LIGAND STRUCTURE-DEPENDENT ACTIVATION OF ESTROGEN RECEPTOR α/Sp BY ESTROGENS AND XENOESTROGENS

Abstract

This study investigated the effects of E2, diethylstilbestrol (DES), antiestrogens, the phytoestrogen resveratrol, and the xenoestrogens octylphenol (OP), nonylphenol (NP), endosulfan, kepone, 2,3,4,5-tetrachlorobiphenyl-4-ol (HO-PCB-Cl₄), bisphenol-A(BPA), and 2,2-bis-(p-hydroxyphenyl)-1,1,1-trichloroethane (HPTE) on induction of luciferase activity in breast cancer cells transfected with a construct (pSp1₃) containing three tandem GC-rich Sp binding sites linked to luciferase and wild-type or variant ERα. The results showed that induction of luciferase activity was highly structure-dependent in both MCF-7 and MDA-MB-231 cells. Moreover, RNA interference assays using small inhibitory RNAs for Sp1, Sp3 and Sp4 also demonstrated structure-dependent differences in activation of ERα/Sp1, ERα/Sp3 and ERα/Sp4. These results demonstrate for the first time that various structural classes of ER ligands differentially activate wild-type and variant ERα/Sp-dependent transactivation, selectively use different Sp proteins, and exhibit selective ER modulator (SERM)-like activity.

1. Introduction

Estrogen receptor α (ERα) and ERβ are the two major ER sub-types, and the classical mechanism of estrogen action involves ligand-induced dimerization of ER which interacts with estrogen responsive elements (EREs) in target gene promoters and results in transcriptional activation [1; 2]. This latter process is complex and involves interactions of the ligand-bound receptor with nuclear coactivators and other coregulatory proteins and components of the basal transcription machinery [3; 4].

Ligand-dependent activation or inhibition of ER-dependent transactivation depends on several factors including ligand structure, cell/tissue-specific expression coactivators/coregulatory proteins, gene promoter and cell context [3]. The development of selective ER modulators (SERMs) such as tamoxifen and raloxifene for treatment of breast cancer and other hormone-related problems is due to this complex pharmacology in which individual SERMs exhibit tissue-specific ER agonist or antagonist activities [5–7]. Several in vitro assays for estrogenic activity using wild-type and variant forms of ERα and ERE-promoter-reporter constructs can distinguish between 17β-estradiol (E2) and different SERMs such as tamoxifen, raloxifene and "pure" antiestrogens such as ICI 164,384 and ICI 182,780 [8; 9]. Moreover, studies in this laboratory have shown that structurally-diverse synthetic industrial estrogenic compounds (xenoestrogens) differentially activate ERE-promoters in cells transfected with wild-type and variant ERα expression plasmids suggesting that these compounds also exhibit SERM-like activity [10–13].

E2-dependent transactivation through nuclear pathways also involves non-classical mechanisms where the liganded ER interacts with other DNA-bound transcription factors including specificity proteins (Sp), activator protein-1 (AP-1), nuclear factor κB (NFκB), and globin transcription factor (GATA) [14–17]. ERα/Sp-dependent transactivation is responsible for activation of several genes in breast cancer cells responsible for cell proliferation, cell signaling, and nucleotide metabolism [17]. Ligand-dependent activation of ERα/Sp has been observed for both estrogens and antiestrogens such as 4′-hydroxytamoxifen (4-OHT) and ICI 182,780; however, in studies using a construct (pSp1₃) containing three GC-rich Sp protein binding sites, activation by estrogens and antiestrogens requires different domains of ERα [17–20]. For example, E2 activates pSp1₃ in cells transfected with wild-type ERα or DNA binding domain (DBD) mutants of ERα containing deletions of zinc finger 1 (ERαΔZF1) or zinc finger 2 (ERαΔZF2), whereas ICI 182,780 or 4-OHT activate ERα but not the DBD mutants [19; 20]. pSp1₃ was not activated in cells transfected with ERβ [19].

In this study, we investigated the structure-dependent activation of ERα/Sp1 by a series of xenoestrogens including octylphenol (OP), nonylphenol (NP), endosulfan, kepone, 2,3,4,5-tetrachlorobiphenyl-4-ol (HO-PCB-Cl₄), bisphenol-A (BPA), and 2,2-bis-(p-hydroxyphenyl)-1,1,1-trichloroethane (HPTE). E2, diethylstilbestrol (DES), 4-OHT, resveratrol (a phytoestrogen), and ICI 182,780 were also used as reference compounds for the study. With the exception of resveratrol, all compounds induced transactivation in breast cancer cells transfected with ERα and pSp1₃; however, activation of pSp1₃ in cells transfected with variant forms of ERα was structure-dependent. Moreover, using RNA interference that selectively decreases Sp1, Sp3 or Sp4 protein expression, we showed that xenoestrogens, E2 and antiestrogen selectively activate ERα/Sp1, ERα/Sp3 and ERα/Sp4.

2. Materials and methods

2.1 Chemicals, biochemicals and plasmids

Fetal bovine serum (FBS) was obtained from JRH Biosciences (Lenexa, KS). Antibiotic antimycotic solution (AAS) (x100) was obtained from Sigma-Aldrich (St Louis, MO). The following test chemicals (and purities) were purchased from Sigma-Aldrich: E2 (≥ 98%), 4-OHT (≥ 98%), resveratrol (> 99%), p-t-octylphenol (97%), p-nonylphenol (98%) and BPA (>99%). HO-PCB-Cl₄ was >98% pure as previously described [21]; HPTE (>98%) was synthesized as previously reported [22]. Kepone (98%) and endosulfan were purchased from Chem-Service (West Chester, PA). ICI 182,780 was provided by Dr. Alan Wakeling (Astra-Zeneca, Macclesfield, UK). Plasmid preparation kits were purchased from Sigma-Aldrich. All other chemicals were obtained from commercial sources at the highest quality available. Human ERα expression plasmid was kindly provided by Dr. Ming-Jer Tsai (Baylor College of Medicine, Houston, TX); ERαΔZF1, ERαΔZF2, ERα(1–537), ERα(1–553), CFP-Sp1, YFP-ERα and CFP-YPF chimeras were made as previously reported [20; 23].

