- Email: [email protected]
Vitamin D-mediated regulation of CYP21A2 transcription — A novel mechanism for vitamin D action
Biochimica et Biophysica Acta 1820 (2012) 1553–1559
Contents lists available at SciVerse ScienceDirect
Biochimica et Biophysica Acta journal homepage: www.elsevier.com/locate/bbagen
Vitamin D-mediated regulation of CYP21A2 transcription — A novel mechanism for vitamin D action Johan Lundqvist ⁎, Kjell Wikvall, Maria Norlin Department of Pharmaceutical Biosciences, Box 591, Uppsala University, SE-751 24 Uppsala, Sweden
a r t i c l e
i n f o
Article history: Received 30 November 2011 Received in revised form 16 March 2012 Accepted 20 April 2012 Available online 27 April 2012 Keywords: Steroidogenesis Vitamin D Calcitriol CYP21A2 Steroid
a b s t r a c t Background: 1α,25-Dihydroxyvitamin D3 has recently been reported to decrease expression and activity of CYP21A2. In this paper, we have studied the mechanisms for the 1α,25-dihydroxyvitamin D3-mediated effect on CYP21A2 transcriptional rate. Methods: We have studied the effects of 1α,25-dihydroxyvitamin D3 using luciferase reporter constructs containing different lengths of the CYP21A2 promoter. These constructs were transfected into cell lines derived from human and mouse adrenal cortex. The mechanism for the effects of vitamin D on the CYP21A2 promoter was studied using chromatin immunoprecipitation assay, mutagenesis and gene silencing by siRNA. Results: 1α,25-Dihydroxyvitamin D3 was found to alter the promoter activity via a VDR-mediated mechanism, including the comodulators VDR interacting repressor (VDIR) and Williams syndrome transcription factor (WSTF). The involvement of comodulator VDIR was confirmed by gene silencing. We identified a vitamin D response element in the CYP21A2 promoter. Interaction between this novel response element and VDR, WSTF and VDIR was shown by chromatin immunoprecipitation assay. When this sequence was deleted, the effect of 1α,25-dihydroxyvitamin D3 was abolished, indicating that this sequence in the CYP21A2 promoter functions as a vitamin D response element. Interestingly, an altered balance between nuclear receptors and comodulators reversed the suppressing effect of vitamin D to a stimulatory effect. General significance: This paper reports data important for the understanding of the mechanisms for vitamin D-mediated suppression of gene expression as well as for the vitamin D-mediated effects on CYP21A2. We report a novel mechanism for effects of 1α,25-dihydroxyvitamin D3. © 2012 Elsevier B.V. All rights reserved.
1. Introduction The steroid 21-hydroxylase CYP21A2 is expressed in the human adrenal cortex and catalyzes the production of deoxycorticosterone and 11-deoxycortisol, which are precursors for the production of corticosterone, aldosterone and cortisol. 21-Hydroxylase deficiency may lead to severe conditions such as congenital adrenal hyperplasia and Addison's disease. CYP21A2 has been reported to be transcriptionally regulated by transcription factors as steroidogenic factor-1 (SF-1) and nur77 [1,2]. Very recently, we reported that 1α,25-dihydroxyvitamin D3 decreases mRNA level and enzyme activity of CYP21A2 as well as the production of corticosterone and aldosterone in human adrenocortical NCI-H295R cells [3]. This is a previously unknown effect of the multifunctional hormone 1α,25-dihydroxyvitamin D3.
Abbreviations: RXR, retinoid X receptor; SF-1, steroidogenic factor-1; VDIR, VDR interacting repressor; VDR, vitamin D receptor; VDRE, vitamin D responsive element; nVDRE, negative vitamin D responsive element; WSTF, Williams syndrome transcription factor ⁎ Corresponding author. Present address: Breast Cancer Group, Unit of Cell Death and Metabolism, Danish Cancer Society Research Center, Strandboulevarden 49, 2100 Copenhagen Ø, Denmark. Tel.: + 46 18 4714144. E-mail address: [email protected] (J. Lundqvist). 0304-4165/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.bbagen.2012.04.017
1α,25-Dihydroxyvitamin D3 may increase the expression of certain genes while it decreases the expression of other genes. Both these effects of 1α,25-dihydroxyvitamin D3 are mediated by activation of the vitamin D receptor (VDR). Extensive research has been conducted to understand the mechanism for 1α,25-dihydroxyvitamin D3-mediated induction of gene expression (e.g. for CYP24A1). The mechanism is now well known and based on a direct interaction between the ligandactivated VDR–RXR complex and a vitamin D responsive element (VDRE) in the gene promoter [4]. However, little is known about the mechanism for 1α,25-dihydroxyvitamin D3-mediated suppression of gene expression [4]. In the present study, we have examined the mechanism for the recently reported 1α,25-dihydroxyvitamin D3-mediated downregulation of CYP21A2 activity and mRNA level [3]. The current study reports a novel mechanism for vitamin D-mediated suppression of gene transcription.
2. Material and methods 2.1. Chemicals and materials 1α,25-Dihydroxyvitamin D3 was obtained from Solvay, Duphar, The Netherlands. The vitamin D receptor antagonist TEI-9647 was a
1554
J. Lundqvist et al. / Biochimica et Biophysica Acta 1820 (2012) 1553–1559
generous gift from Teijin Pharma, Tokyo, Japan. CYP21A2 promoterluciferase reporter constructs were a kind gift from Professor Walter L. Miller, University of California, San Francisco. Expression vectors for VDIR and WSTF were kind gifts from Professor Shigeaki Kato, University of Tokyo. The full-length human vitamin D receptor (VDR) expression vector was a gift from Dr. Leonard Freedman, the Merck Research Laboratories, West Point. The expression vector for the human retinoid X receptor (RXR) was a kind gift from Dr. Ronald M. Evans, Howard Hughes Medical Institute, The Salk Institute for Biological Studies, San Diego. All other chemicals were of analytical grade and purchased from various commercial sources. 2.2. Cell culture and treatment Human adrenocortical carcinoma NCI-H295R cells (ATCC CRL-2128) were cultured in Dulbecco's modified Eagle's medium/Nutrient Mixture F-12 Ham (Sigma) supplemented with 1% ITS Plus premix (BD Biosciences), 2.5% NuSerum (VWR), 1% L-glutamine (Gibco) and 1% antibiotic/ antimycotic (Gibco). Mouse adrenocortical Y-1 (ATCC CCL-79) cells were cultured in F-12K Medium (ATCC) supplemented with 2.5% fetal bovine serum (Gibco), 15% horse serum (ATCC) and 1% antibiotic/ antimycotic (Gibco). All cells were cultured as monolayers in a humidified environment at 37 °C with 5% CO2. Cells were treated with 10 nM 1α,25dihydroxyvitamin D3 dissolved in ethanol for 24 h. The control group was treated with the same amount of ethanol. 2.3. Analysis of CYP21A2 promoter activity by luciferase reporter assay NCI-H295R cells and Y-1 cells were transiently transfected with luciferase reporter constructs containing different lengths of the CYP21A2 promoter, spanning from −9027 to +13, coupled to the luciferase gene [2]. In all experiments, pCMVβ-galactosidase plasmid was cotransfected, in order to standardize for transfection efficiency. The cells were transfected using Lipofectamine 2000 (Invitrogen) in accordance with the manufacturer's recommendations. In some experiments, the cells were cotransfected with expression vectors for human vitamin D receptor (VDR), human retinoid X receptor (RXR), human VDIR and/or human WSTF. The cells were cultured and transfected in 6-well plats. The amount of DNA was (per well) 0.4 μg for the reporter plasmid, 0.1 μg for the pCMVβ-galactosidase plasmid and 0.2 μg respectively for the expression vectors for nuclear receptors and comodulators. Following transfection, the cells were treated with 10 nM of 1α,25-dihydroxyvitamin D3 dissolved in ethanol for 24 h. Control samples were treated with the same amount of ethanol. In some experiments, the cells were also treated with the VDR antagonist TEI-9647 (dissolved in ethanol) in concentrations from 10 nM to 1 μM. After the treatment, cells were lysed and luciferase and βgalactosidase activities were measured as previously described [5,6]. Luciferase reporter activity is expressed as relative light units (RLU) divided by β-galactosidase activity. Reporter activity is presented as fold change compared to the control group. 2.4. In silico-analysis of CYP21A2 promoter The CYP21A2 promoter (9 kb upstream) was in silico-analyzed for putative VDREs. The promoter region was manually searched for consensus sequences for VDRE described by Matilainen et al. [5] and for previously described negative VDREs [6] using Microsoft Word. 2.5. Chromatin Immunoprecipitation assay (ChIP) Chromatin immunoprecipitation assay was performed using EZMagna ChIP kit (Millipore) in accordance with the manufacturer's
recommendations. NCI-H295R cells were cultured as described in Section 2.2 and treated with 10 nM of 1α,25-dihydroxyvitamin D3 dissolved in ethanol for 24 h. Control cells were treated with the same amount of ethanol. The cells were then cross-linked by adding formaldehyde and the chromatin was subjected to immunoprecipitation using antibodies against VDR, WSTF or VDIR (Santa Cruz Biotechnology). Antibodies against RNA Polymerase II and Normal Mouse IgG, included in the EZ-Magna ChIP kit, were used as positive and negative control. Positive control, negative control and input control were analyzed with quantitative PCR in accordance with the manufacturer's recommendations. Immunoprecipitated DNA was used for PCR amplification using primers specific for different parts of the CYP21A2 promoter where the in silico analysis had shown putative VDREs. One primer pair was designed to amplify a 230 bp fragment containing Motif 1 (see Section 3.5) and one primer pair was designed to amplify a 148 bp fragment containing Motif 2. The primer sequences for Motif 1 were (F) CCACTCCCTGGTGAACC and (R) GCCTTCTAACG ACCCTGT. The primer sequences for Motif 2 were (F) CAAGGTGGTG GAGGAGCA and (R) CTGTCGTCCAGGTTCTGC. The primers were designed using the software CLC DNA Workbench. The immunoprecipitated DNA was analyzed using both PCR and real-time PCR. The PCR experiments were performed using AmpliTaq Gold system (Applied Biosystems) in accordance with the manufacturer's recommendations. The real-time PCR experiments were conducted using iQ SYBR Green Supermix (Bio-Rad) and analyzed using an iQ5 Real-Time PCR Detection System (Bio-Rad) in accordance to the manufacturer's recommendations. Each PCR reaction resulted in a single DNA product of the expected length, showing that the primers were gene specific. The specificity of the primers was confirmed using melt curve analysis in the real-time PCR experiments. 2.6. Mutagenesis QuickChange II XL Site-Directed Mutagenesis Kit (Agilent Technologies, CA) was used to produce a deletion mutant where the putative VDRE (Motif 2) in the CYP21A2-promoter luciferase constructs had been deleted. The mutagenesis kit was used in accordance with the manufacturer's recommendations and the primers used had the following sequences: sense 5′-CTACACCGTGTGGTGGGCGCCGGG; antisense 5′-CCCGGCGCCCACCACACGGTGTAG. The primers were purchased from Thermo Fisher Scientific GmbH, Germany. 2.7. VDIR-silencing using siRNA Transfection of siRNA was used to silence the expression of VDIR in NCI-H295R cells. Transfection was carried out with DharmaFECT transfection reagent 1 (Thermo Scientific) in accordance with the manufacturer's recommendations. As siRNA, the siGENOME SMARTpool M-009384-00-0005 (Thermo Scientific) was used. Control cells were transfected with scrambled siRNA (sc-37007, Santa Cruz Biotechnology). NCI-H295R cells were cultured in 6 well plates and transfected with 5 μM siRNA for 24 h. Following transfection, the cells were treated with 10 nM 1α,25-dihydroxyvitamin D3 for 24 h. The cells were then used in a real time PCR experiment (see below). The effectiveness of siRNA silencing was controlled by measuring VDIR mRNA levels using real time RT-PCR (see below). 2.8. Polymerase chain reaction (PCR) PCR was used to measure the expression of comodulators and to measure the level of CYP21A2 mRNA following gene knock down. All RNA preparation and reverse transcription were conducted as previously described [3]. The expression of the key comodulators VDIR and WSTF and the nuclear receptor VDR was analyzed using RT-PCR. The primers for VDIR
