Development and Characterization of a Cell Line That Stably Expresses an Estrogen-Responsive Luciferase Reporter for the Detection of Estrogen Receptor Agonist and Antagonists

Vickie S. Wilson1, Kathy Bobseine and L. Earl Gray, Jr.

U.S. Environmental Protection Agency, ORD, NHEERL, Reproductive Toxicology Division, Research Triangle Park, North Carolina 27711

Received March 17, 2004; accepted May 25, 2004


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Recently several advisory committees (EDSTAC, ICCVAM) have recommended that stable estrogen-dependent gene expression assays be developed for screening chemicals for estrogenic activity because of the high degree of specificity of the response and potential for use in a high-throughput mode. In this paper we describe a specific, sensitive assay developed for screening chemicals for estrogenic and antiestrogenic activities. T47D human breast cancer cells, which naturally express estrogen receptor (ER) alpha and beta, were stably transfected with a triplet ERE (estrogen-responsive elements)–promoter–luciferase reporter gene construct. The transformed cells were named T47D-KBluc. These cells are sensitive to the potent estrogens, 17ß-estradiol, ethynyl estradiol, and diethylstibesterol, and well-characterized weaker environmental estrogens like genistein, HPTE (an estrogenic pesticide metabolite), and 4-nonylphenol. The EC50 for estradiol was about 0.01 nM, reaching maximal induction at 0.1 nM. The antiestrogen, ICI 182,780, was able to completely inhibit the induction of luciferase expression by 0.1 nM estradiol at 10 nM, with an IC50 of 1 nM. In addition, we were able to replicate, in this in vitro assay, the observation that low concentrations of cadmium were able to induce estrogen-dependent gene expression, an effect that was completely inhibited by the potent antiestrogen ICI 182,780. The potent glucocorticoid receptor agonist, dexamethasone, was without effect as an ER agonist at concentrations up to 10 nM, whereas the potent androgen, dihydrotestosterone (DHT), showed no induction at concentration of 50 µM, but was a partial agonist at high concentrations of 0.2 mM and above. In summary, we have developed a specific, sensitive estrogen-responsive gene expression assay in a stable cell line that could possibly be adapted for high throughput screening of large numbers of chemicals for estrogenic and antiestrogenic activity. In addition, herein we also provide key protocol recommendations necessary to identify and eliminate common problems encountered in in vitro screening for estrogenicity.

Key Words: estrogenicity; in vitro screening assay; estrogen-responsive luciferase reporter.


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Estrogens regulate the expression of specific genes through the interaction of the ligand-bound estrogen receptor dimer complex with specific DNA sequences called estrogen-responsive elements (ERE). As a consequence of this action, specific mRNA and subsequently proteins are synthesized by the cells that regulate physiological function. Recently, concern has been raised that both human and wildlife are exposed to environmental chemicals that interfere with normal endocrine function, and that these compounds can elicit adverse effects on development and fertility along with contributing to the increasing rates of certain types of human reproductive disorders (Colborn, 1995Go; Colborn and Clement, 1992Go). An ever-growing list of chemicals has been identified as potentially estrogenic, and a number of in vitro assays have been developed to screen substances for estrogenicity (Legler et al., 1999Go; Pons et al., 1990Go; Rogers and Denison, 2000Go; Zacharewski, 1997Go). Each assay is suited to a particular purpose; however, many of these assays have drawbacks that limit their usefulness or are not freely available to the scientific community as a screening tool. Competitive ligand binding assays assess the ability of a chemical to compete with the endogenous ligand (i.e., 17ß-estradiol) for binding to the steroid receptor (i.e., estrogen receptor). However, they give no insight as to the chemical's ability to initiate or inhibit gene transcription as both estrogens and antiestrogens bind ER similarly in a competitive binding assay. Cell proliferation assays are limited by a lack of specificity, because mitogens other than estrogens can influence cell proliferation (Dickson and Lippman, 1995Go). Furthermore, cell proliferation can be inhibited by a plethora of mechanisms independent of ER. Yeast-based assays are relatively simple to execute and can address gene transcription issues but lack responsiveness to some estrogens and antiestrogens. The yeast assays in general fail to respond appropriately to some chlorinated EDCs (endocrine disrupting compounds). In addition, permeability of compounds through the yeast cell wall differs from that of mammalian cell membranes (Gaido et al., 1997Go). For these reasons, yeast assays were specifically excluded as general estrogen screening assays by both the EDSTAC (Endocrine Disruptors Screening and Testing Advisory Committee) and ICCVAM (Interagency Coordinating Committee on the Validation of Alternative Methods). Transient transfection of reporter gene constructs into mammalian cells can provide similar information, but the process is time consuming, more labor intensive, and the data can be significantly more variable than stably transfected cell-based assays (Wilson et al., 2002Go).