2.2 Cells and transient transfection assays

MCF-7 and MDA-MB-231 breast cancer cells were obtained from American Type Culture Collection (ATCC, Manassas, VA). MCF-7 and MDA-MB-231 cells were maintained in Dulbecco's modified Eagle's medium/nutrient mixture Ham's F-12 (DME/F-12) (Sigma-Aldrich) supplemented with 2.2 g/liter sodium bicarbonate, 5% FBS, and 5 ml/liter AAS. Cells were cultured and grown in a 37°C incubator with humidified 5% CO₂, 95% air. For transient transfection studies, MCF-7 or MDA-MB-231 cells were seeded in 12-well plates in DME/F-12 medium without phenol red supplemented with 2.2 g/liter sodium bicarbonate, and 2.5% charcoal-stripped FBS. After 24 hr, cells were transfected using the calcium phosphate transfection method with 350 ng of luciferase reporter construct (pSp1₃), 100 ng pcDNA3/His/lacZ (Invitrogen, Carlsbad, CA) as a standard reference for transfection efficiency, and 200 ng of the appropriate ER expression plasmid. The pSp1₃ construct and other plasmids containing E2-responsive GC-rich promoter inserts are not responsive to E2 even in MCF-7 cells and this is due to minimal TATA promoter and overexpression of the transfected construct which results in limiting levels of ERα [11; 17–20]. E2- responsiveness requires cotransfection with ERα. Six hr after transfection, cells were shocked with 25% glycerol/PBS for 1 min, washed with PBS, and then treated with dimethylsulfoxide (DMSO, solvent) or different concentrations of estrogens, antiestrogens or xenoestrogens in DMSO for another 20 – 24 hr. Cells were then washed twice in PBS and harvested with 100 µL of reporter lysis buffer (Promega Corp., Madison, MI). After two freeze-thaw cycles, cell lysates were centrifuged for 1 min at 16,000 g, and the supernatant was used for determination of protein activity. Luciferase (Promega Corporation, Madison, MI) and β-galactosidase activity was determined using the Tropix Galacto-Light Plus assay system (Tropix, Bedford, MA). Light emission was detected on a lumicount micro-well plate reader (Packard Instruments, Meriden, CT), and luciferase activity was calculated by normalizing against β-galactosidase activity obtained from the same sample and compared with the DMSO control group (set at 100%) for each set of experiments.

2.3 RNA interference assay in cells transfected with iSp oligonucleotides

MCF-7 cells (5×10⁴) were cultured in phenol red-free DME/F12 supplemented with 2.5% charcoal-stripped FBS without antibiotic in 12-well plates for overnight. siRNA (25 nM) was transfected by Lipofectamine 2000 reagent (Invitrogen), according to the manufacturer’s protocol. After 6 hr, the transfection medium was changed with fresh DME/F-12 and 2.5% serum without phenol red. The next day, following the manufacturer’s instructions, Genejuice transfection reagent (EMD Biosciences Inc., San Diego, CA) was used to transfect cells with 200 ng of luciferase reporter construct (pSp1₃), 50 ng pcDNA3/His/lacZ (Invitrogen) as a standard reference for transfection efficiency, and 100 ng of the human ERα expression plasmid. Six hr later, cells were treated with DMSO or estrogens, antiestrogens, xenoestrogens in antibiotic-free, phenol red-free DMEM/F12 with 2.5% serum. Cells were harvested 48 – 54 hr after siRNA transfection. Cell lysates were assayed for luciferase and β-galactosidase activity as described above.

The siRNA duplexes used in this study are indicated as follows. Silencer® Negative Control #1 siRNA purchased from Ambion (Austin, TX) was used as the non-specific control (iNS). The luciferase GL2 duplex (target sequence, 5′ - CGT ACG CGG AAT ACT TCG A - 3′) RNA from Dharmacon (Lafayette, CO) was used as the positive control in siRNA transfections. The siRNA oligonucleotides for Sp1, Sp3, and Sp4 were also obtained from Dharmacon as follows: Sp1, 5′ - AUC ACU CCA UGG AUG AAA UGA dTdT - 3′; Sp3, 5′ - GCG GCA GGU GGA GCC UUC ACU dTdT - 3′; and Sp4, 5′ - GCA GUG ACA CAU UAG UGA GCdT dT - 3′.

2.4 Western blot analysis

MCF-7 cells were seeded into six-well plates in DMEM/F12 supplemented with 2.5% charcoal-stripped FBS. The next day, cells were transfected with siRNA as described above. Fourty-eight hr after transfection, protein was extracted from the cells by harvesting in a high-salt lysis buffer [50 mM HEPES (pH 7.5), 150 mM NaCl, 10% (vol/vol) glycerol, 1% Triton X-100, 1.5 mM MgCl2, 1 mM EGTA, protease inhibitor cocktail (Sigma)] on ice for 45 – 60 min with frequent vortex and centrifugation at 20,000 g for 10 min at 4°C. Protein concentrations were determined using a Bio-rad (Hercules, CA) protein assay reagent. Protein (60 µg) was diluted with Laemmli’s loading buffer, boiled, and loaded onto 7.5% SDS-PAGE. Samples were resolved using electrophoresis at 150V for 3 – 4 hr and transferred (transfer buffer, 48 mM Tris-HCl, 29 mM glycine, and 0.025% sodium dodecyl sulfate) to a polyvinylidene difluoride membrane (Bio-Rad, Hercules, CA) by electrophoresis at 0.2 A for approximately 12 – 16 hr. Membranes were blocked in 5% TBS-Tween 20-Blotto [10 mmol/L Tris-HCl, 150 mmol/L NaCl (pH 8.0), 0.05% Triton X-100, 5% nonfat dry milk] with gentle shaking for 30 min and incubated in fresh 5% TBS-Tween 20-Blotto with 1:1,000 (for Sp1 and Sp3), 1:500 (for Sp4), and 1:5,000 (for α-actin) primary antibody overnight with gentle shaking at 4°C. The primary antibodies for Sp1 (PEP2), Sp3 (D-20), and Sp4 (V-20) were purchased from Santa Cruz Biotechnology Inc. (Santa Cruz, CA) and antibody for α-actin was purchased from Sigma. Membranes were probed with a horseradish peroxidase-conjugated secondary antibody (1:5000) (Santa Cruz) for 3 – 6 hr at 4°C. Blots were visualized using the chemiluminescence substrate (Perkin-Elmer Life Sciences) and exposure on Image Tek-H x-ray film (American X-ray Supply). Band quantitation was performed by Image (NIH).

