autoregulatory loop controlling orphan nuclear receptor DAX-1 gene exp
http://www.100md.com
《核酸研究医学期刊》
Hormone Research Center, School of Biological Sciences and Technology, Chonnam National University Gwangju, 500-757, Republic of Korea 1Section of Endocrinology, Department of Internal Medicine, Kyungpook National University Hospital, School of Medicine, Kyungpook National University Taegu, Republic of Korea 2Department of Pathology, College of Medicine, Chungnam National University Daejeon, 301-131, Republic of Korea 3Center for Marine Natural Products and Drug Discovery, School of Earth and Environmental Science, Seoul National University Seoul, Republic of Korea
*To whom correspondence should be addressed. Tel: +82 62 530 0503; Fax: +82 62 530 0500; Email: hsc@chonnam.ac.kr
ABSTRACT
The estrogen receptor-related receptor gamma (ERR/ERR3/NR3B3) is a member of the nuclear receptor superfamily that activates transcription in the absence of ligand. However, the detailed mechanism of gene regulation by ERR is not fully understood. In this study we have found that the orphan nuclear receptor ERR activates the DAX-1 promoter, which, in turn, represses transactivation by ERR. Serial deletions of mouse DAX-1 (mDAX-1) gene promoter have revealed that the region responding to ERR is located between –129 and –121 bp and –334 and –326 bp. Gel shift assays and chromatin immunoprecipitation (ChIP) assays demonstrated that ERR binds directly to the mDAX-1 promoter. Site-directed mutagenesis results demonstrated that ERRE1 (–129 to –121 bp) is more important than ERRE2 (–334 to –326 bp) which is not conserved in the human DAX-1 promoter. In addition, adenovirus-mediated overexpression of ERR induced DAX-1 gene expression in MCF-7 breast cancer cells that co-expressed ERR and DAX-1. Moreover, yeast two-hybrid and glutathione S-transferase (GST)-pull down assays demonstrated that DAX-1 physically interacted with ERR and inhibited ERR transactivation, and that this interaction was dependent on the AF-2 domain of ERR. In addition, in vitro competition assays showed that DAX-1 inhibited PGC-1 mediated ERR transactivation, via competition between these two factors for the AF-2 binding domain. We thus propose a novel autoregulatory loop that controls DAX-1 gene expression by ERR.
INTRODUCTION
Nuclear receptors (NRs) are ligand-dependent transcription factors that bind to specific DNA target sequences through their highly conserved DNA-binding domain (DBD) and activate transcription through interaction with coactivators. The latter property is dependent on their moderately conserved ligand-binding domain (LBD) and in particular requires the extreme C-terminal part of it, the so-called AF-2 region (1).
Among NRs, DAX-1 (dosage-sensitive sex reversal, adrenal hypoplasia critical region, on chromosome X, gene 1) (NR0B1) is an atypical orphan nuclear receptor, since it does not contain a DBD (2,3). Instead, the N-terminal half of DAX-1 contains a unique region mainly composed of three repeats, each of which is 65–67 amino acids in length. This region contains LXXLL motif-like sequences (where L and M are leucine and methionine, respectively, and X can be any amino acid) that are necessary for interaction with estrogen receptors (ERs) (4). Since the C-terminal half of DAX-1 containing the putative LBD functions as a repressor domain through interaction with corepressors (NCoR) (5), two mechanisms have been proposed to function cooperatively in the repression of transcription exerted by DAX-1. According to the proposition, DAX-1 would both block the binding of coactivators to NRs and recruit corepressors with its C-terminal half (4–6). The DAX-1 gene was identified through a search for a gene linked to adrenal hypoplasia congenita, a disease affecting the normal development of the adrenal cortex and which is often associated with hypogonadotropic hypogonadism (2,3). Mouse and human DAX-1 promoters have been cloned and characterized. For example, orphan nuclear receptor steroidogenic factor-1 (SF-1) regulates DAX-1 promoter, and the SF-1 response element is located between –129 and –121 bp on the mouse promoter and between –135 and –143 bp on the human promoter (7,8).
Estrogen-related receptors (ERRs-alpha, -beta and -gamma), are orphan NRs closely related to ERs, with which they share identical target response elements and coregulatory proteins (9,10). However, ERRs do not respond to the classical ER ligand. Previous studies have demonstrated that MCAD, pS2 and cyp19 are targets of ERR (11–13). Among the ERR family, ERR, the newest member of the subfamily, is a constitutively active nuclear receptor, and little is known about its in vivo functions (9,10). Previous studies have demonstrated that 4-hydroxytamoxifen (4-OHT) directly binds to and deactivates ERR, suggesting that 4-OHT is an inverse agonist of ERR (14,15). However, the deactivation mechanism of ERR by 4-OHT requires further elucidation. It has been reported that coactivators such as GRIP1 or PGC-1, interact directly with ERR and activate its transcriptional activity (16,17). ERR acts as a monomer and activates transcription constitutively. It has also been suggested in recent reports, that ERR forms dimers through its LBD. In contrast to this, the related receptor, ERR, inhibits the activities of both ERR and ERR through heterodimerization (18). ERR exhibits significant affinities for binding to a wide spectrum of sequences, including inverted and direct repeat sequences composed of AGGTCA half-sites with differing spacings, as well as a monovalent motif of the same sequence carrying an extra T(C/G)A trinucleotide on the 5' side (19). It has been reported that ERR binds to and transactivates both estrogen response elements and SF-1 response elements (SF-1RE). Previous reports have suggested that several monomeric binding partners, such as SF-1, liver receptor homolog-1 (LRH-1) and ERR, can activate and bind directly to the small heterodimer partner (SHP) promoter through SF-1RE (16,20).
DAX-1 is closely related to SHP, which neither contains a DNA-binding domain nor functions as a corepressor (2,3). In addition, both the DAX-1 and SHP promoters are regulated by the monomer binding partner SF-1. Thus, since the DAX-1 promoter contains several SF-1RE, we predicted that ERR could also activate the DAX-1 promoter. Transient transfection studies demonstrated that ERR had a preferential effect on the DAX-1 promoter. The results of deletion and site-directed mutagenesis have revealed that two ERR regulatory elements (ERREs) are responsible for the ERR-mediated activation of the DAX-1 promoter. The results of electrophoretic mobility shift assay (EMSA) and chromatin immunoprecipitation (ChIP) assays strongly suggest that ERR binds directly to the DAX-1 promoter. In addition, the results of overexpression and ERR knock-down experiments demonstrated that ERR regulates DAX-1 gene expression in MCF-7 breast cancer cells and that ERR and DAX-1 are co-expressed in breast cancer tissue. The expression of DAX-1 inhibits ERR transactivation of its own promoter and DAX-1 competes directly with the coactivator, PGC-1, for binding to ERR. Taken together, these results suggest that ERR belongs to a new autoregulatory loop that governs DAX-1 gene expression.
MATERIALS AND METHODS
Plasmids and DNA construction
pcDNA3/HA-ERR, ?, and (AF-2), pcDNA3/HA SF-1, pcDNA3/HA DAX-1, pcDNA3/HA PGC-1, pGEX4T-1 DAX-1, B42 ERR, B42 AF-2 ERR, Lex A DAX-1, and pcDNA3 PGC-1 were described previously (21–23). To generate the –467 bp DAX-1 promoter region, genomic DNA was isolated from a mouse testicular Leydig cell line, K28 and used as template for PCR. The PCR product was cloned into the pGL3-basic plasmid (Promega) between the MluI and XhoI sites. mDAX-1/Luc (–325 and –121 bp) were made by PCR using the primer and cloned into pGL3 at MluI and XhoI. Mutant reporters, mtERRE1/Luc, mtERRE2/Luc and mtERRE1&2/Luc, were constructed by site-directed mutagenesis of the mDAX-1 promoter –467 bp/Luc using the primer. The mutated sequences are shown in Figure 4A. All clones were verified by sequencing analysis.
Figure 4 The ERRE1 (–129 to –121 bp) plays a crucial role in the transactivation of the mDAX-1 promoter by ERR. (A) Schematic diagrams of wild-type and mutant mDAX-1 promoter constructs from –467 to +43 bp as indicated. The putative ERR binding sites are indicated as ERRE1 and ERRE2. (B) ERRE1 is essential for the activation of the mDAX-1 promoter by ERR. 293T cells were co-transfected with 200 ng of wild-type or mutant mDAX-1 promoter constructs together with 200 ng of ERR expression plasmid. Two days after transfection, cells were harvested for luciferase and -galactosidase assays. The representative results are expressed as -fold activation (n-fold) over the value obtained with vector alone with the error bars as indicated.
Cell culture and transient transfection assay
Human embryonic kidney (293T), mouse fibriblast (NIH3T3) and human breast cancer (MCF-7) cells were maintained with DMEM in the presence of 10% fetal bovine serum and antibiotics (Gibco) in humidified air containing 5% CO2 at 37°C. Cells were transfected, using Superfect and Effectene reagents (Qiagen Inc., Germany), according to the manufacturer's instructions. The total DNA used in each transfection was adjusted to 1 μg by adding the appropriate amount of pcDNA3 vector. Approximately 40–48 h post-transfection, cells were harvested and luciferase activity was measured and normalized to ?-galactosidase activity. Experiments were performed three times in duplicate.
Glutathione S-transferase (GST)-pull down assay
The indicated GST-fusion proteins or GST protein alone were expressed in Escherichia coli BL21 (DE3) pLys cultures by adding 0.2 mM isopropyl-?-D-thiogalactopyranoside (IPTG), and the cells were extracted. GST-fusion proteins were pre-bound with a 30 μl aliquot of glutathione-Sepharose beads. The beads were incubated with transcribed and translated methionine-labeled proteins for 3–4 h at 4°C. The beads were then washed three times with the washing buffer, and analyzed by SDS–PAGE and protein bands were visualized by using a phosphorimaging analyzer. Translated proteins in vitro (10%) were used as input.
In vitro translation
ERR and ERR AF-2 were transcribed and translated in vitro using a coupled rabbit reticulocyte system (Promega Corp., Madison, WI) in the presence of methionine according to the manufacturer's instructions. The translated proteins were analyzed by 10% SDS–PAGE and visualized by autoradiography.
Yeast two-hybrid interaction assay
Yeast two-hybrid interaction assays were performed as described previously (22). Briefly, LexA only or LexA fused plasmids, and B42 only or B42 fused plasmids, were transformed into Saccharomyces cerevisae EGY48, containing the galactosidase reporter plasmid 8H18-34, and transformants were selected on plates with the appropriate selection markers. The galactosidase assay on plates was carried out as described previously (22).
Gel mobility shift assays
Double-stranded ERRE oligonucleotides were labeled by filling-in with dCTP, using the Klenow fragment of DNA polymerase I. The oligonucleotide sequences are depicted in Figure 3A. Bacterially expressed GST only and GST fused ERR were purified using GST-Sepharose beads (Amersham Bioscience, Inc.). Gel mobility shift assays (20 μl) contained 10 mM Tris (pH 8.0), 40 mM KCl, 0.05% Nonidet P-40, 6% glycerol, 1 mM DTT, and 1 μg of poly (dI-dC). Either 2 μg of GST only, or GST fused ERR purified proteins, were used in each reaction. Competitor oligonucleotides were included at a 50- to 100-fold molar excess as indicated in the Figure legends. After 15 min incubation on ice, 10 000 c.p.m. of labeled oligonucleotide probe was added and incubation was continued for another 15 min. DNA–protein complexes were analyzed by 5% PAGE in 1x TBE. Gels were dried and analyzed by autoradiography.