In comparison to other existing in vitro assays, reporter gene assays based on stably transfected cell lines provide the most specific, responsive, and relatively quick means to screen substances for potential estrogenic and antiestrogenic activity (Legler et al., 1999Go). Recently, a few such cell lines have been developed (Legler et al., 1999Go; Pons et al., 1990Go). These cell lines are sensitive and responsive to estrogens but are not freely available. Consequently, the goal of this project was to develop an estrogen-dependent stable cell line by stably transfecting an estrogen-responsive luciferase reporter gene construct into T47D human breast adencarcinoma cells expressing endogenous estrogen receptors alpha and beta and to make the cell line easily available to other laboratories. T47D cells express a relatively high number of endogenous ER, reportedly 67.6 ± 6.2 fmol/mg cytosolic protein (Watanabe et al., 1990Go). Western blot analysis indicated that these cells contain both the alpha and beta isoforms of the ER protein with very slightly higher levels of beta than alpha (Power and Thompson, 2003Go). The reporter gene construct consists of three estrogen response elements (ERE) upstream from a TATA box that regulates the expression of a luciferase reporter gene. Stable transfection of this promoter-reporter construct into T47D cells resulted in a sensitive, responsive clone. In very simple terms, compounds enter the cell; estrogen receptor ligands bind to the ER; two ligand-bound receptors dimerize and bind coactivators; then the dimer binds to the ERE on the reporter gene construct and activates the luciferase reporter gene. The presence of the luciferase enzyme can then be assayed by measuring the light produced when the enzyme substrate, luciferin, and appropriate cofactors are added. The amount of light produced is relative to the degree of estrogenic activity of the test chemical. Model compounds, with well-defined mechanisms of action, were used to evaluate the utility of this cell line, T47D-KBluc, and those results are presented here. When testing chemicals using the T47D-KBluc cells, an estrogen was defined as a chemical that induced dose-dependent luciferase activity which could be specifically inhibited by the antiestrogen ICI. Plans include making these cells widely available by depositing the cell line with ATCC for maintenance and distribution.


    MATERIALS AND METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Chemicals. 17ß-Estradiol (E2, 99%), 17{alpha}-ethynylestradiol (EE, >98%), diethylstibestrol (DES, min >99%), 5{alpha}-dihydrotestosterone (DHT, min >99%), dexamethasone (DEX, 100%), genestein (Gen, >98%), tamoxifen (Tam, >99%) and methoxychlor (Meth, >98%) were purchased from Sigma Chemical Company. 4-Nonylphenol (4-NP, a mixture of branched side chains containing 85% p-isomers) was purchased from Fluka Chemical Corp. (Ronkonkoma, NY). The synthesis of 2,2-bis(p-hydroxyphenyl)-1,1,1-trichloroethane (HPTE), a methoxychlor (Sigma, purity >99%) metabolite, was previously published (Waller et al., 1996Go). The antiestrogen, ICI 182,780, was supplied by ICI Pharmaceuticals (Macclesfield, England, Lot #C42710). Cadmium chloride (CdCl, purity 99.5%) was purchased from Fisher Chemical Co. (Fig. 1)



View larger version (13K):
[in this window]
[in a new window]
 
FIG. 1. Schematic representation of the neomycin resistant luciferase reporter plasmid, pGL2.TATA.inr.luc.neo-11. Location of the luc and amp genes are estimated based on restriction digest patterns.

 
Construction of reporter plasmid pGL2.TATA.Inr.luc.neo. Neomycin gene was removed from puc9.neo (provided by Phillip Hartig, U.S. EPA, Research Triangle Park, NC) using a BamHI digest. The 1.8 kb fragment was ligated to pGL2.TATA.Inr.luc containing three estrogen-responsive elements (ERE) (provided by Donald McDonnell, Duke University, Durham, NC) that had been linearized with SmaI. The resulting plasmid was pGL2.TATA.Inr.luc.neo (Fig. 1).

Mammalian cells. The human breast cancer cell line T-47D (ATCC No. HTB 133), ER{alpha}/ß + /GR +, was used for transfection. Cells were screened for sensitivity to the selection antibiotic, geneticin (Gibco/BRL). Concentrations of the antibiotic were selected to produce 100% lethality over a two week culture period (data not shown). Growth media was RPMI (Gibco) supplemented with 2.5 g/l glucose, 10 mM HEPES, 1 mM sodium pyruvate, 1.5 g/l NaHCO2, 0.2 U/ml insulin, 10% FBS (fetal bovine serum, Hyclone, characterized), 100 µg/ml penicillin, 100 U/ml streptomycin, and 0.25 µg/ml amphotericin B (Gibco BRL, purchased as a 100x mixture of penicillin, streptomycin, and amphotericin B).