2.5 Fluorescence resonance energy transfer (FRET)

FRET was performed with the Stallion DDI workstation equipped with a xenon fluorescent light source (300 W) with rapid switching (<2 msec) between excitation wavelengths and two CoolSnap HQ cameras for simultaneous detection of two emission signals. Three types of images were collected for the FRET experiments. The first type was the CFP control (donor only) which was used to calculate the spectral bleed-through of CFP emission visible through the FRET filter set. This type of image was collected using the CFP filter set and the FRET filter set. The second type of image, the YFP control (acceptor only), was used to calculate the spectral bleed-through of normal YFP emission through the FRET filter set. This image was collected using the YFP filter set and the FRET filter set. The third type of image is the raw FRET image (FRETraw) representing the sample under analysis. This type of image was collected using the FRET, CFP and YFP filter sets. Corrected FRET (FRETc), was calculated with consideration of all three images [24] using the following equation:

$$\mathrm{\text{FRETc}}=\mathrm{\text{FRETraw}}-\mathrm{\text{Df}}/\mathrm{\text{Dd}}[\mathrm{\text{CFP}}]-\mathrm{\text{Df}}/\mathrm{\text{Da}}[\mathrm{\text{YFP}}]$$

where FRETraw, [CFP], and [YFP] are the signals visualized through the FRET, CFP, and YFP filter sets, respectively. The constants Df/Dd and Df/Da are the bleed-through constants describing donor emission visible in FRET channel and direct excitation of acceptor, respectively, calculated from the first and second type images. FRET efficiency was then calculated by comparing the constructs under investigation to the CFP-YFP chimera (positive control) [23].

2.6 Real-time PCR

MCF-7 cells (3×10⁶) were seeded into six-well plates in DMEM/F-12 supplemented with 2.5% charcoal-stripped FBS. The next day, cells were transfected with non-specific siRNA or iSp1/iSp3/iSp4 as described above. Forty-eight hr after transfection, cells were treated with 10 nM E2, 20 µM HPTE, 5 µM kepone for 6 hr. Total RNA was isolated using the RNeasy Protection Mini Kit (QIAGEN, Valencia, CA) according to the manufacturer’s protocol and treated with RNase-free DNase set (QIAGEN). RNA was eluted with 50 µl RNase-free water. RNA was reverse transcribed using Superscript III reverse transcriptase (Invitrogen) according to the manufacturer’s protocol. PCR was carried out using SYBR Green PCR Master Mix from PE Applied Biosystems (Warrington, UK) on an ABI Prism 7700 Sequence Detection System (PE Applied Biosystems). Each reaction contained 0.5 µM primer and 2 µl cDNA template and made up to 25 µl. The relative quantitation of samples was carried out using comparative Ct method. TATA binding protein (TBP) was used for normalization. Cad primers were purchased from QIAGEN and TBP primers were synthesized by Integrated DNA Technologies (Coralville, IA). TBP (forward), 5′-TGC ACA GGA GCC AAG AGT GAA-3′; and TBP (reverse), 5′-CAC ATC ACA GCT CCC CAC CA-3′.

2.7 Statistical analysis

For transient transfection studies, results are expressed as means ± SD for at least three separate experiments for each treatment group. Statistical differences (p < 0.05) between control (DMSO) and treatment groups were determined by single factor ANOVA test. For FRET efficiency analysis, data were analyzed using one-way ANOVA followed by Dunnett's multiple comparison test at p < 0.05.

3. Results

3.1 Ligand structure-dependent activation of wild-type ERα/Sp

pSp1₃ is a plasmid containing three consensus GC-rich sites with a minimal TATA promoter linked to the luciferase gene. This construct is not inducible by E2 in ER-positive MCF-7 or ZR-75 cells and E2-responsiveness is only observed after cotransfection with ERα or variant forms of ERα [17–20; 23; 25; 26]. This is due to overexpression of the construct where endogenous ERα in MCF-7 cells becomes limiting. The lack of responsiveness without cotransfection of ERα is also due to the minimal TATA promoter. This has proven to be an excellent model system for determining ligand structure-dependent requirements for different domains of ERα in transactivation studies in a breast cancer cell (ER-positive or ER-negative) context [10–12; 17–20]. ERE-promoter-reporter constructs have been extensively used as models for understanding the classical ERα-mediated transactivation pathways, and pSp1₃ is a model for investigating mechanisms of ERα/Sp-mediated gene expression [17]. The effect of ligand structure on activation of ERα/Sp-dependent transactivation was investigated in MCF-7 cells transfected with pSp1₃ and treated with different concentrations of E2, DES, ICI 182,780 and 4-OHT (Figs. 1A and 1B). Both E2 and DES induced luciferase activity and similar results were observed for the antiestrogens 4-OHT and ICI 182,780. The induced transactivation for the antiestrogens was similar to previous studies using this same construct (pSp1₃) in both ER-positive and -negative cancer cell lines [20]. Several reports show that xenoestrogens and phytoestrogens activate gene expression in MCF-7 cells transfected with ERE constructs [10–13]; however, their activation of pSp1₃ has not previously been reported and is investigated in this study.

