HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

J. Biol. Chem., Vol. 275, Issue 27, 20837-20846, July 7, 2000

This Article Abstract Full Text (PDF) All Versions of this Article: 275/27/20837    most recent M001880200v1 Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Zhang, Z. Articles by Teng, C. T. Search for Related Content PubMed PubMed Citation Articles by Zhang, Z. Articles by Teng, C. T. Social Bookmarking                      What's this?

Estrogen Receptor-related Receptor 1 Interacts with Coactivator and Constitutively Activates the Estrogen Response Elements of the Human Lactoferrin Gene^*

Zhiping Zhang and Christina T. Teng

From the Gene Regulation Group, Laboratory of Reproductive and Developmental Toxicology, NIEHS, National Institutes of Health, Research Triangle Park, North Carolina 27709

Received for publication, March 6, 2000, and in revised form, April 11, 2000

	ABSTRACT

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

The human estrogen receptor-related receptor (ERR1, NR3B1a) was shown to bind a steroidogenic factor binding element (SFRE), TCAAGGTCATC, 26 base pairs upstream from the estrogen response element (ERE) of the human lactoferrin gene promoter. A mutation made at SFRE significantly reduced estrogen-dependent transcription from the lactoferrin ERE in human endometrial cells. In this study, we demonstrated that ERR1 binds both SFRE and ERE elements and constitutively transactivates the lactoferrin gene promoter. In DNase I footprinting protection analysis, both SFRE and ERE regions were protected by glutathione S-transferase-ERR1 fusion protein. The receptor formed two protein-DNA complexes with either SFRE or ERE in electrophoresis mobility shift assay. Homodimerization of ERR1 was confirmed with the mammalian two-hybrid system. ERR1 activates reporter constructs containing various types of estrogen response elements in endometrial and non-endometrial cells in transient transfection experiments. Overexpressing the coactivator, SRC1a or GRIP1, further enhances ERR1-induced transcriptional activity. We demonstrated that the AF2 domain of ERR1 is essential for the transactivation function and that deletion or mutation at this region abrogates the activation capability. Protein-protein interaction between the SRC1a and ERR1 C terminus was confirmed with a GST glutathione S-transferase "pull-down" assay. When comparing ERR1 and the estrogen receptor  (ER) in many of the experiments, we found that ER can also bind SFRE of the lactoferrin gene and transactivate the promoter activity in a ligand-dependent manner. The present study demonstrated that ERR1 binds similar DNA elements as ER and confers its transactivation function constitutively. Therefore, ERR1 may actively modulate the estrogen response of lactoferrin gene as well as other estrogen-responsive genes.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Lactoferrin is a multifunctional iron-binding glycoprotein (for review, see Refs. 1-3 and references therein). It has been demonstrated that the protein is involved in killing bacteria, modulating the immune system and suppressing tumor growth. Presently, it is unclear how lactoferrin exerts its multiple functions. Two facts about lactoferrin make it a primary candidate in the first line of defense; it is expressed in wet surface mucosa, and it is a major protein in the secondary granule of the polymorphic neutrophils. In addition, high levels of lactoferrin in the colostrum and milk suggest a nutritional value for the infant (4). Lactoferrin may also be involved in normal reproduction function, and its expression in the endometrium and uterus is under the influence of estrogen and epidermal growth factor (5). Together, the expression and regulation of lactoferrin in various tissues and under a wide variety of physiological conditions imply that the protein is needed for protection and nutrition. We have extensively studied lactoferrin expression under the influence of estrogen in the reproductive tract (5-9). To understand the molecular mechanisms that are involved in estrogen regulation of the lactoferrin gene in the uterus, we have isolated and characterized a compounded estrogen response element (ERE)¹ of the lactoferrin gene promoter from both human and mouse (7-9). In addition to the compounded ERE, human lactoferrin gene contains a steroidogenic factor 1 binding element, SFRE (10), 26 base pairs upstream from the ERE (11) that participates in ER-mediated responsiveness of the lactoferrin gene promoter (12).

Estrogen receptors (ER and ER) play an important role in the differentiation and development of various organs and in the maintenance of proper cellular function in a wide variety of tissues (13). The estrogen receptor belongs to the nuclear receptor (NR) superfamily (14-16). The structure of the receptors in this family is organized into several domains, including the N-terminal region that contains the activation function 1 (AF1) followed by DNA binding domain and C-terminal ligand binding domain (17-19). Within the ligand binding domain, there is a ligand-dependent transcription activation function 2 (AF2) (20-23). Recent advances in NR research have identified a number of coactivators and corepressors that directly interact with the receptors (24-28). The ligand-bound receptor recruits coactivators to its AF2 domain and exerts transcriptional activation function, whereas the unliganded receptor associates with corepressors and represses transcriptional activity. In addition to the NR whose ligand has been identified, ligands for the majority of NR family members are yet to be defined. These NR members are grouped as the "orphans" (for review, see Refs. 15 and 29).

