Bioassays for estrogenic activity: development and validation of estrogen receptor (ERalpha/ERbeta) and breast cancer proliferation bioassays to measure serum estrogenic activity in clinical studies.

Standard estrogenic prodrugs such as estradiol valerate (E2V) and increasingly popular phytoestrogen formulations are commonly prescribed to improve menopausal health. These drugs are metabolized to numerous bioactive compounds, known or unknown, which may exert combinatorial estrogenic effects in vivo. The aim of this study is to develop and validate estrogen receptor (ER) alpha/ERbeta reporter gene and MCF-7 breast cancer cell proliferation bioassays to quantify serum estrogenic activities in a clinical trial setting. We measured changes in serum estrogenicity following ingestion of E2V and compared this to mass spectrometric measurements of its bioactive metabolites, estrone and 17beta-stradiol. ERalpha bioactivity of the 192 serum samples correlated well (R = 79%) with 17beta-estradiol levels, and adding estrone improved R to 0.83 (likelihood ratio test, P < 0.0001), suggesting that the ERalpha assay reflects summated activity of compounds in serum. ERbeta correlated moderately (R = 0.52) with estrone and 17beta-estradiol, with an estrone/17beta-estradiol coefficient ratio that was twice that of ERalpha, indicating estrone was more active on a molar basis in the ERbeta assay. Unlike the ERalpha and ERbeta bioassays, MCF-7 cell proliferation was driven by 17beta-estradiol, and addition of estrone did not increase the predictive value of the model, suggesting that the driver or drivers for breast cancer cell proliferation were not the same as for ERalpha and ERbeta transactivation. In contrast, a decoction of the traditional Chinese medicinal herb Epimedium pubescens did not induce significant changes in estrogenic bioactivity over baseline. These data indicate that ERalpha/ERbeta reporter gene and MCF-7 breast cancer cell proliferation bioassays reflect different aspects of estrogenic activity and that these assays suggest that the Epimedium formulation tested is unlikely to exert significant estrogenic effects in humans.