J. Lundqvist et al. / Biochimica et Biophysica Acta 1820 (2012) 1553–1559
were 5′-GGACGAGGACGACCTTCTC-3′ (forward) and 5′-AAGGCCTCGTT GATGTCACG-3′ (reverse) as described by Riemenschneider et al. [7]. The primers for WSTF were 5′-GAGCCGCTCTTCACCATCCCG-3′ (forward) and 5′-CTTCACCTTGAGCATTTTCTCCTT-3′ (reverse) as described by Poot et al. [8]. The primers for VDR were 5′-AGCAGCGCATCA TTGCCATA-3′ (forward) and 5′-CAGCATGGAGAGCTGGGACA-3′ (reverse) as described by Pascussi et al. [9]. The mRNA levels for CYP21A2 were measured in cells where the expression of VDIR had been silenced with siRNA (see above) and compared with cells with unaltered VDIR expression. To evaluate the effectiveness of siRNA silencing, the mRNA level of VDIR was measured using real time RT-PCR with primers described by Riemenschneider et al. [7]. Real time RT-PCR experiments for CYP21A2 was conducted as previously described [3]. 2.9. Statistical analysis Analysis of statistical significance was performed using Student's t-test in Microsoft Excel. p-Values b 0.05 were considered statistically significant. 3. Results 3.1. 1α,25-Dihydroxyvitamin D3 suppresses CYP21A2 promoter activity We have previously reported that 1α,25-dihydroxyvitamin D3 suppresses mRNA expression and enzyme activity of CYP21A2 in adrenocortical NCI-H295R cells [3]. The aim of this work was to study the mechanism for this 1α,25-dihydroxyvitamin D3-mediated suppression of CYP21A2. To elucidate if 1α,25-dihydroxyvitamin D3 affects the promoter activity of CYP21A2, we used a luciferase reporter construct where the CYP21A2 promoter (9 kb fragment) had been coupled to the luciferase gene. The effects of 1α,25-dihydroxyvitamin D3 on this long promoter fragment were compared with a luciferase construct where only a short fragment (0.3 kb) of the CYP21A2 promoter had been inserted. The luciferase reporter constructs were transfected into human adrenocortical NCI-H295R cells and mouse adrenocortical Y-1 cells which were then treated with 1α,25-dihydroxyvitamin D3 or vehicle. When cells transfected with the long CYP21A2 promoter fragment were treated with 1α,25-dihydroxyvitamin D3, the promoter activity was decreased in both cell lines. In cells transfected with the short promoter fragment, 1α,25-dihydroxyvitamin D3 treatment did not significantly alter the promoter activity (Fig. 1A,B). These results show that our previous findings [3] on 1α,25dihydroxyvitamin D3-mediated downregulation of CYP21A2 mRNA levels and enzyme activity could be a result of transcriptional suppression of CYP21A2 by 1α,25-dihydroxyvitamin D3. 3.2. Overexpression of VDR and RXR reverses 1α,25-dihydroxyvitamin D3-mediated effects on CYP21A2 promoter activity In order to study the mechanism for the observed 1α,25dihydroxyvitamin D3-mediated effect on CYP21A2 promoter activity, we overexpressed nuclear receptors and comodulators previously reported to be involved in vitamin D-mediated effects on other genes. The vitamin D receptor is crucial for effects of 1α,25dihydroxyvitamin D3 on gene expression. Expression of human VDR has been reported in both cell lines of human origin used in this study [12,13]. We cotransfected cells with CYP21A2 promoter luciferase construct and expression vectors containing cDNA for VDR and RXR, in order to study the effect of altered expression level of nuclear receptors. Surprisingly, when cotransfected cells were treated with 1α,25dihydroxyvitamin D3 the activity of the long promoter fragment was stimulated compared to the control group in both cell lines (Fig. 1C,D). Thus, overexpression of VDR and RXR reversed the
1555
previously observed effect of vitamin D on CYP21A2 promoter activity from suppression to stimulation. Consequently, the expression level of the nuclear receptors VDR and RXR is crucial for the effects of 1α,25-dihydroxyvitamin D3 on CYP21A2 promoter activity. This effect has, so far, only been observed in this in vitro system and it is yet unclear if it has in vivo relevance. 3.3. The effect of 1α,25-dihydroxyvitamin D3 on CYP21A2 promoter activity is mediated by VDR To gain further support for our finding that the effect of 1α,25dihydroxyvitamin D3 on promoter activity of CYP21A2 is mediated by VDR, we used the substance TEI-9647, known to act as an antagonist to human VDR [10]. NCI-H295R cells, transiently transfected with the long CYP21A2 promoter luciferase construct and expression vectors for VDR and RXR, were treated with 1α,25-dihydroxyvitamin D3 and increasing concentrations of the VDR antagonist TEI-9647 (Fig. 1E). As shown above, treatment with 1α,25-dihydroxyvitamin D3 alone increased CYP21A2 promoter activity. Cotreatment with a VDR antagonist decreased the effect of 1α,25-dihydroxyvitamin D3 in a dosedependent manner. This lends further support for a VDR-mediated effect of 1α,25-dihydroxyvitamin D3 on CYP21A2 transcription. 3.4. Comodulators VDIR and WSTF are essential for 1α,25-dihydroxyvitamin D3-mediated suppression of CYP21A2 Although the mechanism for vitamin D-mediated suppression of gene expression remains unclear, it has been shown for the CYP27B1 gene and the parathyroid hormone gene that the VDR interacting repressor (VDIR) and the Williams syndrome transcription factor (WSTF) act as key comodulators [8,9,15]. Using RT-PCR, we examined the expression of VDIR and WSTF in NCI-H295R cells and found that both VDIR and WSTF are expressed in this cell line (data not shown). To examine if these comodulators are involved in the regulation of CYP21A2 transcription, NCI-H295R cells transfected with the long CYP21A2 promoter luciferase construct and expression vectors containing cDNA for VDR and RXR were cotransfected with expression vectors for WSTF and/or VDIR. Interestingly, when VDIR was overexpressed, the 1α,25-dihydroxyvitamin D3-mediated stimulation of promoter activity was abolished (Fig. 2C) indicating a role for VDIR in suppression of CYP21A2 transcription. When both VDIR and WSTF were overexpressed, 1α,25-dihydroxyvitamin D3 suppressed the promoter activity by about 50% (Fig. 2C), which is in the same order of magnitude as the suppression of CYP21A2 mRNA level and enzyme activity that we have reported previously [3]. These results show that both VDIR and WSTF are required for 1α,25-dihydroxyvitamin D3-mediated suppression of CYP21A2 promoter activity. To obtain further evidence regarding the role of VDIR, the expression of VDIR in NCI-H295R cells was silenced using siRNA transfection and the CYP21A2 mRNA level was measured using real time RT-PCR. To control the effectiveness of siRNA silencing of VDIR, we measured the mRNA level of VDIR in cells transfected with siRNA and compared it with a control sample. siRNA silencing decreased the VDIR mRNA level by 70% (Fig. 2B). Similarly as in our previous publication [3], treatment with 1α,25-dihydroxyvitamin D3 decreased mRNA level of CYP21A2 in NCI-H295R cells with unaltered VDIR expression (Fig. 2A). However, when the VDIR expression was silenced, treatment with 1α,25-dihydroxyvitamin D3 did not alter the expression of CYP21A2 (Fig. 2A), showing that VDIR is a key comodulator for the 1α,25dihydroxyvitamin D3-mediated suppression of CYP21A2 mRNA. Furthermore, we measured the expression of VDR in NCI-H295R cells and human liver RNA, known to express VDR, was used as a positive control. We found that the expression of VDR in NCI-H295R cells was in the same order of magnitude as in human liver RNA. Our results show that both VDIR and WSTF play important roles in the 1α,25-dihydroxyvitamin D3-mediated suppression of CYP21A2
1556
J. Lundqvist et al. / Biochimica et Biophysica Acta 1820 (2012) 1553–1559
A
Y-1
-VDR/RXR
900
B
800
120
*
*
100
Control 1,25D3
500
Control 1,25D3
80
400 60
300 200
40
100
20
0 Long promoter fragment Short promoter fragment
C 800 700
NCI-H295R
140
700 600
-VDR/RXR
160
Y-1
+VDR/RXR *
0 Long promoter fragment Short promoter fragment
D
+VDR/RXR
4500 4000
600
3500
500
3000
*
2500
400
Control 1,25D3
300
NCI-H295R
Control 1,25D3
2000 1500
200
1000
100
500
0 Long promoter fragment Short promoter fragment
0 Long promoter fragment Short promoter fragment
E 1600
*
1400
*
1200
*
1000
°
800 600 400
1,25D3 + 1000 nM antagonist
1,25D3 + 100 nM antagonist
1,25D3 + 10 nM antagonist
1,25D3
0
Control
200
Fig. 1. Effects of 1α,25-dihydroxyvitamin D3 on CYP21A2 promoter activity. Experiments were carried out by transient transfection of Y-1 cells (A) and NCI-H295R cells (B) with CYP21A2 promoter luciferase reporter constructs (long promoter fragment, 9 kb or short promoter fragment, 0.3 kb) together with pCMVβ-galactosidase plasmid. In other experiments, Y-1 cells (C) and NCI-H295R cells (D) were transfected with CYP21A2 promoter luciferase reporter constructs (long promoter fragment, 9 kb or short promoter fragment, 0.3 kb), expression vectors for human VDR and human RXR together with pCMVβ-galactosidase plasmid. The cells were treated with 10 nM of 1α,25-dihydroxyvitamin D3. In a separate experiment, NCI-H295R cells were transiently transfected with CYP21A2 promoter luciferase reporter construct (long promoter fragment, 9 kb), expression vectors for human VDR and human RXR together with pCMVβ-galactosidase plasmid. The cells were treated with 10 nM of 1α,25-dihydroxyvitamin D3 and with increasing concentrations of the VDR antagonist TEI9647 (10 nM to 1 μM). Following treatment, all cells were analyzed for luciferase activity after 24 h of incubation as described in the Material and methods section. Ethanol was used as control. Data are presented as reporter activity (RLU/β-gal)± standard deviation (n= 3–7). *Statistically significant difference compared to the control (pb 0.05). °Statistically significant difference compared to 1,25D3 alone (pb 0.05).
