Biochimica et Biophysica Acta 1771 (2007) 864 – 872 www.elsevier.com/locate/bbalip co p y Functional characterization of vitamin D responding regions in the human 5-Lipoxygenase gene Sabine Seuter a,b,c,⁎, Sami Väisänen b , Olof Rådmark d , Carsten Carlberg b,c , Dieter Steinhilber a a Institute of Pharmaceutical Chemistry, University of Frankfurt, D-60438 Frankfurt, Germany b Department of Biochemistry, University of Kuopio, FIN-70211 Kuopio, Finland c Life Sciences Research Unit, University of Luxembourg, L-1511 Luxembourg, Luxembourg d Department of Medical Biochemistry and Biophysics, Division of Physiological Chemistry II, Karolinska Institute, S-17177 Stockholm, Sweden al Received 29 December 2006; received in revised form 30 March 2007; accepted 6 April 2007 Available online 19 April 2007 on Abstract pe rs 5-lipoxygenase (5-LO) is the key enzyme in the biosynthesis of proinflammatory leukotrienes. The 5-LO gene is a primary target of 1α,25dihydroxyvitamin D3 (1α,25(OH)2D3) and its expression is prominently increased during myeloid cell differentiation. Since no functional vitamin D response element (VDRE) has been reported for this gene so far, we performed in silico screening of the whole 5-LO gene area (84 kb, including 10 kb promoter region) and identified 22 putative VDREs. Both gelshift and reporter gene assays identified four of these candidates as functional VDREs. Their approximate positions are −2,250 (promoter), + 21,400 (intron 2), + 42,000 (intron 4) and + 50,600 (intron 5) in relation to the transcription start site (TSS). Remarkably, the VDRE at position +42,000 is one of the strongest known VDREs of the human genome. Chromatin immunoprecipitation (ChIP) assays demonstrated simultaneous association of vitamin D receptor (VDR), retinoid X receptor (RXR) and RNA polymerase II (Pol II) to the 5-LO gene regions containing two of these four putative VDREs. This indicates DNA looping of the TSS to even very distant gene regions. In summary, we suggest that the upregulation of the primary 1α,25(OH)2D3 target 5-LO is mediated in vivo by a prominent VDRE in intron 4. © 2007 Elsevier B.V. All rights reserved. 1. Introduction or 's Keywords: 5-LO; VDR; Chromatin; In silico screening; Vitamin D response elements Au th 5-LO is the key enzyme in the biosynthesis of leukotrienes [1,2], which are involved in inflammatory reactions. Recent data suggest that the 5-LO pathway plays a role in the development of cancer, cardiovascular diseases, adiposity, bone Abbreviations: 1α,25(OH)2D3, 1α,25-dihydroxyvitamin D3; 5-LO, 5lipoxygenase; ANF, atrial natriuretic factor; ARP0, acidic riboprotein P0; ChIP, chromatin immunoprecipitation; CYP24, 24-hydroxylase; DR, direct repeat; ER, everted repeat; FCS, fetal calf serum; MM6, Mono Mac 6; Pol II, RNA polymerase II; RE, response element; RXR, retinoid X receptor; TGFβ, transforming growth factor β; TSS, transcription start site; VDR, vitamin D receptor; VDRE, vitamin D response element ⁎ Corresponding author. Life Sciences Research Unit, University of Luxembourg, 162a, Avenue de la Faïencerie, L-1511 Luxembourg, Luxembourg. Tel.: +352 4666446451; fax: +352 4666446435. E-mail address: sabine.seuter@uni.lu (S. Seuter). 1388-1981/$ - see front matter © 2007 Elsevier B.V. All rights reserved. doi:10.1016/j.bbalip.2007.04.007 traits and neurodegenerative diseases [3–7]. 5-LO is expressed in a variety of immune competent cells including B-lymphocytes, granulocytes, monocytes, mast cells and dendritic cells [8]. 5-LO activity, protein expression and mRNA are strongly increased during differentiation of myeloid cell lines. 1α,25 (OH)2D3 and transforming growth factor-β (TGFβ) were identified as potent inducers of 5-LO gene expression [9,10]. Interestingly, the addition of cycloheximide inhibited the effects of TGFβ but not those of 1α,25(OH)2D3. This indicates a primary genomic effect of 1α,25(OH)2D3 on 5-LO mRNA expression in the presence of TGFβ-induced proteins [11]. Thus, the TGFβ-mediated effect on 5-LO gene expression seems to be a secondary response. Remarkably, for the strong upregulation of 5-LO mRNA expression the 5-LO promoter does not seem to be sufficient, since 1α,25(OH)2D3 and TGFβ had no effect on 5-LO transcription in nuclear run-off assays using nuclear extracts from MM6 cells [12] and on 5-LO 865 S. Seuter et al. / Biochimica et Biophysica Acta 1771 (2007) 864–872 2.2. RNA extraction and real-time quantitative PCR y Total RNA extraction and real-time quantitative PCR were performed as described [27]. The gene-specific primer pairs (and product sizes) were as follows: 5-LO gene forward 5′-ACCATTGAGCAGATCGTGGACACGC-3′ and reverse 5′-GCAGTCCTGCTCTGTGTAGAATGGG-3′ (488 bp) and control gene acidic riboprotein P0 (ARP0, also known as 36B4) forward 5′-AGATGCAGCAGATCCGCAT-3′ and reverse 5′-GTGGTGATACCTAAAGCCTG-3′ (318 bp). co p 2.3. In silico screening for putative VDREs on al Based on the analysis of the in vitro VDR-binding preferences of systematic variations of the consensus DR3-type VDRE [28], VDREs were subdivided into classes I, II and III according to their binding strength in relation to the consensus DR3-type VDRE of the mouse osteopontin gene. In this classification system, the consensus hexameric core sequence RGKTCA (R= G or A, G or T, K = G or T) along with single nucleotide variations with the same strong affinity to in vitro bind VDR (90–100%) are classified as consensus“, while REs containing single nucleotide variations from the consensus sequence with a 60–90%, 30–60% and 0–30% binding strength are grouped into the classes I, II and III, respectively. Consequently, REs containing multiple variations can be grouped into categories according to the number of variations belonging to class I, II or III (e.g. 2/1/0 for REs with two class I and one class II variations). The average binding strength of known REs belonging to each of these categories allows a good estimation, whether a putative RE will in vitro display VDR association. Based on this categorization, we kept in the list we obtained from the in silico screening only candidate VDREs with maximal two deviations from the optimal RGKTCA core binding sequence. For the specific