Epidermal Growth Factor–Induced Signaling in Breast Cancer Cells Results in Selective Target Gene Activation by Orphan Nuclear Receptor Estrogen-Related Receptor α

Introduction

The estrogen-related receptor α (ERRα, NR3B1) is an orphan member of the nuclear receptor superfamily identified on the basis of its structural similarity to the estrogen receptor α (ERα, NR3A1; ref. 1). ERRα is functionally similar to ERα in that it can bind the inverted repeat estrogen response element (ERE) sequence in target gene promoters, although ERRα preferentially binds a nine-nucleotide extended half site sequence, the ERRE ( 2– 4). In contrast to the ligand-dependent activation of ERα, ERRα is constitutively active and does not respond to estradiol or other natural estrogens ( 1, 5), although its activity can be inhibited by the synthetic estrogen diethylstilbestrol ( 6, 7). The constitutive ERRα can interact with and be modulated by members of the SRC and PGC-1 families of coactivators ( 6, 8– 14). The function of ERRα as a transcriptional activator seems to be cell type and promoter specific. This specificity is proposed to be dependent on factors including relative coactivator levels, interactions with other nuclear receptors, and the presence of appropriate activating stimuli or the context of the response element ( 2, 9, 15– 17). ERRα activates the promoter of its own gene, ESRRA, thereby providing positive regulation of its own expression ( 13). Other known ERRα target genes are involved in regulation of mitochondrial biogenesis and energy metabolism ( 2, 14, 18, 19). ERRα also activates the TFF1 (also known as pS2) and aromatase genes, which are implicated in breast cancer ( 17, 20).

ERRα is widely expressed in normal tissues ( 2) and RNA expression studies show the presence of ERRα in a range of breast cancer cell lines ( 17). It has also been shown that ERRα is an unfavorable breast cancer biomarker as its presence in tumor samples associates with an increased risk of disease recurrence or adverse clinical outcome ( 21). In addition, ERRα mRNA levels in primary breast tumors correlate both with ERα-negative tumor status, a predictor of poor prognosis, and with expression of ErbB2, an indicator of aggressive tumor behavior ( 22). ErbB2 is a receptor tyrosine kinase that is overexpressed or amplified in 15% to 30% of malignant breast cancers. ErbB2, in combination with the ErbB family members epidermal growth factor receptor (EGFR), ErbB3 and ErbB4, transmits signals from EGF and EGF-like growth factor ligands to activate a complex array of downstream signaling pathways including the phosphoinositide 3-kinase/Akt and mitogen-activated protein kinase pathways. Such activation leads to phosphorylation of multiple downstream targets and ultimately to regulation of transcription through phosphorylation of specific transcription factors (reviewed in ref. 23).

Nuclear receptors are phosphoproteins and their activity can be modulated by growth factor signaling ( 24, 25). Phosphorylation occurs most often in the AF-1, AF-2, or DNA-binding domains, and is mediated by kinases downstream of various cytokine and growth factor signaling pathways. Phosphorylation regulates receptor functions, including DNA binding, transactivation, and interaction with coactivators, and can result in both activation and inhibition of receptor transcriptional activity (reviewed in refs. 24, 25). Modulation of ERRα by phosphorylation may provide a way to regulate its constitutive activity in the absence of a known ligand and may facilitate understanding the apparent promoter specificity of its transcriptional effects. Given the correlation between overexpression of ERRα and ErbB2 in breast cancer cells, the transcriptional activity of ERRα may thus be influenced by EGFR/ErbB2 signaling pathways.

Protein kinase Cδ (PKCδ) is a member of the novel subfamily of PKC serine/threonine kinases. It is activated by diacylglycerol and phorbol esters and is phosphorylated by various signaling pathways, including EGFR ( 26). PKCδ is ubiquitously expressed, and once activated it functions in a cell- and tissue-specific manner, probably because of differential phosphorylation events. Several reports indicate that PKCδ is a negative regulator of proliferation in certain cell types, and studies from null mice show that PKCδ is not required for proliferation of normal cells ( 27). In contrast, PKCδ is thought to play a proliferative role in transformed cells. It is a prosurvival factor and is important in maintaining cell migration in breast cancer cell lines ( 28, 29). PKCδ may also play a role in breast cancer progression as its expression in breast tumor lines correlates with their metastatic potential ( 30). In addition, enhanced expression of PKCδ leads to increased anchorage-independent growth of highly metastatic cells ( 31). The potential role of PKCδ in breast cancer progression thus makes it an interesting target in the investigation of EGF signaling in breast cancer cells.

In this study, we investigated a possible relationship between phosphorylation and ERRα function through EGF signaling in breast cancer cells. We found that ERRα is phosphorylated in response to EGF, which enhances receptor DNA binding in vivo. Sequence scanning of ERRα indicated the presence of potential PKCδ phosphorylation sites in the DNA-binding domain. We showed that PKCδ does indeed phosphorylate ERRα in the DNA-binding domain, and enhances its DNA binding in vitro. In addition, PKCδ signaling augments ERRα DNA binding and enhances ERRα transactivation in MCF-7 cells. Moreover, we found via chromatin immunoprecipitation that EGF enhances ERRα target promoter occupancy in a gene-specific manner. Our results suggest that ERRα phosphorylation provides a mechanism to enhance receptor DNA-binding and transactivation, events that contribute to the selective activation of target genes in breast cancer cells.

