad. Sci. U. S. A. 95, 3519–352436. Zhang, Y., Iratni, R., Erdjument-Bromage,H., Tempst, P., andReinberg,D. (1997)Cell

89, 357–36437. Zhang, Y., Sun, Z. W., Iratni, R., Erdjument-Bromage, H., Tempst, P., Hampsey, M.,

and Reinberg, D. (1998)Mol. Cell 1, 1021–103138. Sengupta, P., Xu, Y., Wang, L., Widom, R., and Smith, B. D. (2005) J. Biol. Chem. 280,

21004–2101439. Sengupta, P. K., Fargo, J., and Smith, B. D. (2002) J. Biol. Chem. 277, 24926–2493740. Sengupta, P. K., Erhlich, M., and Smith, B. D. (1999) J. Biol. Chem. 274, 36649–3665541. Schreiber, E.,Matthias, P.,Muller,M.M., and Schaffner,W. (1989)Nucleic Acids Res.

17, 641942. Goldberg, H., Helaakoski, T., Garrett, L. A., Karsenty, G., Pellegrino, A., Lozano, G.,

Maity, S., and de Crombrugghe, B. (1992) J. Biol. Chem. 267, 19622–1963043. Zhang, X., Wharton, W., Yuan, Z., Tsai, S. C., Olashaw, N., and Seto, E. (2004)Mol.

Cell. Biol. 24, 5106–511844. Sengupta, S., Smith, E. M., Kim, K., Murnane, M. J., and Smith, B. D. (2003) Cancer

Res. 63, 1789–179745. Sengupta, P. K., and Smith, B. D. (1998) Biochim. Biophys. Acta 1443, 75–8946. Chiba, T., Yokosuka, O., Fukai, K., Hirasawa, Y., Tada, M., Mikata, R., Imazeki, F.,

Taniguchi, H., Iwama, A., Miyazaki, M., Ochiai, T., and Saisho, H. (2005) Eur. J.Cancer 41, 1185–1194

47. Zika, E., Greer, S. F., Zhu, X. S., and Ting, J. P. (2003)Mol. Cell. Biol. 23, 3091–310248. Zika, E., and Ting, J. P. (2005) Curr. Opin. Immunol. 17, 58–6449. Mudhasani, R., and Fontes, J. D. (2005)Mol. Immunol. 42, 673–68250. Wang, A. H., Gregoire, S., Zika, E., Xiao, L., Li, C. S., Li, H., Wright, K. L., Ting, J. P.,

and Yang, X. J. (2005) J. Biol. Chem. 280, 29117–29127

51. Medugno, L., Florio, F., De Cegli, R., Grosso, M., Lupo, A., Costanzo, P., and Izzo, P.(2005) Gene (Amst.) 359, 35–43

52. Matsumura, T., Suzuki, T., Aizawa, K., Munemasa, Y., Muto, S., Horikoshi, M., andNagai, R. (2005) J. Biol. Chem. 280, 12123–12129

53. Kang, J. E., Kim,M. H., Lee, J. A., Park, H.,Min-Nyung, L., Auh, C. K., andHur,M.W.(2005) Cell. Physiol. Biochem. 16, 23–30

54. Zhao, S., Venkatasubbarao, K., Li, S., and Freeman, J. W. (2003) Cancer Res. 63,2624–2630

55. Chiba, T., Yokosuka, O., Fukai, K., Kojima, H., Tada, M., Arai, M., Imazeki, F., andSaisho, H. (2004) Oncology 66, 481–491

56. Niki, T., Rombouts, K., De Bleser, P., De Smet, K., Rogiers, V., Schuppan, D., Yoshida,M., Gabbiani, G., and Geerts, A. (1999) Hepatology 29, 858–867

57. Rombouts, K., Niki, T.,Wielant, A., Hellemans, K., andGeerts, A. (2001)ActaGastro-Enterol. Belg. 64, 239–246

58. Rombouts, K., Niki, T., Greenwel, P., Vandermonde, A., Wielant, A., Hellemans, K.,De Bleser, P., Yoshida, M., Schuppan, D., Rojkind, M., and Geerts, A. (2002) Exp. CellRes. 278, 184–197

59. Brunet, A., Sweeney, L. B., Sturgill, J. F., Chua, K. F., Greer, P. L., Lin, Y., Tran,H., Ross,S. E.,Mostoslavsky, R., Cohen,H. Y., Hu, L. S., Cheng,H. L., Jedrychowski,M. P., Gygi,S. P., Sinclair, D. A., Alt, F. W., and Greenberg, M. E. (2004) Science 303, 2011–2015

