The Center for Alternatives to Animal Testing is an academic center affiliated with the Division of Toxicological Sciences in the Department of Environmental Health Sciences of the Johns Hopkins University Bloomberg School of Public Health.

 

Johns Hopkins School of Public Health

Abstract for TestSmart -- Endocrine Disruptors

Methods for Determining Estrogenic (E-Screen) and Androgenic (A-Screen) Activity of Chemicals Using Cell Proliferation Assays

Ana M. Soto*, Carlos Sonnenschein*, Janine Calabro*, Ruthann Rudel+ and Julie Brody+
University School of Medicine, 136 Harrison Ave, Boston, MA +Silent Spring Institute, 29 Craft Rd, Newton, MA

The 1996 amendments to the Safe Drinking Water Act and the 1996 Food Quality Protection Act required US EPA to develop a program to screen and test chemicals used in large volumes that may contaminate water and food to assess their potential activity as endocrine disruptors. These chemicals are to be tested for their ability to affect estrogenic, androgenic and thyroid functions.

Use of the E-SCREEN assay to assess the estrogen agonists and antagonists: In order to develop a bioassay for estrogens, it is necessary to choose the effect that would be used as end point. There are at least three definitions of estrogens. Hertz proposed that: "estrogens are substances, which elicit the proliferative activity of the organs of the female genital tract." A second definition states that estrogens are substances that, by binding to their receptors, elicit the expression of genes that are controlled by estrogen-responsive elements. Finally, the third definition unifies the biological and biochemical aspects of estrogen action and states that "estrogens are substances that elicit the proliferative activity and the control of expression of specific genes in tissues of the female genital tract." Each definition implies specific end points. The discovery of a second estrogen receptor further complicates the task of defining estrogenicity. The ERa is thought to drive the uterotropic response, since the uteri of the ERa "knock-out" mice do not respond to estrogen administration. It is not known at present which effects that are exclusively mediated by ERb .

Briefly, the E-SCREEN assay requires the completion of the following steps. A comparable number of cells are seeded in multi-well plates, in regular serum supplemented-medium. They are allowed to attach for 24 hours, and then the seeding medium is replaced with experimental medium. Cells are then allowed to proliferate for five days in the presence of CD serum-supplemented medium ("negative" control); in neighboring wells, a range of concentrations of the chemical being tested is added to these cells. In other wells, cells cultured with a range of concentrations of estradiol represent the "positive" controls. The estradiol concentration resulting in half-maximal proliferation (M50) is about 7-12 pM. The E-SCREEN assay is the most sensitive bioassay available so far. The assay also discriminates between partial and full-agonists.

Antagonists are detected in a two-step test by a modification of the E-SCREEN assay. In the first step, the ability of the chemical to inhibit estrogen action is tested. A range of concentrations of the presumptive antagonist is added to medium containing the minimal dose of estradiol that induces maximal proliferation. Once it is found that a compound inhibits estrogen action as shown following the procedure outlined in the previous paragraph (Step-1), it is imperative to verify that this is a receptor-mediated phenomenon; this is verified by increasing the concentration of estrogen which can reverse it. This reversal by estrogen, called "estrogen rescue", is the hallmark of a true antagonist. In this second step, the minimal dose of the antagonist needed for maximal inhibition is tested in the presence of a range of doses of estradiol.

A recent study compared three assays, in vitro estrogen receptor (ER) competitive binding assays, yeast-based reporter gene assays, and the E-SCREEN assay in order to determine their quantitative agreement in identifying structurally diverse estrogens. Assay performance was examined for relative sensitivity, detection of active/inactive chemicals and estrogen/antiestrogen activities. The examination combined individual data sets in a specific, quantitative data mining exercise. Data sets for at least 29 chemicals from five laboratories were analyzed pair-wise by X-Y plots. The ER binding assay was a good predictor for the other two assay results when the antiestrogens were excluded (r2 is 0.78 for the yeast assays and 0.85 for the E-SCREEN assays). Additionally, the examination revealed that biological information that is not apparent from any of the individual assays can be discovered by quantitative pair-wise comparisons among assays. Antiestrogens are identified as outliers in the ER binding/yeast assay, while complete antagonists are identified in the ER binding/E-SCREEN assays. In the E-SCREEN assay, only 4 of 19 (21%) low potency [log Relative Proliferative Potency (RP) <-1.5] chemicals had a Relative Proliferative Effect (RPE) less than 50% of estradiol. In contrast, 12 of 16 (75%) low potency chemicals [log Relative Potency (RP) <-1.5] in the yeast assay had a Relative Inductive Efficiency (RIE) less than 50% of estradiol. Most of these chemicals are androgens, alkylphenols, DDTs and phthalates. This finding, combined with the linear correlation of RIE with log RP, suggests these results are inherent to the reporter gene construct and demonstrates that the yeast assay has lower resolving power measured by the RIE or RPE than the E-SCREEN for low potency chemicals. Although these assays involve different levels of biological complexity, the major conclusion is that they generally provided consistent information in quantitatively determining estrogenic activity.