Figure 3 ERR binds to ERRE in the DAX-1 promoter. (A) The sequences of wild-type (nt –126 to –106 and –337 to –317) and mutant oligonucleotides used for the gel retardation assays are depicted. Italic letters indicate the substituted nucleotides in ERRE1 and ERRE2. (B) Gel mobility shift and unlabeled competition assays were performed with the 32P-labeled probes, and recombinant ERR (2 μg) protein as indicated. Unlabeled oligonucleotides were added as competitors at 50- to 100-fold molar excess where indicated. (C) ChIP assay. 293T cells were transfected with expression vectors for HA epitope alone or HA-ERR with –467 bp DAX-1-Luc. Soluble chromatin from these cells was prepared and immunoprecipitated with monoclonal antibody against HA (lanes 3 and 4). The –350 –100 bp fragment (250 bp) contains the ERR binding site and 10% of the soluble chromatin used in the reaction was used as input (lanes 1 and 2). PCR was performed as described in Figure 3C.
ChIP assay
The ChIP assay was performed according to the instructions of the ChIP Assay Kit (Upstate, USA). 293T cells were transfected with 1 μg of pcDNA3 HA-ERR or pcDNA3 HA and pGL3 –467 bp mDAX-1 using Effectene reagent (Qiagen Inc., Germany). The cells were fixed with formaldehyde 48 h after transfection and harvested. For immunoprecipitation, anti-HA antibody (Roche Molecular Biochemicals) was used. Final DNA extractions were amplified using pairs of primers that cover the ERR response region within the mDAX-1 promoter (nt –350 to –100 bp). The primers used for PCR were as follows: 5'-AAAGGAAAT AAGTTAGAGGTCAGAG-3' (Forward) and 5'-AGCGCGTCCGCCTCCTCCTCTTGGA-5' (Reverse). MCF-7 cells (Figure 5C) were transfected with 1 μg of pcDNA3/HA or pcDNA3/HA-ERR. Soluble chromatin was immunoprecipitated with anti-HA antibody. Final DNA extractions were amplified by 30 cycles of PCR using pairs of primers that cover the ERR response region within the hDAX-1 promoter (nt –260 to +6 bp). The primers used for PCR were as follows: 5'-CAG CAT CCA GGC GCT CGC TCT CC-3' (Forward) and 5'-TTC TGC CCA GTG GCT GCC TCC TGG-5' (Reverse).
Figure 5 Co-expression of ERR and DAX-1 in breast cancer cell. (A) MCF-7 cells were infected with adenoviral vector expressing ERR (100 pfu/cells). Total RNA was isolated from cells and analyzed by RT–PCR. (B) The effect of siRNA-ERR on the mRNA level of DAX-1. Endogenous ERR gene expression was inhibited by transfection with a 21 nt RNA duplex siRNA-ERR/I in MCF-7 cells. The effects of siRNAs on ERR and DAX-1 expression were assayed by realizing RT–PCR for ERR, DAX-1, and ?-actin as a control. (C) MCF-7 cells were transfected with HA or HA-ERR. The –260 +6 bp fragment (266 bp) contains the ERR binding site and 10% of the soluble chromatin used in the reaction was used as input (lanes 1 and 2). PCR was performed as described in Figure 5C. (D) DAX-1 and ERR expression in human breast cancer cells. The DAX-1-expressing tumor cells (top left) were also positive for ERR (top right). The DAX-1-negative cancer cells were also negative for ERR (bottom left and right) (original magnification, 200x).
Preparation of recombinant adenovirus
The recombinant adenovirus was prepared as described previously (24). In brief, the cDNA encoding ERR was cloned into a pAd-YC2 shuttle vector, which (under the control of the cytomegalovirus promoter) contains a bovine growth hormone polyadenylation signal sequence. For homologous recombination, a pAd-YC2 shuttle vector (5 μg) and a rescue vector, pJM17 (5 μg), were co-transfected into human embryonic kidney 293 cells. To purify pure plaques, cell culture supernatant was serially diluted into serum-free media and incubated with 293 cells at 37°C for 1 h. An equal volume mixture of 2x medium and 1% agarose was overlaid on the 293 cells. After 7 days, plaques that were well isolated were purified further and propagated in 293 cells and screened by PCR, using upstream primers derived from the cytomegalovirus promoter, and downstream primers from the bovine growth hormone polyadenylation sequence. Then, the recombinants were amplified in 293 cells and were purified and isolated using CsCl2 (Sigma). The preparations were collected and desalted, and titers were determined by counting the number of plaques.
Immunohistochemistry
Immunohistochemical analysis was performed with nine human breast cancer tissues. Three sections were cut from the paraffin blocks, deparaffinized antigen retrieval by pressure cooker in 10 mM sodium citrate (pH 6.0) at full power for 4 min. The primary antibodies used were rabbit polyclonal anti-ERR (Abcam, 1:50) and anti-DAX-1 (Santa Cruz, 1:150). Immunohistochemistry was performed using an EnVision kit (DAKO, Carpinteria, CA). After preincubation with blocking serum for 15 min, the primary polyclonal antbody for ERR or DAX-1 was added and incubated for 30 min in a humid chamber followed by a wash with TBS buffer. Slides were then incubated for 30 min with EnVision peroxidase reagent (DAKO, Carpinteria, CA).
The slides were then sequentially incubated with DAB (3,3-diaminobenzidine) chromogen for 5 min, counterstained with Meyer's hematoxylin and mounted. Careful rinses with several changes of TBS were performed between each stage of the procedure. To assess the specificity of the immunoreaction, a negative control without the primary antibody or with pre-immune rabbit serum was performed.
RT–PCR analysis
MCF-7 cells were infected with an adenovirus containing the ERR gene (100 pfu/cells). After viral infection for 24 h, cells were harvested for total RNA isolation using the TRIzol reagent (Invitrogen Life Technologies) according to the manufacturer's protocol. The mRNAs of ERR and DAX-1 were analyzed by RT–PCR. First-strand cDNA was synthesized from 1 μg of total RNA utilizing an anchored oligo(dT) primers and reverse transcriptase. The resulting first-strand cDNA was then amplified to measure mRNA levels of ERR, DAX-1 and ?-actin by realizing 28, 28 and 23 cycles of PCR using specific primers, respectively. The mRNA levels of ?-actin served as an internal control for the RT–PCR analysis. The primers used for PCR of ERR, DAX-1 and ?-actin were as follows: ERR, forward 5'-GACTTGACTCGCCACCTCTC-3' and reverse 5'-GTGGTACCCAGAAGCGATGT-3'; DAX-1, forward 5'-CCGCTTGCAGTTCGA GACTGTG-3' and reverse 5'-CTCATGGTGAACTGCACTACTG-3'; ?-actin, forward 5'-GTCATCACCATTGGCAATGAG and reverse 5'-CGTCATACTCCTGCTTGCTG-3'.
siRNA experiment
The siRNAs for ERR (si-ERR/I and /II) were chemically synthesized (Shamchully Pharm. Co., Ltd), deprotected, annealed, and transfected according to the manufacturer's instructions. MCF-7 cells were transfected with siRNA using Oligofectamine reagent (Invitrogen Life Technologies). After transfection for 48 h, total RNA was isolated for RT–PCR of ERR (28 cycles), DAX-1 (27 cycles) and ?-actin (20 cycles) as a control. The sequences of siRNA are as follows: si-ERR/I, sense 5'-UGG CCA UCA GAA CGG ACU UTT-3'; si-ERR/II sense 5'-CGA AGA CCA GUC CAA AUU ATT-3'.
RESULTS
Identification of ERR as a transcriptional activator of the DAX-1 promoter
SF-1 has been identified as an activator of the DAX-1 promoter (7,8). Since SFRE is also a DNA-binding site for ERR, we speculated that an ERR response element (ERRE) exists in the DAX-1 promoter. As shown in Figure 1A, two ERREs were located in the mDAX-1 promoter; ERRE1 (–129 to –121 bp), which was previously identified as an SF-1RE, and a new ERRE2 (–334 to –326 bp). Interestingly, ERRE1 is conserved in both mouse and human promoters, whereas ERRE2 only exists in the mouse promoter (Figure 1B).
Figure 1 ERREs on the DAX-1 promoter. (A) Mouse DAX-1 promoter sequences from –467 to +43 relative to the putative transcription start site. The adenine at the transcription start site (arrow) and TATA-box (underline) are designated +1 and TATA-box. The translation initiation codon (ATG) is underlined with the encoded amino acids shown below the codons. Putative ERREs are lined above the respective sequences. (B) Sequence alignment of ERRE1 and ERRE2 within the mouse and human DAX-1 promoters. The numbers indicate the nucleotide positions: –129 to –21 bp and –135 to –127 bp in mouse and human ERRE1, respectively, and –334 to –326 bp and –347 to –339 bp in mouse and human ERRE2, respectively.
In order to determine whether ERRs regulate the DAX-1 promoter, we performed transient transfection assays using the luciferase reporter gene driven by a –467 bp region from mouse DAX-1 promoter. As shown in Figure 2A and B, there was strong activation by ERR in 293T cells and NIH3T3 cells, whereas ERR and ERR? did not demonstrate any significant effect. As expected, SF-1, the positive control for the transfection assays, activated the DAX-1 gene promoter significantly. These results indicated that the DAX-1 gene is a specific target of ERR, but not of ERR or ERR?. In order to investigate the sequences that were required for ERR mediated activation, a series of 5' deletions of the DAX-1 promoter were constructed, as shown in Figure 2C. Since the DAX-1 promoter contains two ERREs, we designed deletion constructs both with and without an ERRE. Activation by ERR was decreased when the DAX-1 promoter was deleted up to –325 bp, as demonstrated in Figure 2D. Furthermore, the activation of the DAX-1 promoter by ERR, was almost lost with the –121 bp DAX-1 construct. These results suggest that ERREs are required for ERR response.
Figure 2 ERR activates DAX-1 gene promoter. (A and B) Transcriptional activation of the mDAX-1-promoter by ERR, but not by ERR and ?. 293Tcells (A) and NIH3T3 cells (B) cells were transiently transfected with 200 ng of indicated reporter gene together with the indicated concentrations of HA-pcDNA3-ERR , ?, and SF-1. (C) Schematic diagram of the mDAX-1 promoter deletion constructs cloned in the pGL3-basic vector. (D) Mapping the sequences required for activation of the mDAX-1 promoter by ERR. 293T cells were transfected with 200 ng of each reporter plasmid and ERR expression plasmid as indicated. Approximately 40 h after transfection, the cells were harvested and luciferase activity was measured and normalized against -galactosidase activity. One representative experiment is shown. All values represent the mean of duplicate samples, and similar results were obtained in at least three independent experiments. The representative results were expressed as -fold activation (n-fold) over the value obtained with vector alone with the error bars as indicated.
Determination of ERR binding regions
In order to clarify the specific binding of ERR to ERREs, gel mobility shift assays were performed. As demonstrated in Figure 3A, we designed both wild-type and mutant probes that corresponded to ERRE1 and ERRE2. Previously, the TNAAGGTCA sequence was shown to be a bona fide ERRE. Within the sequence, the GG dimer is pivotal for ERR protein binding. Thus, we substituted GG for the TT sequence. Figure 3B demonstrates that ERR formed specific complexes with ERRE1 and ERRE2; a 50- or 100-fold molar excess of unlabeled ERRE1 and ERRE2 competed strongly with the DNA–protein complex. In order to verify that ERR binds to the DAX-1 promoter in vivo, we performed a ChIP assay using PCR primers that flanked both ERRE1 and ERRE2 (–350 to –100 bp). 293T cells were transfected with the DAX-1 promoter, either with hemagglutinin (HA) only or with HA-tagged ERR. As demonstrated in Figure 3C, a 250 bp PCR product (–350 to –100 bp) was observed in cells that were transfected with the expression vector for HA-ERR but not with the vector expressing only the HA epitope. This indicated that HA-ERR formed a specific complex with the DAX-1 promoter in vivo. Taken together, this suggests that ERR regulates the DAX-1 promoter by direct binding to ERRE.
