Biochimica et Biophysica Acta 1803 (2010) 1276–1286

Contents lists available at ScienceDirect

Biochimica et Biophysica Acta

j ourna l homepage: www.e lsev ie r.com/ locate /bbamcr

Primary effect of 1α,25(OH)2D3 on IL-10 expression in monocytes is short-termdown-regulation

Juha M. Matilainen a, Tiia Husso a,b, Sari Toropainen a, Sabine Seuter c, Mikko P. Turunen b, Petra Gynther a,Seppo Ylä-Herttuala b, Carsten Carlberg a,c, Sami Väisänen a,⁎a Department of Biosciences, University of Eastern Finland, FIN-70211 Kuopio, Finlandb Ark Therapeutics Oy, Neulaniementie 2L9, Kuopio, Finlandc Life Sciences Research Unit, University of Luxembourg, L-1511 Luxembourg, Luxembourg

Abbreviations: 1α,25(OH)2D3, 1α,25-dihydroxyvconformation capture; ChIP, chromatin immunoprecipeverted repeat; FBS, fetal bovine serum; IL, interleukin; Lmicrococcal nuclease; NCoR, nuclear co-repressor; Rtranscription start site; VDIR, VDR-interacting repressVDRE, vitamin D response element⁎ Corresponding author. Department of Biosciences, Un

Box 1627, FIN-70211 Kuopio, Finland. Tel.: +358 40 3553E-mail address: Sami.Vaisanen@uef.fi (S. Väisänen).

0167-4889/$ – see front matter © 2010 Elsevier B.V. Adoi:10.1016/j.bbamcr.2010.07.009

a b s t r a c t

a r t i c l e i n f o

Article history:Received 23 March 2010Received in revised form 7 July 2010Accepted 26 July 2010Available online 4 August 2010

Keywords:ChromatinIL-10VDRTranscription

The biologically most active vitamin D compound, 1α,25-dihydroxyvitamin D3 (1α,25(OH)2D3), influencesthe status of inflammation by modulating the expression of several cytokine genes. In this study, we haveexamined the mechanism of transcriptional regulation of interleukin 10 (IL-10) by 1α,25(OH)2D3 inlipopolysaccharide (LPS)-treated human monocytes (THP-1). Quantitative PCR showed that IL-10 mRNAexpression was significantly down-regulated (2.8-fold) during the first 8 h of 1α,25(OH)2D3 treatment,while after 48 h it was up-regulated (3-fold). Gel shift and quantitative chromatin immunoprecipitation(ChIP) assays showed that the vitamin D receptor (VDR) binds in a cyclical fashion to a promoter region1500–1700 bp upstream of the IL-10 transcription start site (TSS) containing two conserved VDR bindingsites. Targeting of VDR binding sites by enhancer specific duplex RNAs revealed that only the more distalelement is functional and chromosome conformation capture analysis suggested that this region loops 1α,25(OH)2D3-dependently to the TSS. Quantitative ChIP and micrococcal nuclease assays also revealed 1α,25(OH)2D3-dependent cyclical epigenetic changes and nucleosome remodeling at this promoter region. Inconclusion, in LPS-treated THP-1 cells the primary effect of 1α,25(OH)2D3 on IL-10 expression is down-regulation, which is achieved via a cyclical recruitment of VDR to the promoter.

itamin D3; 3C, chromosomeitation; DR, direct repeat; ER,PS, lipopolysaccharide; MNase,XR, retinoid X receptor; TSS,or; VDR, vitamin D receptor;

iversity of Eastern Finland, P.O.064; fax: +358 17 2811510.

ll rights reserved.

© 2010 Elsevier B.V. All rights reserved.

1. Introduction

The biologically most active vitamin D compound, 1α,25(OH)2D3,has besides its classical properties on regulation of serum calcium andphosphorus levels and bone formation also an important role as amodulator of immune responses. 1α,25(OH)2D3 regulates thetranscription of several cytokine genes, for example interleukin (IL)2, 10 and 12 and interferon γ in T cells and dendritic cells [1–3]. Therole of 1α,25(OH)2D3 as a regulator of immune responses is furtherunderlined by the fact that the vitamin D receptor (VDR) is expressedin antigen presenting cells [4,5] and that activated macrophages areable to synthesize and secrete 1α,25(OH)2D3 [6].

VDR binds as a heterodimer with the retinoid X receptor (RXR) tospecific genomic loci, so-called vitamin D response elements (VDREs),within the regulative regions of primary 1α,25(OH)2D3 target genes.This results in the activation or the repression of these genes. VDR–RXR heterodimers can bind to VDREs that are formed by direct repeats(DRs) of the hexameric consensus sequence RGKTCA (R=A or G,K=G or T) with three or four intervening nucleotides (DR3 and DR4)or by everted repeats (ERs) with six to nine spacing nucleotides (ER6to ER9) [7–9]. Recent studies have suggested that the association ofVDR with regulatory regions of primary up-regulated 1α,25(OH)2D3

target gene is cyclic and that this cyclical behavior also mirrors to themRNA expression level [10–12]. So far only MYC is known as anexample of a primary down-regulated 1α,25(OH)2D3 target gene thatshows transcriptional cycling [13].

The 1α,25(OH)2D3-dependent repression of VDR target genesappears to be mediated either through E-box type elements [14] orvia consensus VDREs [15–18]. The repression via E-box type elementsinvolves the association of VDR-interacting repressors (VDIRs),which inturn recruit VDR–RXR complexes and other proteins needed fortranscriptional repression, such as the co-repressor NCoR [19,20].However, the repression via consensus VDREs is still poorly understood.

IL-10 primarily dampens the immune response, since it isproduced mainly by regulatory T cells and macrophages and inhibits

1277J.M. Matilainen et al. / Biochimica et Biophysica Acta 1803 (2010) 1276–1286

the synthesis of pro-inflammatory IL-12 in macrophages and of co-stimulatory proteins and major histocompatibility II complexes indendritic cells [21]. However, IL-10 also enhances B cell survival andantibody production and thus it can also stimulate the immuneresponse. Thismay explain inconsistent reports on the effects of 1α,25(OH)2D3 on IL-10 expression, some of which claimed repressiveeffects [22–24] but others stimulatory results [25–28].

In this study, we focused on the primary effects of 1α,25(OH)2D3

on IL-10 expression and their mechanisms. Our data suggest that IL-10is repressed during the first 8 h after onset of 1α,25(OH)2D3

treatment in LPS-treated THP-1 cells. However, 48 h later IL-10mRNA expression is clearly increased suggesting a secondary effect.Highly similar, bidirectional effect of 1α,25(OH)2D3 on IL-10 expres-sion could be observed also in primary monocytes. According to ourpromoter targeting data the primary effect of 1α,25(OH)2D3 appearsto bemediated via an ER8-type VDRE located 1721–1701 bp upstreamof the TSS. Chromosome conformation capture (3C) analysissuggested that this promoter region loops in a ligand- and time-dependent fashion with the TSS of the IL-10 gene. Quantitative ChIPassays revealed that the promoter region and the TSS recruit VDR, RXRand the co-repressor NCoR in a cyclical fashion as well as markers ofactive and repressed chromatin. This could be confirmed byassessment of nucleosome enrichment.