Transfection procedure. Cells were seeded 2 x 105 cells per 60 mm culture dish. Cells were transfected using Fugene 6 (Roche) per manufacturer protocol with 5 µg pGL2.TATA.Inr.luc.neo per dish. The cells were placed in selection media (growth media plus 500 µg/ml gentamycin) 24 h after transfection. Cells were grown in selection media until colony formation was observed. Colonies were transferred by trypsinization to 24-well plates and then to T-25 cm2 flasks for continued culture.

Initial screening of clones. For initial screening of colonies, cells were plated at 104 per well in 96-well luminometer plates and allowed to attach 5–6 h. After attachment, growth media were replaced with fresh media, except 5% dextran-charcoal treated FBS was substituted for 10% regular FBS. After 40 to 48 h cells were dosed with 100 µl dosing media/well (5% dextran-charcoal treated FBS media plus test chemical) and incubated for 24 h. Stock chemicals were prepared in 95% ethanol. Dosing solutions were prepared by diluting the chemical stock with fresh dosing media to the desired concentration. In no case did the ethanol concentration exceed 0.2%. Negative control wells were dosed with media plus 0.1% ethanol. Positive control wells were dosed with 0.1 nM or 1.0 nM 17ß-estradiol (E2). Both controls (vehicle and E2) were also competed with 1 µM ICI, an ER antagonist, to assess ER-specific responsiveness and background. Cells were washed with phosphate buffered saline at room temperature and then 25 µl lysis buffer (Ligand Pharmaceuticals) was added per well and incubated until cells were lysed (15–30 min). Relative light units per well were determined using a 96-well MLX Luminometer (Dynex, Chantilly, VA). The final clone was chosen based on appropriate ligand responsiveness and genetic stability over time and renamed T47D-KBluc.

Chemical screening. Stock cells from the chosen clone, T47D-KBluc, were maintained in standard growth media as detailed above. Cells were placed in growth media modified by replacement of 10% FBS with 10% dextran-charcoal treated FBS (Hyclone) without antibiotic supplement one week prior to assay. Dosing media was further modified by reduction of dextran-charcoal treated FBS to 5%. Cells were seeded at 104 cells per well in 96-well luminometer plates and allowed to attach overnight. Media was then replaced with 100 µl/well of dosing media and the test chemical and incubated 24 h. Ethanol vehicle did not exceed 0.2%. Cells were washed with phosphate buffered saline at room temperature, then harvested in 25 µl lysis buffer (Ligand Pharmaceuticals) per well. Luciferase activity was determined using an MLX microtiter plate luminometer (Dynex, Chantilly, VA) and quantified as relative light units (RLU). Each well received 25 µl reaction buffer (25 mM glycylglycine, 15 mM MgCl2, 5 mM ATP, 0.1 mg/ml BSA, pH 7.8), followed by 25 µl 1 mM D-luciferin 5 s later. Each chemical was assayed independently at least three times (three replicate assay plates) with a minimum of four wells per each replicate assay unless otherwise noted in the text. Cells were screened with a battery of chemicals using agonist positive (E2), negative (vehicle only), antagonist (E2 plus ICI), and background (vehicle plus ICI) controls on every plate. Each chemical was tested both alone and in the presence of an appropriate competitor such as 0.1 nM E2 (suspected antagonist) or ICI (suspected agonists). E2 positive controls were monitored over time as an assessment of the stability of the cells line. In instances where cytotoxicity of a chemical was suspected, duplicate plates were dosed in parallel. Luciferase activity was assayed in one plate as described above and the second plate was tested for cell toxicity. Cytotoxicity was evaluated by determining the mitochondrial function of the cells using the tetrazolium dye 3-[4,5-dimethylthiazol-2-yl]-2,5-diphenyltetrazolium bromide (MTT) following treatment with the compound. MTT is a yellow vital dye that is actively converted by mitochondrial oxidation-reduction reactions into blue formazan crystals. The formation of the blue MTT crystals within the cell decreases in direct proportion to the viability of cells (Li et al., 1999Go).

Statistical analysis. These data were collected from several independent experiments, with three or more replicates (plates) per experiment. A replicate was a 96-well plate which included 4–8 independent observations (wells) each of the vehicle control, positive control (either 0.1 or 1.0 nM E2), antiestrogen control (E2 plus ICI), background control (vehicle plus ICI), and all other treatment groups. Data were analyzed by two-way ANOVA (main effects being replicate [a nuisance or blocking factor] and treatment) using Proc GLM available from SAS version 6.09 (SAS Institute, Cary, NC.) on the U.S. EPA IBM mainframe computer. Relative light units were converted to fold induction above the vehicle control value or as percent of E2 positive control response for each replicate for statistical analysis. Data were analyzed in a GLM model that included the concentrations and replicates. Statistically significant effects (p < 0.05) were examined using the least squares means (LSMEANS) procedure available on SAS. Means and standard errors were calculated using PROC means. For agonists, which stimulate luciferase expression, treatments were compared to the vehicle (media plus ETOH) control group or to the relative response of their respective E2 control (either 0.1 or 1.0 nM E2). Estrogen antagonist, ICI, which blocks E2-induced luciferase expression, was compared to the E2 positive control group. Graphs were prepared using Origin scientific graphing software (OriginLab, Northampton, MA). Best fit curves were generated using logistic fit of the data.