Materials and Methods

Materials. Human recombinant EGF and purified recombinant PKCδ were from Upstate Biotechnology (Waltham, MA). The ERRα polyclonal antibody has been described previously ( 13). The RNA polymerase II and HA polyclonal antibodies were from Upstate Biotechnology and the HA monoclonal antibody was from Covance Research Products (Princeton, NJ). The PKCδ and ERα antibodies were from Santa Cruz Biotechnologies (Santa Cruz, CA) and the phospho-PKCδ (Thr⁵⁰⁵) antibody was from Cell Signaling Technology (Beverly, MA). Phorbol-12-myristate 13-acetate (PMA) was from Calbiochem (San Diego, CA).

In vivo phosphorylation. All cells were plated in 10 cm dishes. COS-1 cells were transiently transfected for 16 hours with 3 μg pCMXhERRα then incubated in serum- and phosphate-free DMEM for 24 hours, followed by addition of 1 mCi [³²P]orthophosphate for 6 hours. MCF-7 cells stably transfected with pOPRShERRα ³ were incubated in serum- and phosphate-free DMEM for 24 hours then induced with 1 mmol/L isopropyl-l-thio-B-d-galactopyranoside (IPTG) and labeled with 1 mCi [³²P]orthophosphate for 6 hours. Cells were treated with 100 ng/mL EGF for the final 30 minutes of the 6-hour incubation. MCF-7 cells were incubated in serum- and phosphate-free DMEM for 24 hours, labeled with 1 mCi of [³²P]orthophosphate for 6 hours and treated with 100 ng/mL EGF for the final 20 minutes of the 6-hour incubation. Cells were harvested in modified radioimmunoprecipitation assay buffer (RIPA) buffer and lysates immunoprecipitated with anti-ERRα or anti-HA polyclonal antibodies. Reactions were electrophoresed, transferred to polyvinylidene difluoride (PVDF) membrane and detected by autoradiography or Western blotting using the anti-ERRα or anti-HA monoclonal antibody.

Electromobility shift assays. Electromobility shift assays (EMSA) were done as described ( 32) using consensus ERRE 5′-TCGACGCTTTCAAGGTCATATCCG-3′ and ERE 5′-TCGACAAAGTCAGGTCACAGTGACCTGATCAAG-3′ probes. MCF-7 cells were grown in serum-free medium for 24 hours before treatment with 100 ng/mL EGF for 20 minutes. Nuclear extracts were prepared ( 33) and 10 μg extract was used per EMSA reaction. For EMSA using acid phosphatase and for PGC-1α supershift experiments, COS-1 cells were transiently transfected with 5 μg pCMXhERRα for 16 hours and nuclear extracts were prepared. Two micrograms of extract were incubated with acid phosphatase at 25°C for 1 hour before the EMSA reaction. For PGC-1α supershift experiments, 2 μg of extract were combined with 1, 2, or 3 μg of purified GST-PGC-1α. This construct was prepared by PCR cloning the cDNA encoding residues 1 to 250 of PGC-1α from pcDNA3/HA-hPGC-1 ( 34) into pGEX-2T. EMSA with recombinant PKCδ was done using equivalent amounts of purified GST-O or GST-ERRαDBD fusion proteins. Samples were incubated with 25 ng purified recombinant PKCδ according to the manufacturer's instructions at 30°C for 30 minutes before the EMSA reaction. For EMSA following PMA treatment, MCF-7 cells were grown in serum-free medium for 24 hours before addition of 100 nmol/L PMA for 5, 10, 20, 30, 40, or 60 minutes. Ten micrograms of nuclear extract were used per EMSA reaction. Four microliters of the ERRα antibody were used per reaction in supershift experiments.

Phosphorylation site prediction. The ERRα amino acid sequence was scanned for prediction of consensus phosphorylation sites using the following programs: Scansite (http://scansite.mit.edu), Netphos 2.0 (http://www.cbs.dtu.dk/services/NetPhos/), and Macvector Version 7.2 (Accelrys, Inc., San Diego, CA).

In vitro kinase assays. In vitro kinase assays were done using equivalent amounts of GST-O, GST-hERRα-A/B, DNA-binding domain, or ligand-binding domain fusion proteins bound to glutathione Sepharose. Reactions were prepared with 25 ng purified recombinant PKCδ and 10 μCi [γ³²P]ATP according to the manufacturer's instructions, and incubated at 30°C for 30 minutes with frequent mixing. Reactions were washed twice in ADBIII according to the manufacturer's instructions then subjected to PAGE. Gels were stained with Coomassie blue, dried, and exposed for ³²P incorporation.