The estrogen receptor-related receptors 1 and 2 (ERR, NR3B1; and ERR, NR3B2 (Ref. 30)) were the earlier orphan receptors identified by screening the cDNA library with radiolabeled DNA binding domain of the ER (31). For the past few years, our laboratory has isolated an isoform of ERR (ERR1) from human endometrial carcinoma RL95-2 cells based on its ability to bind lactoferrin SFRE (12). There are two potential initiation sites for ERR1 that could generate polypeptides of 506 and 422 amino acids (32). The first methionine (Met) of the long form is present in the first exon, whereas the first Met of the short form resides in the second exon of the ERR gene (32). Indeed, the nuclear protein of human endometrium RL95-2 cells shows two proteins at 42- and 57-kDa range that reacted to the antibody generated against the GST-ERR1 fusion protein in Western analysis (12). The short form of ERR1 is identical to the original ERR from the third Met to the end of the protein (31), whereas the major difference between ERR and the long form of ERR1 lies in the N-terminal region. However, there are minor discrepancies between the ERR and ERR1 at the C-terminal region. The region in question arose from a frameshift of the original ERR. The sequence of ERR1 at this region is identical between human and mouse, and the peptide antibody derived from this region detected the same protein as the GST-ERR1 antibody (33). Thus, we concluded that the sequence of ERR1 was correct, and the short form of ERR1 clone was used for functional study. Recently, a third estrogen receptor-related receptor (ERR3) was identified by yeast two-hybrid screening using the glucocorticoid receptor interacting protein 1 (GRIP1) as bait (34). The DNA binding domain region of ERRs and ERs is highly conserved; however, the other parts of the protein share very little homology (31, 34). ERR1 is expressed in the mesodermal tissues and nervous system early in mouse embryonic life (35) and plays a role in bone morphogenesis (36) and lipid metabolism (37, 38). It is expressed selectively in various adult tissues and in a species-specific pattern of expression (31, 32, 38). Recently, a diverse class of genes that possesses ERR1 binding element was identified. These genes were modulated by the receptor either positively or negatively (12, 37-43), suggesting the pivotal roles of ERR1 in a variety of pathways during differentiation, development, and homeostasis.

ERR1 was isolated based on its ability to bind a SFRE in the human lactoferrin gene (11, 12). More extensive studies on the binding requirement of ERR1 showed that its binding is in fact permissive and can bind a wide variety of natural and synthetic estrogen response elements (38, 39, 44, 45). Ligands for the ERR receptors are not clearly identified; however, the pesticides (46) and serum components have been suggested as the potential ligands for ERR1 (44). Contrary to the ligand-dependent receptors, ERR1 was shown as a potent transactivator when its C-terminal region was fused to the progesterone receptor DNA binding domain and tested with a progesterone receptor response element driven reporter in mammalian cells (47). Constitutive activation of transcription by ERR and ERR was due to the binding of coactivators to the receptors in a ligand-independent manner (48). These findings make ERR1 a significant modulator in many signaling pathways that require the same coactivator and similar DNA response elements. In this report, we extended our previous study and further examined the functional role of ERR1 on regulation of the human lactoferrin promoter activity in the endometrial cells. Here we showed homodimer binding of ERR1 to both ERE and SFRE of the lactoferrin gene. Furthermore, ERR1 interacts with coactivator SRC1a in a ligand-independent manner, and the conserved AF2 of the ERR1 is essential for the constitutive transactivation function.

	EXPERIMENTAL PROCEDURES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Materials-- Human endometrial carcinoma cell line HEC-1B (ATCC no. HTB-113) and CV1 (ATCC no. CCL-70) were obtained from ATCC (Rockville, MD). Tissue culture components were obtained from Life Technologies, Inc. Estradiol-17 (E₂) was purchased from Sigma. [³⁵S]Methionine and and S-linkage glutathione-Sepharose beads were acquired from Amersham Pharmacia Biotech, and [¹⁴C]chloramphenicol from NEN Life Science Products.

Plasmids and Oligonucleotides-- The wild type and mutant human lactoferrin hLF-414CAT plasmids have been described previously (11). Oligonucleotides corresponding to the various types of ERE were synthesized by Life Technologies, Inc. and linked to CAT promoter reporter plasmids. The 3xERE-Luc construct was a gift from D. McDonnell (Duke University, Durham, NC). The fragment of ERR1 plasmids used in the study was generated either by restriction enzyme digestion or PCR of the full-length short form of ERR1 and subsequent subcloning into appropriate vectors. The first methionine in the short form ERR1 was designated as 1 and the last amino acid glutamine as 422 (12, 32). The mutations made in the C-terminal region of ERR1 were carried out by Transformer site-directed mutagenesis kit (CLONTECH, Palo Alto, CA), according to the manufacturer's instructions. The plasmids were purified either by CsCl centrifugation or column chromatography (QIAfilter Maxi column, Qiagen, Santa Cruz, CA). Sequence of all the plasmids were verified by automatic sequencing. Reporter plasmid 5x(Gal4)-CAT was obtained from CLONTECH. Expression plasmids pSG5-ER (9) and the recombinant GST fusion protein plasmid pGEX-ERR1 and pGEX-ER (12) have been previously described. Coactivator pCR3.1-hSRC-1a (26) was a gift from M. Tsai (Baylor College of Medicine, Houston, TX), pSG5-GRIP1 (49) was from M. Stallcup (University of Southern California, Los Angeles, CA), and pcDNA3-CBP-Flag2x (50) was from M. Rosenfeld (University of California, San Diego, La Jolla, CA). Plasmids Gal4-ERR1 and VP16-ERR1 were constructed by PCR using the ERR1 cDNA (12) as a template. PCR products were subsequently cloned into Gal4-DB and VP16 vectors (CLONTECH).