Figure 1.

Activation of ERα/Sp by estrogens, antiestrogens, xenoestrogens and resveratrol in MCF-7 cells. Induction by E2/DES (A) and 4-OHT/ICI 182,780 (B). MCF-7 cells were transfected with pSp1₃ and ERα expression plasmid, treated with different concentrations of E2, DES, 4-OHT or ICI 182,780, and luciferase activity determined as described in the Materials and Methods. Each point is the mean ± SD for three replicate determinations and significant (p < 0.05) induction is indicated (*). E2 does not induce luciferase activity in MCF-7 cells transfected with pSp1₃ alone and E2-responsiveness requires cotransfection with ERα. Induction by HPTE/BPA (C), octylphenol/nonylphenol (D), HO-PCB-Cl₄/resveratrol (E), kepone/endosulfan (F). MCF-7 cells were transfected with pSp1₃ and ERα, treated with different concentrations of the various ERα agonists and luciferase activity determined as described in the Materials and Methods. Each point is the mean ± SD for three replicate determinations and significant (p < 0.05) induction is indicated (*).

The results in Figure 1C summarize the concentration-dependent activation of pSp1₃ by two bisphenolic estrogenic compounds HPTE and BPA; both compounds induced luciferase activity, with HPTE being the more potent of the two bisphenols. Octyl- and nonylphenol (Fig. 1D) also induced transactivation in cells transfected with pSp1₃ and, among the bisphenolic and phenolics xenoestrogens, their induction response varied from 4- to 8-fold. A third phenolic compound HO-PCB-Cl₄ also induced approximately a 4-fold increase in luciferase activity, whereas resveratrol, a trihydroxystilbene analog was inactive in this assay (Fig. 1E). Both endosulfan and kepone are chlorinated caged xenoestrogenic compounds and the potency of endosulfan was similar to that observed for the phenolic/bisphenolic compounds (6.5-fold induction); however, a > 25-fold increase in luciferase activity was observed in MCF-7 cells transfected with pSp1₃ and treated with kepone which induced the highest activity among all the 12 compounds tested in this assay (Fig. 1F). All compounds induced a concentration-dependent induction of luciferase activity and, in subsequent studies, we used concentrations that induced maximal activity with no cytotoxicity.

3.2 Ligand structure-dependent activation of wild-type and mutant ERα/Sp

Previous studies showed that although E2 and antiestrogens activate wild-type ERα/Sp-dependent transactivation in MCF-7 cells transfected with pSp1₃; however, activation of deletion and point mutant forms of ERα (transfected) are ligand structure-dependent [19; 20]. We therefore compared the effects of different structural classes of xenoestrogens and resveratrol on induction of luciferase activity in MCF-7 cells transfected with pSp1₃ and wild-type ERα (Fig. 2A), ERαΔZF1 (Fig. 2B), ERαΔZF2 (Fig. 2C), ERα(1–553) (Fig. 2D), and ERα(1–537) (Fig. 2E). These ERα mutants contained deletions of zinc finger 1 (amino acids 185–205), zinc finger 2 (amino acids 218–243), the F domain (amino acids 554–595), and the F domain plus amino acids in helix 12 of the E domain (amino acids 538–595), respectively. With the exception of resveratrol, all compounds induce transactivation in MCF-7 cells transfected with pSp1₃/ERα (Fig. 2A) as observed in Figure 1. The results in Figures 2B and 2C show that the antiestrogens ICI 182,780 and 4-OHT did not induce transactivation in MCF-7 cells transfected with pSp1₃ and ERαΔZF1 or ERαΔZF2 as previously reported [19; 20]. In contrast, both E2 and DES significantly activated luciferase activity, and the xenoestrogens also induced transactivation and resembled E2. Resveratrol was also significantly active, and maximally induced responses were observed for HPTE, kepone and HO-PCB-Cl₄ in cells transfected with both zinc finger deletion mutants. The fold-induction responses were decreased in cells transfected with ERα(1–553) in which the F domain has been deleted; however, in contrast to a previous report [20], E2 and the remaining compounds all significantly induced transactivation (Fig. 2C), and HPTE, HO-PCB-Cl₄ and kepone were the most potent inducers. Major structure-dependent differences were observed in MCF-7 cells transfected with ERα(1–537); E2 and DES did not induce transactivation, whereas the antiestrogens, resveratrol and xenoestrogens were all inducers of luciferase activity with kepone, HO-PCB-Cl₄ and HPTE among the most active compounds. Since ERα(1–537) does not contain an intact helix-12 which is important for coactivator-ERα interactions, this suggests that other nuclear cofactors may be involved and this is currently being investigated.

Figure 2.

Structure-dependent activation of variant ERα/Sp in MCF-7 cells/Transfection with wild-type ERα (A), ERαΔZF1 (B), ERαΔZF2 (C), ERα(1–553) (D), and ERα(1–537) (E). Cells were transfected with pSp1₃ and variant forms of ERα, treated with ERα agonists and luciferase activity determined as described in the Materials and Methods. Results are expressed as means ± SD for three replicate determinations for each treatment group and significant (p < 0.05) induction is indicated (*). The concentrations used in this experiment were those which maximally induced luciferase activity using pSp1₃ and wild-type ERα (Fig. 1 and Fig. 2) without causing any cytotoxicity.