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Stable transfection of T47D breast cancer cells containing both endogenous ER{alpha} and ERß with pGL2.TATA.Inr.luc.neo resulted in a sensitive and responsive cell line, T47D-KBluc, for the assessment of compounds with estrogenic (or antiestrogenic) activity. A variety of model compounds, which included pharmaceutical and environmental estrogens, with well-defined mechanisms of action were used to demonstrate the utility of the cell line. Initially, responsiveness and sensitivity to 17ß-estradiol (E2) were assessed. Concentrations at log intervals from 1fM (10–6 nM) to 100 nM E2 were tested and data plotted as fold induction over vehicle controls. Consistent, statistically significant increases in fold induction were detected at concentrations of 1 pM and above with maximal induction obtained at 0.1 nM E2 (n = 6 replicate plates; four wells per plate, Fig. 2A). To assess antiestrogenic activity, test chemicals were assayed against 0.1 nM E2, the lowest concentration that produced a maximal estrogenic response. To determine the maximum inhibitory response to a potent antiestrogen, ICI 182,780, doses from 10 pM to 10 µM were assayed against 0.1 nM E2 (Fig. 2B). A significant decrease in E2-induced luciferase activity was detected at 1 nM ICI. Maximum inhibition of the E2-induced luciferase response was obtained with 0.1 µM ICI. As a result, controls chosen to be run on all subsequent plates were: 0.1 nM E2 (or 1.0 nM E2 in some cases) as a positive control, 0.1 or 1.0 µM ICI as antiestrogen control, vehicle only control, and vehicle plus 1 µM ICI (to assess background). The antiestrogen, tamoxifin (TAM), was also tested at concentrations ranging from 10 pM to 10 µM in competition with 0.1 nM E2 (Fig. 2B). As expected, TAM reduced E2-induced luciferase activity, albeit at higher concentrations than ICI. The IC50 for TAM was approximately 500 nM compared to about 0.3 nM for ICI.



View larger version (15K):
[in this window]
[in a new window]
 
FIG. 2. Dose response of the T47D clone, T47D-KBluc, to the potent estrogen 17ß-estradiol (top) and the estrogen antagonist, ICI 182,780, in competition with 0.1 nM E2 (bottom). Cells were dosed with increasing concentrations of either agonist or antagonist in 100 µl medium in 96-well luminometer plates. Data are presented as the mean fold induction compared to vehicle controls of three replicate assays (four wells per replicate) ± standard error of the mean.

 
The potent synthetic estrogens, DES and EE, were assayed at concentrations from 10 pM to 100 nM or from 100 pM to 10 nM, respectively. Dose-response data are presented as a percent of their respective 0.1 nM E2 control response (Fig. 3). As expected, both compounds induced luciferase activity at concentrations similar to that seen with E2. Estimated EC50 values were 8.5 and 14 pM for EE and DES, respectively. Maximal induction of luciferase activity was obtained at 0.1 nM for both compounds.



View larger version (14K):
[in this window]
[in a new window]
 
FIG. 3. Responsiveness of the cell line to the synthetic estrogens, DES and EE. Data are presented as the mean percent of the 0.1 nM E2 response compared to E2 controls of three replicate assays (four wells per replicate) ± standard error.

 
To test the specificity of the estrogenic response, DHT and DEX were tested in the T47D-KBluc cells (Fig. 4). DEX at concentration from 1 pM to 10 nM failed to induce luciferase expression above vehicle control. DHT did induce luciferase activity but only at high concentration of 200 nM and above. Preliminary assays indicated that there was no significant response to DHT at concentrations less than 50 nM, so only responses at 50 nM and greater are displayed in the graph. Even at concentrations of 200 nM to 1 µM the degree of induction by DHT was significantly lower that that attained by 0.1 nM E2. This response appeared to be ER mediated, however, because 1.0 µM ICI reduced the luciferase induction by all concentrations of DHT tested to the level of the vehicle control.



View larger version (15K):
[in this window]
[in a new window]
 
FIG. 4. Response of the cells to the glucocorticoid, DEX, and to the androgen, DHT. The potent glucocorticoid, DEX, failed to induce luciferase activity at concentrations up to 10 nM. The androgen receptor agonist, DHT, induced luciferase activity only at high concentrations (200 nM and greater), and even then the degree of induction by DHT was significantly lower than that attained by 0.1 nM E2. Data are presented as the mean fold induction compared to vehicle controls of three replicate assays (with four wells per replicate) ± standard error.