Cell Culture and Transient Transfection Assays-- HEC-1B and CV1 cells were maintained in Eagle's minimum essential medium and supplemented with 10% fetal bovine serum (FBS) (Life Technologies, Inc) under 5% CO₂. Prior to transfection, cells were split into six-well plates and cultured in phenol red-free Eagle's minimum essential medium with 10% dextran-coated charcoal-stripped FBS (DCCS-FBS) for 24 h. To examine the effect of serum on ERR1 activity, cells were cultured in DCCS-FBS for 3 weeks prior to transfection experiments. Transient transfections were performed with Cellphect transfection kit (Amersham Pharmacia Biotech) or Effectene Reagent (Qiagen) according to the manufacturer's instructions. The cells were transfected with 6.0 µg (Cellphect transfection) or 1.0 µg (Effectene Reagent) of total DNA consisting of chloramphenicol acetyltransferase (CAT) or luciferase (Luc) reporter construct, the expression plasmids, the internal control (pCH110, -galactosidase) and the salmon sperm DNA or empty vectors made up the final DNA concentration. Sixteen hours after transfection, cells were washed and 10⁷ M E₂ or vehicle was added. The cells were cultured for an additional 24 h in a serum-free condition before preparation of cell extracts. CAT enzyme and the -galactosidase activities were measured as described previously (12). The CAT activities were normalized against -galactosidase activities. All experiments were repeated at least three times with duplicate samples in each experiment, and data are presented as mean ± S.D.

DNase I Footprinting and EMSA-- DNase I protection of the human lactoferrin 5'-flanking sequence with GST-ERR1 (12) and baculovirus-expressed ER (provided by M. Parker, Imperial Cancer Research Fund, London, United Kingdom) was performed according to the standard method of Ausubel et al. (51). DNA fragment from 441 to 308 of the human lactoferrin gene (7) was used in DNase I protection. SFRE and ERE oligonucleotides and GST fusion proteins used in EMSA were described (9, 12). The antibodies to GST acquired from Amersham Pharmacia Biotech and ERR1 (12) were used to supershift the DNA-bound GST-ERR1 and GST-ER.

GST Pull-down Assay-- GST pull-down experiments with GST-ERR1 fusion protein and SRC1a protein were performed as described (52). GST and GST fusion proteins were prepared from Escherichia coli BL21 cells after isopropyl-1-thio--D-galactopyranoside induction of an overnight culture. The bacteria were disrupted by sonication, and the fusion protein was isolated with 50% slurry of glutathione-Sepharose beads (52). The in vitro translated SRC1a was labeled with [³⁵S]methionine by a TNT wheat germ extract-coupled transcription/translation system (Promega). Binding of the ³⁵S-labeled SRC1a to the full-length protein and fragments of GST-ERR1 fusion protein was examined in SDS-PAGE and visualized by autoradiography.

Western Immunoblot Analysis-- Various Gal4-ERR1 mutant constructs (1.0 µg) were transfected into HEC-1B cells. The cell lysates were prepared 24 h after transfection. Ten µg of protein from each cell lysates were loaded and separated on the NuPAGE 4-12% Bis-Tris gel with MOPS running buffer (Novex, San Diego, CA). The SeeBlue pre-stained standard (Novex) was used as molecular weight marker. After electrophoresis, the proteins were blotted onto polyvinylidene difluoride membrane and visualized with anti-Gal4 binding domain (1-147) rabbit polyclonal IgG (0.2 µg/ml) (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) and ECL detection kit (Amersham Pharmacia Biotech).