ERRE1 (–129 to –121 bp) is essential for transactivation of the DAX-1 promoter by ERR
In order to verify the functional significance of ERR binding sites in the DAX-1 promoter, site-directed mutagenesis was performed on the –467 bp DAX-1 promoter using the primers as shown in Figure 4A. Wild-type –467 bp/Luc and several mutant constructs were transiently transfected with ERR into 293T cells. The ERRE1 mutation (mtERRE1/Luc) and double ERRE1 and ERRE2 mutations (mtERRE1&2/Luc) completely abolished the ERR-mediated transactivation of the DAX-1 promoter whereas the ERRE2 mutation (mtERRE2/Luc) had a slight effect (Figure 4B). These results indicated that ERRE1, rather than ERRE2, was essential for the activation of the DAX-1 promoter by ERR. Thus, we conclude that ERRE1 plays a major role in the activation of the DAX-1 promoter by ERR.
Induction of DAX-1 gene expression by ERR in breast cancer
Previous reports have demonstrated that ERR is a favorable biomarker and, possibly, an indicator of hormonal sensitivity in breast cancer. In addition, DAX-1 has also been detected in breast cancer tissue (25,26). In order to investigate whether ERR can directly regulate DAX-1 gene expression in breast cancer cells, we introduced ERR by adenovirus infection into MCF-7 cells. As demonstrated in Figure 5A, a basal expression of DAX-1 and ERR was detected in MCF-7 cells, and DAX-1 mRNA was induced by overexpression of Ad-ERR 24 h after infection. Since ERR is basally expressed in MCF-7 cells, we further tested whether DAX-1 gene expression is down-regulated by performing knock-down of endogenous ERR expression using small interfering RNA (si-ERR/I or II) in MCF-7 cells. si-ERR/I, but not si-ERR/II significantly reduced endogenous ERR expression (Figure 5B). DAX-1 mRNA level was decreased by knock-down of ERR by si-ERR/I, but not by si-ERR/II, suggesting that endogenous ERR can activate DAX-1 expression. In addition, CHIP assays were performed to confirm the direct binding of ERR to the endogenous DAX-1 promoter in MCF-7 cells. As shown in Figure 5C, ERR can bind to the DAX-1 promoter in MCF-7 cells (Figure 5C). Next, we analyzed for co-expression of DAX-1 and ERR in human breast cancer tissues, using immunohistochemistry (Figure 5D). Nuclear immunoreactivity of DAX-1 was detected in breast cancer tissues. The overexpression of ERR was observed in three out of nine cases. All of the ERR-positive cancers co-expressed the DAX-1 protein (Figure 5D, upper part). Two cancer tissues that were DAX-1-negative were also negative for ERR. No staining was observed in the negative controls using either pre-immune rabbit serum or by omitting the primary antibody (data not shown). These results suggest that ERR can stimulate DAX-1 gene expression in human breast cancer cells.
DAX-1 represses ERR transactivation of its own promoter
DAX-1 interacts with several orphan NRs (ER, AR, LRH-1, SF-1 and Nur77) and represses their transcriptional activity (4–6,21,27). We investigated whether DAX-1 can also repress ERR mediated transactivation of its own promoter. The results of transient transfections demonstrated that DAX-1 inhibits ERR activity of the DAX-1 promoter in a dose-dependant manner (Figure 6A). Since ERR also activates the SHP promoter, and SHP represses ERR transactivation of its own promoter, we examined whether DAX-1 represses the ERR-mediated transactivation of the SHP promoter or vice versa. As demonstrated in Figure 6B and C, DAX-1 and SHP repressed ERR transactivation of both the SHP promoter and the DAX-1 promoter. These results suggest that both DAX-1 and SHP are regulated by ERR and repress the transcriptional activity of ERR through an autoregulatory loop. Since Figure 6A shows that DAX-1 repressed the transcriptional activity of ERR, we examined whether DAX-1 can directly interact with ERR, by using a GST-pull down and yeast two-hybrid interaction assays. As demonstrated in Figure 7A and B, DAX-1 interacted with ERR. Since repression or activation is dependant on the AF-2 domain, we deleted this domain in ERR. As expected, DAX-1 could not interact with ERR AF-2, which demonstrates that the AF-2 domain of ERR is crucial for the repressive effect of DAX-1 (Figure 7A). In the case of DAX-1, an LBD-like region was involved in the interaction with ERR (Figure 7B). These results suggest that DAX-1 interacts directly with the AF-2 surface of ERR.
Figure 6 DAX-1 represses ERR activity. (A–C) 293T cells were co-transfected with 200 ng of mDAX-1 (–467 bp)-Luc, mSHP (–2.2 kb)-Luc, 300 ng of ERR, and increasing amounts of DAX-1 and SHP as indicated in the figure. The luciferase activities were measured and normalized against ?-galactosidase activity. One representative experiment is shown, and similar results were obtained from at least three independent experiments.
Figure 7 ERR physically interacts with DAX-1. (A) GST only or GST fused to DAX-1 was isolated from bacterial culture and immobilized on glutathione-Sepharose beads. In vitro transcribed and translated methionine-labeled ERR or ERR AF-2 was incubated with purified GST fused receptors or GST alone as indicated in the Figure. The complexes were resolved by a 12% denaturing PAGE, and analyzed by autoradiography. (B) A yeast strain EGY48 which contains an integrated ?-galactosidase reporter gene controlled by the LexA-binding site was transformed with the indicated LexA and B42 plasmids. The transformants were selected on plates containing appropriate selection markers, and assayed for ?-galactosidase activity. The results shown are the mean ?-galactosidase values from six independent transformant colonies.
Since the AF-2 surface of ERR is involved in the interaction with DAX-1, we explored the role of coactivator competition in the repression of ERR activity. As demonstrated in Figure 8A, PGC-1, which was described previously as a coactivator of ERR, significantly increased the activity of ERR on the DAX-1 promoter and the co-expression of DAX-1 with PGC-1 repressed this induction in a dose-dependant manner. This result suggests that both PGC-1 and DAX-1 compete for the binding of the AF-2 pocket of ERR, as demonstrated for other receptors. In order to confirm the direct competition between DAX-1 and PGC-1 for binding to ERR, we performed in vitro competition binding assays, using PGC-1 and DAX-1 with GST fused ERR. The results of this study confirmed the findings of a previous report (17). Specifically, 35S-labeled PGC-1 interacted with ERR, and DAX-1 inhibited the interaction between PGC-1 and ERR (Figure 8B). In addition, no interaction between DAX-1 and PGC-1 occurred. Taken together, these observations suggest that DAX-1 physically inhibits the transcriptional activity of ERR by competing with PGC-1 for interaction with ERR.
Figure 8 DAX-1 represses PGC-1-enhanced transactivity of ERR by interfering with PGC-1 binding to ERR. (A) 293T cells were co-transfected with 200 ng of mDAX-1 (–467 bp)-Luc, the indicated concentrations of PGC-1 and ERR, and increasing doses of DAX-1 as indicated in the figure. The luciferase activities were measured and normalized against -galactosidase activity. One representative experiment is shown, and similar results were obtained from at least three independent experiments. (B) GST only or GST fused to ERR was isolated from bacterial cultures and immobilized on glutathione-Sepharose beads. In vitro transcribed and translated methionine-labeled PGC-1 and unlabeled DAX-1 (3 and 6 μl) were incubated with purified GST fused receptors or GST alone as indicated in the Figure. The interaction complexes were resolved by 12% denaturing PAGE, and the gels analyzed by autoradiography.
DISCUSSION
ERR is one of the newest members of the orphan nuclear receptor family. Therefore, the target and molecular mechanisms of ERR have not yet been clearly identified. In this study, we demonstrated that ERR acts as a potential regulator of DAX-1, which, in turn represses ERR at a molecular level. This suggests the existence of an autoregulatory loop.
Although the physiological role of DAX-1 has been clearly identified, regulation of the DAX-1 promoter is not fully understood. A previous report demonstrated that SF-1 regulates the DAX-1 promoter through SFRE (7,8). Even though several SFREs are located on the DAX-1 promoter, the functional property of SFRE on the DAX-1 promoter has been characterized only in the region between –129 and –121 bp, which is conserved in both mouse and human (7). As demonstrated in Figure 1, ERRE1 is identical to sequences previously identified as the SFRE. We also characterized a new ERRE2 that was not previously identified as an SF-1RE. Although both ERRE1 and ERRE2 are required for the binding of ERR, the data from mutation analysis suggest that ERRE1 is more responsible for ERR transactivation than ERRE2 (Figures 3 and 4). In the human DAX-1 promoter, only ERRE1 is present. However, SF-1 can still activate the DAX-1 promoter. This demonstrates that ERRE1 is critical for transactivarion (7). Although all three members (, ? and ) of the ERR family bound to the same DNA response element (data not shown), only ERR was able to activate the DAX-1 promoter (Figure 2). This suggests that DNA-binding is not completely responsible for transactivation.
The basal transcriptional activity of ERR and ? is relatively low compared to the high activity of ERR. This suggests that the basal transcriptional mechanism governing ERR and ? may be distinct from that of ERR. A previous report demonstrated that ERR is constitutively active, because ERR forms a complex with coactivators, such as SRC-1 (28). It has been suggested that, depending on the specific co-regulator recruitment, transcriptional activity can be very different. Furthermore, since NRs can differentially modulate target promoters through the recruitment of different coregulators (29,30), ERR may regulate target promoters, such as DAX-1, in a distinctly unique manner, compared to ERR and ?. Further investigation is necessary to determine whether ERR regulation of target promoters differs with co-regulator recruitment.
Until recently, the ligand for ERR was unknown, although 4-OHT was reported to be an inverse agonist of ERR (14,15). Despite a recent report demonstrating that phenolic acyl hydrazone is a selective agonist of ERR? and ERR, further study will be required to prove that phenolic acyl hydrazone is a bona fide activator of ERR? and (31). 4-OHT is widely used in the treatment of breast cancer and deactivates ERR which was proposed as a favorable biomarker in human breast cancer (25). As demonstrated in Figure 5, both ERR and DAX-1 are expressed in breast cancer cells and tissues. Furthermore, ERR regulated the gene expression of DAX-1 in a positive manner, demonstrating that the DAX-1 promoter is a potential target of ERR in vivo. Although the function of ERR and DAX-1 in breast cancer is not completely understood, we speculate that there is both a positive and negative relationship between DAX-1 and ERR, since DAX-1 acts as both a corepressor of ER and ERR and as a regulator of the DAX-1 promoter. Recent reports indicated that ERR regulates the ERR promoter via a conserved hormone response element and epidermal growth factor signaling which affects ERR transactivation in breast cancer cells (32,33). Since ERR, ERR and PGC-1 are co-expressed in breast cancer, their physiological function may be significantly correlated in breast cancer tissues.
It has been reported that DAX-1 interacts directly with various NRs and represses their transcriptional activity (4–6,21,27). Previous studies have demonstrated that DAX-1 uses its N-terminal repeating region, defined as the LXXLL-containing interaction domain, for interaction with NRs (6). In this study, we demonstrated that the LBD-like region of DAX-1 interacts with ERR. We also found that DAX-1 interacts with ERR and ? (data not shown). The AF-2 domain of ERR was crucial for the interaction with DAX-1 and acted as a binding surface for competition with the coactivator PGC-1 (Figure 7). Depending on the binding region of the nuclear receptor with the corepressor, the repressive mechanism of the corepressor can be quite different. When the DBD of the nuclear receptor is involved in interaction with the corepressor, the DNA-binding of the nuclear receptor is blocked by the corepressor (34). Inversely, if the AF-2 domain of the nuclear receptor interacts with the corepressor, the competition between coactivators and the corepressor occurs on the AF-2 surface of the nuclear receptor (35). In the case of ERR, the AF-2 domain interacts with DAX-1 and acts as a binding surface for interaction with the coactivator PGC-1 (Figure 8).