VDRE search we considered only hexameric core sequence pairs in DR3, DR4, ER6, ER7, ER8 or ER9 orientation. Au th or 's pe rs promoter activity in reporter gene assays in different cell types [13]. In a recent study [14] we demonstrated VDR–RXR heterodimer association to a putative VDRE in the 5-LO promoter region from − 305 to − 8 bp. However, his region failed to show transactivation in reporter gene experiments. Thus, no functional VDRE has been reported for the human 5-LO gene so far. Previously, we have identified Smad binding elements in the distal part of the 5-LO gene [15]. Together, these data suggested that the induction of 5-LO mRNA expression by 1α,25 (OH)2D3 might be mediated by VDREs, which are also located outside of the 5-LO promoter region. The biologically most active vitamin D metabolite, 1α,25 (OH)2D3, is essential for mineral homeostasis and skeletal integrity [16] and is important for the control of the growth and differentiation in normal tissues and malignant cells derived from prostate, breast and bone [17]. 1α,25(OH)2D3 mainly acts as a nuclear hormone and mediates its genomic effects via the nuclear receptor VDR [18]. A direct modulation of transcription by 1α,25 (OH)2D3 is achieved through the specific binding of activated VDR to a VDRE [19] with the optimal RGKTCA (R = A or G, K = G or T) core binding sequence. VDR binds as a dimer to DNA and in most cases the nuclear receptor RXR is its heterodimeric partner [19]. Simple VDREs are therefore formed by two hexameric core binding motifs in a direct repeat (DR) or everted repeat (ER) orientation. The optimal spacing for DR-type VDREs was found to be three and four nucleotides (DR3, DR4) [20,21] and that of ER-type VDREs seven to nine nucleotides (ER7, ER8, ER9) [22,23]. Although individual VDREs are able to induce transactivation on their own, the presence of multiple VDREs in a regulatory region will allow a more flexible and complex regulation of the respective gene. Ligand binding to the VDR causes a conformational change within its ligand-binding domain that results in the replacement of corepressor protein by coactivator protein of the p160-family [24]. The latter provides a link to enzymes displaying histone acetyltransferase activity and lead to chromatin relaxation [25,26]. In this study, we confirmed that the human 5-LO gene is a primary 1α,25(OH)2D3 target. We identified in the whole 5-LO gene area (84 kb, including the 10 kb promoter region) 22 putative VDREs and characterized them by in vitro methods as well as by ChIP analysis in the living cell. At least two of these REs seem to be functional. Remarkably, the VDRE at position + 42,000 is one of the strongest known DR3-type VDREs of the human genome. 2. Materials and methods 2.1. Cell culture MM6 cells were grown in RPMI 1640 medium supplemented with 10% (vol/vol) fetal calf serum (FCS), 1× nonessential amino acids, sodium pyruvate (1 mM), oxalacetate (1 mM) and insulin (10 μg/ml). MCF-7 cells were grown in α-MEM supplemented with 5% FCS. All media also contained 2 mM L-glutamine, 0.1 mg/ml streptomycin and 100 U/ml penicillin and the cells were kept at 37 °C in a humidified 95% air/5% CO2 incubator. Prior to mRNA or chromatin extraction, cells were grown overnight in phenol red-free medium supplemented with charcoal-stripped FCS. Then, cells were treated with solvent (ethanol) or 1α,25 (OH)2D3 (kindly provided by Dr. Lise Binderup, LEO Pharma, Ballerup, Denmark) for RNA extractions and chromatin preparations. 2.4. DNA constructs Full-length cDNAs for human VDR and RXRα were cloned into the pSG5 expression vector (Stratagene, La Jolla, CA, USA) [20]. The same constructs were used for both T7 RNA polymerase-driven in vitro transcription/translation of the respective cDNAs and for SV40 promoter-driven overexpression in mammalian cells. Each two copies of the VDREs derived from the rat ANF and the candidate response elements (REs) of the 5-LO gene (Table 1) were fused with the thymidine kinase promoter driving the firefly luciferase reporter gene. All constructs were verified by sequencing. 2.5. Transient transfections and luciferase reporter gene assays MCF-7 cells were lipofected with a reporter plasmid and the expression vector for human VDR (each 1 μg). Subsequently, the cells were stimulated with 100 nM 1α,25(OH)2D3 or solvent. Transfection and luciferase reporter gene assay were performed as described [27]. MM6 cells were transfected by electroporation with 40 μg of reporter gene plasmid, 5 μg of the expression vectors for human VDR and RXR and 1 μg of the internal standard pCMVSEAP. The cells were incubated with 100 nM 1α,25(OH)2D3 or solvent for 6 h. Electroporation and reporter gene assays were performed as described [29]. 2.6. Gelshift assay In vitro translated VDR and RXR proteins were generated by coupled in vitro transcription/translation using their respective pSG5-based full-length cDNA expression constructs and rabbit reticulocyte lysate as recommended by the supplier (Promega, Madison, WI, USA). Protein batches were quantified by test translation in the presence of [35S]-methionine. The specific concentration of the receptor proteins was adjusted to ∼4 ng/μl (10 ng corresponds approximately to 0.2 pmol) after taking the individual number of methionine residues per protein into account. Gelshift assays were performed with 10 ng of the appropriate in vitro translated proteins. The proteins were incubated for 15 min in a total volume of 20 μl binding buffer (150 mM KCl, 1 mM dithiothreitol, 0.2 μg/μl poly(dI-C), 5% glycerol, 10 mM HEPES, pH 7.9). Constant amounts (1 ng) of [32P]-labeled double-stranded 35- to 41-mer oligonucleotides (50,000 cpm) containing one copy of the respective REs (Table 1) were then added and incubation was continued for 866 S. Seuter et al. / Biochimica et Biophysica Acta 1771 (2007) 864–872 Table 1 Sequence and position of putative VDREs within the human 5-LO gene Sequence Type Position (TSS = +1) Gene region Strand ratANF 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 AG AGGTCA TGA AGGACA GG GGTGTA AATT AGTTCA CA AGTTCA GCA AGTTTG CA TGGCCC TCAGCCA AGTTCA CT TGGCCT TAAGCAAT AGTTTA TG TGACCT TGGGTAA AGTTCA TC