	RESULTS

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

Activation of Human Lactoferrin Promoter by ERR1-- ERR1 enhanced ER-mediated transactivation of the human lactoferrin gene promoter (12). To examine whether ERR1 by itself also activates the human lactoferrin gene promoter, we transfected the ERR1 expression vectors together with the wild type (wt) and mutant (m1, m6, m1/m6) hLF-414CAT reporter plasmids that contain the previously characterized SFRE and ERE into human endometrial carcinoma, HEC-1B, cells. The sequence of wild type and mutant SFRE and ERE in the reporter plasmids is shown in Fig. 1A. Overexpressing ERR1 in HEC-1B cells strongly stimulated both wt and mutant (m1 and m6) reporter activities (Fig. 1B) in an estrogen-independent manner (compare solid and open bars). It was surprising that mutations made at either SFRE (m1) or ERE (m6) only slightly reduced the transactivation capability to ERR1 in the cells. The basal activity of the m6 plasmids was 6-fold higher than the wt (compare open bars of wt and m6 in Fig. 1B) and overexpressing ERR1 enhanced the m6 reporter activity 4-fold; therefore, the overall transactivation activity of ERR1 was slightly reduced. These results demonstrated that the ERE or SFRE mutation reduced ERR1's functional activity to the same degree (Fig. 1C), suggesting that ERR1 can stimulate promoter through SFRE and ERE. To eradicate the ERR1 activity, both ERE and SFRE had to be mutated (m1/m6). These points are better illustrated in the form of -fold of stimulation (Fig. 1C). ERR1 clearly showed E₂-independent transactivation function with either SFRE or ERE. As control, we transfected ER into the cells and obtained a strong stimulation of the wt reporter activity in the presence of E₂. We found that mutation at SFRE (m1) significantly reduced ER activity, whereas mutation at ERE did not abolish ER function and significant level of E₂-dependent transactivation activity was retained if SFRE was left intact (m6). To completely abrogate ER's transactivating ability, both SFRE and ERE had to be destroyed (m1/m6). These results were in agreement with our previous findings (12). There was no ER in the HEC-1B cells (53), and apparently the endogenous level of ERR1 was not high enough to activate the reporter constructs, as shown by the unresponsiveness of the cells overexpressing empty vector (Fig. 1B). Whether ERR1 activates a heterologous promoter such as SV40 through various EREs was studied in HEC-1B (Fig. 2, solid bars) and CV1 cells (Fig. 2, open bars). Among EREs, the synthetic 3xERE (Fig. 2A) and the single consensus ERE (Fig. 2B) were strong elements for ERR1. The activities of the reporter carrying these elements were increased about 10-fold by expressing ERR1 in HEC-1B cells. The human lactoferrin imperfect ERE (hLF-ERE) also served as a strong element for ERR1. When this imperfect ERE was examined in the context of lactoferrin gene sequence as shown in hLF SFRE/ERE (hLF 396/335), which consists of SFRE and ERE, a higher activation by ERR1 was seen (compare 5.3-fold increase with hLF-ERE to 6.8-fold increase with hLF-SFRE/ERE). SFRE mutation weakens the strength of the ERR1 (hLF-SFREm1/ERE) to the level of hLF-ERE, suggesting that the elements together serve as a stronger enhancer for ERR1. A similar pattern of activation by ERR1 with the same reporters was also found in CV1 cells even though the overall activity was lower (open bars). To test whether the activation function of ERR1 was dose-dependent, we cotransfected increasing amount of ERR1 expression plasmids with SFRE-151-CAT (11) into CV1 cells, and the results showed a dose-dependent activation by ERR1 (Fig. 2C). Results from these experiments demonstrated that ERR1 activates a variety of EREs either in the context of lactoferrin promoter or in isolation. To examine whether serum component plays a role in ERR1's function, we cultured the cells in 10% DCCS-FBS supplemented medium for 3 weeks prior to the transfection experiments. The results were identical to the cells maintained in 10% FBS (data not shown), which implies a constitutive activation of ERR1.

View larger version (18K): [in this window] [in a new window]   Fig. 1.   Activation of human lactoferrin promoter by ERR1 and ER in human endometrial HEC-1B cells. A, diagrammatic presentation of the human lactoferrin gene reporter construct, hLF414-CAT, containing the wild-type (wt) or mutated SFRE (m1) and ERE (m6) sequences. Mutated nucleotides are in bold lowercase letters. B, relative CAT activity. Various reporter constructs, expression plasmids and internal control were transiently transfected into the cells. Twenty-four hours later, the cells were collected and assayed for CAT and -galactosidase activity. Data presented as means ± S.D. of duplicated samples from three independent experiments. C, -fold of activation. Overexpression of ERR1 or ER in the cells enhance the reporter activity. The open and solid bars were cells treated with vehicle or E2, respectively.

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Effect of overexpressing ERR1 on reporters containing various EREs. A, Luc activity. B, CAT activity. EREs were cloned into the SV40-CAT reporter and transfected into HEC-1B (solid bars) or CV1 (open bars) cells. 3x ERE, three copies of consensus ERE (9); 1x ERE, one copy of consensus ERE (9); hLF-ERE, human lactoferrin imperfect ERE (7); hLF-SFRE/ERE, human lactoferrin sequence 396/335, which houses the SFRE and ERE elements (11); hLF-SFRE m1/ERE, mutation made within SFRE of 396/335 (12). -Fold activation is indicated on top of the bars. C, effect of increasing concentration of ERR1 on human lactoferrin promoter (151-CAT) containing the SFRE binding element (418/378) (11). The constructs were transfected into CV1 cells for the study. Data are presented as means ± S.D. of duplicated samples from three independent experiments.