Use of the E-SCREEN ASSAY to determine the estrogenic activity in extracts of water samples: Over 70,000 synthetic chemicals are registered for commercial use in US EPA's Toxic Substances Control Act Inventory. Based on a review of the published literature, only about 150 compounds have been assessed for estrogenicity to date. Hence, in assessing exposure to xenoestrogens, there are two approaches. One is to perform chemical analysis of the matrix of interest for the chemicals already known to be estrogenic; this approach risks missing activity due to unidentified estrogens. A second approach is to extract the matrix and test the extract for estrogenic activity. Even better, one may do both things and ask whether the estrogenic chemicals detected in the extract may account for all of the estrogenic activity revealed by the bioassay. If the answer is no, the extract can be fractionated and the unknown estrogen may be identified. In order to do this, it is necessary to develop an extraction method and bioassay that can measure estrogenic activity with accuracy. To overcome this problem, we developed methodology to assess the total estrogen activity of mixtures and the exposure to xenoestrogens in water and in human tissues (blood and fat) which is based on the E-SCREEN bioassay. Using this approach, an estradiol dose-response curve is run with every test sample as a calibration curve and estrogenic activity is expressed in the sample as estradiol equivalents per liter. We chose as test models two chemicals, one non-volatile, bisphenol-A (BPA), and one volatile, nonylphenol (NP). Water was spiked with 100 nM BPA or 100 nM NP and triplicate extractions were carried out using two different methods (liquid-liquid and solid phase). Our results showed that the liquid-liquid extraction was more reliable than the solid phase method.

The estrogenic activity of the spiked water samples was expressed in concentration units of estradiol (estradiol equivalents, EEq) and of the chemical with which it was spiked (BPA or NP). The cell number data from the different dilutions of each extract were interpolated into the dose response curve of estradiol. The estrogenic activity of the 100 nM BPA extract was 4515.1+359.2 fM EEq, and that of the 100 nM NP extract was 13940.9+4177.8 fM Eeq. When the cell numbers were interpolated into the dose response curve of the corresponding chemical. the concentration of the BPA extract was slightly underestimated (89.2 nM), and that of the NP extract was slightly overestimated (132.8 nM). Alternatively, the proliferative effects of these chemicals, measured in estrogen-equivalents, may be converted into BPA and NP concentrations by using the RPP of BPA and NP calculated from the half-maximal point of their dose-response curves (M50). The estimated BPA concentration is 63 nM and the estimated NP concentration is 125 nM. In both instances, the accuracy of the E-SCREEN method is excellent. The minor discrepancies between measured and theoretical concentrations may be due to the fact that these chemicals are not totally pure and may contain estrogenic contaminants that are not measured by GC-MS (recovery is based on the main peak). It is remarkable, however, that a bioassay provides this level of accuracy. Examples of the application of this method to environmental samples will be discussed.

The A-SCREEN assay: The control of the proliferation of androgen target cells occurs by a two-step mechanism. In Step 1, androgens increase the proliferation of the target cells, presumably by canceling the inhibition exerted by a specific plasma-borne protein. In Step 2, androgens directly trigger the expression of AS3, an endogenous inhibitor of cell proliferation. To further investigate the conditions necessary for sex hormone target cells to express Step 1 or Step 2 we transfected a full-length, wild-type, androgen receptor into MCF7 cells. These stably transfected MCF7-AR1 cells only express Step 2.

Contrary to the process for detecting estrogens (measuring the induction of cell proliferation or Step 1), we use the inhibition of cell proliferation as an endpoint to screen for androgen agonists. To screen for androgen antagonists we use the inhibition of this response, which is an increase of the cell number. To avoid interference by estrogen on cell proliferation activity, this assay is performed in serumless medium supplemented with insulin and transferrin. In these conditions, estrogens do not affect MCF7-AR1 cells (or the ancestral MCF7 cells from which they derive). Both DHT and the synthetic androgen R1881 inhibit the proliferation of MCF7-AR1 cells. The cells undergo G0 arrest within 24 hours of androgen treatment. Maximal activity has been observed at 0.1 nM DHT. The point of half-maximal activity (M50) is about 0.01 nM. This assay is being used to assess the androgenicity of water extracts.

Conclusions: "In culture" screens using breast cancer cells are accurate and sensitive for the detection of estrogen and androgen agonists and antagonists.