Cell culture and transfections. MCF-7 cells were maintained in DMEM without phenol red, supplemented with 10% fetal bovine serum (FBS). All transfections were done using FuGene 6 (Roche Diagnostics, Indianapolis, IN). Cells were plated in 12-well plates using DMEM without phenol red with 10% charcoal-stripped FBS. For PKC-CAT experiments, cells were transfected with 100 ng pHACE-PKCδCAT or pHACE-PKCδCAT-KR (provided by Dr. J. Soh, Inha University, Incheon, Korea), along with 400 ng TFF1Luc reporter and 100 ng pCMVβgal transfection efficiency control plasmid. Cells were transfected for 24 hours before harvesting for Luciferase and β-galactosidase activity. For PMA experiments, cells were transfected with 400 ng TFF1Luc, TFF1LucΔERE, TFF1LucΔERRE, TFF1ΔERE/ERRELuc ( 17) or pGL3ESRRA reporter ( 13), and 100 ng pCMVβgal transfection efficiency control plasmid. Cells were transfected for 24 hours then treated with 100 nmol/L PMA for a further 8 hours before harvesting for Luciferase and β-galactosidase activity.

Immunoprecipitation and Western blotting. MCF-7 cells were grown in DMEM without phenol red with 10% charcoal-stripped FBS. Media was changed to serum-free DMEM for 24 hours followed by treatment with 100 ng/mL EGF or 100 nmol/L PMA for 1, 5, 15, 30, 45, or 60 minutes immediately before harvesting. Whole-cell extracts were prepared using modified RIPA buffer and 100 μg extract was electrophoresed then transferred to nitrocellulose membranes. Membranes were blocked with 5% skim milk powder in TBST and blotted overnight with the phospho-PKCδ (Thr⁵⁰⁵) antibody in 5% bovine serum albumin/TBST. Blots were stripped in 10 mmol/L Tris/150 mmol/L NaCl (pH 2.3) for 30 minutes at room temperature before reblotting with the PKCδ antibody. Western blot analysis for HA-ERRα and ERRα was done using PVDF membranes and either the anti-HA monoclonal or anti-hERRα polyclonal antibody, incubated in 5% skim milk powder/PBST, and blotted overnight.

Chromatin immunoprecipitation assay. Chromatin immunoprecipitation assays were done as previously described ( 35). Briefly, MCF-7 cells were cultured in DMEM containing 10% charcoal-stripped FBS for 2 days, followed by serum starvation for 24 hours before treatment with 200 ng/mL EGF for 20 minutes. Cells were cross-linked for 10 minutes at room temperature, collected, centrifuged, and the pellets resuspended in lysis buffer [1% SDS, 10 mmol/L EDTA, 50 mmol/L Tris-HCl (pH 8.1)] supplemented with protease inhibitor cocktail (Roche Diagnostics). The lysate was sonicated seven times for 8 seconds and centrifuged. Soluble chromatin was diluted 10-fold in chromatin immunoprecipitation buffer [1% Triton X-100, 2 mmol/L EDTA, 150 mmol/L NaCl, 20 mmol/L Tris-HCl (pH 8.1)] and immunoprecipitated using either the hERRα polyclonal antibody or preimmune serum or the hERα antibody. Chromatin for immunoprecipitation with the RNA polymerase II antibody was diluted 2.5-fold in chromatin immunoprecipitation buffer [1% Triton X-100, 2 mmol/L EDTA, 500 mmol/L NaCl, 20 mmol/L Tris-HCl (pH 8.1)]. Overnight immunoprecipitation was followed by addition of 40 μg salmon sperm DNA/protein A agarose (Upstate Biotechnology) for 2 hours. Precipitates were washed for 10 minutes each in Buffer I [0.1% SDS, 1% Triton X-100, 2 mmol/L EDTA, 20 mmol/L Tris-HCl (pH 8.1), 150 mmol/L NaCl], Buffer II [0.1% SDS, 1% Triton X-100, 2 mmol/L EDTA, 20 mmol/L Tris-HCl (pH 8.1), 500 mmol/L NaCl] and Buffer III [0.25 mol/L LiCl, 1% NP40, 1% Na deoxycholate, 1 mmol/L EDTA, 10 mmol/L Tris-HCl (pH 8.1)], followed by washing in TE and elution in 1% SDS, 0.1 mol/L NaHCO₃ buffer at 65°C for at least 4 hours. Isolated fragments were purified using the QIAquick spin kit (Qiagen, Valencia, CA). Quantitative PCR was done using the Lightcycler and Fast Start DNA Master SYBG Green 1 (Roche Diagnostics). The primers for the TFF1 promoter were 5′-GGCCATCTCTCACTATGAATCACTTCTGC-3′ and 5′-GGCAGGCTCTGTTTGCTT-AAAGAGCG-3′, which encompass the ERRE. Primers for the ESRRA promoter were 5′-CCATCCGAGTGGAATTTGAG-TCCTAAAG-3′ and 5′-GAACCGTAGACCCAG-TAGCCCCACAGAG-3′, which amplifies the region covering the ERRE element. Enrichment of these regions was compared with a negative control segment located 4 kb upstream of the either the TFF1 or ESSRA promoters, using the following sets of primers: TFF1 promoter 5′-CTGGGCAATGCGAGGAGAGTCAAGACTG-3′ and 5′-GGGAGGGAGGAGTTTGGAGGAAGTGG-3′ and ESRRA promoter 5′-GTGGCCC-ACAGGTGTCGCTCAAGTCTTC-3′ and 5′-GGATGCAGTGTCCTTCTCCCCCAG-ATTG-3′. PCR products were visualized after migration on a 2% agarose gel.