TGGCCT GTGAATCT GGGTTA GA TGACCT TGTTTGAC AGTCTA CC TGAACT AATTTTTC AGGTTA AG TGACCT GCAGTGGAA GGGACA CA AGGTCA AGG AGTTTG AT GGGACA GGC AGGACA TG TGACCC ACC TGGCCC TC TGCCCC TCC TGGCCC GG TGAACC TCAA TGGCCT CG TGACCT CGTC TAACCC TG AGGTCA GGA GGGCCA CA TGGCCC TGGC TGCCCT CC AGGGCA CGT GGGTTA AC AGGTCA AAT AGTACA TG CGCCCC CTGAGCCA GGTTCA GG TGCCCT CCGGCCTT GGGGCA CC AGGTCA CCCC AGGTTA DR3 DR4 DR3 ER7 ER8 ER7 ER8 ER8 ER8 ER9 DR3 DR3 DR3 DR3 DR4 DR4 DR3 DR4 DR3 DR3 ER8 ER8 DR4 − 907 − 6650 − 2239 +1519 +12590 +21427 +29723 +31127 +34481 +35595 +42034 +42318 +42898 +43857 +46610 +48811 +50673 +50688 +52118 +66822 +69186 +69413 +70102 promoter promoter promoter intron 1 intron 2 intron 2 intron 3 intron 3 intron 3 intron 3 intron 4 intron 4 intron 4 intron 4 intron 4 intron 4 intron 5 intron 5 intron 6 intron 8 intron 10 intron 11 intron 13 + + + + + + + + + + + + − − − − + − + + + + + co p al on The following PCR profile was used: preincubation for 5 min at 94 °C, 45 cycles of 30 s at 95 °C, 30 s at 60–65 °C and 30 s at 72 °C and one final incubation for 10 min at 72 °C. The primers are listed in Table 2. PCR products were separated by electrophoresis through 2.0% agarose, stained by SybrGreen and monitored on a Fuji FLA3000 reader using ScienceLab99 software. 2.7. ChIP assays and PCR of chromatin templates pe rs 20 min at room temperature. Protein–DNA complexes were resolved by electrophoresis through 8% non-denaturing polyacrylamide gels (mono- to bisacrylamide ration 19:1) in 0.5× TBE (45 mM Tris, 45 mM boric acid, 1 mM EDTA, pH 8.3) for 90 min at 200 V and quantified on a FLA-3000 reader (Fuji, Tokyo, Japan) using ScienceLab99 software (Fuji). y RE Table 2 Genomic PCR primers Location 2 6 60 °C +29636 to +29984 60 °C CCACTTCCATTTGACCTTGTACTGAGG TGCTTCCCAGGTTCAAGTGATCCATTTAC CTGTTGTTTCTATGCTTCTTCCTTGGCTTC GAAAAGGACAGATGAGTGGAAGGAGGAAG CAGCAGGTAGACGAGGCTGAGAAC GGAGATGGTGGATACTGCTGCATTAC TGGCTGGCGACAGCTTGGTTTC AGCTTTCTCCTGTGCCAAACTGTG GAACAGAGACAGCATCACTCCAGTG TGGGAGGTGATTAGGGTTAGACGAG TGGGCTGAAGGACCTGAGTGCTC GATCAACACAGGGTTGCAGCCATTC GGAGGGCATCTGAGATGTGGAGTG GACCATCATTCTGTAAGGCAGCTG AGGAACAGACACCTCGCTGAGGAG GAGGCTGAGGTAGATGTAGTCGTCAGTG GTCCAGGCTGGGGGTATCTG CGCAGAGGAGGGCGGAGTGG 62 °C +46414 to +46788 62 °C +48473 to +48839 60 °C 16/17 +50501 to +50821 62 °C 19 +66449 to +66858 60 °C TSS − 220 to +142 60 °C CYP24 − 424 to − 64 65 °C 15 MM6 human monocytic cells were treated with 1α,25 (OH)2D3 and the relative changes of 5-LO mRNA expression was monitored by real-time quantitative PCR in relation to the − 2392 to − 2046 +41849 to +42210 14 3.1. The human 5-LO gene is transcriptionally regulated by 1α,25(OH)2D3 Primer sequences (5′–3′) Au 10 3. Results Annealing temperature th Region or 's ChIP assays were performed as described [27]. The only modification from this protocol was the isolation of the nuclei prior to SDS lysis: following the PBS washes, the cell pellets were resuspended in Pipes buffer (5 mM Pipes pH 8.0, 85 mM KCl, 0.5% NP-40, protease inhibitor cocktail (Roche, Mannheim, Germany)) and incubated for 10 min on ice. After centrifugation, the nuclei pellets were resuspended in SDS lysis buffer and further processed according to the protocol. The antibodies against VDR (sc-1008), RXRα (sc-553) and Pol II (sc-899) and IgG (sc-2027) were obtained from Santa Cruz Biotechnologies (Heidelberg, Germany). 867 co p al on Fig. 1. Time course of 5-LO mRNA expression in MM6 cells and association of the 5-LO TSS with VDR. Real-time quantitative PCR was used to determine the relative changes of 5-LO mRNA expression in relation to the control gene ARP0 in MM6 cells in response to 50 nM 1α,25(OH)2D3 over a time period of 24 h (A). The data represent the means of at least three independent cell treatments and the bars show the standard deviations. Two-tailed Student's t-tests were performed to determine the significance of the mRNA induction by 1α,25 (OH)2D3 in reference to untreated cells (time point 0 h) (**p b 0.01) (A). Chromatin was extracted from MM6 and MCF-7 cells, that had been treated for the indicated time periods with 50 nM or 10 nM 1α,25(OH)2D3, respectively (B). ChIP experiments were performed with anti-VDR, anti-RXR or anti-Pol II antibodies. ChIP with IgG and template derived from mock-immunoprecipitated chromatin (no ab) served as negative controls. The association of VDR and its partner proteins was monitored on the TSS of the 5-LO gene and the proximal promoter of the human CYP24 gene. Representative agarose gels are shown. or 's pe rs control gene ARP0 over a time period of 0 to 24 h. Fold inductions were calculated in relation to time point 0 h. Already after 2 h of ligand treatment, 5-LO mRNA expression was induced 1.8-fold and after 24 h, 5-LO mRNA has accumulated to result in a statistically significant 4.0-fold upregulation (Fig. 1A). In previous semiquantitative RT-PCR studies, the addition of cycloheximide did not inhibit the induction of 5-LO mRNA by 1α,25(OH)2D3 [11]. For another approach in verifying that 5-LO is a direct 1α,25 (OH)2D3 target gene, we assessed via ChIP assays a 1α,25 (OH)2D3-dependent association of VDR with the 5-LO TSS (Fig. 1B). In a recent study [14] we demonstrated VDR–RXR heterodimer association to the 5-LO promoter region adjacent to the TSS from − 305 to − 8 bp containing a putative VDRE, for which no functionality could be shown. Here, we demonstrate that VDR, RXR and Pol II interact with the proximal 5-LO promoter region containing the TSS (−220 to + 142 bp) in 5-LO permissive MM6 cells, but not in MCF-7 human breast cancer cells, which do not express 5-LO mRNA (data not shown). Even though the bands are not very strong, which is probably due to the difficult PCR amplification of this extremely GC-rich region, the difference to the no antibody, IgG and MCF-7 controls is obvious. The interaction of VDR and Pol II with the TSS of the 5-LO gene seems to be moderately ligand-dependent and appeared to increase from untreated cells to cells stimulated for 1 and 2 h with 1α,25(OH)2D3, whereas the binding of RXR is apparently ligand-independent. The known primary 1α,25 (OH)2D3 target gene 24-hydroxylase (CYP24) was used as a positive control [30]. A 