 
A number of chemicals identified as environmental estrogens were also tested in this stable cell line. The compounds chosen have previously been shown to be estrogenic in other assays. Data for methoxychlor (Meth), HPTE, genistein (Gen), 4-nonylphenol (4-NP), and cadmium chloride (CdCl) are shown as percent of the 0.1 nM E2 response from the positive control values on their respective plates (Figs. 5 and 6). Meth, as expected, was only weakly estrogenic in this assay, producing slight increases in luciferase activity at concentrations of 200 nM and above. The metabolite of Meth, HPTE, is reported to be the active estrogen. HPTE treatment resulted in significant increases in luciferase activity in the cell reaching 100% of the 0.1 nM E2 response at 500 nM. At 10 µM HPTE (10,000 nM), the cells were showing visible signs of distress and cytotoxicity (i.e., floating cells, blebbing of membranes), and luciferase activity was reduced as would be expected. Genistein produced dose-dependent increases in luciferase activity at concentrations of 1 nM and above. At 10 nM Gen showed luciferase activity exceeding that of 0.1 nM E2 and reached about three times the E2 control level at 1 µM. 4-NP also produced significant increases in luciferase activity at all concentrations tested (0.5 nM to 5 µM), reaching about 115% of the 0.1 nM estradiol response at 5 µM. The relative potencies of both Gen and 4-NP, however, were significantly less than E2, as expected. In addition, 1.0 µM ICI totally inhibited the luciferase activity induced by both Gen and 4-NP (Fig. 5). CdCl was tested at concentrations ranging from 10 fM to 10 µM (Fig. 6). These concentrations of CdCl resulted in a dose-dependent increase in luciferase activity that reached near 100% of the 0.1 nM E2 response at 1 µM. This responsiveness to CdCl appears to be mediated through the ER (either alpha or beta or both) because luciferase activity at all concentrations can be significantly reduced by 1.0 µM ICI. The specific mechanism, however, is unknown.



View larger version (27K):
[in this window]
[in a new window]
 
FIG. 5. Treatment of T47D-KBluc cells with four environmental estrogens. Effects of increasing concentrations of methoxychlor, the methoxychlor metabolite, HPTE, genistein, and 4-nonylphenol. Data are presented as the percent of the response attained by the respective 0.1 nM E2 control for three replicate assays (four wells per assay) ± standard error. Gen and 4-NP data are shown both alone and in competition with 1.0 µM ICI. Asterisks denote statistical significance of the response of the individual chemical as compared to the respective vehicle control (0 value). *p < 0.01, **p < 0.001, ***p < 0.0001.

 


View larger version (26K):
[in this window]
[in a new window]
 
FIG. 6. Responsiveness of T47D-KBluc cells to cadmium chloride alone and in the presence of 1 µM ICI 182,780. Cadmium chloride induced a dose-dependent increase in luciferase activity that could be inhibited by the ER specific antagonist ICI. Data are presented as the percent of the response attained by the respective 0.1 nM E2 control for three replicate plates (four wells per plate) ± standard error.

 
Controls were included on each plate as follows: vehicle control (ethanol solvent), vehicle plus 0.1 µM ICI (for assessing background estrogenicity), positive control (0.1 nM E2), and antiestrogen control (0.1 nM E2 plus 0.1 µM ICI). High background is always a potential problem when working with estrogen-responsive cells; therefore, a control that could be used to assess background was included on each plate (i.e., solvent plus ICI). We found several factors that could contribute to high background in these cells. One of those was the use of an antibiotic and antifungal mixture in the assay medium. Figure 7 illustrates the difference in background levels of luciferase activity with and without antibiotic/antifungals in the medium. These data are the mean of results obtained on 9–10 assay plates with four wells per plate for each with and without antibiotics values (36–40 individual values). When ICI 182,780 plus vehicle (ethanol) is compared to vehicle only controls with antibiotics and antifungals in the medium, ICI reduces the level of luciferase activity about twenty-five fold indicating quite high background levels of estrogenic activity. When cells are cultured for 1 week prior to conducting an assay without antibiotics or antifungals in the medium, background levels were significantly reduced. Because of this, stock cultures were maintained in media with a typical antibiotic–antimycotic mixture, but cells withdrawn from antibiotics or antimycotics for 1 week prior to chemical testing.



View larger version (14K):
[in this window]
[in a new window]
 
FIG. 7. Difference in background levels of luciferase activity of cells cultured with and without antibiotic/antifungals mixture in the medium. When an antibiotic/antifungal mixture is included in the assay medium, high levels of background luciferase activity are present. Background activity becomes apparent when cells are cultured with the addition of the antiestrogen, ICI. When cells are cultured for 1 week without an antibiotic/antifungal mixture in the culture medium, background luciferase activity is significantly reduced to an acceptable level. Background luciferase activity was monitored by including a vehicle plus ICI control on each assay plate (four wells per plate).