Quantitative reverse transcription-PCR. MCF-7 cells were grown in DMEM without phenol red with 10% charcoal-stripped FBS, serum starved for 24 hours, and treated with 100 ng/mL EGF for 8 hours. RNA was extracted from control and treated cells using the RNeasy kit (Qiagen). cDNA synthesis was done using Transcriptor Reverse Transcriptase (Roche Applied Science) according to the manufacturer's instructions. Quantitative PCR was done using the Lightcycler and Fast Start DNA Master SYBG Green 1. Primers used to amplify the TFF1 transcript were 5′-ATGGCCACCATGGAGAACAAGG-3′ and 5′-CTAAAATTCACACTCCTCTTCTGG-3′, for ESRRA were 5′-CGAGAGGAGTATGTTCTA-3′ and 5′-CCCGCCCGCCGCCGCTCAGCA-3′, and for the RPLP control were 5′-TGGAGAAACTGCTGCCTCATA-3′ and 5′-CCCTGGAGATTTTAGTGGTGA-3′. TFF1 and ESRRA transcript levels were corrected for expression of the RPLP control.

Results

Estrogen-related receptor α is phosphorylated in response to epidermal growth factor, which enhances its DNA binding. To investigate whether ERRα can be modified by phosphorylation, COS-1 cells were metabolically labeled and ectopic ERRα immunoprecipitated using a specific ERRα antibody. Phosphorylated ERRα was detected in cells transfected with the ERRα construct but not from control-transfected cells ( Fig. 1A ). MCF-7 human breast cancer cells stably transfected with an IPTG-inducible HA-ERRα construct were also metabolically labeled and lysates immunoprecipitated with the HA tag antibody ( Fig. 1B). HA-ERRα was phosphorylated, and the level of phosphorylation increased after treatment of cells with EGF, while Western blot analysis indicated that total receptor levels remained constant ( Fig. 1B, bottom). To investigate phosphorylation of the endogenous receptor, lysates from metabolically labeled MCF-7 cells were immunoprecipitated using the ERRα antibody. Phosphorylated ERRα was detected and the extent of phosphorylation was also enhanced in response to EGF treatment, while total receptor levels were constant ( Fig. 1C).

• Download figure
• Open in new tab
• Download powerpoint

Figure 1.

ERRα is a phosphoprotein. A, metabolic labeling of COS-1 cells. Cells were transiently transfected with constructs expressing hERRα or negative control and metabolically labeled with ³²P. Cell lysates were immunoprecipitated with the ERRα antibody, transferred, and exposed to film. B, metabolic labeling of MCF-7 cells. MCF-7 stably transfected with an IPTG-inducible HA-ERRα construct were induced and labeled with ³²P then treated with EGF for 30 minutes. Cell lysates were immunoprecipitated with the anti-HA polyclonal antibody; electrophoresed products were transferred and exposed to film before blotting with the anti-HA monoclonal antibody. C, MCF-7 cells were metabolically labeled with ³²P, treated with EGF for 20 minutes, and lysates immunoprecipitated with the ERRα antibody to detect endogenous ERRα phosphorylation. Equivalent ERRα expression in MCF-7 samples was confirmed by Western blotting. ctl, whole cell extract of transiently transfected ERRα in COS-1 cells; *, nonspecific band. Three replicates were done for each metabolic labeling experiment.

Using programs that predict consensus phosphorylation sites (see Materials and Methods for details), the presence of high probability consensus sites were found predominantly in the DNA-binding domain of ERRα. Hence, we assessed potential effects of endogenous ERRα phosphorylation on DNA binding. MCF-7 cells were treated with EGF and the resulting nuclear extracts subjected to EMSA. ERRα binding to a consensus ERRE-containing probe was enhanced after EGF treatment, as illustrated by the representative autoradiogram in Fig. 2A . Western blot analysis confirmed that EGF-treated and untreated nuclear extracts expressed ERRα equivalently ( Fig. 2A, top right). Densitometric analysis of binding for three independent experiments indicated that EGF treatment of MCF-7 cells enhanced ERRα DNA binding by ∼50% over that of untreated cells ( Fig. 2A, bottom right). The effects of phosphorylation on ERRα DNA binding were also assessed using phosphatase treatment ( Fig. 2B). Transfected ERRα is basally phosphorylated ( Fig. 1A) so nuclear extracts prepared from transfected cells were treated with acid phosphatase to remove any phosphorylation before EMSA. Ectopic ERRα binds to the consensus ERRE as a monomer and a dimer, whereas it binds to an ERE element only as a dimer ( Fig. 2B). Interestingly, dephosphorylation of the nuclear extracts converted ERRα from a primarily dimeric to a monomeric DNA-binding protein on the ERRE element ( Fig. 2B, left). In addition, ERRα dimeric DNA binding on the ERE element was decreased after phosphatase treatment ( Fig. 2B, right). Together, these experiments indicate that phosphorylation of both endogenous and transiently transfected ERRα has a positive effect on its DNA binding as a dimer. It has been suggested that ERRα transactivation function may be dependent on dimeric DNA binding, indicating that dimerization is required for functional interaction with coactivator proteins ( 36). To further assess the significance of ERRα coactivator interactions and dimer functionality, EMSA was done using transiently transfected ERRα nuclear extract in the presence of the coactivator PGC-1α ( Fig. 2C). Incubation of ERRα with PGC-1α produced a supershift of the shifted dimer band, while the mobility of the monomer band was unchanged ( Fig. 2C). Together, these data indicate that phosphorylation enhances ERRα DNA binding as a dimer, which, because of its interaction with PGC-1α, is likely to be the transcriptionally competent form of the receptor.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2.