We also investigated ligand structure-dependent activation of ERα/Sp-dependent transactivation in ER-negative MDA-MB-231 breast cancer cells transfected with pSp1₃, wild-type ERα or the same set of ERα mutants ERαΔZF1, ERαΔZF2, ERα(1–554) and ERα(1–537), respectively (Figs. 3A – 3E). The concentrations of xenoestrogens were initially optimized for this cell line, and the estrogens, antiestrogens and xenoestrogens all induced luciferase activity in cells cotransfected with pSp1₃ and wild-type ERα (Fig. 3A). The pattern of induction in MDA-MB-231 cells transfected with the zinc-finger mutants was similar to that observed in MCF-7 cells; E2 and DES but not the antiestrogens ICI 182,780 or 4-OHT induced activity, and with the exception of the alkylphenols, the xenoestrogens and resveratrol were all active with the latter compound exhibiting significantly higher activity in MDA-MB-231 than in MCF-7 breast cancer cells (Figs. 3B and 3C). In MDA-MB-231 cells transfected with the F domain deletion mutant [ERα(1–553)] (Fig. 3D), all compounds with the exception of the alkylphenols and endosulfan induced activity, and these results were similar to those obtained in MCF-7 cells (Fig. 2D) where only limited induction was observed for alkylphenols and endosulfan. Higher concentrations of these compounds may induce transactivation in MDA-MB-231 cells; however, this was not carried out due to cytotoxicity. The structure-dependent induction of luciferase activity in MDA-MB-231 cells transfected with ERα(1–537) (Fig. 3E) was also similar to that observed in MCF-7 cells (Fig. 2E) in which the xenoestrogens and resveratrol but not E2 or DES were active. These results demonstrate that xenoestrogens, resveratrol, antiestrogens and E2/DES differentially activate wild-type and variant ERα/Sp in ER-negative MDA-MB-231 cells transfected with pSp1₃, and the pattern of activation of wild-type and variant ERα/Sp by xenoestrogens was similar in both MCF-7 and MDA-MB-231 cells.

Figure 3.

Structure-dependent activation of wild-type and variant ERα/Sp in MDA-MB-231 cells. Transfection with wild-type ERα (A), ERαΔZF1 (B), ERαΔZF2 (C), ERα(1–553) (D), and ERα(1–537) (E). MDA-MB-231 cells were transfected with pSp1₃, wild-type or variant ERα, treated with ER agonists and luciferase activity determined as described in the Materials and Methods. Results are expressed as means ± SD for three replicate determinations for each treatment group and significant (p < 0.05) induction is indicated (*). Preliminary concentration-dependent studies were carried out in cells transfected with pSp1₃ and wild-type ERα, and the concentrations of each compound that maximally-induced activity without cytotoxicity were used in all experiments.

3.3 Role of Sp1, Sp3 and Sp4 in activation of a GC-rich construct (pSp1₃)

Previous studies show that hormonal activation of other GC-rich promoters involves ERα/Sp1, ERα/Sp3 and ERα/Sp4 [25; 26]. The role of individual Sp proteins on ligand structure-dependent activation of ERα/Sp in MCF-7 cells transfected with pSp1₃ was further investigated using RNA interference and small inhibitory RNAs for Sp1 (iSp1), Sp3 (iSp3), and Sp4 (iSp4). Results illustrated in Figure 4A summarize Western immunoblot analysis of Sp protein levels in whole cell lysates from MCF-7 cells transfected with iSp1, iSp3, iSp4 and binary combination of these reagents. Since transfection efficiencies are in the range of 60–90%, the results demonstrate highly effective and specific Sp protein knockdown as previously reported using these same siRNAs [25–27]. The siRNAs were initially used to investigate the role of individual Sp proteins in mediating E2-, 4-OHT- and HPTE-mediated activation of pSp1₃ in MCF-7 cells and determine possible structure-dependent utilization of one or more Sp proteins. HPTE was selected as a prototypical xenoestrogen that induces transactivation in cells transfected with wild-type and variant ERα expression plasmids. In all experiments, iSp1, iSp3 and iSp4 decreased basal and ligand-induced activity and therefore the results are expressed as fold induction in order to directly compare the role of individual Sp proteins on ligand-dependent activation of ERα/Sp. Results in Figure 4B demonstrate that iSp1 alone or in combination with iSp3 or iSp4 completely decreased inducibility by E2. iSp4 alone or in combination with iSp1 or iSp3 decreased induction by E2, whereas iSp3 did not affect the induction response. These results suggest that ERα/Sp1 is the major complex activated by E2, and ERα/Sp4 also contributes to this response. In contrast, Sp3 did not significantly decrease E2 inducibility, suggesting that ERα/Sp3 doe not play a role in hormonal activation of pSp1₃. These results were observed in several replicate experiments showing that E2-dependent activation of ERα/Sp followed the order ERα/Sp1 > ERα/Sp4 > ERα/Sp3. Using a similar approach, we also investigated HPTE- and 4-OHT-dependent activation of ERα/Sp (Figs. 4C and 4D, respectively) and the results suggest that ERα/Sp1 and ERα/Sp4 but not ERα/Sp3 play a role in 4-OHT/HPTE activation of pSp1₃. The major difference between E2 and 4-OHT/HPTE over several replicate experiments was the increased role of ERα/Sp4 for the latter two compounds.

Figure 4.

RNA interference studies with iSp1, iSp3 and iSp4. (A) Western immunoblot analysis. MCF-7 cells were transfected with iSp1, iSp3, iSp4, iNS (non-specific), or various combinations, and whole cell lysates were analyzed by Western blot analysis as described in the Materials and Methods. Quantitation of Sp knockdown was determined using three replicates for each treatment and levels were normalized to β-actin. Significantly (p < 0.05) decreased levels are indicated (*) and results are compared to those observed in the DMSO group (set at 1.0). Effects of Sp protein knockdown on activation of pSp1₃ by E2 (B), 4-OHT (C), and HPTE (D). MCF-7 cells were transfected with pSp1₃, ERα and small inhibitory RNAs for iNS or Sp proteins and treated with 10 nM E2, 1 µM 4-OHT, or 25 µM HPTE. Luciferase activity was determined as described in the Materials and Methods, and results are means ± SD for at least three replications for each treatment group. Significantly (p < 0.05) induced activity relative to DMSO is indicated (*) and a decrease in induced activity (compared to iNS) by iSp1, iSp3, iSp4 or their combinations is indicated by a double asterisk (**).