2. Material and methods

2.1. Cell culture

THP-1 human acute monocytic leukemia cells were maintained inRPMI-1640 medium (Sigma-Aldrich, St. Louis, MO, USA) and SW-480human colon adenocarcinoma cell line was cultured in Dulbecco'sModified Eagle Medium (DMEM). Both media contained 10% fetalbovine serum (FBS), 2 mM L-glutamine, 0.1 mg/ml streptomycin and100 U/ml penicillin and culturing was in a humidified 95% air/5% CO2

incubator. Before use in experimental procedures, FBS was strippedfrom lipophilic compounds, such as endogenous nuclear receptorligands, by stirring it with 5% activated charcoal (Sigma-Aldrich) for3 h at room temperature. Charcoal was then removed by centrifuga-tion, and the medium was sterilized by filtration (0.2 μm pore size).THP-1 and SW480 cells were maintained for experiments in phenolred-free DMEM, supplemented with 5% charcoal-stripped FBS, 2 mML-glutamine, 0.1 mg/ml streptomycin and 100 U/ml penicillin. Priorto RNA or chromatin extraction, the cells were treatedwith 100 ng/mlLPS for 24 h and then exposed to either solvent (ethanol, 0.1% finalconcentration) or 10 nM 1α,25(OH)2D3 (kindly provided by MilanUskokovic, Bioxell Inc, Nutley, NJ, USA). For ChIP assays THP-1 cellswere first treated with LPS as above, then with 2.5 μM α-amanitin(Sigma-Aldrich) for 2 h, after which it was removed, followed byexposure to 1α,25(OH)2D3. For mRNA stability studies THP-1 cellswere first treated with LPS as above, after which 1 μg/ml actinomycinD (Sigma-Aldrich) was added.

2.2. Human monocyte isolation

To obtain pure primary monocyte preparations from buffy coats(provided by the Red Cross, Luxembourg), first peripheral bloodmononuclear cells (PBMCs) were isolated by density gradientcentrifugation using Ficoll–Paque Plus (GE Healthcare, Diegem,Belgium) and Leucosep tubes (Greiner bio-one, Wemmel, Belgium).The layer containing PBMC and platelets was collected and washedtwice withMACS buffer. The number of PBMCwas determined using ahemacytometer. Then, CD14+ cells are magnetically labeled withcolloidal superparamagnetic CD14 MicroBeads (Miltenyi Biotech,Utrecht, The Netherlands), which are conjugated with monoclonalanti-human CD14 antibodies, according to the manufacturer'sprotocol. Briefly, after incubation of cells and microbeads (15 min at

4 °C), cells were washed with MACS buffer (PBS pH 7.2, 2 mM EDTA,0.5% (w/v) BSA) and resuspended. The cell suspension was loadedonto a MACS LS separation column (Miltenyi Biotech, Utrecht, TheNetherlands) in a magnetic field. Thus, only magnetically labeledCD14+ cells were retained in the column and were eluted with MACSbuffer from the column after washing and removal of the column fromthe magnetic field. The purity of the CD14+ cells was evaluated byflow cytometry using anti-CD14-FITC and anti-IgG2a-FITC antibodies(ImmunoTools) and a FACSCantoII Flow Cytometry System (BectonDickinson). The number of the primary monocytes was determinedusing a hemacytometer. Then, the cells were pelleted and resus-pended in prewarmed RPMI-1640medium (Lonza, Verviers, Belgium)containing 10% FBS, 2 mM L-glutamine, 0.1 mg/ml streptomycin, and100 U/ml penicillin to a density of 0.8–1×106 cells/ml and incubatedovernight in a humidified 95% air/5% CO2 incubator. Then, ligandtreatments were performed followed by total RNA extraction.

2.3. Total RNA extraction, cDNA synthesis, and real-time quantitativePCR

Total RNA was extracted using High Pure RNA Isolation Kit (RocheDiagnostics, Mannheim, Germany) and cDNA synthesis was performedusing Transcriptor High Fidelity cDNA Synthesis Kit (Roche) accordingto the manufacturer's instructions. Real-time quantitative PCR wasperformed with a LightCycler 480 apparatus (Roche) using TaqManGene ExpressionAssays (Applied Biosystems, Carlsbad, CA, USA) for IL-10(Hs00174086_m1), IL-12B (Hs00233688_m1), TLR-4 (Hs00152939_m1)and RPLP0 (4333761F) and Maxima™ Probe qPCR Master Mix (Fer-mentas, Vilnius, Lithuania). PCR cycling conditions were: 10 min at 95 °C,followed by 50 cycles of 15 s at 95 °C and 60 s at 60 °C. Fold changeswerecalculated using the formula 2−(ΔΔCt), where ΔΔCt=ΔCt(1α,25(OH)2D3)−ΔCt(EtOH), and ΔCt=Ct(target gene)−Ct(RPLP0). Ct is the cycle at which thethreshold line is crossed. For statistical analysis, fold changes were log2-transformed and two-tailed Student's t-tests were performed to calculatestatistical significance between solvent treated and ligand treatedsamples.

2.4. siRNA silencing

THP-1 cells were transfected with either non-specific siRNAoligomers or Stealth™ siRNAs targeting the VDR mRNA (Invitrogen,Carlsbad, CA, USA) by using RNAiMAX reagent (Invitrogen) accordingto the manufacturer's instructions. The cells were seeded into 6-wellplates and grown in phenol red-free DMEM supplemented with 5%charcoal-stripped FBS. Liposomes containing control or VDR siRNAwere formed by incubating 100 pmol of each siRNA duplex with 5 μlof RNAiMAX for 20 min at room temperature in a total volume of500 μl of phenol red-free DMEM without antibiotics. The liposomeswere added to the cells and siRNA treatment was continued for 48 h,then the cells were treated with 100 ng/ml LPS for 24 h and finallyexposed to either solvent (ethanol, 0.1% final concentration) or 10 nM1α,25(OH)2D3 for indicated time periods. Silencing of VDR at theprotein level was verified by Western blotting using 25 μg of whole-cell extract from THP-1 cells and anti-VDR antibody (sc-1008, SantaCruz Biotechnology, Santa Cruz, CA, USA). Cellular proteins wereseparated using 12% SDS polyacrylamide gel electrophoresis. Theblotted proteins were detected by using IR800 fluorescence labeledsecondary antibodies (Thermo Fisher Scientific, Waltham, MA, USA)and Odyssey reader (Li-Cor Biosciences, Lincoln, NE, USA).

2.5. Gel shift assay

VDR, RXR, and VDIR proteins were generated by coupled in vitrotranscription/translation reaction using full-length cDNAs subclonedinto the T7/SV40 promoter-driven pSG5 expression vector (Strata-gene, LaJolla, CA, USA) and TNT Quick Coupled Transcription/

1278 J.M. Matilainen et al. / Biochimica et Biophysica Acta 1803 (2010) 1276–1286

Translation Systems according to the manufacturer's instructions(Promega, Madison, WI, USA). Oligonucleotides were labeled withKlenow fragment DNA polymerase (Fermentas) in the presence of anucleotide mixture containing 32P α-dCTP. In gel shift assays, 100 ngof the appropriate in vitro translated proteins was incubated for15 min in a total volume of 20 μl of binding buffer (10 mM Hepes, pH7.9, 150 mM KCl, 1 mM dithiothreitol, 0.2 μg/μl of poly(dI–C), 5%glycerol). 32P-labeled double-stranded oligonucleotidescorresponding one VDRE were then added and incubation wascontinued at room temperature for 20 min. Protein–DNA complexeswere resolved by electrophoresis through non-denaturing 5% (w/v)polyacrylamide gels in 0.5×TBE (45 mM Tris, 45 mM boric acid, and1 mM EDTA, pH 8.3) and quantified on an FLA-3000 reader (Fuji,Tokyo, Japan) using Image Gauge software (Fuji).

2.6. Cell transfections and luciferase reporter gene assays

Human colon adenocarcinoma SW-480 cells were transfectedwiththe pGL4 containing minimal promoter of thymidine kinase (tk) geneand the upstream region of the human IL-10 gene including ER7 andER8 elements. A construct containing proximal promoter of humanCYP24 gene with a cluster of DR3-type VDREs was used as a positivecontrol. SW-480 cells were treated with 100 ng/ml LPS for 24 h priorto transfection. The DOTAP transfection protocol was used with amixture of 1 μg of different reporter constructs and 1 μg of expressionvector for human VDR gene (pSG5-hVDR) in a 6-well plate format.After 5 h, 1α,25(OH)2D3 was added to 100 nM and luciferase reporterassay was performed 16 h post-treatment with BriteLite luciferasesubstrate (Perkin Elmer, Waltham, MA, USA) in a Victor3 multilabelcounter (Perkin Elmer) equipped with automatic dispenser. Lucifer-ase activities were normalized with respect to total proteinconcentration.