Previously, we reported that the SHP promoter is a target of ERR (16). In the current study, we suggest that DAX-1 is a target of ERR. As demonstrated in Figure 6, DAX-1 repressed the transcriptional activity of ERR on the DAX-1 and SHP promoters. In addition, SHP also inhibited ERR on the DAX-1 and SHP promoters. It is very interesting that ERR regulates corepressor proteins, such DAX-1 and SHP, which act as repression modulators. Although both SHP and DAX-1 are involved in the regulation of the corepressor-mediated transactivation of NRs, their physiological functions are quite different from each other. While DAX-1 plays a role in steroidogenic tissues, SHP functions mainly in the liver (1). We demonstrated that there is a correlation between ERR and DAX-1 in breast cancer cells (Figure 5). In the case of SHP, we confirmed using adenovirus that ERR activated SHP gene expression in the pancreas, but not in the liver (data not shown). This demonstrated that tissue-specific target gene expression is modulated by ERR. DAX-1 is also differentially regulated by ERR, depending on the target tissue. The modulation of DAX-1 and SHP by ERR in vivo needs to be examined further, in order to identify the physiological roles of ERR, SHP and DAX-1.
Collectively, as depicted in Figure 9, we hypothesize that ERR activates the DAX-1 promoter, which in turn inhibits ERR through coactivator competition. In this study, we demonstrated that DAX-1 is a potential target of ERR. Our study provides a new insight into understanding the basic mechanisms of DAX-1 and ERR, and is expected to provide clues for the identification of the functions of ERR.
Figure 9 Proposed model of ERR and DAX-1 action. ERR directly binds to the DAX-1 (SHP) promoter, resulting in an increase in DAX-1 (SHP) gene expression. In turn, DAX-1 (SHP) represses ERR through a direct interaction with its AF-2 domain. DAX-1 (SHP) competes with the coactivator for binding to the AF-2 surface of ERR and homeostasis between ERR and DAX-1 (SHP) is maintained through this autoregulatory loop.
ACKNOWLEDGEMENTS
This work was supported by a National Research Laboratory grant (M1-0500-4705J-4710), a KRF grant (C00126 ) and a Marine Bio21 grant. Y.Y.P. and S.W.A. are recipients of the Brain Korea 21 program. Funding to pay the Open Access publication charges for this article was provided by Brain Korea 21 program.
REFERENCES
Giguere, V. (1999) Orphan nuclear receptors: from gene to function Endocr. Rev., 20, 689–725 .
Muscatelli, F., Strom, T.M., Walker, A.P., Zanaria, E., Recan, D., Meindl, A., Bardoni, B., Guioli, S., Zehetner, G., Rabl, W., et al. (1994) Mutations in the DAX-1 gene give rise to both X-linked adrenal hypoplasia congenita and hypogonadotropic hypogonadism Nature, 372, 672–676 .
Zanaria, E., Muscatelli, F., Bardoni, B., Strom, T.M., Guioli, S., Guo, W., Lalli, E., Moser, C., Walker, A.P., McCabe, E.R., et al. (1994) An unusual member of the nuclear hormone receptor superfamily responsible for X-linked adrenal hypoplasia congenita Nature, 372, 635–641 .
Zhang, H., Thomsen, J.S., Johansson, L., Gustafsson, J.A., Treuter, E. (2000) DAX-1 functions as an LXXLL-containing corepressor for activated estrogen receptors J. Biol. Chem., 275, 39855–39859 .
Crawford, P.A., Dorn, C., Sadovsky, Y., Milbrandt, J. (1998) Nuclear receptor DAX-1 recruits nuclear receptor corepressor N-CoR to steroidogenic factor 1 Mol. Cell. Biol., 18, 2949–2956 .
Suzuki, T., Kasahara, M., Yoshioka, H., Morohashi, K., Umesono, K. (2003) LXXLL-related motifs in Dax-1 have target specificity for the orphan nuclear receptors Ad4BP/SF-1 and LRH-1 Mol. Cell. Biol., 23, 238–249 .
Yu, R.N., Ito, M., Jameson, J.L. (1998) The murine Dax-1 promoter is stimulated by SF-1 (steroidogenic factor-1) and inhibited by COUP-TF (chicken ovalbumin upstream promoter-transcription factor) via a composite nuclear receptor-regulatory element Mol. Endocrinol., 12, 1010–1022 .
Vilain, E., Guo, W., Zhang, Y.H., McCabe, E.R. (1997) DAX1 gene expression upregulated by steroidogenic factor 1 in an adrenocortical carcinoma cell line Biochem. Mol. Med., 61, 1–8 .
Giguere, V. (2002) To ERR in the estrogen pathway Trends Endocrinol. Metab., 13, 220–225 .
Horard, B. and Vanacker, J.M. (2003) Estrogen receptor-related receptors: orphan receptors desperately seeking a ligand J. Mol. Endocrinol., 31, 349–357 .
Sladek, R., Bader, J.A., Giguere, V. (1997) The orphan nuclear receptor estrogen-related receptor alpha is a transcriptional regulator of the human medium-chain acyl coenzyme A dehydrogenase gene Mol. Cell. Biol., 17, 5400–5409 .
Lu, D., Kiriyama, Y., Lee, K.Y., Giguere, V. (2001) Transcriptional regulation of the estrogen-inducible pS2 breast cancer marker gene by the ERR family of orphan nuclear receptors Cancer Res., 61, 6755–6761 .
Yang, C., Zhou, D., Chen, S. (1998) Modulation of aromatase expression in the breast tissue by ERR alpha-1 orphan receptor Cancer Res., 58, 5695–5700 .
Tremblay, G.B., Bergeron, D., Giguere, V. (2001) 4-Hydroxytamoxifen is an isoform-specific inhibitor of orphan estrogen-receptor-related (ERR) nuclear receptors beta and gamma Endocrinology, 142, 4572–4575 .
Coward, P., Lee, D., Hull, M.V., Lehmann, J.M. (2001) 4-Hydroxytamoxifen binds to and deactivates the estrogen-related receptor gamma Proc. Natl Acad. Sci. USA, 98, 8880–8884 .
Sanyal, S., Kim, J.Y., Kim, H.J., Takeda, J., Lee, Y.K., Moore, D.D., Choi, H.S. (2002) Differential regulation of the orphan nuclear receptor small heterodimer partner (SHP) gene promoter by orphan nuclear receptor ERR isoforms J. Biol. Chem., 277, 1739–1748 .
Huss, J.M., Kopp, R.P., Kelly, D.P. (2002) Peroxisome proliferator-activated receptor coactivator-1alpha (PGC-1alpha) coactivates the cardiac-enriched nuclear receptors estrogen-related receptor-alpha and -gamma. Identification of novel leucine-rich interaction motif within PGC-1alpha J. Biol. Chem., 277, 40265–40274 .
Huppunen, J. and Aarnisalo, P. (2004) Dimerization modulates the activity of the orphan nuclear receptor ERRgamma Biochem. Biophys. Res. Commun., 314, 964–970 .
Razzaque, M.A., Masuda, N., Maeda, Y., Endo, Y., Tsukamoto, T., Osumi, T. (2004) Estrogen receptor-related receptor gamma has an exceptionally broad specificity of DNA sequence recognition Gene, 340, 275–282 .
Lee, Y.K., Parker, K.L., Choi, H.S., Moore, D.D. (1999) Activation of the promoter of the orphan receptor SHP by orphan receptors that bind DNA as monomers J. Biol. Chem., 274, 20869–20873 .
Song, K.H., Park, Y.Y., Park, K.C., Hong, C.Y., Park, J.H., Shong, M., Lee, K., Choi, H.S. (2004) The atypical orphan nuclear receptor DAX-1 interacts with orphan nuclear receptor Nur77 and represses its transactivation Mol. Endocrinol., 18, 1929–1940 .
Park, Y.Y., Kim, H.J., Kim, J.Y., Kim, M.Y., Song, K.H., Cheol Park, K., Yu, K.Y., Shong, M., Kim, K.H., Choi, H.S. (2004) Differential role of the loop region between helices H6 and H7 within the orphan nuclear receptors small heterodimer partner and DAX-1 Mol. Endocrinol., 18, 1082–1095 .
Knutti, D., Kressler, D., Kralli, A. (2001) Regulation of the transcriptional coactivator PGC-1 via MAPK-sensitive interaction with a repressor Proc. Natl Acad. Sci. USA, 98, 9713–9718 .
Choi, Y.K., Kim, Y.J., Park, H.S., Choi, K., Paik, S.G., Lee, Y.I., Park, J.G. (2003) Suppression of glomerulosclerosis by adenovirus-mediated IL-10 expression in the kidney Gene Ther., 10, 559–568 .
Ariazi, E.A., Clark, G.M., Mertz, J.E. (2002) Estrogen-related receptor alpha and estrogen-related receptor gamma associate with unfavorable and favorable biomarkers, respectively, in human breast cancer Cancer Res., 62, 6510–6518 .
Conde, I., Alfaro, J.M., Fraile, B., Ruiz, A., Paniagua, R., Arenas, M.I. (2004) DAX-1 expression in human breast cancer: comparison with estrogen receptors ER-alpha, ER-beta and androgen receptor status Breast Cancer Res., 6, R140–R148 .
Holter, E., Kotaja, N., Makela, S., Strauss, L., Kietz, S., Janne, O.A., Gustafsson, J.A., Palvimo, J.J., Treuter, E. (2002) Inhibition of androgen receptor (AR) function by the reproductive orphan nuclear receptor DAX-1 Mol. Endocrinol., 16, 515–528 .
Greschik, H., Wurtz, J.M., Sanglier, S., Bourguet, W., van Dorsselaer, A., Moras, D., Renaud, J.P. (2002) Structural and functional evidence for ligand-independent transcriptional activation by the estrogen-related receptor 3 Mol. Cell, 9, 303–313 .
Guan, H.P., Ishizuka, T., Chui, P.C., Lehrke, M., Lazar, M.A. (2005) Corepressors selectively control the transcriptional activity of PPARgamma in adipocytes Genes Dev., 19, 453–461 .
Sanyal, S., Matthews, J., Bouton, D., Kim, H.J., Choi, H.S., Treuter, E., Gustafsson, J.A. (2004) Deoxyribonucleic acid response element-dependent regulation of transcription by orphan nuclear receptor estrogen receptor-related receptor gamma Mol. Endocrinol., 18, 312–325 .
Zuercher, W.J., Gaillard, S., Orband-Miller, L.A., Chao, E.Y., Shearer, B.G., Jones, D.G., Miller, A.B., Collins, J.L., McDonnell, D.P., Willson, T.M. (2005) Identification and structure-activity relationship of phenolic acyl hydrazones as selective agonists for the estrogen-related orphan nuclear receptors ERRbeta and ERRgamma J. Med. Chem., 48, 3107–3109 .
Liu, D., Zhang, Z., Teng, C.T. (2005) Estrogen-related receptor-gamma and peroxisome proliferator-activated receptor-gamma coactivator-1alpha regulate estrogen-related receptor-alpha gene expression via a conserved multi-hormone response element J. Mol. Endocrinol., 34, 473–487 .
Barry, J.B. and Giguere, V. (2005) Epidermal growth factor-induced signaling in breast cancer cells results in selective target gene activation by orphan nuclear receptor estrogen-related receptor alpha Cancer Res., 65, 6120–6129 .
Kim, J.Y., Kim, H.J., Kim, K.T., Park, Y.Y., Seong, H.A., Park, K.C., Lee, I.K., Ha, H., Shong, M., Park, S.C., et al. (2004) Orphan nuclear receptor small heterodimer partner represses hepatocyte nuclear factor 3/Foxa transactivation via inhibition of its DNA binding Mol. Endocrinol., 18, 2880–2894 .