ERRα phosphorylation enhances DNA binding. A, EMSA of MCF-7 nuclear extracts on a consensus ERRE probe, after treatment with EGF or control for 20 minutes. The supershift reaction was done using the ERRα antibody (Ab). EGF-treated and untreated extracts were also subjected to PAGE and blotted with the ERRα antibody to confirm equivalent expression. Shifted ERRα bands from three separate EMSA experiments as represented in the autoradiogram were analyzed densitometrically. Columns, mean percentage binding for control and EGF-treated samples; bars, SD. B, EMSA of transfected ERRα or negative control nuclear extract on ERRE and ERE probes after treatment of extracts with acid phosphatase (AcP). Supershift reactions were done using the ERRα antibody. D and M denote the positions of dimer and monomer bands, respectively. C, EMSA of transfected ERRα nuclear extract on the ERRE probe. ERRα was supershifted using increasing amounts of a purified GST fusion of PGC-1α. S denotes supershifted band, D and M indicate ERRα dimer and monomer, respectively, and NS denotes a nonspecific band.

Estrogen-related receptor α is phosphorylated in the DNA-binding domain by protein kinase Cδ, which enhances its DNA binding and transcriptional activity. Protein sequence analysis of ERRα revealed the presence of three high probability consensus phosphorylation sites for PKCδ in the DNA-binding domain of the receptor. Thus, PKCδ was tested for its ability to phosphorylate equivalent amounts of glutathione S-transferase (GST) fusion constructs of the ERRα A/B domain, DNA-binding domain, or ligand-binding domain ( Fig. 3A , top). Purified recombinant PKCδ phosphorylated ERRα in the DNA-binding domain ( Fig. 3A, middle) and a control peptide ( Fig. 3A, bottom), but not in the A/B domain or ligand-binding domain. In EMSA experiments, binding of GST-ERRαDBD to an ERRE element was enhanced in the presence of purified recombinant PKCδ ( Fig. 3B). Given the effects of PKCδ on ERRα in vitro, we investigated its effects on ERRα transactivation function by transient transfection with constitutively active (PKC-CAT) or kinase dead (PKC-CAT-KR) mutants of PKCδ ( 37). PKC-CAT enhanced the basal activity of endogenous ERRα on the TFF1 promoter 2-fold, while the inactive PKC-CAT-KR had no effect ( Fig. 3C). PKCδ can thus phosphorylate ERRα in the DNA-binding domain, which results in enhanced receptor DNA-binding and transcriptional activity.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3.

PKCδ phosphorylates ERRα in the DNA-binding domain and enhances DNA binding. A, in vitro kinase assay using GST fusions of the ERRα A/B domain (GST-AB), DNA-binding domain (GST-DBD), or ligand-binding domain (GST-LBD), plus GST alone negative control or positive control peptide. Samples were incubated with purified recombinant PKCδ in vitro. After washing, proteins were electrophoresed and stained with Coomassie blue (top) and then exposed to film (bottom) to detect ³²P incorporation. Stars in the bottom panel correspond to the location of Coomassie-stained bands in the top panel. B, equivalent amounts of GST-ERRαDBD or GST control were incubated with purified recombinant PKCδ then subjected to EMSA on a consensus ERRE probe. C, transient transfection of MCF-7 cells with constructs expressing constitutively active (PKC-CAT) or a kinase dead mutant (PKC-CAT-KR) of PKCδ, and the TFF1LUC reporter.

Given that ERRα phosphorylation and DNA binding are enhanced by EGF in MCF-7 cells, and that PKCδ can phosphorylate and enhance the DNA binding of ERRα in vitro, we next investigated a link between ERRα and PKCδ signaling in MCF-7 cells. PKCδ is reportedly phosphorylated in response to EGF in fibroblastic cells and keratinocytes ( 38– 40), so to investigate whether such activation also occurs in breast cancer cells, Western blot analysis was done after treatment of cells with EGF over a 1-hour time course ( Fig. 4A ). PKCδ phosphorylation was enhanced after 15 to 30 minutes of EGF treatment, and the signal was maintained for at least 60 minutes, while total PKCδ levels remained constant. The activation time of PKCδ also corresponded with the time (20 minutes) at which ERRα DNA binding was enhanced after EGF treatment ( Fig. 2A). To confirm that PKCδ was also activated in MCF-7 cells after treatment with the PKC activator PMA, Western blot analysis was done after a 1-hour time course ( Fig. 4B). PKCδ phosphorylation increased steadily from 1 to 60 minutes of treatment, thus confirming its activation, while total PKCδ levels remained constant.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4.