2.7. Targeting of putative VDREs by duplex RNAs

The siRNA molecules for enhancer targeting were designed byDharmagon siDESIGN Center algorithm (www.dharmacon.com/DesignCenter/DesignCenterPage.aspx) to cover either one or bothhalf sites of the consensus sequences of ER7 and ER8-type elementsand obtained from Thermo Fisher (for sequences see Table S1). TheStealth RNAi™ siRNA Negative Control Med GC (Invitrogen) was usedas a control. THP-1 cells (2.5×105 cells) were plated on a 24-wellplate immediately prior to transfection in phenol red-free DMEMsupplemented with 5% charcoal-stripped FBS, 2 mM glutamine,0.1 mg/ml streptomycin and 100 U/ml penicillin. 4 μl TransIT-TKOTransfection Reagent (Mirus Bio LCC, Madison, WI) was diluted with50 μl serum-free DMEM and incubated for 20 min at room temper-ature. siRNA molecules (40 nM final concentration) were added andthe incubationwas continued for 20 min at room temperature and themixture was added dropwise to the cells. Phenol red-free DMEMsupplemented with 5% charcoal-stripped FBS, 2 mM glutamine,0.1 mg/ml streptomycin, and 100 U/ml penicillin was added 6 hafter transfection and 24 h after transfection the cells were treatedwith 100 ng/ml LPS for 24 h after which the cells were exposed toeither solvent (ethanol, 0.1% final concentration) or to 10 nM 1α,25(OH)2D3. After indicated periods of time, total RNA was extracted andRT-PCR was performed as described previously. Viability of the cellswas determined using AlamarBlue reagent according to the instruc-tions of the manufacturer (Invitrogen).

2.8. ChIP assays

ChIP was performed with LPS-treated THP-1 cells as previouslydescribed [29] with few modifications. Briefly, nuclear proteins werecross-linked to DNA by adding formaldehyde for 10 min. Cross-linkingwas stoppedwith glycine at room temperature for 5 min. Themedium

was removed and the cells were washed twice with ice-cold PBS. Thecells were collected and resuspended in lysis buffer containingprotease inhibitors. The chromatin was fragmented by sonicationand non-specific backgroundwas removed using salmon spermDNA/protein A agarose slurry (Millipore, Temecula, CA, USA) at 4 °C for 1 h.The recovered chromatin solutions were diluted 1:10 (v/v) in ChIPdilution buffer and incubated with 1 μg of indicated antibodies at 4 °Covernight. Non-specific IgG (12-370) and antibodies against H3K9acetylated at K9 and K14 (H3Ac, 06-599) and dimethylated H3K9 (07-521) were from Upstate Biotechnology (Lake Placid, NY, USA), theantibodies against the C-terminus of histone H3 (07-690) anddimethylated H3K4 (07-030) from Millipore, the antibodies againstVDR (sc-1008) and RXRα (sc-553) from Santa Cruz Biotechnology andthe antibody against NCoR (ab24552) from Abcam (Cambridge, UK).The immunocomplexes were collected using protein A agarose slurry(Millipore) and reverse cross-linked in the presence of 2 μl ofproteinase K (18.9–20.1 mg/ml) (Fermentas) at 64 °C overnight,after which phenol:chloroform extraction and ethanol precipitationwere performed. ChIP samples were analyzed with quantitative PCRusing BHQ1-FAM6 hydrolysis probes (Eurogentec, Liege, Belgium)and Maxima™ Probe qPCR Master Mix. The sequences of the primersand the hydrolysis probes are listed in Tables S2 and S3, respectively.The qPCR reaction was performed with a LightCycler 480 apparatususing the same PCR profile as with cDNA samples in an appropriateannealing temperature. The results were normalized with respectto input. Non-specific IgG results were subtracted by using theformula 2−(ΔCt)⁎100(specific antibody)−2−(ΔCt)⁎100(non-specific IgG),where ΔCt is Ct(immunoprecipitated DNA)−Ct(input) and Ct is the cycle atwhich the threshold line is crossed. Two-tailed Student's t-tests wereperformed and statistical significance of the percents of inputs wascalculated in reference to time point 0 min.

2.9. Micrococcus nuclease (MNase) assay

Chromatin was fragmented at mono-nucleosome level by MNasedigestion using themethod described by Soutoglou and Talianidis [30]with slight modifications. Prior to the MNase digestion, THP-1 cells(106 cells) were washed twice with PBS and suspended to 1 ml of lysisbuffer (10 mM Tris, pH 7.4, 10 mMNaCl, 5 mMMgCl2, 0.1% NP40) andincubated on ice for 10 min. The nuclei were collected by centrifuga-tion (1500×g, 5 min) and washed once with lysis buffer. The nucleiwere suspended toMNase reaction buffer (50 mMTris, pH 7.5, 15 mMNaCl, 3 mM MgCl2, 5 mM KCl, 1 mM CaCl2, protease inhibitors) and30 U of MNase (Fermentas) was added. After a 1 h incubation at 37 °C,2 μl of proteinase K was added and the samples were incubated at64 °C overnight. Finally, DNA was recovered by phenol–chloroformextraction followed by ethanol precipitation, and analyzed byquantitative PCR using primers listed in Tables S4 and S5 andMaxima™ SYBR Green qPCR Master Mix (Fermentas). PCR cyclingconditionswere: 10 min at 95 °C, followed by 50 cycles of 20 s at 95 °C,20 s at appropriate annealing temperature, and 20 s at 72 °C, endingwith 10 min at 72 °C. The resultswere normalizedwith respect to non-digested DNA (input) by using the formula 2−(ΔCt)⁎100, where ΔCt isCt(MNase-digested DNA)−Ct(input) and Ct is the cycle at which thethreshold line is crossed. Two-tailed Student's t-tests were performedand statistical significance of the percents of inputs was calculated inreference to time point 0 min.

2.10. Chromosome conformation capture assay

8×106 THP-1 cells were treated with LPS and 1α,25(OH)2D3, afterwhich the cells were collected as described for ChIP analysis, with theexception of the cross-linking time (20 min). The samples wereresuspended in 100 μl lysis buffer and sonicated for one 30 s cycle.25 μl aliquots of the lysate were separated as +ligase and −ligasesamples. ChIP dilution buffer, protease inhibitors, 2×Tango buffer

1279J.M. Matilainen et al. / Biochimica et Biophysica Acta 1803 (2010) 1276–1286

(Fermentas) and 25 UMvaI (Fermentas) were added to achieve a totalvolume of 100 μl, after which the samples were incubated at 37 °Covernight. After that 10 μl 10% Triton X-100 (Sigma-Aldrich) wasadded and the samples were incubated at 37 °C for 1 h. To −ligasesamples 500 μl H2O and 2 μl proteinase K were added and the sampleswere incubated at 64 °C overnight. Before proteinase K treatment +ligase samples were incubated 3 h at room temperature in thepresence of T4 ligase buffer (Fermentas), 0.5 mM ATP and 15 U T4ligase (Fermentas) in a total volume of 600 μl. In the end phenol:chloroform extraction and ethanol precipitation were performed. Therecovered DNA was analyzed by PCR using primers listed in Table S6and Maxima™ SYBR Green qPCR Master Mix, after which the PCRproducts were separated on 2% agarose gel. PCR cycling conditionswere the same as with MNase samples.