Lee, Y.K., Dell, H., Dowhan, D.H., Hadzopoulou-Cladaras, M., Moore, D.D. (2000) The orphan nuclear receptor SHP inhibits hepatocyte nuclear factor 4 and retinoid X receptor transactivation: two mechanisms for repression Mol. Cell. Biol., 20, 187–195 .(Yun-Yong Park, Seung-Won Ahn, Hye-Jin Ki)
*To whom correspondence should be addressed. Tel: +82 62 530 0503; Fax: +82 62 530 0500; Email: hsc@chonnam.ac.kr
ABSTRACT
The estrogen receptor-related receptor gamma (ERR/ERR3/NR3B3) is a member of the nuclear receptor superfamily that activates transcription in the absence of ligand. However, the detailed mechanism of gene regulation by ERR is not fully understood. In this study we have found that the orphan nuclear receptor ERR activates the DAX-1 promoter, which, in turn, represses transactivation by ERR. Serial deletions of mouse DAX-1 (mDAX-1) gene promoter have revealed that the region responding to ERR is located between –129 and –121 bp and –334 and –326 bp. Gel shift assays and chromatin immunoprecipitation (ChIP) assays demonstrated that ERR binds directly to the mDAX-1 promoter. Site-directed mutagenesis results demonstrated that ERRE1 (–129 to –121 bp) is more important than ERRE2 (–334 to –326 bp) which is not conserved in the human DAX-1 promoter. In addition, adenovirus-mediated overexpression of ERR induced DAX-1 gene expression in MCF-7 breast cancer cells that co-expressed ERR and DAX-1. Moreover, yeast two-hybrid and glutathione S-transferase (GST)-pull down assays demonstrated that DAX-1 physically interacted with ERR and inhibited ERR transactivation, and that this interaction was dependent on the AF-2 domain of ERR. In addition, in vitro competition assays showed that DAX-1 inhibited PGC-1 mediated ERR transactivation, via competition between these two factors for the AF-2 binding domain. We thus propose a novel autoregulatory loop that controls DAX-1 gene expression by ERR.
INTRODUCTION
Nuclear receptors (NRs) are ligand-dependent transcription factors that bind to specific DNA target sequences through their highly conserved DNA-binding domain (DBD) and activate transcription through interaction with coactivators. The latter property is dependent on their moderately conserved ligand-binding domain (LBD) and in particular requires the extreme C-terminal part of it, the so-called AF-2 region (1).
Among NRs, DAX-1 (dosage-sensitive sex reversal, adrenal hypoplasia critical region, on chromosome X, gene 1) (NR0B1) is an atypical orphan nuclear receptor, since it does not contain a DBD (2,3). Instead, the N-terminal half of DAX-1 contains a unique region mainly composed of three repeats, each of which is 65–67 amino acids in length. This region contains LXXLL motif-like sequences (where L and M are leucine and methionine, respectively, and X can be any amino acid) that are necessary for interaction with estrogen receptors (ERs) (4). Since the C-terminal half of DAX-1 containing the putative LBD functions as a repressor domain through interaction with corepressors (NCoR) (5), two mechanisms have been proposed to function cooperatively in the repression of transcription exerted by DAX-1. According to the proposition, DAX-1 would both block the binding of coactivators to NRs and recruit corepressors with its C-terminal half (4–6). The DAX-1 gene was identified through a search for a gene linked to adrenal hypoplasia congenita, a disease affecting the normal development of the adrenal cortex and which is often associated with hypogonadotropic hypogonadism (2,3). Mouse and human DAX-1 promoters have been cloned and characterized. For example, orphan nuclear receptor steroidogenic factor-1 (SF-1) regulates DAX-1 promoter, and the SF-1 response element is located between –129 and –121 bp on the mouse promoter and between –135 and –143 bp on the human promoter (7,8).
Estrogen-related receptors (ERRs-alpha, -beta and -gamma), are orphan NRs closely related to ERs, with which they share identical target response elements and coregulatory proteins (9,10). However, ERRs do not respond to the classical ER ligand. Previous studies have demonstrated that MCAD, pS2 and cyp19 are targets of ERR (11–13). Among the ERR family, ERR, the newest member of the subfamily, is a constitutively active nuclear receptor, and little is known about its in vivo functions (9,10). Previous studies have demonstrated that 4-hydroxytamoxifen (4-OHT) directly binds to and deactivates ERR, suggesting that 4-OHT is an inverse agonist of ERR (14,15). However, the deactivation mechanism of ERR by 4-OHT requires further elucidation. It has been reported that coactivators such as GRIP1 or PGC-1, interact directly with ERR and activate its transcriptional activity (16,17). ERR acts as a monomer and activates transcription constitutively. It has also been suggested in recent reports, that ERR forms dimers through its LBD. In contrast to this, the related receptor, ERR, inhibits the activities of both ERR and ERR through heterodimerization (18). ERR exhibits significant affinities for binding to a wide spectrum of sequences, including inverted and direct repeat sequences composed of AGGTCA half-sites with differing spacings, as well as a monovalent motif of the same sequence carrying an extra T(C/G)A trinucleotide on the 5' side (19). It has been reported that ERR binds to and transactivates both estrogen response elements and SF-1 response elements (SF-1RE). Previous reports have suggested that several monomeric binding partners, such as SF-1, liver receptor homolog-1 (LRH-1) and ERR, can activate and bind directly to the small heterodimer partner (SHP) promoter through SF-1RE (16,20).
DAX-1 is closely related to SHP, which neither contains a DNA-binding domain nor functions as a corepressor (2,3). In addition, both the DAX-1 and SHP promoters are regulated by the monomer binding partner SF-1. Thus, since the DAX-1 promoter contains several SF-1RE, we predicted that ERR could also activate the DAX-1 promoter. Transient transfection studies demonstrated that ERR had a preferential effect on the DAX-1 promoter. The results of deletion and site-directed mutagenesis have revealed that two ERR regulatory elements (ERREs) are responsible for the ERR-mediated activation of the DAX-1 promoter. The results of electrophoretic mobility shift assay (EMSA) and chromatin immunoprecipitation (ChIP) assays strongly suggest that ERR binds directly to the DAX-1 promoter. In addition, the results of overexpression and ERR knock-down experiments demonstrated that ERR regulates DAX-1 gene expression in MCF-7 breast cancer cells and that ERR and DAX-1 are co-expressed in breast cancer tissue. The expression of DAX-1 inhibits ERR transactivation of its own promoter and DAX-1 competes directly with the coactivator, PGC-1, for binding to ERR. Taken together, these results suggest that ERR belongs to a new autoregulatory loop that governs DAX-1 gene expression.
MATERIALS AND METHODS
Plasmids and DNA construction
pcDNA3/HA-ERR, ?, and (AF-2), pcDNA3/HA SF-1, pcDNA3/HA DAX-1, pcDNA3/HA PGC-1, pGEX4T-1 DAX-1, B42 ERR, B42 AF-2 ERR, Lex A DAX-1, and pcDNA3 PGC-1 were described previously (21–23). To generate the –467 bp DAX-1 promoter region, genomic DNA was isolated from a mouse testicular Leydig cell line, K28 and used as template for PCR. The PCR product was cloned into the pGL3-basic plasmid (Promega) between the MluI and XhoI sites. mDAX-1/Luc (–325 and –121 bp) were made by PCR using the primer and cloned into pGL3 at MluI and XhoI. Mutant reporters, mtERRE1/Luc, mtERRE2/Luc and mtERRE1&2/Luc, were constructed by site-directed mutagenesis of the mDAX-1 promoter –467 bp/Luc using the primer. The mutated sequences are shown in Figure 4A. All clones were verified by sequencing analysis.
Figure 4 The ERRE1 (–129 to –121 bp) plays a crucial role in the transactivation of the mDAX-1 promoter by ERR. (A) Schematic diagrams of wild-type and mutant mDAX-1 promoter constructs from –467 to +43 bp as indicated. The putative ERR binding sites are indicated as ERRE1 and ERRE2. (B) ERRE1 is essential for the activation of the mDAX-1 promoter by ERR. 293T cells were co-transfected with 200 ng of wild-type or mutant mDAX-1 promoter constructs together with 200 ng of ERR expression plasmid. Two days after transfection, cells were harvested for luciferase and -galactosidase assays. The representative results are expressed as -fold activation (n-fold) over the value obtained with vector alone with the error bars as indicated.
Cell culture and transient transfection assay
Human embryonic kidney (293T), mouse fibriblast (NIH3T3) and human breast cancer (MCF-7) cells were maintained with DMEM in the presence of 10% fetal bovine serum and antibiotics (Gibco) in humidified air containing 5% CO2 at 37°C. Cells were transfected, using Superfect and Effectene reagents (Qiagen Inc., Germany), according to the manufacturer's instructions. The total DNA used in each transfection was adjusted to 1 μg by adding the appropriate amount of pcDNA3 vector. Approximately 40–48 h post-transfection, cells were harvested and luciferase activity was measured and normalized to ?-galactosidase activity. Experiments were performed three times in duplicate.
Glutathione S-transferase (GST)-pull down assay
The indicated GST-fusion proteins or GST protein alone were expressed in Escherichia coli BL21 (DE3) pLys cultures by adding 0.2 mM isopropyl-?-D-thiogalactopyranoside (IPTG), and the cells were extracted. GST-fusion proteins were pre-bound with a 30 μl aliquot of glutathione-Sepharose beads. The beads were incubated with transcribed and translated methionine-labeled proteins for 3–4 h at 4°C. The beads were then washed three times with the washing buffer, and analyzed by SDS–PAGE and protein bands were visualized by using a phosphorimaging analyzer. Translated proteins in vitro (10%) were used as input.
In vitro translation
ERR and ERR AF-2 were transcribed and translated in vitro using a coupled rabbit reticulocyte system (Promega Corp., Madison, WI) in the presence of methionine according to the manufacturer's instructions. The translated proteins were analyzed by 10% SDS–PAGE and visualized by autoradiography.
Yeast two-hybrid interaction assay
Yeast two-hybrid interaction assays were performed as described previously (22). Briefly, LexA only or LexA fused plasmids, and B42 only or B42 fused plasmids, were transformed into Saccharomyces cerevisae EGY48, containing the galactosidase reporter plasmid 8H18-34, and transformants were selected on plates with the appropriate selection markers. The galactosidase assay on plates was carried out as described previously (22).
Gel mobility shift assays
Double-stranded ERRE oligonucleotides were labeled by filling-in with dCTP, using the Klenow fragment of DNA polymerase I. The oligonucleotide sequences are depicted in Figure 3A. Bacterially expressed GST only and GST fused ERR were purified using GST-Sepharose beads (Amersham Bioscience, Inc.). Gel mobility shift assays (20 μl) contained 10 mM Tris (pH 8.0), 40 mM KCl, 0.05% Nonidet P-40, 6% glycerol, 1 mM DTT, and 1 μg of poly (dI-dC). Either 2 μg of GST only, or GST fused ERR purified proteins, were used in each reaction. Competitor oligonucleotides were included at a 50- to 100-fold molar excess as indicated in the Figure legends. After 15 min incubation on ice, 10 000 c.p.m. of labeled oligonucleotide probe was added and incubation was continued for another 15 min. DNA–protein complexes were analyzed by 5% PAGE in 1x TBE. Gels were dried and analyzed by autoradiography.
Figure 3 ERR binds to ERRE in the DAX-1 promoter. (A) The sequences of wild-type (nt –126 to –106 and –337 to –317) and mutant oligonucleotides used for the gel retardation assays are depicted. Italic letters indicate the substituted nucleotides in ERRE1 and ERRE2. (B) Gel mobility shift and unlabeled competition assays were performed with the 32P-labeled probes, and recombinant ERR (2 μg) protein as indicated. Unlabeled oligonucleotides were added as competitors at 50- to 100-fold molar excess where indicated. (C) ChIP assay. 293T cells were transfected with expression vectors for HA epitope alone or HA-ERR with –467 bp DAX-1-Luc. Soluble chromatin from these cells was prepared and immunoprecipitated with monoclonal antibody against HA (lanes 3 and 4). The –350 –100 bp fragment (250 bp) contains the ERR binding site and 10% of the soluble chromatin used in the reaction was used as input (lanes 1 and 2). PCR was performed as described in Figure 3C.