 
E2-induced luciferase activity was monitored over several months of continuous culture to assess the stability of the clone. During this time, cells were maintained without selective pressure to gentamycin, and stably transformed cells were assayed over more than 74 passages for their responsiveness to 0.1 nM E2. Once the antibiotic/antimycotic mixture was removed from the test medium and background was at acceptable levels, new cells from the same clone were cultured from frozen stocks and again monitored for 24 passages (from passage 34 to and including passages 57) for stability and responsiveness to 0.1 nM E2. Over these 24 passages, fold induction of 0.1 nM E2 as compared to vehicle control averaged 11.7 ± 6.2 (mean ± standard deviation). Additionally, cryopreserved cell stocks stored in liquid nitrogen for over 2 years were successfully restored into culture with only nominal loss of cell viability and retained similar responsiveness to E2.


    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
The stable cell line, T47D-KBluc, constructed and described in this paper resulted in the generation of a sensitive and responsive tool for the screening of chemicals for estrogenic activity. T47D human breast cancer cells, which contain both endogenous ER{alpha} and ERß, were transfected with an ERE- luciferase reporter gene construct. It provides an in vitro system that can be used to evaluate the ability of chemicals to modulate the activity of estrogen-dependent gene transcription. This cell line has the potential to be used both for screening chemicals and as an aid in defining mechanism of action of chemicals with estrogenic and antiestrogenic activity. The presence of both ER alpha and beta in these cells was an asset. It would be expected that if a highly specific ER ligand is tested the luciferase response would be driven by both isoforms of the receptor. This is valuable for a first-pass type in vitro assay, as a ligand for either receptor could drive the luciferase reporter gene. Thus two different assays (one for alpha and one for beta) would not be needed.

The performance of this assay was evaluated using known ER agonists and antagonists. We have shown, herein, that the responsiveness of this stable cell line to potent estrogens, including E2, EE, and DES, is in the pM to nM range, which would imply sufficient sensitivity for most applications. At E2 concentrations between 0.1 and 1 pM luciferase activity rapidly increased, illustrating the ability of the assay to detect very low E2 concentrations. The EC50 of E2 following a 24-h exposure was approximately 3 pM, with maximum induction occurring near 100 pM (0.1 nM) and above. These results are similar to those obtained in the ER-CALUX assay for E2, which also used T47D cells as the parent cell line, where an EC50 of 6 pM was reported (Legler et al., 1999Go). They reported maximum induction at 30 pM, while we report maximum induction near 100 pm. We chose to test concentrations at log intervals; however, it is possible that if intermediate doses were tested that number could be refined. Fold induction due to estrogens in our stable cell line is less than that obtained with the ER-CALUX assay, but the sensitivity and predictability of the assays is very similar when testing both potent estrogens and weaker environmental estrogenic compounds.

Similar results were obtained for both EE and DES, as would be expected for potent estrogenic compounds. EE-induced luciferase activity dramatically increased at concentrations greater than 1 pM, with an EC50 of approximately 8.5 pM. The EC50 value for DES was estimated at 14 pM. Maximal induction was achieved for all three compounds, E2, EE, and DES, at 0.1 nM. Together these results indicate their similar potency in this assay. To test the specificity of the assay, both DEX and DHT were tested. DEX did not activate gene transcription at any of the concentrations tested (1 pM to 10 nM). DHT did activate gene transcription to about half the maximal E2 response, but only at very high concentrations (statistically significant at 200 nM and above). The response to DHT was reduced by the addition of ICI, indicating it was likely ER mediated. T47D breast cancer cells have been shown to express mRNA for aromatase (Sadekova et al., 1994Go); however, since DHT cannot be aromatized to estrogen, this is an unlikely explanation for the induction of luciferase activity by DHT at high concentrations.

Several environmental estrogens were also tested in the T47D-KBluc cells, and results were generally as expected. The lowest observed effect concentrations (LOEC) for genistein and 4-NP were 10 nM and 0.5 nM, respectively. This is somewhat lower that the LOEC (100 nM for both compounds) reported for the ER-CALUX assay. This is an interesting observation, given that the two assays are similar and were developed from the same parent cell line, T47D cells. Interestingly, some compounds such as genistein and 4-NP elicited luciferase induction higher than the maximal response attained by E2. Similar responses have been seen by others when these compounds are tested at high concentrations (Legler et al., 1999Go; Routledge and Sumpter, 1996Go). Even though these are "simplified" systems compared to the whole animal, they still contain complex biological processes. The mechanism behind this action is not clear, but some have hypothesized that it may be due to stimulated receptor and/or coactivation factor renewal or to effects on luciferase stability (Legler et al., 1999Go). Most importantly, the goal was to correctly identify estrogenic compounds, and as an initial first-pass screen, both compounds would have been correctly identified for their estrogenic activity and their relative potency based upon their EC50. Before one could conclude that a compound truly had some sort of supramaximal activity, estrogenic activity would need to be confirmed in vivo.