Fig. 1. Expression profiling of the human IL-10 gene. Relative mRNA expression of IL-10wasm(B) for indicated time periods after stimulation with 10 nM 1α,25(OH)2D3. (C) siRNA silencfrom LPS-treated THP-1 cells. (D) Relative mRNA expression of IL-10 was measured after 6was studied by treating the cells with 1 μg/ml actinomycin D prior to RNA extraction. The resperformed to calculate p-values in reference to vehicle treatment (*pb0.05, **pb0.01, ***p

3. Results

3.1. IL-10 is a down-regulated primary VDR target gene

LPS-treated human acute monocytic leukemia THP-1 cells servethroughout this study as a model system for human macrophages.Prior to treatment with 1α,25(OH)2D3 the cells were treated for 24 hwith LPS. Quantitative PCR using TaqMan probes showed that thisincreased the basal expression of IL-10 2.9-fold (Fig. S1). The effect of1α,25(OH)2D3 on IL-10 expressionwasmeasured for 8 h every 15 minand after 24 h and 48 h (Fig. 1A). IL-10 mRNA expression started todecrease 90 min after onset of ligand treatment and reached with 36%of basal expression the lowest level after 255–270 min (Fig. 1A). Attime point 300 min IL-10 mRNA expression turned back to basal

easured by quantitative PCR in LPS-treated THP-1 cells (A) and primarymonocytic cellsing of VDR at protein level was verified by Western blotting using whole-cell extractsh of 10 nM 1α,25(OH)2D3 treatment after silencing of VDR. (E) Stability of IL-10 mRNAults were normalized to the housekeeping gene RPLP0. Two-tailed Student's t-tests wereb0.001). In each panel, n is at least 3. Error bars indicate S.D.

1280 J.M. Matilainen et al. / Biochimica et Biophysica Acta 1803 (2010) 1276–1286

levels, but later on showed approximately 2-fold repression at leastfor another 3 h. At time point 24 h IL-10 mRNA expression hadreturned to basal levels, while at time point 48 h it had increased 3-fold. For comparison we examined the effect of 1α,25(OH)2D3 on IL-10 expression also in primary human monocytes. According to ourdata, the ligand effect was not as clear as in THP-1 cells, since at timepoint 2 h no repression could be observed and at time point 4 h onlyminor repression was visible. However, at time point 6 h a significant1.6-fold repression and at 24 h a 2.3-fold induction could be observed(Fig. 1B). Please note that the primary cells were not pre-treated withLPS.

In order to confirm that the observed repression of IL-10was VDR-dependent, we silenced the VDR gene using siRNA. The efficiency ofVDR silencing at the protein level was verified by Western blottingusing whole-cell extracts (Fig. 1C). 6 h after 1α,25(OH)2D3 stimula-tion in the presence of VDR specific siRNA, no IL-10 repression wasobserved, while non-specific siRNA did not affect to the repressive

Fig. 2. Functionality of ER7- and ER8-type VDREs at the IL-10 promoter. (A) Gel shift analysistype VDREs in the presence and absence of 1α,25(OH)2D3. A DR4-type element from the ratprotein-complexed DNA was quantified using a Fuji FLA-3000 reader. NS indicates non-specelements was cloned in front of thymidine kinase (tk) promoter driving the luciferase (Luc) gVDREs was used as a positive control and the empty reporter vector as a negative control. Luctreatment. Columns indicate the means of three independent experiments and the tips of tcalculate p-values in reference to solvent treated samples (*pb0.05, **pb0.01, ***pb0.001)

effect of the VDR ligand (Fig. 1D). The stability of IL-10 mRNA wasdetermined by treating LPS-stimulated THP-1 cells with actinomycinD (Fig. 1E). The half-life of the IL-10 mRNA was found to be 0.9 h,which is in agreement with the repression observed in the time-course experiments.

Taken together, our data suggest that IL-10 is a primary VDR targetgene, which is initially down-regulated and later after 24 h up-regulated.

3.2. Putative VDREs within the IL-10 gene

In order to locate the genomic regions within the human IL-10gene that associate with VDR, we performed a ChIP scanning analysisusing anti-VDR antibody (Fig. S2). From the 27 genomic regionsspanning over the region from−10037 to +1862 relative to the IL-10TSS ten recruited VDR ligand-dependently. Regions 12, 19 and 22located at repetitive sequences and regions 14 and 17 did not contain

was used to study the ability of in vitro translated VDR and RXR to bind to ER7- and ER8-Pit-1 served as a positive control. Representative gels are shown. (B) The percentage ofific complexes. (C) The upstream region of the human IL-10 gene including ER7 and ER8ene. The proximal promoter of the human CYP24 gene containing a cluster of DR3-typeiferase activity was determined from LPS-treated SW480 cells after a 16 h 1α,25(OH)2D3

he bars represent standard deviations. Two-tailed Student's t-tests were performed to. In each bar, n is at least 3. Error bars indicate S.D.

0 10

20

30

40

50

60

70

80

90

100

PCR

ctr

l 1 ,25(OH)2D3 treatment time (min)

ligase+ligase

+ligaseligase

IL-10 1 kb

a b c d

c to d

b to e

e f

+ ligaseligase a to e

+ ligaseligase f to e

MvaI

E-b

ox

ER

8 E

R7

Fig. 3. Chromatin looping connects the distal VDREs to the TSS. 3C analysiswas performedto detect the looping between TSS anddifferent chromatin regions at the IL-10promoter inresponse to 10 nM1α,25(OH)2D3 treatment. GenomicDNAof LPS-treated THP-1 cellswasdigested with MvaI (recognition sites represented by vertical lines) and ligated with T4DNA ligase. The primers used in PCR are shown as horizontal arrowheads. Digestions andligations performedwith clonedDNA fragments of the IL-10 promoter served as a positivecontrol. Representative agarose gels of the PCR products are shown.

1281J.M. Matilainen et al. / Biochimica et Biophysica Acta 1803 (2010) 1276–1286

any VDREs based on our recent in silico screen [31]; all five regionswere therefore not further investigated. Region 5 (+99 to −134)contained the TSS and therefore it was not surprising to find VDRthere due to looping effects from other regions. An in silico screeningusing the net-based program RESearch [32] revealed that region 2(+1433 to +1006) contains E-box elements, where VDR may bindbackpack on VDIR, and that region 9 (−1432 to −1986) containedER7- and ER8-type VDREs for direct binding of VDR (Fig. S2). Weassume that the observed recruitment of VDR to the regions 1 and 10resulted in flanking effect from regions 2 and 9, respectively. Aconservation analysis based on 28 species by PhyloP [33] as well asmanual sequence comparison between human, mouse and dogsuggested that the genomic regions containing the above putativeVDREs and E-box elements were conserved across different species(Fig. S2).

3.3. Functionality of the putative VDREs

The potential of the putative regulative regions to associate with invitro translated VDR and RXR was examined using gel shift analysis.The ER7-type VDRE (−1502 to −1484) and the ER8-type VDRE(−1721 to −1701) were able to bind VDR–RXR heterodimercomplexes, but less prominent than the established DR4-type VDREof the rat Pit-1 gene (gaAGTTCAtgagAGTTCA) [34]. Interestingly, theER8 type VDRE of the IL-10 gene aswell as the control VDRE showed inthe presence of 1α,25(OH)2D3 more prominent VDR–RXR hetero-dimer binding than in its absence (Fig. 2A and B). Compared with thecontrol negative VDRE (nVDRE) of the human CYP27B1 gene(ccattaaccCACCTGcCATCTGcc) [12], the E-box type elements(+1079 to +1097) did not show direct binding of VDR–RXRheterodimers and associated only weakly with VDIR (Fig. S3A).