ChIP assay
The ChIP assay was performed according to the instructions of the ChIP Assay Kit (Upstate, USA). 293T cells were transfected with 1 μg of pcDNA3 HA-ERR or pcDNA3 HA and pGL3 –467 bp mDAX-1 using Effectene reagent (Qiagen Inc., Germany). The cells were fixed with formaldehyde 48 h after transfection and harvested. For immunoprecipitation, anti-HA antibody (Roche Molecular Biochemicals) was used. Final DNA extractions were amplified using pairs of primers that cover the ERR response region within the mDAX-1 promoter (nt –350 to –100 bp). The primers used for PCR were as follows: 5'-AAAGGAAAT AAGTTAGAGGTCAGAG-3' (Forward) and 5'-AGCGCGTCCGCCTCCTCCTCTTGGA-5' (Reverse). MCF-7 cells (Figure 5C) were transfected with 1 μg of pcDNA3/HA or pcDNA3/HA-ERR. Soluble chromatin was immunoprecipitated with anti-HA antibody. Final DNA extractions were amplified by 30 cycles of PCR using pairs of primers that cover the ERR response region within the hDAX-1 promoter (nt –260 to +6 bp). The primers used for PCR were as follows: 5'-CAG CAT CCA GGC GCT CGC TCT CC-3' (Forward) and 5'-TTC TGC CCA GTG GCT GCC TCC TGG-5' (Reverse).
Figure 5 Co-expression of ERR and DAX-1 in breast cancer cell. (A) MCF-7 cells were infected with adenoviral vector expressing ERR (100 pfu/cells). Total RNA was isolated from cells and analyzed by RT–PCR. (B) The effect of siRNA-ERR on the mRNA level of DAX-1. Endogenous ERR gene expression was inhibited by transfection with a 21 nt RNA duplex siRNA-ERR/I in MCF-7 cells. The effects of siRNAs on ERR and DAX-1 expression were assayed by realizing RT–PCR for ERR, DAX-1, and ?-actin as a control. (C) MCF-7 cells were transfected with HA or HA-ERR. The –260 +6 bp fragment (266 bp) contains the ERR binding site and 10% of the soluble chromatin used in the reaction was used as input (lanes 1 and 2). PCR was performed as described in Figure 5C. (D) DAX-1 and ERR expression in human breast cancer cells. The DAX-1-expressing tumor cells (top left) were also positive for ERR (top right). The DAX-1-negative cancer cells were also negative for ERR (bottom left and right) (original magnification, 200x).
Preparation of recombinant adenovirus
The recombinant adenovirus was prepared as described previously (24). In brief, the cDNA encoding ERR was cloned into a pAd-YC2 shuttle vector, which (under the control of the cytomegalovirus promoter) contains a bovine growth hormone polyadenylation signal sequence. For homologous recombination, a pAd-YC2 shuttle vector (5 μg) and a rescue vector, pJM17 (5 μg), were co-transfected into human embryonic kidney 293 cells. To purify pure plaques, cell culture supernatant was serially diluted into serum-free media and incubated with 293 cells at 37°C for 1 h. An equal volume mixture of 2x medium and 1% agarose was overlaid on the 293 cells. After 7 days, plaques that were well isolated were purified further and propagated in 293 cells and screened by PCR, using upstream primers derived from the cytomegalovirus promoter, and downstream primers from the bovine growth hormone polyadenylation sequence. Then, the recombinants were amplified in 293 cells and were purified and isolated using CsCl2 (Sigma). The preparations were collected and desalted, and titers were determined by counting the number of plaques.
Immunohistochemistry
Immunohistochemical analysis was performed with nine human breast cancer tissues. Three sections were cut from the paraffin blocks, deparaffinized antigen retrieval by pressure cooker in 10 mM sodium citrate (pH 6.0) at full power for 4 min. The primary antibodies used were rabbit polyclonal anti-ERR (Abcam, 1:50) and anti-DAX-1 (Santa Cruz, 1:150). Immunohistochemistry was performed using an EnVision kit (DAKO, Carpinteria, CA). After preincubation with blocking serum for 15 min, the primary polyclonal antbody for ERR or DAX-1 was added and incubated for 30 min in a humid chamber followed by a wash with TBS buffer. Slides were then incubated for 30 min with EnVision peroxidase reagent (DAKO, Carpinteria, CA).
The slides were then sequentially incubated with DAB (3,3-diaminobenzidine) chromogen for 5 min, counterstained with Meyer's hematoxylin and mounted. Careful rinses with several changes of TBS were performed between each stage of the procedure. To assess the specificity of the immunoreaction, a negative control without the primary antibody or with pre-immune rabbit serum was performed.
RT–PCR analysis
MCF-7 cells were infected with an adenovirus containing the ERR gene (100 pfu/cells). After viral infection for 24 h, cells were harvested for total RNA isolation using the TRIzol reagent (Invitrogen Life Technologies) according to the manufacturer's protocol. The mRNAs of ERR and DAX-1 were analyzed by RT–PCR. First-strand cDNA was synthesized from 1 μg of total RNA utilizing an anchored oligo(dT) primers and reverse transcriptase. The resulting first-strand cDNA was then amplified to measure mRNA levels of ERR, DAX-1 and ?-actin by realizing 28, 28 and 23 cycles of PCR using specific primers, respectively. The mRNA levels of ?-actin served as an internal control for the RT–PCR analysis. The primers used for PCR of ERR, DAX-1 and ?-actin were as follows: ERR, forward 5'-GACTTGACTCGCCACCTCTC-3' and reverse 5'-GTGGTACCCAGAAGCGATGT-3'; DAX-1, forward 5'-CCGCTTGCAGTTCGA GACTGTG-3' and reverse 5'-CTCATGGTGAACTGCACTACTG-3'; ?-actin, forward 5'-GTCATCACCATTGGCAATGAG and reverse 5'-CGTCATACTCCTGCTTGCTG-3'.
siRNA experiment
The siRNAs for ERR (si-ERR/I and /II) were chemically synthesized (Shamchully Pharm. Co., Ltd), deprotected, annealed, and transfected according to the manufacturer's instructions. MCF-7 cells were transfected with siRNA using Oligofectamine reagent (Invitrogen Life Technologies). After transfection for 48 h, total RNA was isolated for RT–PCR of ERR (28 cycles), DAX-1 (27 cycles) and ?-actin (20 cycles) as a control. The sequences of siRNA are as follows: si-ERR/I, sense 5'-UGG CCA UCA GAA CGG ACU UTT-3'; si-ERR/II sense 5'-CGA AGA CCA GUC CAA AUU ATT-3'.
RESULTS
Identification of ERR as a transcriptional activator of the DAX-1 promoter
SF-1 has been identified as an activator of the DAX-1 promoter (7,8). Since SFRE is also a DNA-binding site for ERR, we speculated that an ERR response element (ERRE) exists in the DAX-1 promoter. As shown in Figure 1A, two ERREs were located in the mDAX-1 promoter; ERRE1 (–129 to –121 bp), which was previously identified as an SF-1RE, and a new ERRE2 (–334 to –326 bp). Interestingly, ERRE1 is conserved in both mouse and human promoters, whereas ERRE2 only exists in the mouse promoter (Figure 1B).
Figure 1 ERREs on the DAX-1 promoter. (A) Mouse DAX-1 promoter sequences from –467 to +43 relative to the putative transcription start site. The adenine at the transcription start site (arrow) and TATA-box (underline) are designated +1 and TATA-box. The translation initiation codon (ATG) is underlined with the encoded amino acids shown below the codons. Putative ERREs are lined above the respective sequences. (B) Sequence alignment of ERRE1 and ERRE2 within the mouse and human DAX-1 promoters. The numbers indicate the nucleotide positions: –129 to –21 bp and –135 to –127 bp in mouse and human ERRE1, respectively, and –334 to –326 bp and –347 to –339 bp in mouse and human ERRE2, respectively.
In order to determine whether ERRs regulate the DAX-1 promoter, we performed transient transfection assays using the luciferase reporter gene driven by a –467 bp region from mouse DAX-1 promoter. As shown in Figure 2A and B, there was strong activation by ERR in 293T cells and NIH3T3 cells, whereas ERR and ERR? did not demonstrate any significant effect. As expected, SF-1, the positive control for the transfection assays, activated the DAX-1 gene promoter significantly. These results indicated that the DAX-1 gene is a specific target of ERR, but not of ERR or ERR?. In order to investigate the sequences that were required for ERR mediated activation, a series of 5' deletions of the DAX-1 promoter were constructed, as shown in Figure 2C. Since the DAX-1 promoter contains two ERREs, we designed deletion constructs both with and without an ERRE. Activation by ERR was decreased when the DAX-1 promoter was deleted up to –325 bp, as demonstrated in Figure 2D. Furthermore, the activation of the DAX-1 promoter by ERR, was almost lost with the –121 bp DAX-1 construct. These results suggest that ERREs are required for ERR response.
Figure 2 ERR activates DAX-1 gene promoter. (A and B) Transcriptional activation of the mDAX-1-promoter by ERR, but not by ERR and ?. 293Tcells (A) and NIH3T3 cells (B) cells were transiently transfected with 200 ng of indicated reporter gene together with the indicated concentrations of HA-pcDNA3-ERR , ?, and SF-1. (C) Schematic diagram of the mDAX-1 promoter deletion constructs cloned in the pGL3-basic vector. (D) Mapping the sequences required for activation of the mDAX-1 promoter by ERR. 293T cells were transfected with 200 ng of each reporter plasmid and ERR expression plasmid as indicated. Approximately 40 h after transfection, the cells were harvested and luciferase activity was measured and normalized against -galactosidase activity. One representative experiment is shown. All values represent the mean of duplicate samples, and similar results were obtained in at least three independent experiments. The representative results were expressed as -fold activation (n-fold) over the value obtained with vector alone with the error bars as indicated.
Determination of ERR binding regions
In order to clarify the specific binding of ERR to ERREs, gel mobility shift assays were performed. As demonstrated in Figure 3A, we designed both wild-type and mutant probes that corresponded to ERRE1 and ERRE2. Previously, the TNAAGGTCA sequence was shown to be a bona fide ERRE. Within the sequence, the GG dimer is pivotal for ERR protein binding. Thus, we substituted GG for the TT sequence. Figure 3B demonstrates that ERR formed specific complexes with ERRE1 and ERRE2; a 50- or 100-fold molar excess of unlabeled ERRE1 and ERRE2 competed strongly with the DNA–protein complex. In order to verify that ERR binds to the DAX-1 promoter in vivo, we performed a ChIP assay using PCR primers that flanked both ERRE1 and ERRE2 (–350 to –100 bp). 293T cells were transfected with the DAX-1 promoter, either with hemagglutinin (HA) only or with HA-tagged ERR. As demonstrated in Figure 3C, a 250 bp PCR product (–350 to –100 bp) was observed in cells that were transfected with the expression vector for HA-ERR but not with the vector expressing only the HA epitope. This indicated that HA-ERR formed a specific complex with the DAX-1 promoter in vivo. Taken together, this suggests that ERR regulates the DAX-1 promoter by direct binding to ERRE.