1α,25(OH)2D3-dependent association of VDR, RXR and Pol II to the proximal promoter of the CYP24 gene (containing two DR3-type VDREs) was demonstrated in both cell lines, but most prominently in the highly 1α,25(OH)2D3-responsive MCF-7 cells. These findings provide another indication that the 5-LO gene is regulated by VDR and its ligand. Taken together, both short-term mRNA upregulation as well as the occupation of the TSS with VDR confirmed that the 5-LO gene is a primary target of 1α,25(OH)2D3. y S. Seuter et al. / Biochimica et Biophysica Acta 1771 (2007) 864–872 3.2. In silico screening of the human 5-LO gene for putative VDREs Au th Based on a list of more than 15 known natural VDREs [31] we used the hexameric core sequence RGDKYR (R = G or A, D = A, G or T, K = G or T, Y = C or T) for in silico screening of the 5-LO gene area (10 kB upstream of the TSS until the end of the last exon) for putative VDREs. For the specific VDRE search we considered only hexameric core sequence pairs in DR3, DR4, ER6, ER7, ER8 or ER9 orientation and kept in our list only candidates with maximal two deviations from the optimal RGKTCA core binding sequence. These restrictions resulted in 22 putative VDREs, two of which (REs 1 and 2) were located in the promoter region of the 5-LO gene, one (RE3) in intron 1, two (REs 4 and 5) in intron 2, four (REs 6 to 9) in intron 3, six (REs 10 to 15) in intron 4, two (REs 16 and 17) in intron 5 and one each in introns 6, 8, 10, 11 and 13 (REs 18, 19, 20, 21 and 22, respectively) (Fig. 2). No REs were found in the exon sequences. Eight DR3-type, five DR4-type, six ER8-type, two ER7-type and one ER9-type VDREs have been identified (Table 1). Interestingly, introns 4 to 6 and introns 8 to 13 represent regions with a higher RE density. Moreover, screening for repetitive sequence (www.girinst.org/censor) showed that these “hot spots” are also areas of rather low density of interspersed elements, whereas in general the 5-LO gene contains a rather high amount of repetitive sequences (Fig. 2). In summary, the in silico screening resulted in 22 putative VDREs within the whole 5-LO gene area. 3.3. In vitro characterization of putative VDREs within the 5-LO gene To assess the relative VDR–RXR heterodimer binding on the 22 putative VDREs of the 5-LO gene, gelshift assays were performed using in vitro translated VDR and RXR proteins (Fig. 3). The conditions were identical to our earlier DR3-type VDRE comparative study [31] and the DR3-type VDRE of the rat atrial natriuretic factor (ANF) gene was chosen as a reference. Only VDR and RXR in combination but not on their own created specific shifts (data not shown) making competition 868 co p y S. Seuter et al. / Biochimica et Biophysica Acta 1771 (2007) 864–872 Fig. 2. Genomic organization of the 5-LO gene and in silico screening for putative VDREs. The whole 5-LO gene area comprising 10 kb promoter region and the sequence until the end of the last exon (84 kB) was screened in silico for putative VDREs. 22 candidate VDREs (red lines) were identified. The sequences of the REs are shown in Table 1. Below the distribution of repetitive sequence is indicated. al 3.4. Functionality of putative VDREs within the human 5-LO gene on The functionality of 10 of the 22 putative VDREs, which displayed reasonable to high VDR–RXR heterodimer binding capacity in vitro, four putative VDREs with low in vitro VDR– RXR association, two putative REs without VDR–RXR heterodimer binding property as controls and the reference VDRE of the rat ANF gene [32] was assessed by reporter gene assays in transiently transfected MCF-7 cells (Fig. 4A). Two copies of each of the VDREs were fused with the thymidine kinase promoter driving the firefly luciferase reporter gene. MCF-7 cells were transfected with these constructs and an expression vector for the human VDR and the response to 1α,25 (OH)2D3 was monitored after 16 h. The best four VDREs plus controls were also used with MM6 cells, VDR and RXR were cotransfected and the stimulation time was only 6 h (Fig. 4B). In the established 1α,25(OH)2D3-responding MCF-7 cells [33] high inductions by 1α,25(OH)2D3 were obtained (Fig. 4A). Compared with the 20.2-fold inducibility of the reference VDRE, RE10 was prominently more potent with a 118.2-fold induction of reporter gene activity (Fig. 4A). The 3.0-, 2.8- and 1.9-fold induction mediated by the REs 2, 5 and 16, respectively, indicate that they are also able to function as VDREs. Interestingly, the only functional VDRE located in the 5-LO promoter region, RE2, and the not inducible RE3 Au th or 's pe rs experiments or site directed mutagenesis dispensable. The DR3type REs 2 and 10 bound VDR–RXR heterodimers 2.1- and 2.5fold more effectively than the reference DR3-type VDRE, respectively. Also, the ER7-type RE5 showed a 1.4-fold stronger binding. The DR4-type RE15 (66% of the binding of the reference RE) and the DR3-type RE19 (37%) can be considered as VDREs as well. Finally, the DR3-type RE16 (11% heterodimer binding capacity), the DR4-type REs 1, 14, and 22 (19%, 17%, and 26%, respectively) and the ER8-type RE20 (10%) display weak VDR–RXR heterodimer binding. In contrast, the binding of VDR–RXR heterodimers to the REs 3, 4, 6–9, 11–13, 17, 18, and 21 was clearly too low under our stringent assay conditions to consider them as efficient VDREs. The non-specific bands with REs 3 and 18 occurred regardless of the presence or absence of VDR–RXR heterodimers and are thus not considered to have any relevance for 1α,25(OH)2D3 signalling. The strongest REs 2, 5, 10, 15 and 19 belong to the category of REs with consensus sequence, whereas all other REs showed weaker in vitro association with VDR–RXR heterodimers, because they carry one or two variations from the consensus sequence. Taken together, according to standardized in vitro criteria, of the 22 candidate REs, REs 2 and 10 can be considered as high affinity VDREs and REs 5, 15 and 19 as good VDREs (See supplementary table for the grouping of the putative VDREs into variation categories; compare Materials and Methods 2.2). Fig. 3. In vitro analysis of putative VDREs derived from the 5-LO gene. Gelshift experiments were performed with in vitro translated human VDR and RXR in the presence of different [32P]-labeled REs representing the 22 candidate VDREs of the 5-LO gene and the rat ANF DR3-type reference VDRE. Protein–DNA complexes were resolved from free probe through non-denaturing 8% polyacrylamide gels. Representative gels are shown. The relative amounts of VDR–RXR heterodimer complex formation below the gels indicate the means of at least three independent gel shift experiments. NS indicates non-specific complexes. 