The effects of several xenoestrogens on activation of ERα/Sp-dependent transactivation were also investigated in MCF-7 cells cotransfected with pSp1₃ and iSp1, iSp3 or iSp4 (Fig. 5 and Fig. 6). Figure 5 summarizes the effects of various iSps on the induction of luciferase activity in MCF-7 cells transfected with pSp1₃ and treated with DES, HO-PCB-Cl₄, BPA, NP and endosulfan (Figs. 5A – 5E, respectively). The results show that like E2, DES and HO-PCB-Cl₄ primarily activate ERα/Sp1 and also activate ERα/Sp4 (< ERα/Sp1) but not ERα/Sp3 (Figs. 5A and 5B). BPA, endosulfan and NP activate ERα/Sp1 and ERα/Sp4 but not ERα/Sp3 (Figs. 5C – 5E) and resemble 4-OHT in their differential activation of ERα/Sp. The effects of iSps on activation of pSp1₃ by ICI 182,780 (Figs. 6A and 6B) and kepone (Figs. 6C and 6D) were different from those observed for the other ERα agonists and antagonists. Induction of luciferase activity by ICI 182,780 was decreased 40 – 60% by individual and combined iSps (Fig. 6A); however, basal activity was decreased in the order iSp4 > iSp3 > iSp1, and the lowest basal activity was observed in cells transfected with iSp4 alone or in combination with other iSps. When the luciferase activity is plotted as fold induction, the results show that ICI 182,780 activates ERα/Sp1 and ERα/Sp3 since iSp3, iSp1 and iSp1 plus iSp3 significantly decreased transactivation (Fig. 6B). Similar results were observed for kepone (Figs. 6C and 6D), although neither iSp3 or iSp4 were as effective in decreasing fold inducibility by kepone compared to ICI 182,780. The results demonstrate that activation of ERα/Sp-dependent transactivation by ICI 182,780 and kepone differs from the other estrogens, xenoestrogens and 4-OHT with respect to the role of Sp1, Sp3 and Sp4 in mediating this response.

Figure 5.

Differential role of Sp proteins in activation of ERα/Sp by ERα agonists. Effects of Sp protein knockdown on activation of pSp1₃ by DES (A), HO-PCB-Cl₄ (B), BPA (C), nonylphenol (D), and endosulfan (E). MCF-7 cells were transfected with pSp1₃/ERα and small inhibitory RNAs, treated with various ERα ligands, and luciferase activity determined as described in the Materials and Methods. Results are expressed as means ± SD for three replicate determinations for each treatment group. Significantly (p < 0.05) induced activity relative to DMSO is indicated (*) and a decrease in induced activity (compared to iNS) by iSp1, iSp3 or iSp4 is indicated by a double asterisk (**).

Figure 6.

Differential role of Sp proteins in activation of ERα/Sp by ICI 182,780 (A and B) or kepone (C and D). MCF-7 cells were transfected with pSp1₃/ERα and various iSps, treated with 1 µM ICI 182,780 or 5 µM kepone, and luciferase activity determined as described in the Materials and Methods. Results are expressed as means ± SD for three replicate determinations for each treatment group. Significantly (p < 0.05) induced activity relative to DMSO is indicated (*) and a decrease in induced activity (compared to iNS) by iSp1, iSp3 or iSp4 is indicated by a double asterisk (**).

3.4 Result of FRET studies

Previous studies showed that E2, 4-OHT and ICI 182,780 induced interactions of chimeric CFP-Sp1 and YFP-ERα in MCF-7 cells transfected with expression plasmids for these chimeras and analyzed by fluorescence resonance energy transfer (FRET) [23]. FRET efficiencies could be determined by comparing results to the YFP-CFP chimera which was used as a positive control [23]. Using this approach, we compared FRET efficiencies for E2 and selected xenoestrogens in MCF-7 cells transfected with CFP-Sp1 and YFP-ERα. Figure 7A illustrates the pseudocolor CFP and YFP fluorescence intensities in MCF-7 cells treated with DMSO or 10 µM kepone. The FRET signal in the kepone-treated cells indicates an increase in FRET efficiency compared to DMSO and that kepone induces ERα-Sp1 interactions at a distance of < 10 Å and thereby allows energy transfer. The FRET efficiency values were determined 20 – 30 min after treatment with 10 nM E2, 25 µM HO-PCB-Cl₄, 10 µM kepone, and 75 µM BPA, and FRET efficiencies were significantly increased compared to solvent (DMSO) control (Fig. 7B). These results demonstrate that xenoestrogens not only activate ERα/Sp-dependent transactivation but also induce ERα-Sp1 interactions in living breast cancer cells. Currently, we are investigating ligand structure-dependent differences in the rate of ERα interactions with Sp1, Sp3 and Sp4 using the FRET technique.

Figure 7.