PKCδ signaling regulates ERRα DNA binding in MCF-7 cells. A, Western blot analysis of MCF-7 cell extracts after treatment with EGF over a 1-hour time course. Membranes were blotted with the phospho-Thr⁵⁰⁵–specific PKCδ antibody then with the total PKCδ antibody. B, Western blot analysis of MCF-7 cell extracts after treatment with PMA over a 1-hour time course. Membranes were blotted as in (A). C, EMSA of MCF-7 nuclear extracts after treatment with PMA over a 1-hour time course on the consensus ERRE probe. The supershift was done using the ERRα antibody. Shifted bands were analyzed densitometrically and graphed as percentage increase binding over untreated control. Results are representative of three individual EMSAs, each of which gave at least a doubling of ERRα DNA binding after 10 minutes of PMA treatment.

Because PKCδ is activated by both EGF and PMA in MCF-7 cells, and EGF stimulates ERRα DNA binding in these cells, we did an EMSA experiment to assess the effects of PMA on ERRα DNA binding over a 1-hour time course ( Fig. 4C). ERRα binding to the ERRE was enhanced to double that of the untreated control after 10 minutes PMA treatment, and this increase was maintained for the length of the time course. Maintenance of enhanced ERRα binding was consistent with the progressive activation of PKCδ from 1 to 60 minutes of PMA treatment. In summary, PKCδ is activated in MCF-7 cells by both EGF and PMA, and this activation produces a concomitant enhancement of ERRα DNA binding.

Estrogen-related receptor α activation by protein kinase Cδ signaling induces selective target promoter activation. The effects of PKCδ signaling on ERRα function were further examined using transient transfection experiments in MCF-7 cells. Cells were transfected with reporter constructs containing the promoter of either the breast cancer marker TFF1 or the ERRα gene, ESRRA. Cells were then treated with PMA to activate endogenous PKCδ ( Fig. 5A ). PMA treatment greatly enhanced the activity of endogenous ERRα on the TFF1 promoter ( Fig. 5A, left), whereas in contrast it did not change its activity on the ESRRA promoter ( Fig. 5A, right). The TFF1 promoter contains both ERRE and ERE elements ( 17); thus, to further investigate PMA-induced ERRα transactivation on this promoter, transfections were done using constructs with either or both of these elements deleted ( Fig. 5B). Deletion of the ERE did not change the PMA-induced activation, while deletion of the ERRE or both elements together inhibited the response by ∼40%. These data indicate that the effects of PMA on the TFF1 promoter are partly through enhanced ERRα binding to the ERRE, but not through the binding of ERRα or ERα to the ERE. The partial response on the TFF1ΔERE/ERRE construct likely relates to the observation that PMA can activate proteins other than PKC depending on the functional concentration of the compound in distinct cellular systems ( 26). Such activation could result in the binding of other proteins to the TFF1 promoter through sites distinct from the ERRE and ERE.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5.

ERRα activation by PKCδ selectively induces target promoter activation. A, MCF-7 cells were transiently transfected with either the TFF1LUC or pGL3ESRRA reporter then treated with PMA for 8 hours before harvesting. B, MCF-7 cells were transiently transfected with the TFF1LUC, TFF1ΔERELUC, TFF1ΔERRELUC, or TFF1ΔERE/ERRELUC reporter. The deletion constructs are the same as TFF1LUC except that they lack either the ERE or ERRE or both binding sequences. Cells were treated with PMA as in (A). In all transfections, the measured Luciferase activity was corrected using the β-galactosidase transfection efficiency control. All results are representative of at least three independent transfections done in duplicate.