and that altered expression levels of these comodulators may shift a suppressing effect of 1α,25-dihydroxyvitamin D3 to a stimulatory effect. 3.5. Identification of a functional VDRE in the CYP21A2 promoter To identify regions of the CYP21A2 promoter containing putative vitamin D response elements (VDREs), experiments were conducted with progressive deletion constructs of the CYP21A2 promoter. When NCI-H295R cells transfected with these deletion constructs were
treated with 1α,25-dihydroxyvitamin D3, the long promoter construct (9 kb) was markedly upregulated by the treatment (Fig. 3). The other deletion constructs were not affected to the same extent. The results indicate that the promoter sequence most important for 1α,25dihydroxyvitamin D3-mediated effects on CYP21A2 transcription is located in the region between −5.6 kb and −9 kb. In order to obtain more information on the location of the VDRE, we performed in silico analysis of the sequence for the CYP21A2 promoter in the region −5.6 kb to −9 kb upstream of the transcription
J. Lundqvist et al. / Biochimica et Biophysica Acta 1820 (2012) 1553–1559
1557
4000
mRNA level CYP21A2 - fold change vs control
1,2 1 0,8
Control
0,6
1,25D3
*
0,4 0,2 0
mRNA level VDIR - fold change vs control
1 Control VDIR silenced
0,8 0,6
*
0,4 0,2 0
Reporter activity (RLU/β-gal)
80 70 60 Control
50
1,25D3
*
40 30
* *
20 10 0
3000 2500
WSTF
Control
2000
1,25D3
1500 1000 500 9 kb
5,6 kb
5,0 kb
4,6 kb
3,8 kb
3,4 kb
0,3 kb
Fig. 3. Effects of 1α,25-dihydroxyvitamin D3 on the activity of CYP21A2 promoter fragments of different lengths. Experiments were carried out by transient transfection of NCI-H295R cells with CYP21A2 promoter luciferase reporter constructs of different lengths (−9 kb, −5.6 kb, −5.0 kb, −4.6 kb, −3.8 kb, −3.4 kb, −0.3 kb), expression vectors for human VDR and human RXR together with pCMVβ-galactosidase plasmid. The cells were treated with 10 nM of 1α,25-dihydroxyvitamin D3 and analyzed for luciferase activity after 24 h of incubation as described in the Material and methods section. Ethanol was used as control. Data are presented as reporter activity (RLU/β-gal)± standard deviation (n= 4). *Statistically significant difference compared to the control (pb 0.05).
1,2
C
3500
0
VDIR silenced
Unaltered VDIR expression
B
NCI-H295R
NCI-H295R Reporter activity (RLU/β-gal)
A
VDIR
WSTF+VDIR
Fig. 2. Role for comodulators in the 1α,25-dihydroxyvitamin D3-mediated alterations of CYP21A2 gene expression. (A) Effects of 1α,25-dihydroxyvitamin D3 treatment on CYP21A2 mRNA level after VDIR silencing. VDIR expression in NCI-H295R cells was silenced with siRNA as described in the Material and methods section. In the control sample (transfected with scrambled siRNA), 1α,25-dihydroxyvitamin D3 suppressed the CYP21A2 mRNA level. When VDIR was silenced, the suppressing effect of 1α,25dihydroxyvitamin D3 was abolished. (B) The effect of siRNA transfection on VDIR expression in NCI-H295R cells. mRNA levels were measured using real time RT-PCR as described in the Material and methods section. Data are presented as fold change compared to the control group ± standard deviation (n= 3–4). *Statistically significant difference compared to the control (p b 0.05). (C) The role of key comodulators for the 1α,25dihydroxyvitamin D3-mediated alterations in CYP21A2 promoter activity. Experiments were carried out by transient transfection of NCI-H295R cells with CYP21A2 promoter luciferase reporter construct (long promoter fragment, 9 kb), expression vectors for human VDR and human RXR together with pCMVβ-galactosidase plasmid. In some of the experiments, cells were cotransfected with expression vectors for the comodulators VDIR and/or WSTF. The cells were treated with 10 nM of 1α,25-dihydroxyvitamin D3 and were analyzed for luciferase activity after 24 h of incubation as described in the Material and methods section. Ethanol was used as control. Data are presented as fold change compared to the control group ± standard deviation (n= 3–9). *Statistically significant difference compared to the control (pb 0.05).
start site. The promoter region was analyzed for consensus sequences for VDRE described by Matilainen et al. [5] and for previously described negative VDREs [6]. We found two sequences that are identical to previously described negative VDREs (Fig. 4E) in the promoters for human CYP27B1 and human PTHrP. The sequences are identical in all three promoters but the distance between the two sequences differs between the three promoters.
In order to study if the vitamin D receptor and/or the comodulators VDIR and WSTF interact with either of these putative VDREs, we performed ChIP assay with NCI-H295R cells treated with 1α,25dihydroxyvitamin D3 or ethanol. The chromatin was cross-linked and immunoprecipitated using antibodies against VDR, WSTF and VDIR. The immunoprecipitated DNA was then used for PCR amplification with primers specific for the two putative VDREs found in the in silicoanalysis. The ChIP assay showed that VDR, WSTF and VDIR interact with Motif 2 in the ethanol treated cells (Fig. 4A–D). Treatment with 1α,25-dihydroxyvitamin D3 dissociated WSTF from Motif 2. This indicates that 1α,25-dihydroxyvitamin D3 is able to alter the composition of a nuclear receptor–comodulator complex, binding to the CYP21A2 promoter at a 148 bp region around Motif 2. The results from the ChIP assay indicate only a weak binding of VDIR to Motif 1. In order to confirm this finding of a VDRE in the CYP21A2 promoter, the long CYP21A2 promoter luciferase reporter construct was mutated. A mutated construct was produced, where the putative VDRE located at −6377 to −6382 was deleted (Motif 2). NCI-H295R cells were then transiently transfected with the mutated constructs or the construct with the wild type CYP21A2 promoter. All cells were cotransfected with expression vectors for VDR and RXR. As expected, the wild type CYP21A2 promoter was stimulated when treated with 1α,25-dihydroxyvitamin D3 (Fig. 4F). However, the mutated construct where Motif 2 had been deleted, was not stimulated by 1α,25-dihydroxyvitamin D3 treatment. We conclude from this, that the sequence CATCTG located at −6377 to −6382 in the CYP21A2 promoter is important for the effects of 1α,25-dihydroxyvitamin D3 on the promoter activity and that it functions as a VDRE. 4. Discussion The current data reveal a novel mechanism for regulation of the human CYP21A2 gene and add important knowledge regarding the mechanisms for vitamin D-mediated suppression of gene expression. Extensive research has been conducted to understand the mechanism for 1α,25-dihydroxyvitamin D3-mediated induction of gene expression [11]. The mechanism for vitamin D-mediated downregulation of gene expression, on the other hand, has in large part remained unclear [4]. Studies on the mechanism for 1α,25-dihydroxyvitamin D3-mediated downregulation of gene expression have been performed for only a few genes. The proposed mechanism for those genes includes recruitment of corepressors such as VDR interacting repressor (VDIR) and Williams syndrome transcription factor (WSTF) as well as epigenetic mechanisms such as histone deacetylation and DNA methylation [6,12–15]. The
1558
J. Lundqvist et al. / Biochimica et Biophysica Acta 1820 (2012) 1553–1559
B Amount DNA (fold change vs input control)
A
Controls
45 40 35 30 25 20 15 10 5 0 Input
Fold change
Positive control
1.0
Negative control
21.7
0.76
Input 1.0
Control
Amount DNA (fold change vs input control)
D
Motif 1
70 60 50 40
EtOH 1,25D3
30 20 10 0
* VDR
WSTF
Amount DNA (fold change vs input control)
C
F Reporter activity (RLU/beta/gal) (fold change vs input control)
E
39.7
0.02
1,25D3
Motif 2
3.5 3.0 2.5 2.0 1.5
EtOH
*
*
1,25D3
1.0
*
0.5 0.0
VDIR
Positive Negative control control
18
VDR
WSTF
VDIR
*
16 14 12 10 EtOH
8
1,25D3
6 4 2 0 Wild type
Motif 2 deleted
Fig. 4. Analysis of interaction of VDR, WSTF and VDIR with the CYP21A2 promoter. NCI-H295R cells were cross-linked with formaldehyde and chromatin immunoprecipitation was performed with antibodies against VDR, WSTF and VDIR. Recruitment of VDR, WSTF and VDIR to the CYP21A2 promoter was analyzed by PCR (A) and quantitative PCR (C and D) with primers for the two putative VDREs found in the in silico analysis. Input control, positive control and negative control (B) were analyzed in accordance with the manufacturer's recommendations. (E) In silico analysis of the CYP21A2 promoter revealed two motifs with sequences identical to previously described nVDREs. The in silico-analysis was performed as described in the Material and methods section. (F) Effects of 1α,25-dihydroxyvitamin D3 on a deletion mutant of the CYP21A2 promoter. Experiments were carried out by transient transfection of NCI-H295R cells with CYP21A2 promoter luciferase reporter constructs, expression vectors for human VDR and human RXR together with pCMVβ-galactosidase plasmid. The CYP21A2 promoter was either wild type or with the putative nVDRE deleted using mutagenesis as described in the Material and methods section. The cells were treated with 10 nM of 1α,25-dihydroxyvitamin D3 and analyzed for luciferase activity after 24 h incubation as described in the Material and methods section. Ethanol was used as control. Data are presented as fold change compared to the control group± standard deviation (n= 4–8). *Statistically significant difference compared to the control (pb 0.05).