The functionality of the putative VDREs was examined by reportergene assays (Fig. 2C). The reporter construct containing the proximalpromoter of human CYP24 gene was 13-fold inducible by 1α,25(OH)2D3. In contrast, the basal activity of the IL-10 construct wasreduced by 25% by 1α,25(OH)2D3 treatment. The construct containingE-box type elements showed less response to stimulation with 1α,25(OH)2D3 than the ER7- and ER8-type elements (Fig. S3B).

In summary, within 12 kb of the IL-10 gene we identified atpositions −1500 to −1700 two VDR–RXR heterodimer bindingVDREs and at position +1000 an E-box sequence associating withVDIR.

3.4. The distal VDREs loop to the TSS

For proving the functionality of the promoter region containingthe ER7- and ER8-type VDREs, we used 3C analysis [35] andinvestigated the chromatin looping between the region of theVDREs and the TSS of the IL-10 gene (Fig. 3). The results suggestthat the VDRE region can loop to the TSS in a ligand- and time-dependent fashion, since the distal promoter region was connectedwith the TSS at time points 10–20 min, 50 min and 80–90 min afteronset of ligand treatment. In contrast, the control regions or the regionof the E-box elements showed no interaction with the TSS. Takentogether, 1α,25(OH)2D3 specifically induces looping between thedistal promoter region and the TSS. Therefore, we focused furtherinvestigations on this region.

3.5. IL-10 repression by 1α,25(OH)2D3 is mediated via ER8-type element

In order to determine if one or both distal elements participate tothe 1α,25(OH)2D3-dependent regulation of IL-10 expression, weapplied the recently reportedmethod of promoter targeting by duplexRNAs [36,37]. We reasoned that targeting of the putative VDREswould prevent the observed ligand effects assuming that the elementswere functional. According to our results, targeting of the ER7-type

element did not show significant effect on 1α,25(OH)2D3-dependentrepression of IL-10 mRNA expression at early time points 2 and 4 h(Fig. 4). Instead, targeting of the ER8-type element prevented therepression and even lead to significant induction of IL-10, suggestingthat the primary effects of 1α,25(OH)2D3 are mediated via the ER8-type element. Interestingly, both ER7 and ER8 appeared to participateto the 1α,25(OH)2D3-dependent up-regulation of IL-10 expression,since targeting of any of the two elements prevented the ligand effectafter a 48 h hormone treatment. To verify that the observed targetingeffects were specific to the IL-10 expression, we repeated the RT-PCRanalysis with IL-12B and TLR-4 genes, that are known to be repressedby 1α,25(OH)2D3 [2,38]. Our data showed that neither targeting withnon-specific siRNA nor ER7 or ER8 specific siRNAs affected to theligand-dependent repression of above genes. The viability studiesshowed that both non-specific siRNA and specific siRNA treatmentsclearly decreased the cell's viability in a treatment–time-dependentfashion (Fig. S5). However, it is notable that the viability was similarafter all treatments. Thus, the observed promoter targeting resultswere reliable.

3.6. VDR, RXR and NCoR associate cyclically with the promoter and theTSS regions

The periodic looping between the VDRE containing promoterregion and the TSS suggests that this region may recruit VDR in acyclical fashion and that via looping VDR may also be detectable onthe TSS. Therefore, we used time-resolved quantitative ChIP analysis(every 10 min over 100 min) and measured the recruitment of VDR,RXR and the co-repressor NCoR to the promoter and the TSS regions(Fig. 5). VDR associated with both the promoter region and the TSS attime points 10 and 50 min after onset of ligand treatment (Fig. 5A).The recruitment of RXR to the TSS followed that of VDR, while itassociatedwith the promoter regionmore evenly andwas only absentat time point 40 min (Fig. 5B). Also NCoR bound to the promoterregion rather evenly, but was enriched on the TSS at time points 20, 50and 70 min (Fig. 5C).

Rel

ativ

e m

RN

A f

old

chan

geR

elat

ive

mR

NA

fol

d ch

ange

Rel

ativ

e m

RN

A f

old

chan

ge

1α,25(OH)2D3 treatment time (h)

1α,25(OH)2D3 treatment time (h)

1α,25(OH)2D3 treatment time (h)

IL-10

IL-12B

TLR4

A

B

C

Fig. 4. Targeting of the ER7- and ER8-type VDREs by duplex RNAs. THP-1 cells, whichhad been pre-treated for 24 h with 100 ng/ml LPS, were transfected with control siRNAor ER7 or ER8 specific siRNAs. After 24 h the cells were exposed to either solvent(ethanol, 0.1% final concentration) or to 10 nM 1α,25(OH)2D3. After the indicatedperiods of time, total RNA was extracted and mRNA expression of IL-10 (A), IL-12B(B) or TLR-4 (C) was determined as described previously. Two-tailed Student's t-testswere performed to calculate p-values in reference to solvent treated samples (*pb0.05,**pb0.01, ***pb0.001). In each bar, n is at least 3. Error bars indicate S.D.

A

B

C

IL-10

1 kb

ER

8

ER

7

Fig. 5. Ligand treatment leads to cyclical recruitment of transcription factors to the IL-10promoter. ChIP assays were performed using antibodies against VDR (A), RXR (B) andNCoR (C) with 10 min intervals of 10 nM 1α,25(OH)2D3 treatment up to 100 min inLPS-treated THP-1 cells. The IL-10 TSS is indicated in black and the VDRE region in red.Two-tailed Student's t-tests were performed to calculate p-values in reference to timepoint 0 (*pb0.05, **pb0.01, ***pb0.001). In each panel, n is at least 3. Error barsindicate S.D.

1282 J.M. Matilainen et al. / Biochimica et Biophysica Acta 1803 (2010) 1276–1286

In summary, VDR shows a clear cyclical association both with thepromoter region and the TSS with a periodicity of 40 min, while RXRand NCoR demonstrate significant peaks of association only with theTSS.

3.7. Ligand-dependent epigenetic changes at the promoter and TSSregions

In order to monitor changes of chromatin activity related to the1α,25(OH)2D3-dependent repression of IL-10 transcription, weperformed quantitative ChIP analysis with antibodies against acety-lated or methylated histone H3 (Fig. 6). As for VDR, RXR and NCoRbinding (Fig. 5), the changes in the chromatin status were assessedevery 10 min up to 100 min after onset of ligand treatment. In the

absence of ligand both the promoter region and the TSS showedhistone H3 acetylation (H3ac), which was lost at time points 50 and80 min, but was maximal at 60–70 min (Fig. 6A). In the absence ofligand dimethylation of histone H3 at K9 (H3K9me2), a marker forinactive chromatin, was not detectable both at the promoter regionand the TSS (Fig. 6B). After ligand treatment the H3K9me2 levelspeaked at the promoter region at time points 10, 50 and 90 min. At theTSS region the H3K9me2 levels were also maximal at time point10 min, but stayed high also in the period of 40–90 min.

B

A

IL-10

1 kb

ER

8

ER

7

Fig. 6. Ligand-dependent epigenetic changes at the IL-10 promoter and the TSS. ChIPassays were performed using antibodies against H3ac (A) and H3K9me2 (B) with10 min intervals of 10 nM 1α,25(OH)2D3 treatment up to 100 min in LPS-treated THP-1cells. The IL-10 TSS is indicated in black and the VDRE region in red. Two-tailedStudent's t-tests were performed to calculate p-values in reference to time point 0(*pb0.05, **pb0.01, ***pb0.001). In each panel, n is at least 3. Error bars indicate S.D.

1283J.M. Matilainen et al. / Biochimica et Biophysica Acta 1803 (2010) 1276–1286

Taken together, both on the level of active chromatin (H3ac) andon inactive chromatin (H3K9me2) the promoter region and the TSSshowed a cyclical behavior with a periodicity of 30–50 min betweenthe minima and maxima.