ERRE1 (–129 to –121 bp) is essential for transactivation of the DAX-1 promoter by ERR
In order to verify the functional significance of ERR binding sites in the DAX-1 promoter, site-directed mutagenesis was performed on the –467 bp DAX-1 promoter using the primers as shown in Figure 4A. Wild-type –467 bp/Luc and several mutant constructs were transiently transfected with ERR into 293T cells. The ERRE1 mutation (mtERRE1/Luc) and double ERRE1 and ERRE2 mutations (mtERRE1&2/Luc) completely abolished the ERR-mediated transactivation of the DAX-1 promoter whereas the ERRE2 mutation (mtERRE2/Luc) had a slight effect (Figure 4B). These results indicated that ERRE1, rather than ERRE2, was essential for the activation of the DAX-1 promoter by ERR. Thus, we conclude that ERRE1 plays a major role in the activation of the DAX-1 promoter by ERR.
Induction of DAX-1 gene expression by ERR in breast cancer
Previous reports have demonstrated that ERR is a favorable biomarker and, possibly, an indicator of hormonal sensitivity in breast cancer. In addition, DAX-1 has also been detected in breast cancer tissue (25,26). In order to investigate whether ERR can directly regulate DAX-1 gene expression in breast cancer cells, we introduced ERR by adenovirus infection into MCF-7 cells. As demonstrated in Figure 5A, a basal expression of DAX-1 and ERR was detected in MCF-7 cells, and DAX-1 mRNA was induced by overexpression of Ad-ERR 24 h after infection. Since ERR is basally expressed in MCF-7 cells, we further tested whether DAX-1 gene expression is down-regulated by performing knock-down of endogenous ERR expression using small interfering RNA (si-ERR/I or II) in MCF-7 cells. si-ERR/I, but not si-ERR/II significantly reduced endogenous ERR expression (Figure 5B). DAX-1 mRNA level was decreased by knock-down of ERR by si-ERR/I, but not by si-ERR/II, suggesting that endogenous ERR can activate DAX-1 expression. In addition, CHIP assays were performed to confirm the direct binding of ERR to the endogenous DAX-1 promoter in MCF-7 cells. As shown in Figure 5C, ERR can bind to the DAX-1 promoter in MCF-7 cells (Figure 5C). Next, we analyzed for co-expression of DAX-1 and ERR in human breast cancer tissues, using immunohistochemistry (Figure 5D). Nuclear immunoreactivity of DAX-1 was detected in breast cancer tissues. The overexpression of ERR was observed in three out of nine cases. All of the ERR-positive cancers co-expressed the DAX-1 protein (Figure 5D, upper part). Two cancer tissues that were DAX-1-negative were also negative for ERR. No staining was observed in the negative controls using either pre-immune rabbit serum or by omitting the primary antibody (data not shown). These results suggest that ERR can stimulate DAX-1 gene expression in human breast cancer cells.
DAX-1 represses ERR transactivation of its own promoter
DAX-1 interacts with several orphan NRs (ER, AR, LRH-1, SF-1 and Nur77) and represses their transcriptional activity (4–6,21,27). We investigated whether DAX-1 can also repress ERR mediated transactivation of its own promoter. The results of transient transfections demonstrated that DAX-1 inhibits ERR activity of the DAX-1 promoter in a dose-dependant manner (Figure 6A). Since ERR also activates the SHP promoter, and SHP represses ERR transactivation of its own promoter, we examined whether DAX-1 represses the ERR-mediated transactivation of the SHP promoter or vice versa. As demonstrated in Figure 6B and C, DAX-1 and SHP repressed ERR transactivation of both the SHP promoter and the DAX-1 promoter. These results suggest that both DAX-1 and SHP are regulated by ERR and repress the transcriptional activity of ERR through an autoregulatory loop. Since Figure 6A shows that DAX-1 repressed the transcriptional activity of ERR, we examined whether DAX-1 can directly interact with ERR, by using a GST-pull down and yeast two-hybrid interaction assays. As demonstrated in Figure 7A and B, DAX-1 interacted with ERR. Since repression or activation is dependant on the AF-2 domain, we deleted this domain in ERR. As expected, DAX-1 could not interact with ERR AF-2, which demonstrates that the AF-2 domain of ERR is crucial for the repressive effect of DAX-1 (Figure 7A). In the case of DAX-1, an LBD-like region was involved in the interaction with ERR (Figure 7B). These results suggest that DAX-1 interacts directly with the AF-2 surface of ERR.
Figure 6 DAX-1 represses ERR activity. (A–C) 293T cells were co-transfected with 200 ng of mDAX-1 (–467 bp)-Luc, mSHP (–2.2 kb)-Luc, 300 ng of ERR, and increasing amounts of DAX-1 and SHP as indicated in the figure. The luciferase activities were measured and normalized against ?-galactosidase activity. One representative experiment is shown, and similar results were obtained from at least three independent experiments.
Figure 7 ERR physically interacts with DAX-1. (A) GST only or GST fused to DAX-1 was isolated from bacterial culture and immobilized on glutathione-Sepharose beads. In vitro transcribed and translated methionine-labeled ERR or ERR AF-2 was incubated with purified GST fused receptors or GST alone as indicated in the Figure. The complexes were resolved by a 12% denaturing PAGE, and analyzed by autoradiography. (B) A yeast strain EGY48 which contains an integrated ?-galactosidase reporter gene controlled by the LexA-binding site was transformed with the indicated LexA and B42 plasmids. The transformants were selected on plates containing appropriate selection markers, and assayed for ?-galactosidase activity. The results shown are the mean ?-galactosidase values from six independent transformant colonies.
Since the AF-2 surface of ERR is involved in the interaction with DAX-1, we explored the role of coactivator competition in the repression of ERR activity. As demonstrated in Figure 8A, PGC-1, which was described previously as a coactivator of ERR, significantly increased the activity of ERR on the DAX-1 promoter and the co-expression of DAX-1 with PGC-1 repressed this induction in a dose-dependant manner. This result suggests that both PGC-1 and DAX-1 compete for the binding of the AF-2 pocket of ERR, as demonstrated for other receptors. In order to confirm the direct competition between DAX-1 and PGC-1 for binding to ERR, we performed in vitro competition binding assays, using PGC-1 and DAX-1 with GST fused ERR. The results of this study confirmed the findings of a previous report (17). Specifically, 35S-labeled PGC-1 interacted with ERR, and DAX-1 inhibited the interaction between PGC-1 and ERR (Figure 8B). In addition, no interaction between DAX-1 and PGC-1 occurred. Taken together, these observations suggest that DAX-1 physically inhibits the transcriptional activity of ERR by competing with PGC-1 for interaction with ERR.
Figure 8 DAX-1 represses PGC-1-enhanced transactivity of ERR by interfering with PGC-1 binding to ERR. (A) 293T cells were co-transfected with 200 ng of mDAX-1 (–467 bp)-Luc, the indicated concentrations of PGC-1 and ERR, and increasing doses of DAX-1 as indicated in the figure. The luciferase activities were measured and normalized against -galactosidase activity. One representative experiment is shown, and similar results were obtained from at least three independent experiments. (B) GST only or GST fused to ERR was isolated from bacterial cultures and immobilized on glutathione-Sepharose beads. In vitro transcribed and translated methionine-labeled PGC-1 and unlabeled DAX-1 (3 and 6 μl) were incubated with purified GST fused receptors or GST alone as indicated in the Figure. The interaction complexes were resolved by 12% denaturing PAGE, and the gels analyzed by autoradiography.
DISCUSSION
ERR is one of the newest members of the orphan nuclear receptor family. Therefore, the target and molecular mechanisms of ERR have not yet been clearly identified. In this study, we demonstrated that ERR acts as a potential regulator of DAX-1, which, in turn represses ERR at a molecular level. This suggests the existence of an autoregulatory loop.
Although the physiological role of DAX-1 has been clearly identified, regulation of the DAX-1 promoter is not fully understood. A previous report demonstrated that SF-1 regulates the DAX-1 promoter through SFRE (7,8). Even though several SFREs are located on the DAX-1 promoter, the functional property of SFRE on the DAX-1 promoter has been characterized only in the region between –129 and –121 bp, which is conserved in both mouse and human (7). As demonstrated in Figure 1, ERRE1 is identical to sequences previously identified as the SFRE. We also characterized a new ERRE2 that was not previously identified as an SF-1RE. Although both ERRE1 and ERRE2 are required for the binding of ERR, the data from mutation analysis suggest that ERRE1 is more responsible for ERR transactivation than ERRE2 (Figures 3 and 4). In the human DAX-1 promoter, only ERRE1 is present. However, SF-1 can still activate the DAX-1 promoter. This demonstrates that ERRE1 is critical for transactivarion (7). Although all three members (, ? and ) of the ERR family bound to the same DNA response element (data not shown), only ERR was able to activate the DAX-1 promoter (Figure 2). This suggests that DNA-binding is not completely responsible for transactivation.
The basal transcriptional activity of ERR and ? is relatively low compared to the high activity of ERR. This suggests that the basal transcriptional mechanism governing ERR and ? may be distinct from that of ERR. A previous report demonstrated that ERR is constitutively active, because ERR forms a complex with coactivators, such as SRC-1 (28). It has been suggested that, depending on the specific co-regulator recruitment, transcriptional activity can be very different. Furthermore, since NRs can differentially modulate target promoters through the recruitment of different coregulators (29,30), ERR may regulate target promoters, such as DAX-1, in a distinctly unique manner, compared to ERR and ?. Further investigation is necessary to determine whether ERR regulation of target promoters differs with co-regulator recruitment.
Until recently, the ligand for ERR was unknown, although 4-OHT was reported to be an inverse agonist of ERR (14,15). Despite a recent report demonstrating that phenolic acyl hydrazone is a selective agonist of ERR? and ERR, further study will be required to prove that phenolic acyl hydrazone is a bona fide activator of ERR? and (31). 4-OHT is widely used in the treatment of breast cancer and deactivates ERR which was proposed as a favorable biomarker in human breast cancer (25). As demonstrated in Figure 5, both ERR and DAX-1 are expressed in breast cancer cells and tissues. Furthermore, ERR regulated the gene expression of DAX-1 in a positive manner, demonstrating that the DAX-1 promoter is a potential target of ERR in vivo. Although the function of ERR and DAX-1 in breast cancer is not completely understood, we speculate that there is both a positive and negative relationship between DAX-1 and ERR, since DAX-1 acts as both a corepressor of ER and ERR and as a regulator of the DAX-1 promoter. Recent reports indicated that ERR regulates the ERR promoter via a conserved hormone response element and epidermal growth factor signaling which affects ERR transactivation in breast cancer cells (32,33). Since ERR, ERR and PGC-1 are co-expressed in breast cancer, their physiological function may be significantly correlated in breast cancer tissues.
It has been reported that DAX-1 interacts directly with various NRs and represses their transcriptional activity (4–6,21,27). Previous studies have demonstrated that DAX-1 uses its N-terminal repeating region, defined as the LXXLL-containing interaction domain, for interaction with NRs (6). In this study, we demonstrated that the LBD-like region of DAX-1 interacts with ERR. We also found that DAX-1 interacts with ERR and ? (data not shown). The AF-2 domain of ERR was crucial for the interaction with DAX-1 and acted as a binding surface for competition with the coactivator PGC-1 (Figure 7). Depending on the binding region of the nuclear receptor with the corepressor, the repressive mechanism of the corepressor can be quite different. When the DBD of the nuclear receptor is involved in interaction with the corepressor, the DNA-binding of the nuclear receptor is blocked by the corepressor (34). Inversely, if the AF-2 domain of the nuclear receptor interacts with the corepressor, the competition between coactivators and the corepressor occurs on the AF-2 surface of the nuclear receptor (35). In the case of ERR, the AF-2 domain interacts with DAX-1 and acts as a binding surface for interaction with the coactivator PGC-1 (Figure 8).