Recently, studies have shown that the heavy metal cadmium mimics the effects of E2 both in vitro and in vivo (Johnson et al., 2003Go; Stoica et al., 2000Go), and our results confirm the in vitro estrogenic activity of cadmium chloride (CdCl). The metal appears to bind to the ER with high affinity with estrogenic activity detected at very low concentrations of CdCl (10 fM). Activity continued to increase up to the highest concentration tested (10 µM) and reached near 100% of the 0.1 nM E2 response at 1 µM CdCl. The fact that this response can be blocked by ICI reinforces the indication that the effect is ER mediated. Work done by Stoica et al. (2000)Go indicated that the effect is mediated through ER{alpha}, and the interaction of cadmium with the receptor appears to involve several specific amino acids in the ligand binding domain of the receptor.

A comprehensive summary of the published data on in vitro assays used to test substances for their ability to initiate ER-dependent gene transcription was recently compiled by ICCVAM (Interagency Coordinating Committee on the Validation of Alternative Methods) at the request of the U.S. EPA as part of the methods validation process for the Endocrine Disruptor Screening Program (EDSP). These background review documents along with the Expert Panel review are available on the ICCVAM web site (http://iccvam.niehs.nih.gov/docs/docs.htm). The ICCVAM document for estrogen-dependent transcriptional activation assays and the expert panel review recommended testing a range of doses for test chemicals at log intervals over at least seven orders of magnitude when using transcriptional activation assays for screening purposes. As they recommended, we conducted most of these assays using doses at log intervals. This is a good approach when using transcriptional activation assays for screening of chemicals, especially when little or no information on the relative potency of the test chemical is available, as the probability of detecting a positive is increased when a broad range of concentrations is tested. Admittedly, the determination of EC50 values could be refined by using several intermediate concentrations within the linear section of the dose-response curve, but further refinement is not always necessary when used only for screening purposes.

We found that the use of appropriate controls on every assay plate was essential for monitoring the function of the cells and for defining which data sets are appropriate to include in a response analysis. Always of concern when utilizing estrogen-dependent gene transcriptional activation assays is the issue of high background activation levels. We found that increased background could be variable and influenced by many factors. Because of this, we recommend always testing levels of background activity on every assay plate by adding a vehicle plus ICI control in addition to the standard positive and negative controls. Several factors can influence background activation, such as serum that is improperly stripped of steroid hormones. Each new batch of dextran-coated charcoal stripped fetal bovine serum needs to be tested for any residual steroid activity. We also found that some antibiotic mixtures, added to help avoid contamination, can contribute to high background levels of estrogenicity. Figure 7 illustrates the difference in background levels of luciferase activity with and without antibiotic/antifungals in the medium. High background levels of estrogenicity blunt the responsiveness of the cells and may lead to false conclusions about a compound. During any assay, if background was unusually high on a particular plate, those assay results were discarded and the assay repeated. The portion of the graph shown in Figure 7 without antibiotics illustrates what we considered to be acceptable levels of background luciferase activity. In addition, for the purpose of screening, runs/plates were discarded when the stimulation induced by the 0.1 nM E2 control was less than 2–3 fold when compared to the vehicle control. Once high background problems were reduced by weaning cells from the antibiotic/antimycotic mixture for 1 week prior to conducting an assay, lack of responsiveness to E2 was only a minor concern (less than 1% of plates were discarded).

In summary, in vitro assays are becoming increasingly attractive as screening tools because they are rapidly done and fairly easy to perform, some are amenable to high-throughput systems, and they have the potential to reduce the number of animals needed for chemical testing. Stably transfected cell lines offer several advantages in comparison to other in vitro systems such as binding assays, repeated transient transfections, yeast based assays, and cell proliferation assays. They are also an excellent aid in defining mechanism of action. Assays utilizing this ER stable cell line are sensitive and generate reproducible data. Our plans include making the T47D-KBluc cell line widely available by depositing it with ATCC as we have done with the MDA-kb2 cells (an AR dependent stable cell line) (Wilson et al., 2002Go)


    NOTES
 
The research described in this article has been reviewed by the National Health and Environmental Effects Research Laboratory, ORD, U. S. Environmental Protection Agency, and approved for publication. Approval does not signify that the contents necessarily reflect the views and policies of the Agency nor does the mention of trade names or commercial products constitute endorsement or recommendation for use.

1 To whom correspondence should be addressed at U.S. Environmental Protection Agency, ORD, NHEERL, Reproductive Toxicology Division, MD-72, Research Triangle Park, NC 27711. Fax: (919) 541-4017. E-mail: wilson.vickie{at}epa.gov.


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Colborn, T. (1995). Environmental estrogen: Health implications for humans and wildlife. Environ. Health Perspec. 103, 135–136.[ISI][Medline]

Colborn, T., and Clement, C. (1992). Chemically-induced alterations in sexual and functional development. In Advances in Modern Environmental Toxicology, (T. Colborn and C. Clements, Eds.), Vol. 21. Princeton Scientific Publishing Co., New York, 1–8.