3.8. Ligand-dependent chromatin remodeling at the promoter and TSSregions

In order to investigate, whether the periodic appearance ofrepressive histone marks result in condensation of the chromatin,we performed limited MNase digestion assays (Fig. 7). The changes inthe chromatin status were assessed every 10 min over a period of100 min after onset of ligand treatment. A schematic presentation inFig. 7A illustrates the possible locations of the nucleosomes thatwould protect the chromatin within the studied amplicons againstdigestion by MNase. The efficiency of MNase digestion was confirmedby agarose gel electrophoresis to ensure that the chosen digestiontime resulted in chromatin fragments in a mono-nucleosome(b200 bp) size (Fig. 7B). The nucleosome density increased ligand-dependently at all four studied regions. On regions a and b (aroundthe ER8-type VDRE) nucleosome enrichment was maximal at timepoints 20–30 min and 60–80 min (Fig. 7C and D). The nucleosomeenrichment around the ER7-type VDRE (at regions c and d) was

clearly different, as the highest enrichment could be observed at timepoints 50–80 min at region c and at 70–90 min at region d (Fig. 7E andF). Thus, our data suggest that chromatin remodeling follows thecyclic histone modifications with a 10–20 min delay. The firstrepressive cycle appears to increase the nucleosome density only atthe ER8-type VDRE, but the second repressive cycle results in moregeneral condensation of the chromatin, since increased nucleosomedensity could be observed at both VDREs as well as at the TSS (Fig. S4).

In summary, the nucleosome density assay could confirm theresults of the ChIP assays concerning the periodic changes ofchromatin activity at the promoter region and the TSS.

4. Discussion

IL-10 is a cytokine that showsmainly suppressive effects on innateimmunity, but has also stimulatory effects on adaptive immunity. Thecritical role of IL-10 is underlined by the fact that it is expressedubiquitously among innate and adaptive immune cells includingmonocytes, macrophages, dendritic cells, B cells, CD8+ and regulatoryT cells, TH1, TH2 and TH17 cells [39]. IL-10 limits the ability of antigenpresenting cells to promote the differentiation and proliferationof CD4+ T cells and therefore affects to both initiation and progress ofadaptive T cell responses. In addition, IL-10 inhibits the expression ofa number of pro-inflammatory cytokines and thus further suppressesthe ability of effector T cells to prolong the inflammatory responses[40,41].

Several studies showed that IL-10 expression is induced by 1α,25(OH)2D3 in different cells of the immune system mediating in thisway the known immunosuppressive effects of 1α,25(OH)2D3 [25–28]. However, opposite effects of 1α,25(OH)2D3 on the production ofIL-10 both in vitro and in vivo have also been reported [22–24,42,43].The contrary ligand effects on IL-10 expression may be due todifferent cell types being studied, but interestingly, the reportsdescribing up-regulation of IL-10 expression seem to be related totreatments with 1α,25(OH)2D3 for several days, whereas reports onIL-10 down-regulation by VDR ligands use shorter stimulation times.This let us hypothesize that 1α,25(OH)2D3-dependent regulation ofIL-10 expression could be bidirectional depending on the time ofmeasurement.

In this study we showed a clear, ligand-dependent repression ofIL-10mRNAwithin the first 8 h of ligand treatment, whereas after 48 ha 3-fold induction was observed. Highly similar ligand effect could beobserved also in primarymonocytes. Thus, the primary effect of 1α,25(OH)2D3 on the expressionof IL-10 is repression,which is later overrunby an up-regulating secondary effect. This may explain in part thephysiological role of 1α,25(OH)2D3 in the regulation of adaptiveimmune response. The initial down-regulation of IL-10 immediatelyafter infection would ensure an effective immune response, since lowIL-10 levels allow the production of pro-inflammatory cytokines inhigher quantities. The delayed, secondary effects that result in up-regulation of IL-10 by 1α,25(OH)2D3 then suppress the innate immuneresponse in order to avoid effects of long-lasting inflammation, such astissue damage.

An absolute requirement for direct transcriptional regulation ofany gene by 1α,25(OH)2D3 is that the TSS of the respective gene isconnected to at least one genomic region containing positive ornegative VDREs. Our 3C results suggest that only the promoter regioncontaining an ER7- and an ER8-type VDRE (at positions −1700 to−1500) loops to the TSS. In addition, this region was able tosignificantly repress the luciferase activity in reporter gene analysis.These data indicate that one or both of the above VDREs is responsiblefor the 1α,25(OH)2D3-dependent repression of IL-10 expression. Ourobservations agree with the data of Tavera-Mendoza et al., whoreported that in human SCC25 head and neck squamous cellcarcinoma cells VDR ligand-dependently associates with a region ofan ER8-type VDRE (−1791 to −1464) of the IL-10 gene [44].

Region a

2

4

6

8

% o

f un

dige

sted

0 10 20 30 40 50 60 70 80 90 100

0

**

** **

**

*

**

*

Region b

0 10 20 30 40 50 60 70 80 90 100

1 ,25(OH)2D3 treatment (min)1 ,25(OH)2D3 treatment (min)

1 ,25(OH)2D3 treatment (min) 1 ,25(OH)2D3 treatment (min)

0

10

20

30

40

50

% o

f un

dige

sted

*

*

* *

0 10 20 30 40 50 60 70 80 90 100

0

5

10

20

15

25

% o

f un

dige

sted

**

**

*

Region d

0 10 20 30 40 50 60 70 80 90 100

0

5

15

10

% o

f un

dige

sted

**

Region c

A

F

C D

E

B

-1900bp -1300 bp

a b c d

ER8 ER7

100 bp

Mar

ker

Und

iges

ted

MN

ase

(30

U)

200 bp

Fig. 7. Ligand-dependent chromatin remodeling at the VDRE region. (A) PCR products and the possible locations of nucleosomes around the VDREs at the IL-10 promoter.(B) A representative gel figure shows the digestion of chromatin to mono-nucleosome level. (C–F) Limited MNase treatment and quantitative PCR were used to study theenrichment of nucleosomes at different genomic regions in response to 10 nM 1α,25(OH)2D3 treatment in LPS-treated THP-1 cells. Two-tailed Student's t-tests were performed tocalculate p-values in reference to time point 0 (*pb0.05, **pb0.01). In each panel, n is at least 3. Error bars indicate S.D.

1284 J.M. Matilainen et al. / Biochimica et Biophysica Acta 1803 (2010) 1276–1286

To clarify the role of distal elements to the ligand-dependentregulation of IL-10 expression, we utilized promoter targeting byduplex RNAs [36,37]. According to our results, the ER8-type VDRE iscritical for short-term repression and both ER7 and ER8 elements areneeded for up-regulation after longer exposure to the hormone. Thesedata suggest that the VDREs are used differently during differenthormone exposure times. This can be due to changes in the cell'sproteome in response to hormone treatment [45].

It has been recently reported that the transcriptional activation ofsome nuclear receptor target genes involves cyclic association oftranscription factors and their co-regulators to the regulatory regionsof the respective genes and that this periodicity reflects also to themRNA expression levels [10,12]. However, the available data concernsexclusively up-regulated genes. Moreover, until now the mechanismsof transcriptional repression is not sufficiently known. Therefore, weused the primary down-regulating effects of 1α,25(OH)2D3 on IL-10expression as a model for examining the effects of transcription factor

cycling on gene repression.We found that the repressive transcriptionfactor complex associates in a ligand- and time-depend fashion bothwith the TSS and the promoter region containing the VDREs. Thiscaused periodic epigenetic changes and chromatin remodeling onboth chromatin regions.