Interaction of ERR1 and ER with the Human Lactoferrin Gene SFRE and ERE-- The physical interactions between ERR1 and ER to either SFRE or ERE of the lactoferrin gene were explored. In the DNase I footprinting protection analysis, baculovirus-expressed ER (BV-ER) and GST-ERR1 fusion protein protected both SFRE and ERE (Fig. 3A). However, the area protected by ER was slightly shifted toward the 5' and the ERR1-protected region toward the 3'. This points to the subtly different binding requirements of the two receptors. GST-receptor fusion proteins were able to bind labeled human lactoferrin SFRE (Fig. 3B) and ERE (Fig. 3C) oligonucleotides in an EMSA. ERR1 formed a broad and diffused band with two noticeable complexes with SFRE and ERE, albeit the bindings were much weaker with SFRE (Fig. 3, compare panel B (lane 1) to panel C (lanes 3 and 4)). After the EMSA condition for ERR1 binding was optimized, two bands of protein-DNA complexes were obtained and the majority of the complexes were in the slow moving band (Fig. 3D, lanes 2 and 3). Interestingly, ER formed one major complex with SFRE and two complexes with ERE (Fig. 3, compare panel B (lane 3) to panel C (lanes 5 and 6)). These receptor-DNA complexes were specific for they could be competed off (data not shown) and supershifted by antibodies to GST (Fig. 3, B (lanes 5 and 6) and C (lanes 7 and 8)) and ERR1 (Fig. 3D, lanes 4 and 5). GST alone did not bind these DNA elements (Fig. 3, B (lanes 2 and 4), C (lane 2), and D (lane 1)). We have investigated the possibility of heterodimerization between the two receptors. We kept either ERR1 or ER constant and mixed with varying amounts of the other receptor in EMSA; moreover, we varied the amount of DNA probes in the study. Despite the effort, we were unable to demonstrate the presence of heterodimerization between ERR1 and ER in the current EMSA condition (data not shown). It is possible that the heterodimer of the receptors will occur in vivo and in the process of co-translation in the in vitro system (44). The two complexes produced by ERR1 binding implied that this receptor forms homodimer. To provide further evidence for the protein-protein interaction between the ERR1, we established the mammalian two-hybrid system. When chimeric expression constructs for Gal4-ERR1 and VP16-ERR1 were cotransfected into CV1 cells, strong activation of the 5x(Gal4)-CAT reporter was observed (Fig. 4). The activation reached the level of a positive control (Gal4-53/VP16-T) for the mammalian two-hybrid system. This result confirmed that the ERR1 monomers are able to form homodimers. The observation was repeated in HEC-1B cells (data not shown). Since ER was able to bind SFRE, we examined whether the receptor can transactivate the lactoferrin promoter through SFRE alone. The SFRE-151-CAT and the 1xERE-SV40-CAT reporter constructs were cotransfected with ER expression plasmids into HEC-1B cells. In the presence of ligand, ER stimulated the activity of 1xERE reporter 50-fold and SFRE reporter 3-fold (Fig. 5). Despite the low activity of SFRE-151-CAT, the ligand-dependent activation of ER is significant. The results support the binding studies that ER binds SFRE and activates the lactoferrin promoter in a ligand-dependent manner.

View larger version (63K): [in this window] [in a new window]   Fig. 3.   ERR1 and ER bind to similar DNA elements of the human lactoferrin gene. A, DNase I footprinting protection analysis with baculovirus-expressed ER (BV-ER) and GST-ERR1. Lane 1, standard; lanes 2 and 3, BSA (10 µg); lane 4, BV-ER; lanes 5-7, GST-ERR 1 at 0.5, 1.0, and 1.5 µg, respectively. The sequence and the position of human lactoferrin SFRE and ERE are indicated. B, EMSA of ERR1 and ER binding to hLF-SFRE. C, EMSA of ERR1 and ER binding to hLF-ERE. D, EMSA of ERR1 binding to hLF-SFRE. The protein-DNA complexes were resolved on a 5% (B and C) or a 3.5% (D) nondenaturing gel. One µg of GST-ERR1 or GST-ER were used in the binding reaction. Thin arrows, DNA-protein complexes; bold arrow, GST (B and C) and ERR1 (D) antiserum supershifted band.

View larger version (10K): [in this window] [in a new window]   Fig. 4.   Homodimerization of ERR1 in vivo. CV1 cells were transfected with the 5x(Gal4)-CAT reporter constructs, internal control, and the expression plasmids of Gal4-DB, VP16, Gal4-ERR1, and VP16-ERR1 as indicated. For the positive control, the cells were transfected with the above reporter constructs along with the expression plasmids of Gal4-53 and VP16-T (CLONTECH). CAT activities were assayed 24 h after transfection. Relative activity was normalized with the internal control, -galactosidase. Data are from three experiments with duplicated samples and presented as mean ± S.D.

View larger version (8K): [in this window] [in a new window]   Fig. 5.   Activation of SFRE by ER. The human lactoferrin SFRE (418/378) was cloned upstream of the lactoferrin promoter reporter, 151-CAT (11). One µg of pSG5-ER expression plasmids (9) along with the reporter and internal control plasmids were transfected into HEC-1B cells with or without the presence of 107 M E2 and the CAT activities were measured. The consensus 1xERE-SV40-CAT was included as positive control. Data are from three experiments with duplicated samples and presented as mean ± S.D.

Mapping the AF2 of ERR1-- Since ERR1 was constitutively active in the HEC-1B and CV1 cells, a domain that houses the transactivation function of the receptor must exist. To map the location of the transactivation function, the N- and C-terminal regions of the receptor were linked to Gal4-DNA binding domain and tested with 5x(Gal4)-CAT reporters in HEC-1B cells. Structure of the chimeric plasmids were presented in Fig. 6A and the CAT activity in Fig. 6B. N-terminal region of ERR1 (positions 1-55 and 1-78) slightly inhibited the reporter activity. In contrast, 10-fold activation of the reporter activity was obtained with Gal4-ERR1 C terminus (positions 174-422).