869 S. Seuter et al. / Biochimica et Biophysica Acta 1771 (2007) 864–872 on al co p y by the VDR ligand (1.8-, 1.4-, 1.2- and 0.6-fold, respectively). Since the rat ANF VDRE shows a similarly low induction as RE10 in this cell line, the shorter stimulation time, which was due to a different transfection method, may have caused this attenuated transactivation. The observed stimulations are low, but they are statistically significant and show that at least RE10, which was the strongest RE in MCF-7 cells (Fig. 4A), is responsive in MM6 cells, too. Overall, there is a good correlation between the functional assays in different cell lines (Fig. 4) and the in vitro VDR–RXR heterodimer complex formation (Fig. 3). REs 2 and 10 were confirmed as very good VDREs and RE16 showed a weak response in both assays. RE15 and the in vitro moderately VDR–RXR heterodimer binding RE19 did not display 1α,25 (OH)2D3-mediated transactivation in the functional assay. Due to its location in vast repetitive sequence (Fig. 2), RE5 was considered to be unlikely functional and not examined further. In summary, four VDREs (REs 2, 5, 10, and 16) were shown to respond effectively to 1α,25(OH)2D3 and its receptor in the evaluation in MCF-7 cells. 3.5. Functionality of putative VDREs in chromatin context of living cells pe rs We next examined in living cells, whether VDR was present at the genomic regions containing the REs. Chromatin was extracted from MM6 cells, which had been grown overnight in the presence of 10% charcoal-treated FCS, stimulated for 0, 1 and 2 h with 50 nM 1α,25(OH)2D3 and then cross-linked for 10 min in the presence of formaldehyde. ChIP assays were performed using antibodies against VDR, RXR and Pol II. The DNA fragments that were recovered from reverse-crosslinked chromatin served as templates for PCR reactions (for primer sequences see Table 2). For ChIP experiments those genomic regions were chosen, which contain the two strongest REs 2 and 10. Also, the regions containing REs 15 and 19 were characterized due to their considerable VDR–RXR heterodimer binding affinity in vitro and the region containing RE16 due to its activity in the functional assay. Since REs 16 and 17 are in very close proximity, they cannot be detected separately and are represented together. The regions containing REs 6 and 14 served as negative controls. Representative agarose gels are shown (Fig. 5). The input lane serves as a reference to compare the detection sensitivity for the seven genomic regions within the 5-LO gene and ChIP assays using IgG or no antibody served as specificity controls. Prominent VDR, RXR and Pol II binding to the regions containing the two potent REs 2 and 10 and to a lesser extent also to REs 15, 16/17 and 19 could be detected. VDR, RXR and Pol II binding to these regions was found in the absence of ligand and 1α,25(OH)2D3 treatment did not significantly increase VDR association to these regions. Only in the region of RE 19 a slight 1α,25(OH)2D3-dependent modulation of the VDR and RXR association was observed. It has to be noted, that for regions 15, 16/17 and 19 some background signal is obtained with the no antibody and IgG controls. For the two negative control regions of REs 6 and 14 no association of VDR, RXR or Pol II was obtained. Taken Au th or 's Fig. 4. Functionality of putative VDREs derived from the 5-LO gene. Reporter gene assays were performed with extracts from MCF-7 (A) and MM6 (B) cells that were transiently transfected with luciferase reporter constructs each containing two copies of one of the candidate REs or of the rat ANF DR3type VDRE and an expression vector for human VDR (MCF-7) or human VDR and RXR (MM6). Cells were treated for 16 h (MCF-7) or 6 h (MM6) with either solvent or 100 nM 1α,25(OH)2D3. Relative luciferase activity is shown and fold inductions are indicated above the columns. Columns represent means of triplicate transfections and bars indicate standard deviations. Representatives of at least three independent experiments are shown. Two-tailed Student's t-tests were performed to determine the significance of reporter gene induction by 1α,25(OH)2D3 in reference to solvent controls (*p b 0.05, **p b 0.01, ***p b 0.001). (intron 1) display a rather high basal luciferase activity. Neither the absolute reporter gene activity nor the inducibility of the other putative VDREs was sufficient for a functional VDRE. Reporter gene assays in HeLa human cervix carcinoma cells gave similar results (data not shown). The 5-LO positive cell line MM6 was less responsive to 1α,25(OH)2D3 (Fig. 4B). Both the rat ANF VDRE and RE10 showed a low but statistically significant induction of reporter gene activity by 2.5-fold following the treatment with 1α,25 (OH)2D3. REs 2, 5, 16 and 19 were not significantly activated 870 co p y S. Seuter et al. / Biochimica et Biophysica Acta 1771 (2007) 864–872 Fig. 5. Association of VDRE-containing genomic regions of the 5-LO gene with VDR and other nuclear proteins. Chromatin was extracted from MM6 cells that had been treated for the indicated time periods with 50 nM 1α,25(OH)2D3. ChIP experiments were performed with anti-VDR, anti-RXR or anti-Pol II antibodies. ChIP with IgG and template derived from mock-immunoprecipitated chromatin (no ab) served as negative controls. The association of VDR and its partner proteins was monitored on seven of the 22 RE containing genomic regions of the 5-LO gene. Representative agarose gels are shown. Au th or 's pe rs This study contributes to the understanding of the regulation of the human 5-LO gene by the nuclear hormone 1α,25(OH)2D3. Confirming previous studies that 5-LO is a primary VDR target [11,12], we observed primary ligand