Dickson, R. B., and Lippman, M. E. (1995). Growth factors in breast cancer. Endocr. Rev. 16, 559–589.[ISI][Medline]

Gaido, K. W., Leonard, L. S., Lovell, S., Gould, J. C., Babai, D., Portier, C. J., and McDonnell, D. P. (1997). Evaluation of chemicals with endocrine modulating activity in a yeast-based steroid hormone receptor gene transcription assay. Toxicol. Appl. Pharmacol. 143, 205–212.[CrossRef][ISI][Medline]

Johnson, M. D., Kenney, N., Stoica, A., Hilakivi-Clarke, L., Singh, B., Chepko, G., Clarke, R., Sholler, P. F., Lirio, A. A., Foss, C., et al. (2003). Cadmium mimics the in vivo effects of estrogen in the uterus and mammary gland. Nat. Med. 9, 1081–1084.[CrossRef][ISI][Medline]

Legler, J., van den Brink, C. E., Brouwer, A., Murk, A. J., van der Saag, P. T., Vethaak, A. D., and van der Burg, B. (1999). Development of a stably transfected estrogen receptor-mediated luciferase reporter gene assay in the human T47D breast cancer cell line. Toxicol. Sci. 48, 55–66.[Abstract/Free Full Text]

Li, A. P., Lu, C., Brent, J. A., Pham, C., Fackett, A., Ruegg, C. E., and Silber, P. M. (1999). Cryopreserved human hepatocytes: Characterization of drug-metabolizing enzyme activities and applications in higher throughput screening assays for hepatotoxicity, metabolic stability, and drug-drug interaction potential. Chem. Biol. Interact. 121, 17–35.[CrossRef][ISI][Medline]

Pons, M., Gagne, D., Nicolas, J. C., and Mehtali, M. (1990). A new cellular model of response to estrogens: A bioluminescent test to characterize (anti) estrogen molecules. Biotechniques 9, 450–459.[ISI][Medline]

Power, K. A., and Thompson, L. U. (2003). Ligand-induced regulation of ER{alpha} and ERß is indicative of human breast cancer cell proliferation. Breast Cancer Res. Treat. 81, 209–221.[CrossRef][ISI][Medline]

Rogers, J. M., and Denison, M. S. (2000). Recombinant cell bioassays for endocrine disruptors: Development of a stably transfected human ovarian cell line for the detection of estrogenic and anti-estrogenic chemicals. In Vitr. Mol. Toxicol. 13(1), 67–82.[ISI][Medline]

Routledge, E. J., and Sumpter, J. P. (1996). Estrogenic activity of surfactants and some of their degradation products assessed using a recombinant yeast screen. Environ. Toxicol. Chem. 15, 241–248.[ISI]

Sadekova, S. I., Tan, L., and Chow, T. Y. (1994) Identification of the aromatase gene in the breast carcinoma cell lines T47D and MCF-7. Anticancer Res. 14(2A), 507–511.[ISI][Medline]

Stoica, A., Katzenellenbogen, B. S., and Martin, M. B. (2000). Activation of estrogen receptor-alpha by the heavy metal cadmium. Mol. Endocrinol. 14, 545–553.[Abstract/Free Full Text]

Waller, C. L., Booker, W. J., Gray, L. E., Jr., and Kelce, W. R. (1996). Three dimensional quantitative structure-activity relationships for androgen receptor ligands. Toxicol. Appl. Phamacol. 137, 219–227.[CrossRef][ISI][Medline]

Watanabe, T., Wu, J. Z., Morikawa, K., Fuchigami, M., Kuranami, M., Adachi, I., Yamaguchi, K., and Abe, K. (1990). In vitro sensitivity test of breast cancer cells to hormonal agents in a radionucleotide-incorporation assay. Jpn. J. Cancer Res. 81, 536–543.[ISI][Medline]

Wilson, V. S., Bobseine, K., Lambright, C. R., and Gray, L. E., Jr. (2002). A novel cell line, MDA-kb2, that stably expresses an androgen- and glucocorticoid-responsive reporter for the detection of hormone receptor agonist and antagonist. Toxicol. Sci. 66, 69–81.[Abstract/Free Full Text]

Zacharewski, T. R. (1997). In vitro bioassays for assessing estrogenic substances. Environ. Sci. Technol. 31, 613–623.[CrossRef][ISI]





This Article
Abstract
FREE Full Text (PDF)
All Versions of this Article:
81/1/69    most recent
kfh180v1
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (1)
Disclaimer
Request Permissions
Google Scholar
Articles by Wilson, V. S.
Articles by Gray, L. E.
PubMed
PubMed Citation
Articles by Wilson, V. S.
Articles by Gray, L. E., Jr.