negative vitamin D responsive elements (nVDRE) that have been described only vaguely resemble the pVDRE described [6]. It has been suggested that the mechanism for vitamin D-mediated downregulation of gene expression does not include a direct interaction between the ligand-activated VDR–RXR complex and the nVDRE, but rather an interaction between nVDRE and VDIR [12]. We found a sequence in the CYP21A2 promoter that was identical to a previously described negative VDRE in the promoter for the 1αhydroxylase CYP27B1 [6]. When this sequence was deleted using mutagenesis, the effect of 1α,25-dihydroxyvitamin D3 on promoter activity was abrogated. This indicates that the sequence in question may function as a negative vitamin D responsive element (nVDRE) in the CYP21A2 promoter.
Our data indicate that the recently reported suppression of CYP21A2 gene transcription by 1α,25-dihydroxyvitamin D3 is mediated by a complex formed by the nuclear receptor VDR and the comodulators VDIR and WSTF. Our results suggest that the expression levels of nuclear receptor and key comodulators are crucial for the effect of vitamin D on the CYP21A2 promoter. Thus, the effect of vitamin D on CYP21A2 may be dependent on the expression levels of VDIR and WSTF in cells. This conclusion is supported by the fact that when VDIR was silenced using siRNA, the vitamin D-mediated effect on CYP21A2 mRNA level was abolished. This work has been conducted in cell culture-based in vitro assays such as reporter assay, chromatin immunoprecipitation, gene knock down et cetera. Even though these methods provide novel and
J. Lundqvist et al. / Biochimica et Biophysica Acta 1820 (2012) 1553–1559
1559
generous gift of the VDR antagonist TEI-9647. The authors are grateful to Dr Nathalie Luna Persson, Karolinska Institute, for technical advice on primer design. References
Fig. 5. Hypothesis for the vitamin D-induced effects on the CYP21A2 promoter. Based on our ChIP and EMSA assays, the following hypothesis for the vitamin D-induced effects on CYP21A2 transcription can be proposed. Our results indicate that VDR, WSTF and VDIR interact with a VDRE in the CYP21A2 promoter in the absence of 1α,25dihydroxyvitamin D3 (A). In the presence of 1α,25-dihydroxyvitamin D3, WSTF dissociates and the transcriptional activity of CYP21A2 is decreased (B).
important insights in the regulation of the transcription of CYP21A2 it cannot be excluded that also other mechanisms may be involved in the complex in vivo regulation of CYP21A2. VDIR and WSTF have previously been described as key components in vitamin D-mediated suppression of the CYP27B1 gene and the parathyroid hormone gene [8,9]. However, the mechanism that can be proposed based on our experiments (Fig. 5) has not been reported previously. We conclude that 1α,25-dihydroxyvitamin D3 affects CYP21A2 transcription via a novel mechanism for vitamin D action, where VDR, VDIR and WSTF interact with a VDRE in the promoter in the absence of 1α,25-dihydroxyvitamin D3. When cells are treated with 1α,25dihydroxyvitamin D3, WSTF dissociates from the complex and the promoter activity of CYP21A2 is decreased. Further research is needed to investigate if this complex is composed also of other comodulators, and which of the receptors/comodulators that interacts with the promoter sequence. Accumulated data during recent years have indicated that vitamin D may have many more targets than previously known [11,16,17]. It may be speculated that the expression levels of comodulators may be of importance also for other vitamin D-related genes. If so, such comodulators may be future target to manipulate vitamin D-dependent functions. In conclusion, this paper reports data important for the understanding of the mechanisms for vitamin D-mediated suppression of gene expression as well as for the vitamin D-mediated effects on CYP21A2. Acknowledgements This work was supported by the Swedish Research Council — Medicine. Teijin Pharma, Tokyo, Japan is acknowledged for the
[1] S.N. Kelly, T.J. McKenna, L.S. Young, Modulation of steroidogenic enzymes by orphan nuclear transcriptional regulation may control diverse production of cortisol and androgens in the human adrenal, J. Endocrinol. 181 (2004) 355–365. [2] S.D. Wijesuriya, G. Zhang, A. Dardis, W.L. Miller, Transcriptional regulatory elements of the human gene for cytochrome P450c21 (steroid 21-hydroxylase) lie within intron 35 of the linked C4B gene, J. Biol. Chem. 274 (1999) 38097–38106. [3] J. Lundqvist, M. Norlin, K. Wikvall, 1α,25-Dihydroxyvitamin D3 affects hormone production and expression of steroidogenic enzymes in human adrenocortical NCI-H295R cells, Biochim. Biophys. Acta (BBA) — Mol. Cell Biol. Lipids 1801 (2010) 1056–1062. [4] J.W. Pike, M.B. Meyer, The vitamin D receptor: new paradigms for the regulation of gene expression by 1,25-dihydroxyvitamin D3, Endocrinol. Metab. Clin. North Am. 39 (2010) 255–269. [5] M. Matilainen, M. Malinen, K. Saavalainen, C. Carlberg, Regulation of multiple insulin-like growth factor binding protein genes by 1α,25-dihydroxyvitamin D3, Nucl. Acids Res. 33 (2005) 5521–5532. [6] M. Kim, R. Fujiki, A. Murayama, H. Kitagawa, K. Yamaoka, Y. Yamamoto, M. Mihara, K.-I. Takeyama, S. Kato, 1α,25(OH)2D3-Induced transrepression by vitamin D receptor through E-box-type elements in the human parathyroid hormone gene promoter, Mol. Endocrinol. 21 (2007) 334–342. [7] M. Riemenschneider, T. Koy, G. Reifenberger, Expression of oligodendrocyte lineage genes in oligodendroglial and astrocytic gliomas, Acta Neuropathol. 107 (2004) 277–282. [8] R.A. Poot, L. Bozhenok, D.L.C. van den Berg, S. Steffensen, F. Ferreira, M. Grimaldi, N. Gilbert, J. Ferreira, P.D. Varga-Weisz, The Williams syndrome transcription factor interacts with PCNA to target chromatin remodelling by ISWI to replication foci, Nat. Cell Biol. 6 (2004) 1236–1244. [9] J. Pascussi, A. Robert, M. Nguyen, O. Walrant-Debray, M. Garabedian, P. Martin, T. Pineau, J. Saric, F. Navarro, P. Maurel, M. Vilarem, Possible involvement of pregnane X receptor-enhanced CYP24 expression in drug-induced osteomalacia, J. Clin. Invest. 115 (2005) 177–186. [10] S. Ishizuka, N. Kurihara, Y. Hiruma, D. Miura, J.-I. Namekawa, A. Tamura, Y. Kato-Nakamura, Y. Nakano, K. Takenouchi, Y. Hashimoto, K. Nagasawa, G.D. Roodman, 1α,25-Dihydroxyvitamin D3-26,23-lactam analogues function as vitamin D receptor antagonists in human and rodent cells, J. Steroid Biochem. Mol. Biol. 110 (2008) 269–277. [11] A.W. Norman, From vitamin D to hormone D: fundamentals of the vitamin D endocrine system essential for good health, Am. J. Clin. Nutr. 88 (2008) 491S–499S. [12] R. Fujiki, M.-S. Kim, Y. Sasaki, K. Yoshimura, H. Kitagawa, S. Kato, Ligand-induced transrepression by VDR through association of WSTF with acetylated histones, EMBO J. 24 (2005) 3881–3894. [13] A. Murayama, M.-S. Kim, J. Yanagisawa, K.-I. Takeyama, S. Kato, Transrepression by a liganded nuclear receptor via a bHLH activator through co-regulator switching, EMBO J. 23 (2004) 1598. [14] M. Kim, T. Kondo, I. Takada, M.-Y. Youn, Y. Yamamoto, S. Takahashi, T. Matsumoto, S. Fujiyama, Y. Shirode, I. Yamaoka, H. Kitagawa, K.-I. Takeyama, H. Shibuya, F. Ohtake, S. Kato, DNA demethylation in hormone-induced transcriptional derepression, Nature 461 (2009) 1007–1012. [15] M.-S. Kim, R. Fujiki, H. Kitagawa, S. Kato, 1α,25(OH)2D3-induced DNA methylation suppresses the human CYP27B1 gene, Mol. Cell. Endocrinol. 265–266 (2007) 168–173. [16] H.F. DeLuca, Overview of general physiologic features and functions of vitamin D, Am. J. Clin. Nutr. 80 (2004) 1689S–1696S. [17] S. Ramagopalan, A. Heger, A. Berlanga, N. Maugeri, M. Lincoln, A. Burrell, L. Handunnetthi, A. Handel, G. Disanto, S. Orton, C. Watson, J. Morahan, G. Giovannoni, C. Ponting, G. Ebers, J. Knight, A ChIP-seq defined genome-wide map of vitamin D receptor binding: associations with disease and evolution, Genome Res. 20 (2010) 1352–1360.