responses in the myeloid cell line MM6, which has been reported before to express prominent amounts of 5-LO upon treatment with 1α,25(OH)2D3 and TGFβ [9,10,15]. The 1.8-fold induction of 5-LO mRNA after 2 h of treatment with 1α,25(OH)2D3 is not a very strong upregulation, but it is in the order of what was observed with many other 1α,25 (OH)2D3 target genes [34,35]. The prominent 4.1-fold induction of the mRNA amounts after 24 h may reflect an additional secondary effect of 1α,25(OH)2D3 on 5-LO mRNA expression. Previous studies demonstrated that the effects of TGFβ and 1α,25(OH)2D3 are not mediated via the 5-LO promoter [14,15]. Instead, we found that Smad binding elements and TGFβ responsive elements located in exon 10 and in intron 13 of the 5-LO gene are able to mediate TGFβ effects in reporter gene assays [15]. In order to identify putative VDREs within the 5-LO gene, we performed an in silico screening for VDREs that covered the whole 84 kb of the 5-LO genomic sequence. After a number of carefully selected restrictions (see Results) the screen resulted in 22 candidate VDREs. Five of these 22 REs (REs 2, 5, 10, 15, 19) showed considerable VDR binding affinity in vitro and another five weakly bound VDR–RXR heterodimers. ChIP assays demonstrated that two of these putative REs are bound by VDR and RXR in intact MM6 cells (REs 2 and 10). The functional response in reporter gene assay was tested for the putative VDREs. In these assays two copies of the putative VDREs were inserted into the reporter gene plasmid, each VDRE was thus tested separately. Induction of luciferase activity by 1α,25 (OH)2D3 was obtained for REs 2, 5, 10, and 16. The response was very strong for RE 10 in MCF-7 cells. The rather low inductions of reporter gene activity by the VDR ligand in the 5-LO on 4. Discussion permissive MM6 cells is probably due to the shorter stimulation time and/or to the harsh transfection conditions required to obtain acceptable transfection rates that might reduce cellular functionality [36]. In addition, the expression of coactivator and corepressor protein in this cell line is different from the highly 1α,25(OH)2D3responsive MCF-7 cells (data not shown). In summary, clearly positive results in all assays (gelshift, ChIP, reporter gene assay) were obtained for REs 2 and 10. The success rate of the in silico prediction of VDR binding sites is comparable with similar screenings [37], although it is lower than in our previous study [38]. Based on a more permissive consensus core binding motif, we previously suggested additional VDREs in the 5-LO gene promoter (from −309 to −262). VDR– RXR heterodimers indeed bind to this sequence in vitro and also weakly in the living cell, but there was no functional response [14]. The putative VDRE at +4,414 bp [28] did not match our restrictive search parameters and was not included in this study. Thus, in addition to the here described VDREs, which are functional separately, it appears possible that other VDREs of minor impact, may contribute to a more complex regulation by 1α,25(OH)2D3 for the intact 5-LO gene. The observation that the 5-LO gene contains several functional VDR associated regions is in accordance with the model of multiple VDREs per primary 1α,25(OH)2D3 target gene, which we developed during our analysis of the CYP24, cyclin C, p21 and insulin-like growth factor binding protein genes [27,30, 33,38]. With the in silico screening in addition to DR3- and DR4- also a number of putative ER7-, ER8-, and ER9-type VDREs have been identified. Four of the experimentally confirmed VDREs were of DR3 orientation. Three of these REs were shown to strongly bind VDR–RXR heterodimers in vitro and are potent VDREs in reporter gene assays. Interestingly, the strongest VDRE (RE10) is not located in the promoter region, but far downstream at + 42 kb. Remarkably, RE10 is one of the strongest known DR3-type VDREs in the human genome. The weaker RE2 at − 2250 bp was slightly active in MCF-7 but not in MM6 cells. Also, when this RE is was tested in the natural promoter context with reporter gene assays (plasmids pN6–pN0), it did not mediate 1α,25(OH)2D3 effects in MM6 and HeLa cells. Still, RE 2 and the proximal VDR–RXR binding region − 309 to − 262 [14] may contribute to the regulation of the gene by 1α,25(OH)2D3. al together, the genomic sequence of the 5-LO gene contains two prominent VDR- and RXR-associated regions, one being located in the promoter (RE 2) and the other 42 kb downstream of the TSS in intron 4 (RE 10). Three additional regions seem to display weaker VDR and RXR binding (REs 15, 16/17 and 19 in introns 4, 5 and 8, respectively). 871 S. Seuter et al. / Biochimica et Biophysica Acta 1771 (2007) 864–872 Appendix A. Supplementary data Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.bbalip.2007.04.007. y References on al co p [1] B. Samuelsson, S.-E. Dahlén, J.-Å. Lindgren, C.A. Rouzer, C.N. Serhan, Leukotrienes and lipoxins: structures, biosynthesis, and biological effects, Science 237 (1987) 1171–1176. [2] R.A. Lewis, K.F. Austen, R.J. Soberman, Leukotrienes and other products of the 5-lipoxygenase pathway. Biochemistry and relation to pathobiology in human diseases, N. Engl. J. Med. 323 (1990) 645–655. [3] A. Helgadottir, A. Manolescu, G. Thorleifsson, S. Gretarsdottir, H. Jonsdottir, U. Thorsteinsdottir, N.J. Samani, G. Gudmundsson, S.F. Grant, G. Thorgeirsson, S. Sveinbjornsdottir, E.M. Valdimarsson, S.E. Matthiasson, H. Johannsson, O. Gudmundsdottir, M.E. Gurney, J. Sainz, M. Thorhallsdottir, M. Andresdottir, M.L. Frigge, E.J. Topol, A. Kong, V. Gudnason, H. Hakonarson, J.R. Gulcher, K. Stefansson, The gene encoding 5-lipoxygenase activating protein confers risk of myocardial infarction and stroke, Nat. Genet. 36 (2004) 233–239. [4] S.F. Lockwood, M.S. Penn, S.L. Hazen, Z. Bikadi, F. Zsila, The effects of oral Cardaxtrade mark (disodium disuccinate astaxanthin) on multiple independent oxidative stress markers in a mouse peritoneal inflammation model: influence on 5-lipoxygenase in vitro and in vivo, Life Sci. (2006). [5] H. Manev, T. Uz, K. Sugaya, T. Qu, Putative role of neuronal 5-lipoxygenase in an aging brain, FASEB J. 14 (2000) 1464–1469. [6] M. Mehrabian, H. Allayee, 5-lipoxygenase and atherosclerosis, Curr. Opin. Lipidol. 