To further investigate the apparent selectivity in target gene activation by ERRα after phosphorylation, chromatin immunoprecipitation experiments were done on the TFF1 and ESRRA promoters. MCF-7 cells were treated with EGF and chromatin was immunoprecipitated with the ERRα antibody. PCR analysis was done using primers encompassing either the ERRE region of the TFF1 promoter or a negative control region 4 kb upstream ( Fig. 6A ). ERRα recruitment to the ERRE region of the TFF1 promoter was enhanced after treatment with EGF compared with control ( Fig. 6A, top), but it was not recruited to the upstream negative control region. Quantitative real-time PCR analysis ( Fig. 6A, graph on top) showed that EGF produced a 5-fold enhancement of ERRα recruitment to the TFF1 promoter compared with control. Similarly, recruitment of RNA polymerase II to the TFF1 promoter ERRE region was augmented after EGF treatment compared with control ( Fig. 6A, middle). Quantitative analysis ( Fig. 6A, graph on middle) showed that EGF produced a 3- to 4-fold enrichment of RNA polymerase II at the ERRE region. PCR analysis after ERRα immunoprecipitation was next done using primers that cover the ERRE region of the ESRRA promoter or a negative control region 4 kb upstream ( Fig. 6B). In contrast to the TFF1 promoter, ERRα recruitment to the ESRRA promoter did not change after EGF treatment compared with control ( Fig. 6B, top). There was also no recruitment of ERRα to the negative control region of the ESRRA promoter, confirming the specificity of the chromatin immunoprecipitation. Quantitative real-time PCR analysis ( Fig. 6B, graph on top) confirmed the steady-state of ERRα recruitment to the ESRRA promoter in that the fold enrichment was the same without or with EGF. In agreement with this observation, recruitment of RNA polymerase II to the ERRE region was altered only slightly by EGF, as estimated both by electrophoretic and quantitative analysis ( Fig. 6B, bottom). Finally, chromatin immunoprecipitation analysis was used to confirm that EGF does not affect recruitment of ERα to the TFF1 promoter. EGF-treated chromatin was immunoprecipitated with the ERα antibody and PCR done using the TFF1 promoter primers, as the ERE is located only 124 bp upstream of the ERRE. ERα recruitment to the TFF1 promoter was unchanged by EGF treatment compared with control ( Fig. 6A, bottom). These data indicate that EGF enhances recruitment of both ERRα and RNA polymerase II to the TFF1 promoter, but not to the ESRRA promoter, and confirm that the EGF-induced effects on the TFF1 promoter are specifically through ERRα, and are not influenced by ERα. The presence of more RNA polymerase II at the TFF1 promoter suggests that the TFF1 gene is activated for transcription after EGF treatment, while the ESRRA gene is not. Quantitative reverse transcription-PCR (RT-PCR) analysis ( Fig. 6C) was used to show that TFF1 transcript levels were increased 2.3-fold after EGF treatment compared with untreated cells, whereas levels of ESRRA were not significantly affected. Together, these data indicate that EGF selectively up-regulates ERRα recruitment to the TFF1 promoter, leading to enhanced transcription of this gene.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6.

EGF signaling differentially regulates ERRα target promoter recruitment in vivo. Chromatin immunoprecipitation of MCF-7 cells treated with EGF for 20 minutes or untreated control. A, schematic of the TFF1 promoter indicates the location of primers for positive ERRE or 4 kb upstream negative control amplicons. Chromatin was immunoprecipitated with the ERRα, RNA polymerase II, or ERα antibodies (+) or preimmune serum (−) followed by PCR analysis with TFF1 promoter ERRE (top) or negative control (bottom) primers. i, input. Graph, quantitative RT-PCR analysis of control or EGF-treated chromatin using TFF1 ERRE primers. B, schematic of the ESRRA promoter indicates the location of primers for positive ERRE or 4 kb upstream negative control amplicons. Chromatin was immunoprecipitated with the ERRα or RNA polymerase II antibodies (+) or preimmune serum (−) followed by PCR analysis with ESRRA promoter ERRE (top) or negative control (bottom) primers. Graph, quantitative RT-PCR analysis of chromatin using ESRRA ERRE primers. C, RT-PCR analysis of TFF1 and ESRRA expression in untreated (−) or EGF-treated MCF-7 cells. Expression levels of both genes were corrected for that of the RPLP control.

Discussion

Phosphorylation of nuclear receptors is an important mechanism for regulation of their transcriptional activity. In the case of classic steroid receptors, phosphorylation provides ligand-independent regulation in addition to their hormone-dependent activation, while for orphan nuclear receptors it provides regulation in the absence of cognate ligand. In this report, we investigated a link between ERRα phosphorylation and its transcriptional function in breast cancer cells. ERRα was found to be phosphorylated in vivo in response to EGF signaling, which enhanced receptor DNA binding. In addition, ERRα was phosphorylated in the DNA-binding domain in vitro by PKCδ, an event that also enhanced its DNA binding. PKCδ signaling selectively enhanced ERRα transactivation function in MCF-7 cells, stimulating ERRα-dependent activation of the TFF1 promoter, but not affecting receptor activity on its own promoter. The selective promoter activation by ERRα was confirmed using chromatin immunoprecipitation experiments. EGF stimulation resulted in enhanced occupation of the TFF1 promoter by ERRα and RNA polymerase II, and led to enhanced TFF1 transcription. In contrast, neither ERRα nor RNA polymerase II recruitment to the ESRRA promoter was changed after EGF treatment. We have hence identified a signaling pathway by which ERRα is phosphorylated and provides promoter-specific activation of target genes.