60. van der Horst, A., Tertoolen, L. G., de Vries-Smits, L. M., Frye, R. A., Medema, R. H.,and Burgering, B. M. (2004) J. Biol. Chem. 279, 28873–28879

61. Sakaguchi, K., Herrera, J. E., Saito, S., Miki, T., Bustin, M., Vassilev, A., Anderson,C. W., and Appella, E. (1998) Genes Dev. 12, 2831–2841

62. You, H., and Mak, T. W. (2005) Cell Cycle 4, 37–3863. Furia, B., Deng, L., Wu, K., Baylor, S., Kehn, K., Li, H., Donnelly, R., Coleman, T., and

Kashanchi, F. (2002) J. Biol. Chem. 277, 4973–498064. Braun, H., Koop, R., Ertmer, A., Nacht, S., and Suske, G. (2001)Nucleic Acids Res. 29,

4994–500065. Vervoorts, J., Luscher-Firzlaff, J. M., Rottmann, S., Lilischkis, R., Walsemann, G.,

Dohmann, K., Austen, M., and Luscher, B. (2003) EMBO Rep. 4, 484–49066. Boyes, J., Byfield, P., Nakatani, Y., and Ogryzko, V. (1998) Nature 396, 594–59867. Simonsson,M., Heldin, C. H., Ericsson, J., and Gronroos, E. (2005) J. Biol. Chem. 280,

21797–2180368. Barlev, N. A., Liu, L., Chehab, N. H., Mansfield, K., Harris, K. G., Halazonetis, T. D.,

and Berger, S. L. (2001)Mol. Cell 8, 1243–125469. Ito, A., Lai, C.H., Zhao, X., Saito, S., Hamilton,M.H., Appella, E., andYao, T. P. (2001)

EMBO J. 20, 1331–134070. Luo, J., Li, M., Tang, Y., Laszkowska, M., Roeder, R. G., and Gu,W. (2004) Proc. Natl.

Acad. Sci. U. S. A. 101, 2259–226471. Accili, D., and Arden, K. C. (2004) Cell 117, 421–42672. Gajiwala, K. S., and Burley, S. K. (2000) Curr. Opin. Struct. Biol. 10, 110–11673. Blanchard, F., and Chipoy, C. (2005) Drug Discov. Today 10, 197–20474. Barnes, P. J., Adcock, I. M., and Ito, K. (2005) Eur. Respir. J. 25, 552–56375. Nebbioso, A., Clarke, N., Voltz, E., Germain, E., Ambrosino, C., Bontempo, P., Al-

varez, R., Schiavone, E. M., Ferrara, F., Bresciani, F., Weisz, A., de Lera, A. R., Grone-meyer, H., and Altucci, L. (2005) Nat. Med. 11, 77–84

76. Ito, K., Ito, M., Elliott, W. M., Cosio, B., Caramori, G., Kon, O. M., Barczyk, A.,Hayashi, S., Adcock, I. M., Hogg, J. C., and Barnes, P. J. (2005) N. Engl. J. Med. 352,1967–1976

77. Zhang, Z., Yamashita, H., Toyama, T., Sugiura,H., Ando, Y.,Mita, K., Hamaguchi,M.,Hara, Y., Kobayashi, S., and Iwase, H. (2005) Breast Cancer Res. Treat. 94, 11–16

78. Hatamochi, A., Paterson, B., and de Crombrugghe, B. (1986) J. Biol. Chem. 261,11310–11314

RFX Proteins Interact with HDACs

9270 JOURNAL OF BIOLOGICAL CHEMISTRY VOLUME 281 • NUMBER 14 • APRIL 7, 2006

by guest on March 22, 2020

http://ww

w.jbc.org/

Dow

nloaded from

Yong Xu, Pritam K. Sengupta, Edward Seto and Barbara D. Smith) ExpressionCOL1A22(I) Gene (αDeacetylases to Repress Collagen

Regulatory Factor for X-box Family Proteins Differentially Interact with Histone

doi: 10.1074/jbc.M511724200 originally published online February 6, 20062006, 281:9260-9270.J. Biol. Chem. 

  10.1074/jbc.M511724200Access the most updated version of this article at doi:

 Alerts:

  When a correction for this article is posted• 

When this article is cited• 

to choose from all of JBC's e-mail alertsClick here

  http://www.jbc.org/content/281/14/9260.full.html#ref-list-1

This article cites 78 references, 39 of which can be accessed free at

by guest on March 22, 2020

http://ww

w.jbc.org/

Dow

nloaded from