14 (2003) 447–457. [7] M. Mehrabian, H. Allayee, J. Stockton, P.Y. Lum, T.A. Drake, L.W. Castellani, M. Suh, C. Armour, S. Edwards, J. Lamb, A.J. Lusis, E.E. Schadt, Integrating genotypic and expression data in a segregating mouse population to identify 5-lipoxygenase as a susceptibility gene for obesity and bone traits, Nat. Genet. 37 (2005) 1224–1233. [8] D. Steinhilber, 5-Lipoxygenase: a target for antiinflammatory drugs revisited, Curr. Med. Chem. 6 (1999) 71–85. [9] M. Brungs, O. Rådmark, B. Samuelsson, D. Steinhilber, On the induction of 5-lipoxygenase expression and activity in HL-60 cells: effects of vitamin D3, retinoic acid, DMSO and TGF β, Biochem. Biophys. Res. Commun. 205 (1994) 1572–1580. [10] M. Brungs, O. Rådmark, B. Samuelsson, D. Steinhilber, Sequential induction of 5-lipoxygenase gene expression and activity in Mono Mac 6 cells by transforming growth factor β and 1,25-dihydroxyvitamin D3, Proc. Natl. Acad. Sci. U. S. A. 92 (1995) 107–111. [11] D. Härle, O. Rådmark, B. Samuelsson, D. Steinhilber, Calcitriol and transforming growth factor-beta upregulate 5-lipoxygenase mRNA expression by increasing gene transcription and mRNA maturation, Eur. J. Biochem. 254 (1998) 275–281. [12] D. Härle, O. Rådmark, B. Samuelsson, D. Steinhilber, Transcriptional and posttranscriptional regulation of 5-lipoxygenase mRNA expression in the human monocytic cell line Mono Mac 6 by transforming growth factorbeta and 1,25-dihydroxyvitamin D3, Adv. Exp. Med. Biol. 469 (1999) 105–111. [13] J. Uhl, N. Klan, M. Rose, K.D. Entian, O. Werz, D. Steinhilber, The 5-lipoxygenase promoter is regulated by DNA methylation, J. Biol. Chem. 277 (2002) 4374–4379. [14] B.L. Sorg, N. Klan, S. Seuter, D. Dishart, O. Radmark, A. Habenicht, C. Carlberg, O. Werz, D. Steinhilber, Analysis of the 5-lipoxygenase promoter and characterization of a vitamin D receptor binding site, Biochim. Biophys. Acta 1761 (2006) 686–697. [15] S. Seuter, B.L. Sorg, D. Steinhilber, The coding sequence mediates induction of 5-lipoxygenase expression by Smads3/4, Biochem. Biophys. Res. Commun. 348 (2006) 1403–1410. [16] A.L. Sutton, P.N. MacDonald, Vitamin D: more than a “bone-a-fide” hormone, Mol. Endocrinol. 17 (2003) 777–791. Au th or 's pe rs It is important to note that our in silico screening was not restricted to regions comprising only sequences of maximal − 10 kb to + 5 kb in relation to the TSS, as in recent whole genome screens for regulatory elements [28,39], but involved up to 10 kb upstream sequence plus the complete gene sequence including the introns. Therefore, our approach also identified candidate REs that are located in a distance of up to + 50 kb relative to the TSS and demonstrated their functionality. Based on the present understanding of enhancers, DNA looping and chromatin units being flanked by insulators or matrix attachment sites [40], these distances do not preclude a regulatory function in transcription initiation and/or elongation. Interestingly, apart from the TSS, VDR association to the genomic regions of the 5-LO gene appears to be independent from 1α,25(OH)2D3 treatment. This is in contrast to observations with other 1α,25(OH)2D3 target genes [27,30,33,38,41]. However, VDR is able and has been shown to bind to DNA without its ligand [42], but in this case it is associated with corepressors and thus does not activate transcription [43]. Upon ligand binding, the switch from corepressors to coactivators then leads to transcriptional activation [24]. Recently, we gained more insight into the synergistic regulation of 5-LO gene expression by 1α,25(OH)2D3 and TGFβ [15] and demonstrated that TGFβ signalling via Smad transcription factors is mediated by TGFβ responsive elements and Smad binding elements in exon 10 and intron 13. A crosstalk between the 1α,25(OH)2D3 and TGFβ signalling cascades has been described before [44–48]. However, concerning the prominent induction of 5-LO gene expression, the lack of transcriptional effects of 1α,25(OH)2D3 and TGFβ is surprising [11,12] by Previous data suggest that 5-LO mRNA induction is related to effects of 1α,25(OH)2D3 and TGFβ on transcript elongation and maturation and that this regulatory mechanism involves the more distal part of the 5-LO gene. Interestingly, we found Smad binding elements at the distal part of the 5-LO gene and a strongly functional RE in intron 4. Moreover, as demonstrated by the ChIP experiments, the VDR seems to be present at several sites of the 5-LO gene. At the moment, we can only speculate about the function the TGFβresponsive sites and the VDREs in the natural gene context. One possibility is that these REs interact with the 5-LO promoter and/or with each other via DNA looping. Another explanation which is favoured by the previous data is, that these sites enhance 5-LO transcript elongation and maturation. In conclusion, our study provided novel insight into the regulation of the 5-LO gene by 1α,25(OH)2D3. We suggest that the upregulation of this primary 1α,25(OH)2D3 target gene is mediated in vivo by a prominent DR3-type VDRE in intron 4. Acknowledgments We thank Astrid Brüggerhoff and Maija Hiltunen for expert technical assistance and Merja Matilainen for help with the in silico screening. Grants from the European Research Training Group GRK 757 of the Deutsche Forschungsgemeinschaft, the EU (LSHM-CT-2004-0050333, EICOSANOX) and the Finnish Academy supported this research. 872 S. Seuter et al. / Biochimica et Biophysica Acta 1771 (2007) 864–872 on al co p y [34] S. Swami, N. Raghavachari, U.R. Muller, Y.P. Bao, D. Feldman, Vitamin D growth inhibition of breast cancer cells: gene expression patterns assessed by cDNA microarray, Breast Cancer Res. Treat. 80 (2003) 49–62. [35] H.G. Palmer, M. Sanchez-Carbayo, P. Ordonez-Moran, M.J. Larriba, C. Cordon-Cardo, A. Munoz, Genetic signatures of differentiation induced by 1α,25-dihydroxyvitamin D3 in human colon cancer cells, Cancer Res. 63 (2003) 7799–7806. [36] N. Klan, D. Steinhilber, Transient transfection of the human myeloid cell line Mono Mac 6 using electroporation, Biotechniques 34 (2003) 142–147. [37] W.W. Wasserman, A. Sandelin, Applied bioinformatics for the identification of regulatory elements, Nat. Rev., Genet. 