Since there are no VDR binding sites in vicinity of the TSS [31], weassume that the observed association of the repressive complex withthe TSS results from the looping of the VDRE region to the TSS asobserved in our 3C assays. This suggests that the mechanisms oftranscriptional activation and repression share common features ofcyclic association of transcription factors and that they both involvechanges in higher order chromatin hierarchy. However, in contrast toup-regulated genes, such as CYP24 and CDKN1A [10,12], no clearperiodicity in the repression of IL-10 mRNA could be detected.Therefore, at least in the case of the IL-10 gene, the cyclical associationof transcription factors does not reflect to the mRNA expression level.The reason for this is most likely an insufficient degradation of IL-10

1285J.M. Matilainen et al. / Biochimica et Biophysica Acta 1803 (2010) 1276–1286

mRNA during one transcriptional cycle and the involvement ofactivating transcriptions factors, which have not been characterizedin this study. In addition, to see the cycling also at mRNA level wouldrequire that the mRNA synthesis would happen relatively fast. Thismay not be the case in the LPS induced IL-10 mRNA synthesis that ismediated via cell surface receptor TLR-4.

In conclusion, this study indicates that in LPS-treated THP-1 cellsthe primary effect of 1α,25(OH)2D3 on IL-10 expression is repression,which occurs during the first hours after onset of ligand treatment.We propose that this repression occurs in order to ensure an efficientimmune response against microbial infections. We found that the1α,25(OH)2D3-dependent repression is achieved through cyclicrecruitment of VDR to ER7- and ER8-type VDREs within a promoterregion of the IL-10 gene. Thus, both in transcriptional activation andrepression cyclic association of transcription factors is involved.However, at least for the IL-10 gene, the cycling of transcriptionfactors and chromatin marks does not translate to changes in mRNAexpression levels.

Supplementarymaterials related to this article can be found onlineat doi:10.1016/j.bbamcr.2010.07.009.

Acknowledgements

We would like to thank Dr. Shigeagi Kato for VDIR cDNA and Mrs.Hanna Eskelinen for her technical assistance and help in cell culture.This work was supported by The Academy of Finland.

References

[1] I. Alroy, T.L. Towers, L.P. Freedman, Transcriptional repression of the interleukin-2gene by vitamin D3: direct inhibition of NFATp/AP-1 complex formation by anuclear hormone receptor, Mol. Cell. Biol. 15 (1995) 5789–5799.

[2] D. D'Ambrosio, M. Cippitelli, M.G. Cocciolo, D. Mazzeo, P. Di Lucia, R. Lang, F.Sinigaglia, P. Panina-Bordignon, Inhibition of IL-12 production by 1,25-dihydrox-yvitamin D3. Involvement of NF-kappaB downregulation in transcriptionalrepression of the p40 gene, J. Clin. Invest. 101 (1998) 252–262.

[3] M. Cippitelli, A. Santoni, Vitamin D3: a transcriptional modulator of the interferon-gamma gene, Eur. J. Immunol. 28 (1998) 3017–3030.

[4] D.M. Provvedini, C.D. Tsoukas, L.J. Deftos, S.C. Manolagas, 1,25-dihydroxyvitaminD3 receptors in human leukocytes, Science 221 (1983) 1181–1183.

[5] A. Brennan, D.R. Katz, J.D. Nunn, S. Barker, M. Hewison, L.J. Fraher, J.L. O'Riordan,Dendritic cells from human tissues express receptors for the immunoregulatoryvitamin D3 metabolite, dihydroxycholecalciferol, Immunology 61 (1987)457–461.

[6] L. Overbergh, B. Decallonne, D. Valckx, A. Verstuyf, J. Depovere, J. Laureys, O.Rutgeerts, R. Saint-Arnaud, R. Bouillon, C. Mathieu, Identification and immuneregulation of 25-hydroxyvitamin D-1α-hydroxylase inmurine macrophages, Clin.Exp. Immunol. 120 (2000) 139–146.

[7] M. Noda, R.L. Vogel, A.M. Craig, J. Prahl, H.F. DeLuca, D.T. Denhardt, Identificationof a DNA sequence responsible for binding of the 1,25-dihydroxyvitamin D3receptor and 1,25-dihydroxyvitamin D3 enhancement of mouse secretedphosphoprotein 1 (SPP-1 or osteopontin) gene expression, Proc. Natl. Acad. Sci.U. S. A. 87 (1990) 9995–9999.

[8] S.J. Rhodes, R. Chen, G.E. DiMattia, K.M. Scully, K.A. Kalla, S.C. Lin, V.C. Yu, M.G.Rosenfeld, A tissue-specific enhancer confers Pit-1-dependent morphogeninducibility and autoregulation on the pit-1 gene, Genes Dev. 7 (1993) 913–932.

[9] M. Schräder, S. Nayeri, J.P. Kahlen, K.M. Müller, C. Carlberg, Natural vitamin D3

response elements formed by inverted palindromes: polarity-directed ligandsensitivity of vitamin D3 receptor-retinoid X receptor heterodimer-mediatedtransactivation, Mol. Cell. Biol. 15 (1995) 1154–1161.

[10] J.M. Matilainen, M. Malinen, M.M. Turunen, C. Carlberg, S. Väisänen, The numberof vitamin D receptor binding sites defines the different vitamin D responsivenessof the CYP24 gene in malignant and normal mammary cells, J. Biol. Chem. 285(2010) 24174–24183.

[11] S. Kangaspeska, B. Stride, R. Métivier, M. Polycarpou-Schwarz, D. Ibberson, R.P.Carmouche, V. Benes, F. Gannon, G. Reid, Transient cyclical methylation ofpromoter DNA, Nature 452 (2008) 112–115.

[12] A. Saramäki, S. Diermeier, R. Kellner, H. Laitinen, S. Väisänen, C. Carlberg, Cyclicalchromatin looping and transcription factor association on the regulatory regionsof the p21 (CDKN1A) gene in response to 1α,25-dihydroxyvitamin D3, J. Biol.Chem. 284 (2009) 8073–8082.

[13] S. Toropainen, S. Väisänen, S. Heikkinen, C. Carlberg, The down-regulation of thehuman MYC gene by the nuclear hormone 1a,25-dihydroxyvitamin D3 isassociated with cycling of corepressors and histone deacetylases, J. Mol. Biol.400 (2010) 284–294.

[14] A. Murayama, M.S. Kim, J. Yanagisawa, K. Takeyama, S. Kato, Transrepression by aliganded nuclear receptor via a bHLH activator through co-regulator switching,EMBO J. 23 (2004) 1598–1608.

[15] K.R. McGaffin, S.A. Chrysogelos, Identification and characterization of a responseelement in the EGFR promoter that mediates transcriptional repression by 1,25-dihydroxyvitamin D3 in breast cancer cells, J. Mol. Endocrinol. 35 (2005) 117–133.

[16] M.B. Demay, M.S. Kiernan, H.F. DeLuca, H.M. Kronenberg, Sequences in the humanparathyroid hormone gene that bind the 1,25-dihydroxyvitamin D3 receptor andmediate transcriptional repression in response to 1,25-dihydroxyvitamin D3,Proc. Natl. Acad. Sci. U. S. A. 89 (1992) 8097–8101.

[17] M. Falzon, DNA sequences in the rat parathyroid hormone-related peptide generesponsible for 1,25-dihydroxyvitamin D3-mediated transcriptional repression,Mol. Endocrinol. 10 (1996) 672–681.

[18] R. Kremer, M. Sebag, C. Champigny, K. Meerovitch, G.N. Hendy, J. White, D.Goltzman, Identification and characterization of 1,25-dihydroxyvitamin D3-responsive repressor sequences in the rat parathyroid hormone-related peptidegene, J. Biol. Chem. 271 (1996) 16310–16316.

[19] M.S. Kim, R. Fujiki, A. Murayama, H. Kitagawa, K. Yamaoka, Y. Yamamoto, M.Mihara, K. Takeyama, S. Kato, 1α,25(OH)2D3-induced transrepression by vitaminD receptor through E-box-type elements in the human parathyroid hormone genepromoter, Mol. Endocrinol. 21 (2007) 334–342.