View larger version (14K): [in this window] [in a new window]   Fig. 6.   Mapping the activation domain of ERR1. A, schematic presentation of the truncated and mutated ERR 1 that were linked to the GAL4-DB (Gal4) expression plasmids and used in the transfection experiments. B, chimeric expression constructs (1.5 µg) were transfected along with 5x(Gal4)-CAT reporter constructs (0.3 µg) and carrier plasmid DNA to the HEC-1B cells. CAT activities were assayed 24 h after transfection. The relative activity was normalized with the -galactosidase activity and divided by the activity of Gal4-DB alone. Data are from three experiments with duplicated samples and presented as mean ± S.D. C, various ERR1 mutants were transfected and expressed in HEC-1B cells. Twenty-four hours later, the cell lysates were prepared and 10 µg of protein from each lysates were analyzed by anti-Gal4 binding domain polyclonal IgG in the Western immunoblot analysis.

To fine map the location of the activation function, a series of C-terminal deletion and mutation fragments were linked to the Gal4-DNA binding plasmid. The ability of these constructs to transactivate the 5x(Gal4)-CAT reporter was measured. The results showed that C-terminal deletion mutants 174-294, 174-343, and 174-406 do not transactivate the reporter. Thus, the last 16 amino acids (407-422) of the receptor were required for the transactivation activity. Within this region resides the conserved AF2 domain (20-23). To confirm that the AF2 domain of ERR1 is responsible for the transactivation, site-directed mutations were introduced into the highly conserved AF2 AH domain (20). When phenylalanine 413 and leucine 414 were changed to alanine (F413A and L414A), the transactivating activity of ERR1 C terminus (174-422) was severely compromised but not abrogated. This finding was not surprising because only the Phe-413 is the conserved amino acid in AF2 domain, and is therefore critical for the AF2 function. As the conserved methionine and leucine at amino acid 416 and 417 of the AF2 domain were converted to alanine (M416A and L417A), the transactivating activity was completely blocked (Fig. 6B). To ensure that the various mutant constructs were expressed at equivalent levels in the cells, we examined the fusion protein with anti-Gal4 binding domain IgG in a Western analysis (Fig. 6C). The results showed that the mutants were expressed equally in the transfected HEC-1B cells. These data strongly support that the AF2 AH domain in ERR1 is a functional transactivation domain.

Coactivator SRC1a Interacts with ERR1 and Potentiates Its Transcriptional Activity-- Since AF2 of the steroid receptors recruits coactivators for ligand-dependent activation, it is reasonable to assume that ERR1 also recruits coactivators in a ligand-independent manner. To explore this possibility, we used GST pull-down to examine the protein-protein interaction between coactivators and ERR1. One of the candidate coactivators is steroid receptor coactivator 1a (SRC1a) (26). By in vitro transcription and translation, we produced [³⁵S]methionine-labeled protein from pCR3.1-hSRC1a plasmids (26). The interactions of labeled SRC1a to the full-length (1-422), the N-terminal plus the DNA binding domain (1-173), and the C-terminal (174-422) GST-ERR1 fusion protein were examined (Fig. 7A). The pull-down results demonstrated that the C-terminal ERR1 is the primary region for interaction (compare lanes 4 and 5 to lanes 6 and 7), although the N-terminal region also shows some weak interaction (lanes 4 and 5). Control GST did not interact with SRC1a (lanes 8 and 9). In addition, the interaction between SRC1a and ERR1 was independent from components present in the serum (compare lanes 2 and 3, lanes 4 and 5, and lanes 6 and 7). Increasing concentration of SRC1a and GRIP1 but not CBP further enhanced the transactivation function of ERR1 C-terminal region in the absence of serum and exogenous hormones. These results suggested that the AF2 domain of ERR1 is readily available to interact with coactivators and thus is constantly active in HEC-1B cells (Fig. 7B). Nonetheless, the present experiments cannot rule out the possibility that there is an unknown endogenous ligand that keeps ERR1 in an activated state in HEC-1B and CV1 cells.

View larger version (26K): [in this window] [in a new window]   Fig. 7.   Coactivator SRC1a interacts with ERR1 and potentiates its activity. A, pull-down assay. Full-length and truncated GST-ERR1 fusion protein were isolated from the bacterial culture and immobilized on glutathione-Sepharose beads. In vitro transcribed and translated [35S]methionine-labeled SRC-1a were incubated in the GST fusion proteins in the absence or presence of 5% serum for 1 h at 4 °C. After incubation, protein pull-down by the Sepharose beads were washed, eluted, and resolved on SDS-PAGE and the 35S-labeled SRC-1a was visualized by autoradiography. B, CAT activity. Increasing concentrations of SRC-1a, GRIP1, and CBP with constant level of Gal4-ERR1 (174-422) expression plasmids were transfected into HEC-1B cells. The 5x(Gal4)-CAT reporter activities were measured. Data are from three experiments with duplicated samples and presented as mean ± S.D.