The function of several nuclear hormone receptors is regulated by phosphorylation in the DNA-binding domain, leading to both positive and negative effects on receptor DNA binding. Phosphorylation of the HNF4 (NR2A1) DNA-binding domain enhances DNA binding ( 41), while phosphorylation of both NGF1-B (NR4A1) and T₃Rα1 (NR1A1) in the DNA-binding domain by PKA inhibits DNA binding ( 42, 43). PKC family members directly phosphorylate vitamin D receptor (VDR, NR1I1), retinoic acid receptor α (RARα, NR1B1), and COUP-TF1 (NR2F1) in the DNA-binding domain, with distinct transcriptional outcomes. VDR phosphorylation by PKCβ inhibits binding to the VDRE, reportedly because of a receptor conformation change in the region adjacent to the phosphorylation site, as well as through overall altered tertiary structure of the receptor ( 44). RARα is phosphorylated by PKCα and PKCγ in the extended carboxyl-terminal region of the DNA-binding domain, which inhibits dimerization with the retinoid X receptors, potentially by producing a direct conformational constraint in the dimerization interface or through a conformational change in the receptor DNA-binding domain and ligand-binding domain that reorientates the dimerization interface ( 45). COUP-TF1 phosphorylation by PKC in the DNA-binding domain enhances its DNA binding but does not change receptor dimerization, so the increased DNA binding is thought to be via direct phosphorylation-induced steric or electrostatic changes leading to enhanced receptor affinity for DNA ( 46). In this study, we have shown that EGF-induced ERRα phosphorylation enhances its DNA binding, and phosphatase treatment of the receptor suggests that this occurs through enhanced dimer binding ( Fig. 2). A distinctive feature of our findings is that phosphorylation of the DNA-binding domain of ERRα seems to change not only the DNA-binding capacity of the receptor, but also how the protein interacts with DNA, as a monomer or as a dimer, suggesting a mechanism for selective gene regulation by a receptor possessing the ability to bind DNA in both forms.

We have shown that PKCδ directly phosphorylates ERRα in vitro and increases its DNA binding ( Fig. 3), and that PKCδ activation augments ERRα transactivation of the TFF1 promoter in breast cancer cells ( Fig. 5). Transfections with TFF1 promoter deletion constructs indicate the effects of PMA on TFF1 activation are specifically mediated through ERRα and not ERα. While it is possible that the effects on the ERRE could be mediated via ERα, this is unlikely as it preferentially binds the inverted repeat ERE element and its activation through the ERRE consensus sequence is promoter specific ( 47). Evidence suggests that PKCδ could be an in vivo regulator of the EGF-induced effects on ERRα phosphorylation and transcriptional activity. PKCδ is activated via EGFR/ErbB2 signaling, an effect that is blocked by EGFR inhibitors ( 38– 40). We have also shown that PKCδ is activated in MCF-7 cells by EGF ( Fig. 4). In addition, activated PKCδ localizes to the nucleus ( 48) and can associate with nuclear receptors in vivo. PKCδ has previously been shown to associate with ligand-activated RARα nuclear complexes that bind to retinoic acid response elements ( 49). Together, these data indicate that PKCδ has the potential to be an in vivo kinase for ERRα, acting downstream of the EGF signal to provide phosphorylation and promoter-specific activation of ERRα.

Chromatin immunoprecipitation experiments showed that EGF signaling enhanced ERRα enrichment on the TFF1 promoter, but not on the ESRRA promoter ( Fig. 6), indicating that receptor phosphorylation status selectively regulates promoter occupancy. This selectivity likely involves structural and conformational constraints of both the receptor and target gene promoter. As discussed above, ERRα likely undergoes phosphorylation-induced conformational changes that affect the tertiary structure of the DNA-binding domain itself, as well as the receptor as a whole. In addition, the individual promoter structure will likely be dictated by both the sequence in the immediate vicinity of the ERRE elements and by the nature of the flanking regions. The specific conformation of phosphorylated ERRα in combination with distinct promoter structural features likely contributes to the preferred affinity of the receptor for the TFF1 promoter over the ESRRA promoter. The distinct conformation of the TFF1-bound ERRα may also selectively enhance coactivator recruitment, thus dictating the extent of activation of the specific promoter. The selective enhancement of RNA polymerase II occupancy on the TFF1 promoter, and increased TFF1 RNA expression after EGF treatment ( Fig. 6) provide evidence for such selective promoter activation.

A potential role for ERRα in breast cancer is becoming more evident, as its expression correlates with an aggressive tumor phenotype in paralleling ErbB2 overexpression ( 22). In addition, its presence in tumor samples associates with an increased risk of disease recurrence or adverse clinical outcome ( 21). There is also some evidence to suggest that PKCδ could be important in tumor cells with an aggressive phenotype, as its expression correlates with breast tumor line metastatic potential ( 30). It could be hypothesized (see model Fig. 7 ) that enhanced EGF/ErbB2 signaling in tumor cells activates PKCδ and promotes its nuclear localization. This may lead to ERRα phosphorylation, which would promote its dimeric DNA binding and subsequent activation of distinct target promoters. Such events may provide a mechanism to selectively modulate ERRα target genes involved in breast cancer. The phosphorylation-induced selective promoter activation by ERRα may have important downstream transcriptional consequences and, therefore, it will be of interest to identify the full range of genes up-regulated by EGF-induced activation of ERRα in MCF-7 cells to help understand the potential role played by this receptor in breast cancer cells.

• Download figure
• Open in new tab
• Download powerpoint

Figure 7.

Schematic model of ERRα phosphorylation-induced selective target gene activation. In this model, the EGF signal is transmitted through the EGFR/ErbB2 receptor tyrosine kinases, leading to activation of downstream signaling pathways and phosphorylation of PKCδ. PKCδ activation enhances its nuclear localization and promotes phosphorylation of ERRα. This event produces a conformational change in the receptor and selectively enhances ERRα dimeric DNA binding, resulting in promoter-specific activation of ERRα target genes.