FRET analysis and induction of E2F1 in MCF-7 cells. (A) An example of FRET in MCF-7 cells treated with DMSO (upper panel) and 10 µM kepone (lower panel) for 20 – 30 min. MCF-7 cells were transfected with CFP-Sp1 and YFP-ERα, and each panel represents the pseudocolor intensity of CFP, YFP and FRET signals, respectively. (B) FRET efficiencies. MCF-7 cells were transfected with CFP-Sp1 and YFP-ERα, treated with DMSO, 10 nM E2, 20 µM HO-PCB-Cl₄, 10 µM kepone, 75 µM BPA and 1 µM 4-OHT for 20 – 30 min, and FRET efficiencies were determined as described in the Materials and Methods. Significant differences (p < 0.05) in treated cells from DMSO (control) are indicated (*). (C) Role of Sp proteins in activation of CAD mRNA. MCF-7 cells were transfected with non-specific RNA (iNS) or iSp1, iSp3 and iSp4 (combined), treated with DMSO or 10 nM E2, 5 µM kepone, or 20 µM HPTE for 6 hr, and CAD mRNA levels were analyzed by real time-PCR as described in the Materials and Methods. Experiments were carried out in triplicate. Results are expressed as means ± SD and the CAD mRNA levels (normalized to TBP mRNA) in the DMSO treatment group was set at 1.0. Significant (p < 0.05) induction is indicated by an asterisk, and significant inhibition after transfection with iSp1, iSp3 and iSp4 (combined; 25 nM for each construct) is also indicated (**).

E2 induces carbamoylphosphate synthetase / aspartate transcarbamylase / dihydroorotase (CAD) gene expression in MCF-7 cells through interactions of ERα/Sp with proximal GC-rich motifs [28]. The role of Sp protein in mediating induction of CAD by three prototypical compounds, E2, HPTE and kepone, in MCF-7 cells is illustrated in Figure 7C. E2, HPTE, and kepone induced CAD gene expression, and induction was significantly inhibited in cells cotransfected with iSp1/iSp3/iSp4 (combined). This confirms that activation of CAD by E2 and xenoestrogens is also Sp-dependent, and current studies are investigating ligand structure-dependent activation and inhibition of other E2-response genes through the ERα/Sp pathway.

4. Discussion

E2 is the major endogenous hormone for both ERα and ERβ, and there is increasing evidence that activation of both receptors is complex and dependent on multiple factors including relative expression of each specific ER-subtype and their corresponding coactivators and coregulatory proteins [1–4]. The complexity of estrogen signaling is also linked the multiplicity of signaling pathways which include both genomic and non-genomic mechanisms where the later pathway involves extranuclear interactions of ER with multiple cytosolic or membrane-associated proteins that cooperatively activate kinase signaling [29; 30]. Genomic pathways of estrogen action include not only activation through direct binding of ER (dimeric) to EREs in target gene promoter but also activation of genes through interactions of ER with other DNA bound proteins including specificity proteins, AP1, GATA-1 and NFκB [14–17]. There is also evidence showing that hormonal activation of genes through ERE promoter sequences is structure-dependent and the tissue specific ER agonist and ER antagonist activities of SERMs has led to their applications for treating breast cancer and other hormone-related diseases [5–7].

Several in vitro assays clearly distinguish between the activities of E2 and antiestrogens such as ICI 182,780 and tamoxifen to activate ERE-dependent genes/promoter constructs [8–13]. Structurally-diverse synthetic industrial estrogenic compounds (xenoestrogens) and phytoestrogens also differentially activate various ERE-dependent constructs and related genes [31–40], suggesting that these compounds also exhibit SERM-like activity. The genomic ERα/Sp1 pathway for activating E2-responsive genes containing GC-rich motifs is critical for regulating growth-promoting genes and G₀/G₁ to S phase progression in MCF-7 cells [17], and this study was carried out to determine whether structurally-diverse xenoestrogens also differentially activate ERα/Sp. Previous studies showed that E2 induced transactivation in MCF-7, ZR-75 or MDA-MB-231 cells transfected with pSp1₃ and cotransfection with wild-type ERα or DBD deletion mutants [19; 20], whereas ICI 182,780 or 4-OHT induced transactivation only in cells transfected with wild-type ERα. The pSp1₃ construct is a prototypical E2-responsive GC-rich promoter which has been used to investigate differences between E2 and the antiestrogens 4-OHT and ICI 182,780 to activate ERα/Sp in breast cancer cells. In this study, pSp1₃ was used to investigate the SERM-like activity of structurally-diverse xenoestrogens and resveratrol by determining the effects of cell context (MCF-7 vs. MDA-MB-231), wild-type and variant ERα and different Sp proteins (Sp1, Sp3 and Sp4) on their estrogenic activities. However, in all cases, the relative potencies of the xenoestrogens for activating ERα/Sp were greater than 1000-fold lower than E2. With the exception of resveratrol in MCF-7 cells, all compounds including the antiestrogens ICI 182,780 and 4-OHT, E2, DES and the xenoestrogens induced wild-type ERα/Sp-dependent transactivation in MCF-7 and MDA-MB-231 cells (Fig. 1 – Fig. 3). E2 and DES, but not ICI 182,780 or 4-OHT, induced transactivation in MCF-7 cells transfected with ERαΔZF1 or ERαΔZF2. The pattern of xenoestrogen activation of pSp1₃ in cells transfected with ERα, ERαΔZF1 and ERαΔZF2 (Fig. 2) was similar to that observed for E2 and DES, and the most active compounds were kepone, HPTE, HO-PCB-Cl₄, endosulfan and nonylphenol. Activation of ERαΔZF1/Sp or ERαΔZF2/Sp by xenoestrogens distinguished between these compounds and ICI 182,780 and 4-OHT which were inactive in cells transfected with the DBD mutants.

Previous studies [20] showed that E2 and the antiestrogens ICI 182,780 and 4-OHT differentially induced activity in MCF-7 cells transfected with pSp1₃ and the C-terminal deletion mutants ERα(1–553) and ERα(1–537). ERα(1–537) does not express the F domain and part of helix 12 which interacts with coactivators. Not surprisingly, E2 and DES did not induce activity in MCF-7 (Fig. 2E) or MDA-MB-231 (Fig. 3E) cells transfected with ERα(1–537) and pSp1₃; however, significant induction was observed for both 4-OHT and ICI 182–780. Moreover, in cells transfected with pSp1₃ and ERα(1–537), like the antiestrogens, the xenoestrogens and resveratrol also induced transactivation. Thus, deletion of a region of helix 12 in the C-terminal domain of ERα, which totally abrogates E2-responsiveness, does not affect xenoestrogen-induced transactivation in cells transfected with pSp1₃ demonstrating a fundamental mechanistic difference between activation of ERα/Sp by E2 and the xenoestrogens/resveratrol used in this study.