	DISCUSSION

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES

In the context of the lactoferrin gene promoter, liganded ER binds to an imperfect ERE and collaborates with an adjacent SFRE that binds ERR1 to confer a strong estrogen responsiveness (see Figs. 1 and 2 and Ref. 12). The organization of these EREs in lactoferrin gene provides a unique opportunity to study the regulation of a natural estrogen-responsive gene by ER and ERR1 individually or in combination. The present study demonstrated that ERR1 binds a variety of EREs, including SFRE and the imperfect ERE of lactoferrin gene, and constitutively transactivates the reporters that carry these EREs. Moreover, we identified the critical amino acids in ERR1 required for interaction with the coactivator SRC1a and formed the transactivation function.

ERR1 Constitutively Transactivates Lactoferrin Promoter through SFRE and ERE-- Results from the present study demonstrated that ERR1 is an activator for the lactoferrin promoter. The receptor confers transcriptional activity by either SFRE or ERE of the lactoferrin gene, and neither synergism nor additive effect was detected (Fig. 1B). The ability to activate lactoferrin promoter by ERR1 through either SFRE or ERE was considerably lower than that of the E₂-ER acting through ERE. Nonetheless, the reporter activity in ERR1-transfected HEC-1B cells was 6-8-fold higher than the cells transfected with empty expression vectors. Since this cell type does not contain endogenous ER (53), the high level of reporter activity was the result of overexpressing ERR1. We demonstrated a dose-dependent transactivation by ERR1 through SFRE in CV1 cells, which have low levels of ERR1.² The synthetic EREs, such as the commonly used consensus 1xERE and 3xERE, as well as the lactoferrin SFRE and ERE that were cloned upstream from SV40-promoter reporter were strong elements in conferring ERR1's function. This study is consistent with the report from other laboratories that ERR1 can transactivate a wide variety of EREs in a promoter and reporter insensitive manner (45, 48). These findings show a complicated picture of mechanisms of estrogen action and make the interpretation of estrogen response in different cell types difficult. In contrast to the synthetic SFRE, which required multiple copies to confer ERR1 response (38, 44), a single copy of the lactoferrin SFRE is functioning in our current study (Fig. 2). It is interesting to find that a single SFRE site in the thyroid hormone receptor- (40) is enough, whereas two imperfect SFRE sites in osteopontin gene (42) are required for maximal activation of the promoter by ERR1. Thus, in the context of a natural gene, a SFRE site collaborated with the surrounding sequences and acted as a strong element in ERR1 response. It was surprising to see that ER can exert its function through the lactoferrin SFRE upon addition of estrogen even though the clear choice for ER is ERE (Fig. 5). Recently, the osteopontin gene promoter, stimulated by ERR as well as estrogen-bound ER through SFRE, was found (41). These studies have broadened the spectrum of genes potentially regulated by ER.

To demonstrate that ERR1 and ER activate the reporter activity by binding to SFRE and ERE directly, we examined the receptor and DNA interaction in both DNase I footprinting protection assay and EMSA. It was clear that both receptors interacted with SFRE and ERE of the lactoferrin gene. This finding was further supported by the EMSA (Fig. 3). In EMSA, ERR1 formed two complexes on either element, whereas ER formed only one complex with SFRE. Homodimerization of ER was well documented with EREs (18), and a recent report also showed homodimer binding with SFRE (44). Therefore, the single complex of ER-SFRE in EMSA (Fig. 3B, lane 3) could consist of ER homodimer with one receptor binding to the element, while the other one does not. Indeed, our footprinting data showed that both ERR1 and ER protect primarily a single conserved core element of ERE (AGGTCA) embedded within the SFRE. It was possible that only one of the dimerized receptors is required to make contact with the core element and the other receptor could anchor on the surrounding sequence, including the cryptic core element for stabilization. This argument is supported by the fact that the majority of ERE in natural genes have one conserved and one cryptic core element separated by the space of three nucleotides. ERR1 was originally suggested to bind DNA elements as a monomer (39, 54). However, homodimer binding of ERR to SFRE was demonstrated by co-translation of ERR and truncated ERR that generated an intermediate complex in EMSA (44, 45). Our EMSA study also suggested homodimer binding (Fig. 3D), and we confirmed ERR1 homodimerization by the mammalian two-hybrid system (Fig. 4). Based on the binding nature of ERR1, it is expected that this receptor can bind a wide variety of ERE and be actively involved in ER's signaling pathway.