Contents lists available at SciVerse ScienceDirect
Biochimica et Biophysica Acta journal homepage: www.elsevier.com/locate/bbagen
Vitamin D-mediated regulation of CYP21A2 transcription — A novel mechanism for vitamin D action Johan Lundqvist ⁎, Kjell Wikvall, Maria Norlin Department of Pharmaceutical Biosciences, Box 591, Uppsala University, SE-751 24 Uppsala, Sweden
a r t i c l e
i n f o
Article history: Received 30 November 2011 Received in revised form 16 March 2012 Accepted 20 April 2012 Available online 27 April 2012 Keywords: Steroidogenesis Vitamin D Calcitriol CYP21A2 Steroid
a b s t r a c t Background: 1α,25-Dihydroxyvitamin D3 has recently been reported to decrease expression and activity of CYP21A2. In this paper, we have studied the mechanisms for the 1α,25-dihydroxyvitamin D3-mediated effect on CYP21A2 transcriptional rate. Methods: We have studied the effects of 1α,25-dihydroxyvitamin D3 using luciferase reporter constructs containing different lengths of the CYP21A2 promoter. These constructs were transfected into cell lines derived from human and mouse adrenal cortex. The mechanism for the effects of vitamin D on the CYP21A2 promoter was studied using chromatin immunoprecipitation assay, mutagenesis and gene silencing by siRNA. Results: 1α,25-Dihydroxyvitamin D3 was found to alter the promoter activity via a VDR-mediated mechanism, including the comodulators VDR interacting repressor (VDIR) and Williams syndrome transcription factor (WSTF). The involvement of comodulator VDIR was confirmed by gene silencing. We identified a vitamin D response element in the CYP21A2 promoter. Interaction between this novel response element and VDR, WSTF and VDIR was shown by chromatin immunoprecipitation assay. When this sequence was deleted, the effect of 1α,25-dihydroxyvitamin D3 was abolished, indicating that this sequence in the CYP21A2 promoter functions as a vitamin D response element. Interestingly, an altered balance between nuclear receptors and comodulators reversed the suppressing effect of vitamin D to a stimulatory effect. General significance: This paper reports data important for the understanding of the mechanisms for vitamin D-mediated suppression of gene expression as well as for the vitamin D-mediated effects on CYP21A2. We report a novel mechanism for effects of 1α,25-dihydroxyvitamin D3. © 2012 Elsevier B.V. All rights reserved.
1. Introduction The steroid 21-hydroxylase CYP21A2 is expressed in the human adrenal cortex and catalyzes the production of deoxycorticosterone and 11-deoxycortisol, which are precursors for the production of corticosterone, aldosterone and cortisol. 21-Hydroxylase deficiency may lead to severe conditions such as congenital adrenal hyperplasia and Addison's disease. CYP21A2 has been reported to be transcriptionally regulated by transcription factors as steroidogenic factor-1 (SF-1) and nur77 [1,2]. Very recently, we reported that 1α,25-dihydroxyvitamin D3 decreases mRNA level and enzyme activity of CYP21A2 as well as the production of corticosterone and aldosterone in human adrenocortical NCI-H295R cells [3]. This is a previously unknown effect of the multifunctional hormone 1α,25-dihydroxyvitamin D3.
Abbreviations: RXR, retinoid X receptor; SF-1, steroidogenic factor-1; VDIR, VDR interacting repressor; VDR, vitamin D receptor; VDRE, vitamin D responsive element; nVDRE, negative vitamin D responsive element; WSTF, Williams syndrome transcription factor ⁎ Corresponding author. Present address: Breast Cancer Group, Unit of Cell Death and Metabolism, Danish Cancer Society Research Center, Strandboulevarden 49, 2100 Copenhagen Ø, Denmark. Tel.: + 46 18 4714144. E-mail address: [email protected] (J. Lundqvist). 0304-4165/$ – see front matter © 2012 Elsevier B.V. All rights reserved. doi:10.1016/j.bbagen.2012.04.017
1α,25-Dihydroxyvitamin D3 may increase the expression of certain genes while it decreases the expression of other genes. Both these effects of 1α,25-dihydroxyvitamin D3 are mediated by activation of the vitamin D receptor (VDR). Extensive research has been conducted to understand the mechanism for 1α,25-dihydroxyvitamin D3-mediated induction of gene expression (e.g. for CYP24A1). The mechanism is now well known and based on a direct interaction between the ligandactivated VDR–RXR complex and a vitamin D responsive element (VDRE) in the gene promoter [4]. However, little is known about the mechanism for 1α,25-dihydroxyvitamin D3-mediated suppression of gene expression [4]. In the present study, we have examined the mechanism for the recently reported 1α,25-dihydroxyvitamin D3-mediated downregulation of CYP21A2 activity and mRNA level [3]. The current study reports a novel mechanism for vitamin D-mediated suppression of gene transcription.
2. Material and methods 2.1. Chemicals and materials 1α,25-Dihydroxyvitamin D3 was obtained from Solvay, Duphar, The Netherlands. The vitamin D receptor antagonist TEI-9647 was a
1554
J. Lundqvist et al. / Biochimica et Biophysica Acta 1820 (2012) 1553–1559
generous gift from Teijin Pharma, Tokyo, Japan. CYP21A2 promoterluciferase reporter constructs were a kind gift from Professor Walter L. Miller, University of California, San Francisco. Expression vectors for VDIR and WSTF were kind gifts from Professor Shigeaki Kato, University of Tokyo. The full-length human vitamin D receptor (VDR) expression vector was a gift from Dr. Leonard Freedman, the Merck Research Laboratories, West Point. The expression vector for the human retinoid X receptor (RXR) was a kind gift from Dr. Ronald M. Evans, Howard Hughes Medical Institute, The Salk Institute for Biological Studies, San Diego. All other chemicals were of analytical grade and purchased from various commercial sources. 2.2. Cell culture and treatment Human adrenocortical carcinoma NCI-H295R cells (ATCC CRL-2128) were cultured in Dulbecco's modified Eagle's medium/Nutrient Mixture F-12 Ham (Sigma) supplemented with 1% ITS Plus premix (BD Biosciences), 2.5% NuSerum (VWR), 1% L-glutamine (Gibco) and 1% antibiotic/ antimycotic (Gibco). Mouse adrenocortical Y-1 (ATCC CCL-79) cells were cultured in F-12K Medium (ATCC) supplemented with 2.5% fetal bovine serum (Gibco), 15% horse serum (ATCC) and 1% antibiotic/ antimycotic (Gibco). All cells were cultured as monolayers in a humidified environment at 37 °C with 5% CO2. Cells were treated with 10 nM 1α,25dihydroxyvitamin D3 dissolved in ethanol for 24 h. The control group was treated with the same amount of ethanol. 2.3. Analysis of CYP21A2 promoter activity by luciferase reporter assay NCI-H295R cells and Y-1 cells were transiently transfected with luciferase reporter constructs containing different lengths of the CYP21A2 promoter, spanning from −9027 to +13, coupled to the luciferase gene [2]. In all experiments, pCMVβ-galactosidase plasmid was cotransfected, in order to standardize for transfection efficiency. The cells were transfected using Lipofectamine 2000 (Invitrogen) in accordance with the manufacturer's recommendations. In some experiments, the cells were cotransfected with expression vectors for human vitamin D receptor (VDR), human retinoid X receptor (RXR), human VDIR and/or human WSTF. The cells were cultured and transfected in 6-well plats. The amount of DNA was (per well) 0.4 μg for the reporter plasmid, 0.1 μg for the pCMVβ-galactosidase plasmid and 0.2 μg respectively for the expression vectors for nuclear receptors and comodulators. Following transfection, the cells were treated with 10 nM of 1α,25-dihydroxyvitamin D3 dissolved in ethanol for 24 h. Control samples were treated with the same amount of ethanol. In some experiments, the cells were also treated with the VDR antagonist TEI-9647 (dissolved in ethanol) in concentrations from 10 nM to 1 μM. After the treatment, cells were lysed and luciferase and βgalactosidase activities were measured as previously described [5,6]. Luciferase reporter activity is expressed as relative light units (RLU) divided by β-galactosidase activity. Reporter activity is presented as fold change compared to the control group. 2.4. In silico-analysis of CYP21A2 promoter The CYP21A2 promoter (9 kb upstream) was in silico-analyzed for putative VDREs. The promoter region was manually searched for consensus sequences for VDRE described by Matilainen et al. [5] and for previously described negative VDREs [6] using Microsoft Word. 2.5. Chromatin Immunoprecipitation assay (ChIP) Chromatin immunoprecipitation assay was performed using EZMagna ChIP kit (Millipore) in accordance with the manufacturer's
recommendations. NCI-H295R cells were cultured as described in Section 2.2 and treated with 10 nM of 1α,25-dihydroxyvitamin D3 dissolved in ethanol for 24 h. Control cells were treated with the same amount of ethanol. The cells were then cross-linked by adding formaldehyde and the chromatin was subjected to immunoprecipitation using antibodies against VDR, WSTF or VDIR (Santa Cruz Biotechnology). Antibodies against RNA Polymerase II and Normal Mouse IgG, included in the EZ-Magna ChIP kit, were used as positive and negative control. Positive control, negative control and input control were analyzed with quantitative PCR in accordance with the manufacturer's recommendations. Immunoprecipitated DNA was used for PCR amplification using primers specific for different parts of the CYP21A2 promoter where the in silico analysis had shown putative VDREs. One primer pair was designed to amplify a 230 bp fragment containing Motif 1 (see Section 3.5) and one primer pair was designed to amplify a 148 bp fragment containing Motif 2. The primer sequences for Motif 1 were (F) CCACTCCCTGGTGAACC and (R) GCCTTCTAACG ACCCTGT. The primer sequences for Motif 2 were (F) CAAGGTGGTG GAGGAGCA and (R) CTGTCGTCCAGGTTCTGC. The primers were designed using the software CLC DNA Workbench. The immunoprecipitated DNA was analyzed using both PCR and real-time PCR. The PCR experiments were performed using AmpliTaq Gold system (Applied Biosystems) in accordance with the manufacturer's recommendations. The real-time PCR experiments were conducted using iQ SYBR Green Supermix (Bio-Rad) and analyzed using an iQ5 Real-Time PCR Detection System (Bio-Rad) in accordance to the manufacturer's recommendations. Each PCR reaction resulted in a single DNA product of the expected length, showing that the primers were gene specific. The specificity of the primers was confirmed using melt curve analysis in the real-time PCR experiments. 2.6. Mutagenesis QuickChange II XL Site-Directed Mutagenesis Kit (Agilent Technologies, CA) was used to produce a deletion mutant where the putative VDRE (Motif 2) in the CYP21A2-promoter luciferase constructs had been deleted. The mutagenesis kit was used in accordance with the manufacturer's recommendations and the primers used had the following sequences: sense 5′-CTACACCGTGTGGTGGGCGCCGGG; antisense 5′-CCCGGCGCCCACCACACGGTGTAG. The primers were purchased from Thermo Fisher Scientific GmbH, Germany. 2.7. VDIR-silencing using siRNA Transfection of siRNA was used to silence the expression of VDIR in NCI-H295R cells. Transfection was carried out with DharmaFECT transfection reagent 1 (Thermo Scientific) in accordance with the manufacturer's recommendations. As siRNA, the siGENOME SMARTpool M-009384-00-0005 (Thermo Scientific) was used. Control cells were transfected with scrambled siRNA (sc-37007, Santa Cruz Biotechnology). NCI-H295R cells were cultured in 6 well plates and transfected with 5 μM siRNA for 24 h. Following transfection, the cells were treated with 10 nM 1α,25-dihydroxyvitamin D3 for 24 h. The cells were then used in a real time PCR experiment (see below). The effectiveness of siRNA silencing was controlled by measuring VDIR mRNA levels using real time RT-PCR (see below). 2.8. Polymerase chain reaction (PCR) PCR was used to measure the expression of comodulators and to measure the level of CYP21A2 mRNA following gene knock down. All RNA preparation and reverse transcription were conducted as previously described [3]. The expression of the key comodulators VDIR and WSTF and the nuclear receptor VDR was analyzed using RT-PCR. The primers for VDIR