[20] M.M. Turunen, T.W. Dunlop, C. Carlberg, S. Väisänen, Selective use ofmultiple vitaminD response elements underlies the 1α,25-dihydroxyvitamin D3-mediated negativeregulation of the human CYP27B1 gene, Nucleic Acids Res. 35 (2007) 2734–2747.

[21] C.M. Hawrylowicz, A. O'Garra, Potential role of interleukin-10-secreting regula-tory T cells in allergy and asthma, Nat. Rev. Immunol. 5 (2005) 271–283.

[22] L. Giovannini, V. Panichi, M. Migliori, S. De Pietro, A.A. Bertelli, A. Fulgenzi, C.Filippi, I. Sarnico, D. Taccola, R. Palla, A. Bertelli, 1,25-dihydroxyvitamin D3 dose-dependently inhibits LPS-induced cytokines production in PBMC modulatingintracellular calcium, Transplant. Proc. 33 (2001) 2366–2368.

[23] C.J. Bemiss, B.D. Mahon, A. Henry, V.Weaver, M.T. Cantorna, Interleukin-2 is one ofthe targets of 1,25-dihydroxyvitamin D3 in the immune system, Arch. Biochem.Biophys. 402 (2002) 249–254.

[24] H. Fujita, A. Asahina, M. Komine, K. Tamaki, The direct action of 1α,25(OH)2-vitamin D3 on purified mouse Langerhans cells, Cell. Immunol. 245 (2007) 70–79.

[25] G. Penna, L. Adorini, 1α,25-dihydroxyvitamin D3 inhibits differentiation,maturation, activation, and survival of dendritic cells leading to impairedalloreactive T cell activation, J. Immunol. 164 (2000) 2405–2411.

[26] A. Boonstra, F.J. Barrat, C. Crain, V.L. Heath, H.F. Savelkoul, A. O'Garra, 1α,25-Dihydroxyvitamin D3 has a direct effect on naive CD4(+) T cells to enhance thedevelopment of Th2 cells, J. Immunol. 167 (2001) 4974–4980.

[27] F.J. Barrat, D.J. Cua, A. Boonstra, D.F. Richards, C. Crain, H.F. Savelkoul, R. de Waal-Malefyt, R.L. Coffman, C.M. Hawrylowicz, A. O'Garra, In vitro generation ofinterleukin 10-producing regulatory CD4(+) T cells is induced by immunosup-pressive drugs and inhibited by T helper type 1 (Th1)- and Th2-inducingcytokines, J. Exp. Med. 195 (2002) 603–616.

[28] G. Heine, U. Niesner, H.D. Chang, A. Steinmeyer, U. Zügel, T. Zuberbier, A.Radbruch, M. Worm, 1,25-dihydroxyvitamin D3 promotes IL-10 production inhuman B cells, Eur. J. Immunol. 38 (2008) 2210–2218.

[29] S. Väisänen, T.W. Dunlop, L. Sinkkonen, C. Frank, C. Carlberg, Spatio-temporalactivation of chromatin on the human CYP24 gene promoter in the presence of1α,25-dihydroxyvitamin D3, J. Mol. Biol. 350 (2005) 65–77.

[30] E. Soutoglou, I. Talianidis, Coordination of PIC assembly and chromatin remodelingduring differentiation-induced gene activation, Science 295 (2002) 1901–1904.

[31] J.M. Matilainen, A. Räsänen, P. Gynther, S. Väisänen, The genes encoding cytokinesIL-2, IL-10 and IL-12B are primary 1α,25(OH)2D3 target genes, J. Steroid Biochem.Mol. Biol. 121 (2010) 142–145.

[32] M. Heinäniemi, J.O. Uski, T. Degenhardt, C. Carlberg, Meta-analysis of primarytarget genes of peroxisome proliferator-activated receptors, Genome Biol.8 (2007) R147.

[33] W.J. Kent, C.W. Sugnet, T.S. Furey, K.M. Roskin, T.H. Pringle, A.M. Zahler, D. Haussler,The human genome browser at UCSC, Genome Res. 12 (2002) 996–1006.

[34] S.J. Rhodes, R. Chen, G.E. DiMattia, K.M. Scully, K.A. Kalla, S.C. Lin, V.C. Yu, M.G.Rosenfeld, A tissue-specific enhancer confers Pit-1-dependent morphogeninducibility and autoregulation on the pit-1 gene, Genes Dev. 7 (1993) 913–932.

[35] A. Miele, N. Gheldof, T.M. Tabuchi, J. Dostie, J. Dekker, Mapping chromatininteractions by chromosome conformation capture, Curr. Protoc. Mol. Biol. 74(2006) 21.11.1–21.11.20.

[36] B.A. Janowski, S.T. Younger, D.B. Hardy, R. Ram, K.E. Huffman, D.R. Corey,Activating gene expression in mammalian cells with promoter-targeted duplexRNAs, Nat. Chem. Biol. 3 (2007) 166–173.

[37] M.P. Turunen, T. Lehtola, S.E. Heinonen, G.S. Assefa, P. Korpisalo, R. Girnary, C.K.Glass, S. Väisänen, S. Ylä-Herttuala, Efficient regulation of VEGF expression bypromoter targeted lentiviral shRNAs based on epigenetic mechanism: a novelexample of epigenetherapy, Circ. Res. 105 (2009) 604–609.

[38] K. Sadeghi, B. Wessner, U. Laggner, M. Ploder, D. Tamandl, J. Friedl, U. Zügel, A.Steinmeyer, A. Pollak, E. Roth, G. Boltz-Nitulescu, A. Spittler, Vitamin D3 down-regulates monocyte TLR expression and triggers hyporesponsiveness to patho-gen-associated molecular patterns, Eur. J. Immunol. 36 (2006) 361–370.

[39] A. O'Garra, P. Vieira, T(H)1 cells control themselves by producing interleukin-10,Nat. Rev. Immunol. 76 (2007) 425–428.

[40] A. Wakkach, N. Fournier, V. Brun, J.P. Breittmayer, F. Cottrez, H. Groux,Characterization of dendritic cells that induce tolerance and T regulatory 1 celldifferentiation in vivo, Immunity 18 (2003) 605–617.

[41] C.L. Maynard, C.T. Weaver, Diversity in the contribution of interleukin-10 to T-cell-mediated immune regulation, Immunol. Rev. 226 (2008) 219–233.

1286 J.M. Matilainen et al. / Biochimica et Biophysica Acta 1803 (2010) 1276–1286

[42] L.A. Lyakh, M. Sanford, S. Chekol, H.A. Young, A.B. Roberts, TGF-beta and vitaminD3 utilize distinct pathways to suppress IL-12 production and modulate rapiddifferentiation of human monocytes into CD83+ dendritic cells, J. Immunol. 174(2005) 2061–2070.

[43] E. van Etten, O. Dardenne, C. Gysemans, L. Overbergh, C. Mathieu, 1,25-Dihydroxyvitamin D3 alters the profile of bone marrow-derived dendritic cellsof NOD mice, Ann. N. Y. Acad. Sci. 1037 (2004) 186–192.

[44] L. Tavera-Mendoza, T.-T. Wang, B. Lallemant, R. Zhang, Y. Nagai, V. Bourdeau, R.M.Calderon, J. Desbarats, S. Mader, J.H. White, Convergence of vitamin D and retinoicacid signalling at a common hormone response element, EMBO Rep. 7 (2006)180–185.

[45] T.W. Dunlop, S. Väisänen, C. Frank, C. Carlberg, The genes of the coactivator TIF2and the corepressor SMRT are primary 1α,25(OH)2D3 targets, J. Steroid Biochem.Mol. Biol. 89 (2004) 257–260.