Coactivator Is Required for the Constitutively Activated Function of ERR1-- The fact that ERR1 can activate lactoferrin gene promoter through both SFRE and ERE has made this receptor an active player in the regulation of the lactoferrin gene expression. In contrast to the ER, activity of ERR1 was observed in the absence of added ligand. We have cultured the cells in phenol red-free medium supplemented with dextran-charcoal-treated serum for several days to several weeks and found the same ERR1 activity in both RL 95-2 and HEC-1B cells.³ These results did not support the notion that ligand for ERR1 is present in the serum (44). To support the constitutive activation of ERR1, we demonstrated that the receptor interacts with coactivator SRC1a independent of serum component (Fig. 7A). Nevertheless, we cannot exclude the possibility of an endogenous ligand that is present in specific cell types or the culture condition that does not deplete the regulating factor(s) of the serum. From the deletion and mutation study, we identified the critical amino acids in AF2 domain of ERR1 that are needed for SRC1a or GRIP1 interaction (Fig. 6). During the course of these studies, other laboratories also demonstrated the coactivators and ERR interaction. Among the coactivators examined, GRIP1 was the strongest one for ERR, , and 3 (34, 46, 48). Our present study did not show such disparity between SRC1a and GRIP1 (Fig. 7B). Together, the present findings suggest that ERR1 is functioning very much like the liganded receptor by recruiting coactivator to the AF2 domain in the ligand binding domain for its activity. There is precedent for constitutive activation of steroid receptors. For example, SRC1a augments the transcriptional activity of human ER in both presence and the absence of ligand (55, 56). Furthermore, a single mutation of the tyrosine in the AF2 AH region of ER made this receptor constitutively active (57, 58). This tyrosine residue is highly conserved among the ER from various species (57) and is implicated in the correct realignment of the H12 helix to prevent the recruitment of coactivators. Thus, mutation at this residue will prevent the realignment and render the receptor active without ligand. The corresponding position of ERR1 is a histidine instead of a tyrosine (Fig. 8). This change from a hydrophobic amino acid to a basic amino acid could realign the AF2 AH region from the conformation of an unliganded state to the conformation of a liganded state and allow interaction with coactivator(s). The reason for the constant activation of AF2 domain in ERR1 remains elusive unless a ligand is identified or the crystal structure of the receptor is obtained. A recent report on the existence of antagonists of ERR1 is intriguing (46). Two organochlorine pesticides, toxaphene and chlordane were found to have antagonistic effects against the activity of ERR1. Yang and Chen (46) showed that the toxaphene blocked the GRIP1 enhanced ERR1 activity. Although the direct proof of toxaphene and chlordane binding to ERR1 is lacking, the potential that ERR1 is a ligand-dependent repressor has physiological significance. CAR is an example of ligand-dependent repressor (59). In the absence of ligand, CAR recruits the coactivators and constitutively active. Upon binding to its ligand, androstane metabolites, CAR releases the coactivator and deactivates.

View larger version (24K): [in this window] [in a new window]   Fig. 8.   Sequence comparison of the AF2 region in ERs and ERRs. The amino acids essential for AF2 function are boxed. The conserved tyrosine in ERs that is changed to histidine in ERRs is indicated by arrow. Sequences are from the following references: hERR1 (12), mERR1 (33), hERR and hERR  (31), mERR3 (34), hER (60), mER (61), rER (62), and mER (55).

In addressing the possible transcription properties of the ERR1, we must consider the potential role of the N-terminal region of the receptor. The presence of different N-terminal regions of this receptor may modify the functional aspect of the receptor accordingly. In conclusion, ERR1 is a functional receptor that contains typical steroid receptor characteristics. The receptor has the ability to bind diverse EREs that contain single or multiple steroid receptor binding core elements; thus, ERR1 may have a broader role in steroid hormone-induced transactivation.

	ACKNOWLEDGEMENTS

We thank Drs. Masahiko Negishi and Anton Jetten for critically reading the manuscript and supplying useful comments. The many discussions with Dr. D. Swope (NIEHS, Research Triangle Park, NC) and the contribution of DNase I footprinting data by N. Y. Yang of the Gene Regulation Group are appreciated. We also thank Drs. M. J. Tsai, D. McDonnell, and M. Parker, M. G. Rosenfeld, M. T. Stallcup, and D. Swope for reagents. The paper was edited by L. Moore and C. Beard.

	FOOTNOTES

^* The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

To whom correspondence should be addressed: Laboratory of Reproductive and Developmental Toxicology, NIEHS, National Institutes of Health, P. O. Box 12233, MD E-201, Research Triangle Park, NC 27709. Tel.: 919-541-0344; Fax: 919-541-0696; E-mail: teng@niehs.nih.gov.

Published, JBC Papers in Press, April 21, 2000, DOI 10.1074/jbc.M001880200

² N. Yang and C. T. Teng, unpublished data.

³ Z. Zhang and C. T. Teng, unpublished data.

	ABBREVIATIONS

The abbreviations used are: ERE, estrogen response element; ER, estrogen receptor; SFRE, steroidogenic factor binding element; ERR, estrogen receptor-related receptor; GST, glutathione S-transferase; wt, wild-type; m, mutant; CAT, chloramphenicol acetyltransferase; PCR, polymerase chain reaction; FBS, fetal bovine serum; EMSA, electrophoretic mobility shift assay; NR, nuclear receptor; DCCS, dextran-coated charcoal-stripped; E₂, estradiol-17; MOPS, 4-morpholinepropanesulfonic acid; Bis-Tris, bis(2-hydroxyethyl)iminotris(hydroxymethyl)methane.

	REFERENCES

TOP ABSTRACT INTRODUCTION EXPERIMENTAL PROCEDURES RESULTS DISCUSSION REFERENCES