J. Lundqvist et al. / Biochimica et Biophysica Acta 1820 (2012) 1553–1559
were 5′-GGACGAGGACGACCTTCTC-3′ (forward) and 5′-AAGGCCTCGTT GATGTCACG-3′ (reverse) as described by Riemenschneider et al. [7]. The primers for WSTF were 5′-GAGCCGCTCTTCACCATCCCG-3′ (forward) and 5′-CTTCACCTTGAGCATTTTCTCCTT-3′ (reverse) as described by Poot et al. [8]. The primers for VDR were 5′-AGCAGCGCATCA TTGCCATA-3′ (forward) and 5′-CAGCATGGAGAGCTGGGACA-3′ (reverse) as described by Pascussi et al. [9]. The mRNA levels for CYP21A2 were measured in cells where the expression of VDIR had been silenced with siRNA (see above) and compared with cells with unaltered VDIR expression. To evaluate the effectiveness of siRNA silencing, the mRNA level of VDIR was measured using real time RT-PCR with primers described by Riemenschneider et al. [7]. Real time RT-PCR experiments for CYP21A2 was conducted as previously described [3]. 2.9. Statistical analysis Analysis of statistical significance was performed using Student's t-test in Microsoft Excel. p-Values b 0.05 were considered statistically significant. 3. Results 3.1. 1α,25-Dihydroxyvitamin D3 suppresses CYP21A2 promoter activity We have previously reported that 1α,25-dihydroxyvitamin D3 suppresses mRNA expression and enzyme activity of CYP21A2 in adrenocortical NCI-H295R cells [3]. The aim of this work was to study the mechanism for this 1α,25-dihydroxyvitamin D3-mediated suppression of CYP21A2. To elucidate if 1α,25-dihydroxyvitamin D3 affects the promoter activity of CYP21A2, we used a luciferase reporter construct where the CYP21A2 promoter (9 kb fragment) had been coupled to the luciferase gene. The effects of 1α,25-dihydroxyvitamin D3 on this long promoter fragment were compared with a luciferase construct where only a short fragment (0.3 kb) of the CYP21A2 promoter had been inserted. The luciferase reporter constructs were transfected into human adrenocortical NCI-H295R cells and mouse adrenocortical Y-1 cells which were then treated with 1α,25-dihydroxyvitamin D3 or vehicle. When cells transfected with the long CYP21A2 promoter fragment were treated with 1α,25-dihydroxyvitamin D3, the promoter activity was decreased in both cell lines. In cells transfected with the short promoter fragment, 1α,25-dihydroxyvitamin D3 treatment did not significantly alter the promoter activity (Fig. 1A,B). These results show that our previous findings [3] on 1α,25dihydroxyvitamin D3-mediated downregulation of CYP21A2 mRNA levels and enzyme activity could be a result of transcriptional suppression of CYP21A2 by 1α,25-dihydroxyvitamin D3. 3.2. Overexpression of VDR and RXR reverses 1α,25-dihydroxyvitamin D3-mediated effects on CYP21A2 promoter activity In order to study the mechanism for the observed 1α,25dihydroxyvitamin D3-mediated effect on CYP21A2 promoter activity, we overexpressed nuclear receptors and comodulators previously reported to be involved in vitamin D-mediated effects on other genes. The vitamin D receptor is crucial for effects of 1α,25dihydroxyvitamin D3 on gene expression. Expression of human VDR has been reported in both cell lines of human origin used in this study [12,13]. We cotransfected cells with CYP21A2 promoter luciferase construct and expression vectors containing cDNA for VDR and RXR, in order to study the effect of altered expression level of nuclear receptors. Surprisingly, when cotransfected cells were treated with 1α,25dihydroxyvitamin D3 the activity of the long promoter fragment was stimulated compared to the control group in both cell lines (Fig. 1C,D). Thus, overexpression of VDR and RXR reversed the
1555
previously observed effect of vitamin D on CYP21A2 promoter activity from suppression to stimulation. Consequently, the expression level of the nuclear receptors VDR and RXR is crucial for the effects of 1α,25-dihydroxyvitamin D3 on CYP21A2 promoter activity. This effect has, so far, only been observed in this in vitro system and it is yet unclear if it has in vivo relevance. 3.3. The effect of 1α,25-dihydroxyvitamin D3 on CYP21A2 promoter activity is mediated by VDR To gain further support for our finding that the effect of 1α,25dihydroxyvitamin D3 on promoter activity of CYP21A2 is mediated by VDR, we used the substance TEI-9647, known to act as an antagonist to human VDR [10]. NCI-H295R cells, transiently transfected with the long CYP21A2 promoter luciferase construct and expression vectors for VDR and RXR, were treated with 1α,25-dihydroxyvitamin D3 and increasing concentrations of the VDR antagonist TEI-9647 (Fig. 1E). As shown above, treatment with 1α,25-dihydroxyvitamin D3 alone increased CYP21A2 promoter activity. Cotreatment with a VDR antagonist decreased the effect of 1α,25-dihydroxyvitamin D3 in a dosedependent manner. This lends further support for a VDR-mediated effect of 1α,25-dihydroxyvitamin D3 on CYP21A2 transcription. 3.4. Comodulators VDIR and WSTF are essential for 1α,25-dihydroxyvitamin D3-mediated suppression of CYP21A2 Although the mechanism for vitamin D-mediated suppression of gene expression remains unclear, it has been shown for the CYP27B1 gene and the parathyroid hormone gene that the VDR interacting repressor (VDIR) and the Williams syndrome transcription factor (WSTF) act as key comodulators [8,9,15]. Using RT-PCR, we examined the expression of VDIR and WSTF in NCI-H295R cells and found that both VDIR and WSTF are expressed in this cell line (data not shown). To examine if these comodulators are involved in the regulation of CYP21A2 transcription, NCI-H295R cells transfected with the long CYP21A2 promoter luciferase construct and expression vectors containing cDNA for VDR and RXR were cotransfected with expression vectors for WSTF and/or VDIR. Interestingly, when VDIR was overexpressed, the 1α,25-dihydroxyvitamin D3-mediated stimulation of promoter activity was abolished (Fig. 2C) indicating a role for VDIR in suppression of CYP21A2 transcription. When both VDIR and WSTF were overexpressed, 1α,25-dihydroxyvitamin D3 suppressed the promoter activity by about 50% (Fig. 2C), which is in the same order of magnitude as the suppression of CYP21A2 mRNA level and enzyme activity that we have reported previously [3]. These results show that both VDIR and WSTF are required for 1α,25-dihydroxyvitamin D3-mediated suppression of CYP21A2 promoter activity. To obtain further evidence regarding the role of VDIR, the expression of VDIR in NCI-H295R cells was silenced using siRNA transfection and the CYP21A2 mRNA level was measured using real time RT-PCR. To control the effectiveness of siRNA silencing of VDIR, we measured the mRNA level of VDIR in cells transfected with siRNA and compared it with a control sample. siRNA silencing decreased the VDIR mRNA level by 70% (Fig. 2B). Similarly as in our previous publication [3], treatment with 1α,25-dihydroxyvitamin D3 decreased mRNA level of CYP21A2 in NCI-H295R cells with unaltered VDIR expression (Fig. 2A). However, when the VDIR expression was silenced, treatment with 1α,25-dihydroxyvitamin D3 did not alter the expression of CYP21A2 (Fig. 2A), showing that VDIR is a key comodulator for the 1α,25dihydroxyvitamin D3-mediated suppression of CYP21A2 mRNA. Furthermore, we measured the expression of VDR in NCI-H295R cells and human liver RNA, known to express VDR, was used as a positive control. We found that the expression of VDR in NCI-H295R cells was in the same order of magnitude as in human liver RNA. Our results show that both VDIR and WSTF play important roles in the 1α,25-dihydroxyvitamin D3-mediated suppression of CYP21A2
1556
J. Lundqvist et al. / Biochimica et Biophysica Acta 1820 (2012) 1553–1559
A
Y-1
-VDR/RXR
900
B
800
120
*
*
100
Control 1,25D3
500
Control 1,25D3
80
400 60
300 200
40
100
20
0 Long promoter fragment Short promoter fragment
C 800 700
NCI-H295R
140
700 600
-VDR/RXR
160
Y-1
+VDR/RXR *
0 Long promoter fragment Short promoter fragment
D
+VDR/RXR
4500 4000
600
3500
500
3000
*
2500
400
Control 1,25D3
300
NCI-H295R
Control 1,25D3
2000 1500
200
1000
100
500
0 Long promoter fragment Short promoter fragment
0 Long promoter fragment Short promoter fragment
E 1600
*
1400
*
1200
*
1000
°
800 600 400
1,25D3 + 1000 nM antagonist
1,25D3 + 100 nM antagonist
1,25D3 + 10 nM antagonist
1,25D3
0
Control
200
Fig. 1. Effects of 1α,25-dihydroxyvitamin D3 on CYP21A2 promoter activity. Experiments were carried out by transient transfection of Y-1 cells (A) and NCI-H295R cells (B) with CYP21A2 promoter luciferase reporter constructs (long promoter fragment, 9 kb or short promoter fragment, 0.3 kb) together with pCMVβ-galactosidase plasmid. In other experiments, Y-1 cells (C) and NCI-H295R cells (D) were transfected with CYP21A2 promoter luciferase reporter constructs (long promoter fragment, 9 kb or short promoter fragment, 0.3 kb), expression vectors for human VDR and human RXR together with pCMVβ-galactosidase plasmid. The cells were treated with 10 nM of 1α,25-dihydroxyvitamin D3. In a separate experiment, NCI-H295R cells were transiently transfected with CYP21A2 promoter luciferase reporter construct (long promoter fragment, 9 kb), expression vectors for human VDR and human RXR together with pCMVβ-galactosidase plasmid. The cells were treated with 10 nM of 1α,25-dihydroxyvitamin D3 and with increasing concentrations of the VDR antagonist TEI9647 (10 nM to 1 μM). Following treatment, all cells were analyzed for luciferase activity after 24 h of incubation as described in the Material and methods section. Ethanol was used as control. Data are presented as reporter activity (RLU/β-gal)± standard deviation (n= 3–7). *Statistically significant difference compared to the control (pb 0.05). °Statistically significant difference compared to 1,25D3 alone (pb 0.05).