Thus, results of transfection studies using wild-type and variant ERα show that xenoestrogen activation of ERα/Sp required different domains of ERα than either E2/DES or ICI 182,780/4-OHT. Recent studies in ZR-75 cells showed that ERα/Sp-dependent activation of vascular endothelial growth factor receptor 2 (VEGFR2) by E2 was primarily dependent on ERα/Sp3 and ERα/Sp4 but not ERα/Sp1 [25]. Thus, activation of VEGFR2 by E2 required interactions of ERα with only two of the Sp proteins and, in this study, we investigated structure-dependent activation of ERα/Sp1, ERα/Sp3 and ERα/Sp4 by E2, antiestrogens and xenoestrogens. RNA interference was used to individually knockdown Sp1, Sp3 and Sp4 or their combinations (Fig. 4A) in MCF-7 cells and, the effects of Sp protein knockdown on ligand-dependent activation of pSp1₃ was investigated. Basal activity was decreased in cells transfected with iSp1, iSp3 and iSp4; however, fold induction by estrogens (E2 and DES), xenoestrogens and antiestrogens exhibited three patterns which differentially relied upon ERα/Sp1, ERα/Sp3, ERα/Sp4 or their combinations (Fig. 4 – Fig. 6). For E2, HPTE, DES and HO-PCB-Cl₄, ERα/Sp1 > ERα/Sp4, and ERα/Sp3 had minimal to no effect on activation of ERα/Sp by these compounds. The pattern of ERα/Sp activation for BPA, endosulfan, NP and 4-OHT was ERα/Sp1 ≈ ERα/Sp4 with minimal contributions by ERα/Sp3. In contrast, both ERα/Sp1 and ERα/Sp3 play role in activation of pSp1₃ by ICI 182780 and kepone, but ERα/Sp4 tends to cause an inhibitory effect since iSp4 enhances the fold induction by these compounds. These results demonstrate that for E2 and other ER agonists/antagonists, activation of the GC-rich pSp1₃ construct by ERα/Sp involves different combinations of ERα/Sp1, ERα/Sp3 and ERα/Sp4. Since E2-induced transactivation in MCF-7 cells transfected with pSp1₃ required ERα/Sp1 and ERα/Sp4, these results are in contrast with the requirement of ERα/Sp3 and ERα/Sp4 but not ERα/Sp1 for activation of the GC-rich VEGFR2 promoter by E2 [25]. This suggests that both cell- and promoter-context may also dictate which ERα/Sp complexes are required for ligand-induced transactivation. Therefore, the relative levels of Sp proteins expressed in a cell/tissue may also influence hormonal activation of specific genes, and this is currently being investigated using both cell culture and in vivo models.

In cells transfected with pSp1₃, xenoestrogens, estrogens and antiestrogens all activated ERα/Sp1. Therefore, xenoestrogen-induced interactions of ERα and Sp1 were investigated by FRET in MCF-7 cells transfected with CFP-Sp1 and YPF-ERα expression plasmids. This technique was previously used in this laboratory to show that E2 induced interactions of ERα and Sp1 in living cells [23]. We selected five compounds representative of estrogens, antiestrogens and xenoestrogens, and observed that E2, HO-PCB-Cl₄, kepone, BPA and 4-OHT all significantly induced FRET efficiencies (Figs. 7A and 7B). These data demonstrate a comparable pattern of ligand-dependent ERα-Sp1 interactions, and future studies will focus on ligand structure-dependent differences and similarities in the FRET efficiencies and kinetics of ERα interactions with Sp3 and Sp4.

CAD is induced by E2 in MCF-7 cells through ERα/Sp-dependent activation of proximal GC-rich promoter sequences [28; 41]. Figure 7C illustrates that E2, HPTE and kepone induce CAD mRNA levels in MCF-7 cells, and these responses are abrogated in cells cotransfected with iSp1, iSp3 and iSp4 (combined). We have also observed similar results with CAD promoter constructs (data not shown) which have also been used for studies on xenoestrogens. In contrast, there are structure-dependent differences in activation of other putative ERα/Sp-dependent genes such as E2F1 and these are currently being investigated. These ongoing studies are also using E2-responsive genes that are repressed by E2 since this is a major pathway of estrogen action in breast cancer cells and other hormone-responsive cells and tissues.

In summary, results of this study show that xenoestrogens activate wild-type and variant ERα/Sp-dependent transactivation in MCF-7 cells transfected with pSp1₃. However, the results indicate that for some variant forms of ERα (i.e. ERαΔZF2/ERαΔZF1), most of the xenoestrogens resembled E2, whereas, in cells transfected with ERα(1–537), these compounds resembled the antiestrogens ICI 182,780 and 4-OHT. We also show for the first time that ligand structure-dependent activation of ERα/Sp in MCF-7 cells transfected with pSp1₃ was mediated by multiple Sp proteins (i.e. Sp1, Sp3 and Sp4) and utilization of these proteins was also ligand structure-dependent. There is evidence that Sp1 and Sp3 are differentially organized within the nucleus of MCF-7 cells [42], and this may contribute to the differences between ERα/Sp1 and ERα/Sp3 in activating pSp1₃ and we are also investigating the intranuclear distribution and organization of Sp1, Sp3 and Sp4. Results of this study, coupled with other reports of in vitro and in vivo assays with xenoestrogens [10–13; 31–40], suggest that these compounds are SERMs and their ERα agonist and antagonist activities cannot necessarily be inferred from simple ER binding or transactivation assays.