Previously, we reported that the SHP promoter is a target of ERR (16). In the current study, we suggest that DAX-1 is a target of ERR. As demonstrated in Figure 6, DAX-1 repressed the transcriptional activity of ERR on the DAX-1 and SHP promoters. In addition, SHP also inhibited ERR on the DAX-1 and SHP promoters. It is very interesting that ERR regulates corepressor proteins, such DAX-1 and SHP, which act as repression modulators. Although both SHP and DAX-1 are involved in the regulation of the corepressor-mediated transactivation of NRs, their physiological functions are quite different from each other. While DAX-1 plays a role in steroidogenic tissues, SHP functions mainly in the liver (1). We demonstrated that there is a correlation between ERR and DAX-1 in breast cancer cells (Figure 5). In the case of SHP, we confirmed using adenovirus that ERR activated SHP gene expression in the pancreas, but not in the liver (data not shown). This demonstrated that tissue-specific target gene expression is modulated by ERR. DAX-1 is also differentially regulated by ERR, depending on the target tissue. The modulation of DAX-1 and SHP by ERR in vivo needs to be examined further, in order to identify the physiological roles of ERR, SHP and DAX-1.
Collectively, as depicted in Figure 9, we hypothesize that ERR activates the DAX-1 promoter, which in turn inhibits ERR through coactivator competition. In this study, we demonstrated that DAX-1 is a potential target of ERR. Our study provides a new insight into understanding the basic mechanisms of DAX-1 and ERR, and is expected to provide clues for the identification of the functions of ERR.
Figure 9 Proposed model of ERR and DAX-1 action. ERR directly binds to the DAX-1 (SHP) promoter, resulting in an increase in DAX-1 (SHP) gene expression. In turn, DAX-1 (SHP) represses ERR through a direct interaction with its AF-2 domain. DAX-1 (SHP) competes with the coactivator for binding to the AF-2 surface of ERR and homeostasis between ERR and DAX-1 (SHP) is maintained through this autoregulatory loop.
ACKNOWLEDGEMENTS
This work was supported by a National Research Laboratory grant (M1-0500-4705J-4710), a KRF grant (C00126 ) and a Marine Bio21 grant. Y.Y.P. and S.W.A. are recipients of the Brain Korea 21 program. Funding to pay the Open Access publication charges for this article was provided by Brain Korea 21 program.
REFERENCES
Giguere, V. (1999) Orphan nuclear receptors: from gene to function Endocr. Rev., 20, 689–725 .
Muscatelli, F., Strom, T.M., Walker, A.P., Zanaria, E., Recan, D., Meindl, A., Bardoni, B., Guioli, S., Zehetner, G., Rabl, W., et al. (1994) Mutations in the DAX-1 gene give rise to both X-linked adrenal hypoplasia congenita and hypogonadotropic hypogonadism Nature, 372, 672–676 .
Zanaria, E., Muscatelli, F., Bardoni, B., Strom, T.M., Guioli, S., Guo, W., Lalli, E., Moser, C., Walker, A.P., McCabe, E.R., et al. (1994) An unusual member of the nuclear hormone receptor superfamily responsible for X-linked adrenal hypoplasia congenita Nature, 372, 635–641 .
Zhang, H., Thomsen, J.S., Johansson, L., Gustafsson, J.A., Treuter, E. (2000) DAX-1 functions as an LXXLL-containing corepressor for activated estrogen receptors J. Biol. Chem., 275, 39855–39859 .
Crawford, P.A., Dorn, C., Sadovsky, Y., Milbrandt, J. (1998) Nuclear receptor DAX-1 recruits nuclear receptor corepressor N-CoR to steroidogenic factor 1 Mol. Cell. Biol., 18, 2949–2956 .
Suzuki, T., Kasahara, M., Yoshioka, H., Morohashi, K., Umesono, K. (2003) LXXLL-related motifs in Dax-1 have target specificity for the orphan nuclear receptors Ad4BP/SF-1 and LRH-1 Mol. Cell. Biol., 23, 238–249 .
Yu, R.N., Ito, M., Jameson, J.L. (1998) The murine Dax-1 promoter is stimulated by SF-1 (steroidogenic factor-1) and inhibited by COUP-TF (chicken ovalbumin upstream promoter-transcription factor) via a composite nuclear receptor-regulatory element Mol. Endocrinol., 12, 1010–1022 .
Vilain, E., Guo, W., Zhang, Y.H., McCabe, E.R. (1997) DAX1 gene expression upregulated by steroidogenic factor 1 in an adrenocortical carcinoma cell line Biochem. Mol. Med., 61, 1–8 .
Giguere, V. (2002) To ERR in the estrogen pathway Trends Endocrinol. Metab., 13, 220–225 .
Horard, B. and Vanacker, J.M. (2003) Estrogen receptor-related receptors: orphan receptors desperately seeking a ligand J. Mol. Endocrinol., 31, 349–357 .
Sladek, R., Bader, J.A., Giguere, V. (1997) The orphan nuclear receptor estrogen-related receptor alpha is a transcriptional regulator of the human medium-chain acyl coenzyme A dehydrogenase gene Mol. Cell. Biol., 17, 5400–5409 .
Lu, D., Kiriyama, Y., Lee, K.Y., Giguere, V. (2001) Transcriptional regulation of the estrogen-inducible pS2 breast cancer marker gene by the ERR family of orphan nuclear receptors Cancer Res., 61, 6755–6761 .
Yang, C., Zhou, D., Chen, S. (1998) Modulation of aromatase expression in the breast tissue by ERR alpha-1 orphan receptor Cancer Res., 58, 5695–5700 .
Tremblay, G.B., Bergeron, D., Giguere, V. (2001) 4-Hydroxytamoxifen is an isoform-specific inhibitor of orphan estrogen-receptor-related (ERR) nuclear receptors beta and gamma Endocrinology, 142, 4572–4575 .
Coward, P., Lee, D., Hull, M.V., Lehmann, J.M. (2001) 4-Hydroxytamoxifen binds to and deactivates the estrogen-related receptor gamma Proc. Natl Acad. Sci. USA, 98, 8880–8884 .
Sanyal, S., Kim, J.Y., Kim, H.J., Takeda, J., Lee, Y.K., Moore, D.D., Choi, H.S. (2002) Differential regulation of the orphan nuclear receptor small heterodimer partner (SHP) gene promoter by orphan nuclear receptor ERR isoforms J. Biol. Chem., 277, 1739–1748 .
Huss, J.M., Kopp, R.P., Kelly, D.P. (2002) Peroxisome proliferator-activated receptor coactivator-1alpha (PGC-1alpha) coactivates the cardiac-enriched nuclear receptors estrogen-related receptor-alpha and -gamma. Identification of novel leucine-rich interaction motif within PGC-1alpha J. Biol. Chem., 277, 40265–40274 .
Huppunen, J. and Aarnisalo, P. (2004) Dimerization modulates the activity of the orphan nuclear receptor ERRgamma Biochem. Biophys. Res. Commun., 314, 964–970 .
Razzaque, M.A., Masuda, N., Maeda, Y., Endo, Y., Tsukamoto, T., Osumi, T. (2004) Estrogen receptor-related receptor gamma has an exceptionally broad specificity of DNA sequence recognition Gene, 340, 275–282 .
Lee, Y.K., Parker, K.L., Choi, H.S., Moore, D.D. (1999) Activation of the promoter of the orphan receptor SHP by orphan receptors that bind DNA as monomers J. Biol. Chem., 274, 20869–20873 .
Song, K.H., Park, Y.Y., Park, K.C., Hong, C.Y., Park, J.H., Shong, M., Lee, K., Choi, H.S. (2004) The atypical orphan nuclear receptor DAX-1 interacts with orphan nuclear receptor Nur77 and represses its transactivation Mol. Endocrinol., 18, 1929–1940 .
Park, Y.Y., Kim, H.J., Kim, J.Y., Kim, M.Y., Song, K.H., Cheol Park, K., Yu, K.Y., Shong, M., Kim, K.H., Choi, H.S. (2004) Differential role of the loop region between helices H6 and H7 within the orphan nuclear receptors small heterodimer partner and DAX-1 Mol. Endocrinol., 18, 1082–1095 .
Knutti, D., Kressler, D., Kralli, A. (2001) Regulation of the transcriptional coactivator PGC-1 via MAPK-sensitive interaction with a repressor Proc. Natl Acad. Sci. USA, 98, 9713–9718 .
Choi, Y.K., Kim, Y.J., Park, H.S., Choi, K., Paik, S.G., Lee, Y.I., Park, J.G. (2003) Suppression of glomerulosclerosis by adenovirus-mediated IL-10 expression in the kidney Gene Ther., 10, 559–568 .
Ariazi, E.A., Clark, G.M., Mertz, J.E. (2002) Estrogen-related receptor alpha and estrogen-related receptor gamma associate with unfavorable and favorable biomarkers, respectively, in human breast cancer Cancer Res., 62, 6510–6518 .
Conde, I., Alfaro, J.M., Fraile, B., Ruiz, A., Paniagua, R., Arenas, M.I. (2004) DAX-1 expression in human breast cancer: comparison with estrogen receptors ER-alpha, ER-beta and androgen receptor status Breast Cancer Res., 6, R140–R148 .
Holter, E., Kotaja, N., Makela, S., Strauss, L., Kietz, S., Janne, O.A., Gustafsson, J.A., Palvimo, J.J., Treuter, E. (2002) Inhibition of androgen receptor (AR) function by the reproductive orphan nuclear receptor DAX-1 Mol. Endocrinol., 16, 515–528 .
Greschik, H., Wurtz, J.M., Sanglier, S., Bourguet, W., van Dorsselaer, A., Moras, D., Renaud, J.P. (2002) Structural and functional evidence for ligand-independent transcriptional activation by the estrogen-related receptor 3 Mol. Cell, 9, 303–313 .
Guan, H.P., Ishizuka, T., Chui, P.C., Lehrke, M., Lazar, M.A. (2005) Corepressors selectively control the transcriptional activity of PPARgamma in adipocytes Genes Dev., 19, 453–461 .
Sanyal, S., Matthews, J., Bouton, D., Kim, H.J., Choi, H.S., Treuter, E., Gustafsson, J.A. (2004) Deoxyribonucleic acid response element-dependent regulation of transcription by orphan nuclear receptor estrogen receptor-related receptor gamma Mol. Endocrinol., 18, 312–325 .
Zuercher, W.J., Gaillard, S., Orband-Miller, L.A., Chao, E.Y., Shearer, B.G., Jones, D.G., Miller, A.B., Collins, J.L., McDonnell, D.P., Willson, T.M. (2005) Identification and structure-activity relationship of phenolic acyl hydrazones as selective agonists for the estrogen-related orphan nuclear receptors ERRbeta and ERRgamma J. Med. Chem., 48, 3107–3109 .
Liu, D., Zhang, Z., Teng, C.T. (2005) Estrogen-related receptor-gamma and peroxisome proliferator-activated receptor-gamma coactivator-1alpha regulate estrogen-related receptor-alpha gene expression via a conserved multi-hormone response element J. Mol. Endocrinol., 34, 473–487 .
Barry, J.B. and Giguere, V. (2005) Epidermal growth factor-induced signaling in breast cancer cells results in selective target gene activation by orphan nuclear receptor estrogen-related receptor alpha Cancer Res., 65, 6120–6129 .
Kim, J.Y., Kim, H.J., Kim, K.T., Park, Y.Y., Seong, H.A., Park, K.C., Lee, I.K., Ha, H., Shong, M., Park, S.C., et al. (2004) Orphan nuclear receptor small heterodimer partner represses hepatocyte nuclear factor 3/Foxa transactivation via inhibition of its DNA binding Mol. Endocrinol., 18, 2880–2894 .
Lee, Y.K., Dell, H., Dowhan, D.H., Hadzopoulou-Cladaras, M., Moore, D.D. (2000) The orphan nuclear receptor SHP inhibits hepatocyte nuclear factor 4 and retinoid X receptor transactivation: two mechanisms for repression Mol. Cell. Biol., 20, 187–195 .(Yun-Yong Park, Seung-Won Ahn, Hye-Jin Ki)