and that altered expression levels of these comodulators may shift a suppressing effect of 1α,25-dihydroxyvitamin D3 to a stimulatory effect. 3.5. Identification of a functional VDRE in the CYP21A2 promoter To identify regions of the CYP21A2 promoter containing putative vitamin D response elements (VDREs), experiments were conducted with progressive deletion constructs of the CYP21A2 promoter. When NCI-H295R cells transfected with these deletion constructs were
treated with 1α,25-dihydroxyvitamin D3, the long promoter construct (9 kb) was markedly upregulated by the treatment (Fig. 3). The other deletion constructs were not affected to the same extent. The results indicate that the promoter sequence most important for 1α,25dihydroxyvitamin D3-mediated effects on CYP21A2 transcription is located in the region between −5.6 kb and −9 kb. In order to obtain more information on the location of the VDRE, we performed in silico analysis of the sequence for the CYP21A2 promoter in the region −5.6 kb to −9 kb upstream of the transcription
J. Lundqvist et al. / Biochimica et Biophysica Acta 1820 (2012) 1553–1559
1557
4000
mRNA level CYP21A2 - fold change vs control
1,2 1 0,8
Control
0,6
1,25D3
*
0,4 0,2 0
mRNA level VDIR - fold change vs control
1 Control VDIR silenced
0,8 0,6
*
0,4 0,2 0
Reporter activity (RLU/β-gal)
80 70 60 Control
50
1,25D3
*
40 30
* *
20 10 0
3000 2500
WSTF
Control
2000
1,25D3
1500 1000 500 9 kb
5,6 kb
5,0 kb
4,6 kb
3,8 kb
3,4 kb
0,3 kb
Fig. 3. Effects of 1α,25-dihydroxyvitamin D3 on the activity of CYP21A2 promoter fragments of different lengths. Experiments were carried out by transient transfection of NCI-H295R cells with CYP21A2 promoter luciferase reporter constructs of different lengths (−9 kb, −5.6 kb, −5.0 kb, −4.6 kb, −3.8 kb, −3.4 kb, −0.3 kb), expression vectors for human VDR and human RXR together with pCMVβ-galactosidase plasmid. The cells were treated with 10 nM of 1α,25-dihydroxyvitamin D3 and analyzed for luciferase activity after 24 h of incubation as described in the Material and methods section. Ethanol was used as control. Data are presented as reporter activity (RLU/β-gal)± standard deviation (n= 4). *Statistically significant difference compared to the control (pb 0.05).
1,2
C
3500
0
VDIR silenced
Unaltered VDIR expression
B
NCI-H295R
NCI-H295R Reporter activity (RLU/β-gal)
A
VDIR
WSTF+VDIR
Fig. 2. Role for comodulators in the 1α,25-dihydroxyvitamin D3-mediated alterations of CYP21A2 gene expression. (A) Effects of 1α,25-dihydroxyvitamin D3 treatment on CYP21A2 mRNA level after VDIR silencing. VDIR expression in NCI-H295R cells was silenced with siRNA as described in the Material and methods section. In the control sample (transfected with scrambled siRNA), 1α,25-dihydroxyvitamin D3 suppressed the CYP21A2 mRNA level. When VDIR was silenced, the suppressing effect of 1α,25dihydroxyvitamin D3 was abolished. (B) The effect of siRNA transfection on VDIR expression in NCI-H295R cells. mRNA levels were measured using real time RT-PCR as described in the Material and methods section. Data are presented as fold change compared to the control group ± standard deviation (n= 3–4). *Statistically significant difference compared to the control (p b 0.05). (C) The role of key comodulators for the 1α,25dihydroxyvitamin D3-mediated alterations in CYP21A2 promoter activity. Experiments were carried out by transient transfection of NCI-H295R cells with CYP21A2 promoter luciferase reporter construct (long promoter fragment, 9 kb), expression vectors for human VDR and human RXR together with pCMVβ-galactosidase plasmid. In some of the experiments, cells were cotransfected with expression vectors for the comodulators VDIR and/or WSTF. The cells were treated with 10 nM of 1α,25-dihydroxyvitamin D3 and were analyzed for luciferase activity after 24 h of incubation as described in the Material and methods section. Ethanol was used as control. Data are presented as fold change compared to the control group ± standard deviation (n= 3–9). *Statistically significant difference compared to the control (pb 0.05).
start site. The promoter region was analyzed for consensus sequences for VDRE described by Matilainen et al. [5] and for previously described negative VDREs [6]. We found two sequences that are identical to previously described negative VDREs (Fig. 4E) in the promoters for human CYP27B1 and human PTHrP. The sequences are identical in all three promoters but the distance between the two sequences differs between the three promoters.
In order to study if the vitamin D receptor and/or the comodulators VDIR and WSTF interact with either of these putative VDREs, we performed ChIP assay with NCI-H295R cells treated with 1α,25dihydroxyvitamin D3 or ethanol. The chromatin was cross-linked and immunoprecipitated using antibodies against VDR, WSTF and VDIR. The immunoprecipitated DNA was then used for PCR amplification with primers specific for the two putative VDREs found in the in silicoanalysis. The ChIP assay showed that VDR, WSTF and VDIR interact with Motif 2 in the ethanol treated cells (Fig. 4A–D). Treatment with 1α,25-dihydroxyvitamin D3 dissociated WSTF from Motif 2. This indicates that 1α,25-dihydroxyvitamin D3 is able to alter the composition of a nuclear receptor–comodulator complex, binding to the CYP21A2 promoter at a 148 bp region around Motif 2. The results from the ChIP assay indicate only a weak binding of VDIR to Motif 1. In order to confirm this finding of a VDRE in the CYP21A2 promoter, the long CYP21A2 promoter luciferase reporter construct was mutated. A mutated construct was produced, where the putative VDRE located at −6377 to −6382 was deleted (Motif 2). NCI-H295R cells were then transiently transfected with the mutated constructs or the construct with the wild type CYP21A2 promoter. All cells were cotransfected with expression vectors for VDR and RXR. As expected, the wild type CYP21A2 promoter was stimulated when treated with 1α,25-dihydroxyvitamin D3 (Fig. 4F). However, the mutated construct where Motif 2 had been deleted, was not stimulated by 1α,25-dihydroxyvitamin D3 treatment. We conclude from this, that the sequence CATCTG located at −6377 to −6382 in the CYP21A2 promoter is important for the effects of 1α,25-dihydroxyvitamin D3 on the promoter activity and that it functions as a VDRE. 4. Discussion The current data reveal a novel mechanism for regulation of the human CYP21A2 gene and add important knowledge regarding the mechanisms for vitamin D-mediated suppression of gene expression. Extensive research has been conducted to understand the mechanism for 1α,25-dihydroxyvitamin D3-mediated induction of gene expression [11]. The mechanism for vitamin D-mediated downregulation of gene expression, on the other hand, has in large part remained unclear [4]. Studies on the mechanism for 1α,25-dihydroxyvitamin D3-mediated downregulation of gene expression have been performed for only a few genes. The proposed mechanism for those genes includes recruitment of corepressors such as VDR interacting repressor (VDIR) and Williams syndrome transcription factor (WSTF) as well as epigenetic mechanisms such as histone deacetylation and DNA methylation [6,12–15]. The
1558
J. Lundqvist et al. / Biochimica et Biophysica Acta 1820 (2012) 1553–1559
B Amount DNA (fold change vs input control)
A
Controls
45 40 35 30 25 20 15 10 5 0 Input
Fold change
Positive control
1.0
Negative control
21.7
0.76
Input 1.0
Control
Amount DNA (fold change vs input control)
D
Motif 1
70 60 50 40
EtOH 1,25D3
30 20 10 0
* VDR
WSTF
Amount DNA (fold change vs input control)
C
F Reporter activity (RLU/beta/gal) (fold change vs input control)
E
39.7
0.02
1,25D3
Motif 2
3.5 3.0 2.5 2.0 1.5
EtOH
*
*
1,25D3
1.0
*
0.5 0.0
VDIR
Positive Negative control control
18
VDR
WSTF
VDIR
*
16 14 12 10 EtOH
8
1,25D3
6 4 2 0 Wild type
Motif 2 deleted
Fig. 4. Analysis of interaction of VDR, WSTF and VDIR with the CYP21A2 promoter. NCI-H295R cells were cross-linked with formaldehyde and chromatin immunoprecipitation was performed with antibodies against VDR, WSTF and VDIR. Recruitment of VDR, WSTF and VDIR to the CYP21A2 promoter was analyzed by PCR (A) and quantitative PCR (C and D) with primers for the two putative VDREs found in the in silico analysis. Input control, positive control and negative control (B) were analyzed in accordance with the manufacturer's recommendations. (E) In silico analysis of the CYP21A2 promoter revealed two motifs with sequences identical to previously described nVDREs. The in silico-analysis was performed as described in the Material and methods section. (F) Effects of 1α,25-dihydroxyvitamin D3 on a deletion mutant of the CYP21A2 promoter. Experiments were carried out by transient transfection of NCI-H295R cells with CYP21A2 promoter luciferase reporter constructs, expression vectors for human VDR and human RXR together with pCMVβ-galactosidase plasmid. The CYP21A2 promoter was either wild type or with the putative nVDRE deleted using mutagenesis as described in the Material and methods section. The cells were treated with 10 nM of 1α,25-dihydroxyvitamin D3 and analyzed for luciferase activity after 24 h incubation as described in the Material and methods section. Ethanol was used as control. Data are presented as fold change compared to the control group± standard deviation (n= 4–8). *Statistically significant difference compared to the control (pb 0.05).