5 (2004) 276–287. [38] M. Matilainen, M. Malinen, K. Saavalainen, C. Carlberg, Regulation of multiple insulin-like growth factor binding protein genes by 1α,25dihydroxyvitamin D3, Nucleic Acids Res. 33 (2005) 5521–5532. [39] X. Xie, J. Lu, E.J. Kulbokas, T.R. Golub, V. Mootha, K. Lindblad-Toh, E.S. Lander, M. Kellis, Systematic discovery of regulatory motifs in human promoters and 3′ UTRs by comparison of several mammals, Nature 434 (2005) 338–345. [40] K. Ogata, K. Sato, T.H. Tahirov, Eukaryotic transcriptional regulatory complexes: cooperativity from near and afar, Curr. Opin. Struck. Biol. 13 (2003) 40–48. [41] F. Jiang, P. Li, A.J. Fornace Jr., S.V. Nicosia, W. Bai, G2/M arrest by 1,25dihydroxyvitamin D3 in ovarian cancer cells mediated through the induction of GADD45 via an exonic enhancer, J. Biol. Chem. 278 (2003) 48030–48040. [42] C. Zhang, D.R. Dowd, A. Staal, C. Gu, J.B. Lian, A.J. van Wijnen, G.S. Stein, P.N. MacDonald, Nuclear coactivator-62 kDa/Ski-interacting protein is a nuclear matrix-associated coactivator that may couple vitamin D receptor-mediated transcription and RNA splicing, J. Biol. Chem. 278 (2003) 35325–35336. [43] P. Polly, M. Herdick, U. Moehren, A. Baniahmad, T. Heinzel, C. Carlberg, VDR-Alien: a novel, DNA-selective vitamin D3 receptor-corepressor partnership, FASEB J. 14 (2000) 1455–1463. [44] J.K. Aschenbrenner, H.W. Sollinger, B.N. Becker, D.A. Hullett, 1,25(OH)2D3 alters the transforming growth factor beta signaling pathway in renal tissue, J. Surg. Res. 100 (2001) 171–175. [45] N. Subramaniam, G.M. Leong, T.A. Cock, J.L. Flanagan, C. Fong, J.A. Eisman, A.P. Kouzmenko, Cross-talk between 1,25-dihydroxyvitamin D3 and transforming growth factor-beta signaling requires binding of VDR and Smad3 proteins to their cognate DNA recognition elements, J. Biol. Chem. 276 (2001) 15741–15746. [46] Y. Yanagi, M. Suzawa, M. Kawabata, K. Miyazono, J. Yanagisawa, S. Kato, Positive and negative modulation of vitamin D receptor function by transforming growth factor-beta signaling through smad proteins, J. Biol. Chem. 274 (1999) 12971–12974. [47] J. Yanagisawa, Y. Yanagi, Y. Masuhiro, M. Suzawa, M. Watanabe, K. Kashiwagi, T. Toriyabe, M. Kawabata, K. Miyazono, S. Kato, Convergence of transforming growth factor-beta and vitamin D signaling pathways on SMAD transcriptional coactivators, Science 283 (1999) 1317–1321. [48] L. Yang, J. Yang, S. Venkateswarlu, T. Ko, M.G. Brattain, Autocrine TGFβ signaling mediates vitamin D3 analog-induced growth inhibition in breast cells, J. Cell. Physiol. 188 (2001) 383–393. Au th or 's pe rs [17] C. Mørk Hansen, L. Binderup, K.J. Hamberg, C. Carlberg, Vitamin D and cancer: effects of 1,25(OH)2D3 and its analogs on growth control and tumorigenesis, Front. Biosci. 6 (2001) D820–D848. [18] G. Jones, S.A. Strugnell, H.F. DeLuca, Current understanding of the molecular actions of vitamin D, Physiol. Rev. 78 (1998) 1193–1231. [19] C. Carlberg, P. Polly, Gene regulation by vitamin D3, Crit. Rev. Eukaryot. Gene Expr. 8 (1998) 19–42. [20] C. Carlberg, I. Bendik, A. Wyss, E. Meier, L.J. Sturzenbecker, J.F. Grippo, W. Hunziker, Two nuclear signalling pathways for vitamin D, Nature 361 (1993) 657–660. [21] M. Quack, C. Carlberg, Ligand-triggered stabilization of vitamin D receptor/retinoid X receptor heterodimer conformations on DR4-type response elements, J. Mol. Biol. 296 (2000) 743–756. [22] M. Schräder, S. Nayeri, J.P. Kahlen, K.M. Müller, C. Carlberg, Natural vitamin D3 response elements formed by inverted palindromes: polaritydirected ligand sensitivity of vitamin D3 receptor-retinoid X receptor heterodimer-mediated transactivation, Mol. Cell. Biol. 15 (1995) 1154–1161. [23] M. Schräder, K.M. Müller, S. Nayeri, J.P. Kahlen, C. Carlberg, VDR-T3R receptor heterodimer polarity directs ligand sensitivity of transactivation, Nature 370 (1994) 382–386. [24] C. Leo, J.D. Chen, The SRC family of nuclear receptor coactivators, Gene 245 (2000) 1–11. [25] J.D. Fondell, M. Guermah, S. Malik, R.G. Roeder, Thyroid hormone receptor-associated proteins and general positive cofactors mediate thyroid hormone receptor function in the absence of the TATA box-binding protein-associated factors of TFIID, Proc. Natl. Acad. Sci. U. S. A. 96 (1999) 1959–1964. [26] C. Rachez, B.D. Lemon, Z. Suldan, V. Bromleigh, M. Gamble, A.M. Näär, H. Erdjument-Bromage, P. Tempst, L.P. Freedman, Ligand-dependent transcription activation by nuclear receptors requires the DRIP complex, Nature 398 (1999) 824–828. [27] A. Saramäki, C.M. Banwell, M.J. Campbell, C. Carlberg, Regulation of the human p21(waf1/cip1) gene promoter via multiple binding sites for p53 and the vitamin D3 receptor, Nucleic Acids Res. 34 (2006) 543–554. [28] T.T. Wang, L.E. Tavera-Mendoza, D. Laperriere, E. Libby, N.B. MacLeod, Y. Nagai, V. Bourdeau, A. Konstorum, B. Lallemant, R. Zhang, S. Mader, J.H. White, Large-scale in silico and microarray-based identification of direct 1,25-dihydroxyvitamin D3 target genes, Mol. Endocrinol. 19 (2005) 2685–2695. [29] D. Dishart, N. Schnur, N. Klan, O. Werz, D. Steinhilber, B. Samuelsson, O. Rådmark, GC-rich sequences in the 5-lipoxygenase gene promoter are required for expression in Mono Mac 6 cells, characterization of a novel Sp1 binding site, Biochim. Biophys. Acta 1738 (2005) 37–47. [30] S. Väisänen, T.W. Dunlop, L. Sinkkonen, C. Frank, C. Carlberg, Spatiotemporal activation of chromatin on the human CYP24 gene promoter in the presence of 1α,25-dihydroxyvitamin D3, J. Mol. Biol. 350 (2005) 65–77. [31] A. Toell, P. Polly, C. Carlberg, All natural DR3-type vitamin D response elements show a similar functionality in vitro, Biochem. J. 352 (2000) 301–309. [32] J.P. Kahlen, C. Carlberg, Functional characterization of a 1,25-dihydroxyvitamin D3 receptor binding site found in the rat atrial natriuretic factor promoter, Biochem. Biophys. Res. Commun. 218 (1996) 882–886. [33] L. Sinkkonen, M. Malinen, K. Saavalainen, S. Väisänen, C. Carlberg, Regulation of the human cyclin C gene via multiple vitamin D3-responsive regions in its promoter, Nucleic Acids Res. 33 (2005) 2440–2451.