negative vitamin D responsive elements (nVDRE) that have been described only vaguely resemble the pVDRE described [6]. It has been suggested that the mechanism for vitamin D-mediated downregulation of gene expression does not include a direct interaction between the ligand-activated VDR–RXR complex and the nVDRE, but rather an interaction between nVDRE and VDIR [12]. We found a sequence in the CYP21A2 promoter that was identical to a previously described negative VDRE in the promoter for the 1αhydroxylase CYP27B1 [6]. When this sequence was deleted using mutagenesis, the effect of 1α,25-dihydroxyvitamin D3 on promoter activity was abrogated. This indicates that the sequence in question may function as a negative vitamin D responsive element (nVDRE) in the CYP21A2 promoter.
Our data indicate that the recently reported suppression of CYP21A2 gene transcription by 1α,25-dihydroxyvitamin D3 is mediated by a complex formed by the nuclear receptor VDR and the comodulators VDIR and WSTF. Our results suggest that the expression levels of nuclear receptor and key comodulators are crucial for the effect of vitamin D on the CYP21A2 promoter. Thus, the effect of vitamin D on CYP21A2 may be dependent on the expression levels of VDIR and WSTF in cells. This conclusion is supported by the fact that when VDIR was silenced using siRNA, the vitamin D-mediated effect on CYP21A2 mRNA level was abolished. This work has been conducted in cell culture-based in vitro assays such as reporter assay, chromatin immunoprecipitation, gene knock down et cetera. Even though these methods provide novel and
J. Lundqvist et al. / Biochimica et Biophysica Acta 1820 (2012) 1553–1559
1559
generous gift of the VDR antagonist TEI-9647. The authors are grateful to Dr Nathalie Luna Persson, Karolinska Institute, for technical advice on primer design. References
Fig. 5. Hypothesis for the vitamin D-induced effects on the CYP21A2 promoter. Based on our ChIP and EMSA assays, the following hypothesis for the vitamin D-induced effects on CYP21A2 transcription can be proposed. Our results indicate that VDR, WSTF and VDIR interact with a VDRE in the CYP21A2 promoter in the absence of 1α,25dihydroxyvitamin D3 (A). In the presence of 1α,25-dihydroxyvitamin D3, WSTF dissociates and the transcriptional activity of CYP21A2 is decreased (B).
important insights in the regulation of the transcription of CYP21A2 it cannot be excluded that also other mechanisms may be involved in the complex in vivo regulation of CYP21A2. VDIR and WSTF have previously been described as key components in vitamin D-mediated suppression of the CYP27B1 gene and the parathyroid hormone gene [8,9]. However, the mechanism that can be proposed based on our experiments (Fig. 5) has not been reported previously. We conclude that 1α,25-dihydroxyvitamin D3 affects CYP21A2 transcription via a novel mechanism for vitamin D action, where VDR, VDIR and WSTF interact with a VDRE in the promoter in the absence of 1α,25-dihydroxyvitamin D3. When cells are treated with 1α,25dihydroxyvitamin D3, WSTF dissociates from the complex and the promoter activity of CYP21A2 is decreased. Further research is needed to investigate if this complex is composed also of other comodulators, and which of the receptors/comodulators that interacts with the promoter sequence. Accumulated data during recent years have indicated that vitamin D may have many more targets than previously known [11,16,17]. It may be speculated that the expression levels of comodulators may be of importance also for other vitamin D-related genes. If so, such comodulators may be future target to manipulate vitamin D-dependent functions. In conclusion, this paper reports data important for the understanding of the mechanisms for vitamin D-mediated suppression of gene expression as well as for the vitamin D-mediated effects on CYP21A2. Acknowledgements This work was supported by the Swedish Research Council — Medicine. Teijin Pharma, Tokyo, Japan is acknowledged for the
[1] S.N. Kelly, T.J. McKenna, L.S. Young, Modulation of steroidogenic enzymes by orphan nuclear transcriptional regulation may control diverse production of cortisol and androgens in the human adrenal, J. Endocrinol. 181 (2004) 355–365. [2] S.D. Wijesuriya, G. Zhang, A. Dardis, W.L. Miller, Transcriptional regulatory elements of the human gene for cytochrome P450c21 (steroid 21-hydroxylase) lie within intron 35 of the linked C4B gene, J. Biol. Chem. 274 (1999) 38097–38106. [3] J. Lundqvist, M. Norlin, K. Wikvall, 1α,25-Dihydroxyvitamin D3 affects hormone production and expression of steroidogenic enzymes in human adrenocortical NCI-H295R cells, Biochim. Biophys. Acta (BBA) — Mol. Cell Biol. Lipids 1801 (2010) 1056–1062. [4] J.W. Pike, M.B. Meyer, The vitamin D receptor: new paradigms for the regulation of gene expression by 1,25-dihydroxyvitamin D3, Endocrinol. Metab. Clin. North Am. 39 (2010) 255–269. [5] M. Matilainen, M. Malinen, K. Saavalainen, C. Carlberg, Regulation of multiple insulin-like growth factor binding protein genes by 1α,25-dihydroxyvitamin D3, Nucl. Acids Res. 33 (2005) 5521–5532. [6] M. Kim, R. Fujiki, A. Murayama, H. Kitagawa, K. Yamaoka, Y. Yamamoto, M. Mihara, K.-I. Takeyama, S. Kato, 1α,25(OH)2D3-Induced transrepression by vitamin D receptor through E-box-type elements in the human parathyroid hormone gene promoter, Mol. Endocrinol. 21 (2007) 334–342. [7] M. Riemenschneider, T. Koy, G. Reifenberger, Expression of oligodendrocyte lineage genes in oligodendroglial and astrocytic gliomas, Acta Neuropathol. 107 (2004) 277–282. [8] R.A. Poot, L. Bozhenok, D.L.C. van den Berg, S. Steffensen, F. Ferreira, M. Grimaldi, N. Gilbert, J. Ferreira, P.D. Varga-Weisz, The Williams syndrome transcription factor interacts with PCNA to target chromatin remodelling by ISWI to replication foci, Nat. Cell Biol. 6 (2004) 1236–1244. [9] J. Pascussi, A. Robert, M. Nguyen, O. Walrant-Debray, M. Garabedian, P. Martin, T. Pineau, J. Saric, F. Navarro, P. Maurel, M. Vilarem, Possible involvement of pregnane X receptor-enhanced CYP24 expression in drug-induced osteomalacia, J. Clin. Invest. 115 (2005) 177–186. [10] S. Ishizuka, N. Kurihara, Y. Hiruma, D. Miura, J.-I. Namekawa, A. Tamura, Y. Kato-Nakamura, Y. Nakano, K. Takenouchi, Y. Hashimoto, K. Nagasawa, G.D. Roodman, 1α,25-Dihydroxyvitamin D3-26,23-lactam analogues function as vitamin D receptor antagonists in human and rodent cells, J. Steroid Biochem. Mol. Biol. 110 (2008) 269–277. [11] A.W. Norman, From vitamin D to hormone D: fundamentals of the vitamin D endocrine system essential for good health, Am. J. Clin. Nutr. 88 (2008) 491S–499S. [12] R. Fujiki, M.-S. Kim, Y. Sasaki, K. Yoshimura, H. Kitagawa, S. Kato, Ligand-induced transrepression by VDR through association of WSTF with acetylated histones, EMBO J. 24 (2005) 3881–3894. [13] A. Murayama, M.-S. Kim, J. Yanagisawa, K.-I. Takeyama, S. Kato, Transrepression by a liganded nuclear receptor via a bHLH activator through co-regulator switching, EMBO J. 23 (2004) 1598. [14] M. Kim, T. Kondo, I. Takada, M.-Y. Youn, Y. Yamamoto, S. Takahashi, T. Matsumoto, S. Fujiyama, Y. Shirode, I. Yamaoka, H. Kitagawa, K.-I. Takeyama, H. Shibuya, F. Ohtake, S. Kato, DNA demethylation in hormone-induced transcriptional derepression, Nature 461 (2009) 1007–1012. [15] M.-S. Kim, R. Fujiki, H. Kitagawa, S. Kato, 1α,25(OH)2D3-induced DNA methylation suppresses the human CYP27B1 gene, Mol. Cell. Endocrinol. 265–266 (2007) 168–173. [16] H.F. DeLuca, Overview of general physiologic features and functions of vitamin D, Am. J. Clin. Nutr. 80 (2004) 1689S–1696S. [17] S. Ramagopalan, A. Heger, A. Berlanga, N. Maugeri, M. Lincoln, A. Burrell, L. Handunnetthi, A. Handel, G. Disanto, S. Orton, C. Watson, J. Morahan, G. Giovannoni, C. Ponting, G. Ebers, J. Knight, A ChIP-seq defined genome-wide map of vitamin D receptor binding: associations with disease and evolution, Genome Res. 20 (2010) 1352–1360.