Estrogen receptor (ER)-mediated activation by endocrine disrupting chemicals (EDCs)

Abstract

Studies have shown that many synthetic and natural chemicals, classified as endocrine-disrupting chemicals (EDCs), disrupt the normal endocrine system function in both humans and animal species. In our recent study, we used human HepG2 (hepatocellular cancer), HeLa (cervical cancer), and Ishikawa (endometrial cancer) cell lines to analyze the estrogenic effects of a set of known EDCs on the ERs. Our studies demonstrate the mechanistic importance of chemical structure similarities and cell type/promoter specificity when evaluating the potential activities of EDCs. The synthetic EDCs, such as BPA and its analogs, mainly affect ERα and can act as both agonists and antagonists in a dose-dependent manner. The natural EDCs (daidzein and genistein) are stronger estrogens due to their estrogenic activations for both ERα and ERβ. Most importantly, the dosage determines whether a specific endocrine disrupting compound is beneficial or toxic to the human population and one mechanism cannot explain the complexity of these compounds actions.

Keywords: :

• EDCs
• ER
• ER target genes
• ERE
• estrogenic

For decades, studies have shown that many synthetic and natural chemicals, classified as endocrine-disrupting chemicals (EDCs), disrupt the normal endocrine system function in both humans and animal species.^1,2 One of the first activities shown to be affected was functions involving estrogens, and these compounds were termed xenoestrogens or phytoestrogens.^3,4 Since that early time, however, the designation has become much broader, as research has shown that these compounds, more appropriately termed xenobiotics, have effects on a broad range of organ systems.⁵ Estrogen is a key regulator of growth, differentiation, and physiological functions via estrogen receptor (ER) α and β.⁶ The classical ER mechanism involves ERs directly binding to estrogen response elements (EREs) located in the promoter region of target genes. The non-classical mechanism is the “tethered” mechanism, which involves the ERs regulating gene expression by associating with other transcription factors such as c-Jun or c-Fos, which bind to AP-1/Sp1 sites without direct ER-DNA binding.⁶

In our recent study, we used human HepG2 (hepatocellular cancer), HeLa (cervical cancer), and Ishikawa (endometrial cancer) cell lines to analyze the estrogenic effects of a set of known EDCs on the ERs. We categorized these EDCs into three groups on the basis of product class and similarity of chemical structure.⁷ We demonstrated that: (1) For ERE-mediated activation (the classical mechanism), bisphenol A (BPA), bisphenol AF (BPAF), and HPTE more strongly activated ERα than ERβ. Daidzein, genistein, kaempferol, and coumestrol activated both ERα and ERβ to a similar degree. Endosulfan and kepone weakly activated ERα (2). For the AP-1/Sp1-mediated activation (the “tethered” mechanism), the EDCs induced weak ERα or ERβ activity in a cell-type and promoter-specific manner (3). BPA and BPAF consistently activated endogenous ER target genes, but the activities of other EDCs on ER target gene expression changes were compound-specific.⁷ From this study, we concluded that EDCs with similar chemical structures tended to have similar effects on the classical mechanism of ERs. Importantly, their activities for the ERs do not, however, correlate with their previously reported ligand binding affinities.⁷

Most EDCs are synthetic chemicals used in a variety of industries. Human and animal studies indicate that BPA exposure is associated with many health problems, including infertility, early-onset puberty, obesity, diabetes, cardiovascular effects, and prostate, uterine, and breast cancers.² BPA is able to impede the activity of endogenous estrogens by disrupting the proper activity of ERs in a diverse set of target tissues.⁸ BPA and its isomeric analogs are widely used in manufacturing and consumer products (i.e., plastics, cash register recepits, linings of aluminum cans).^8,9 BPAF, a fluorinated derivative of BPA, is potentially more toxic than BPA due to the electronegative effects of the CF3 group.¹⁰ HPTE is an analog of BPA and BPAF (with the addition of a CCl3 group), and is nearly as potent as BPAF.¹⁰ Using microarray analysis from the mouse uterus, we classified that BPA and HPTE are weak estrogens.¹¹ From our more recent studies to evaluate the estrogenic activity of BPA and BPAF on the ERs, we conclude that BPA and BPAF can function as EDCs by acting as cell-specific agonists (≥ 10 nM) or antagonists (≤ 10 nM) for ERα and ERβ.¹² These findings demonstrate the further need to study these isomeric compounds in a variety of tissues and/or cell lines to define their full mechanisms of action.

Over 10 y ago, a study by the National Toxicology Program indicated evidence for low-dose effects of a select number of well-known EDCs.¹³ More recent studies conclusively demonstrate that some EDCs exhibit nonmonotonic dose-response curves; therefore, the effect at low doses cannot be predicted by the effect(s) observed at high doses.¹⁴ Due to the pleiotropic actions of BPA, it is defined as a selective estrogen receptor modulator (SERM), which binds to ERs and to act as an agonist or antagonist in a tissue-specific manner.¹⁴ In vivo studies show that the antagonistic effects of BPA at low concentrations (0.1–1 nM) inhibit key adipokines, which are proposed to protect humans from developing metabolic syndrome.¹⁵ On the other hand, a study also using low dose of BPA treatment found in mouse β TC-6 cells that treatment decreases insulin secretion though ERβ.¹⁶ Also, in vitro studies in HeLa cells report that BPAF is a full agonist for ERα, but is a highly specific antagonist at low doses for ERβ.¹⁰ To investigate mechanistically the antagonistic effect of EDCs, we treated cells with low doses of BPA or BPAF (1 or 10 nM) and 17β-Estradiol (E2, 10 nM). Significant inhibition of E2-mediated ERE reporter activity was obtained from co-treatment with BPA for ERα in Ishikawa cells and from co-treatment with BPAF for ERβ in HeLa cells.¹² These results confirm the tissue selective spectrum of effects in vivo with BPA and BPAF as possible antagonists in the presence of E2. The in vitro cell observations have been confirmed by recent studies that have shown in vivo uterotrophic antagonistic activity of BPA.¹⁷

As concern regarding the toxic effects of BPA increases, BPA found in many consumer products is now being replaced with analogs such as bisphenol S (BPS), bisphenol B (BPB), and bisphenol F (BPF).¹⁸ Limited studies showed that these compounds had similar estrogenic activities to those of BPA.¹⁸ To investigate the agonistic activity of the bisphenol analogs via ERα, we compared ERE-mediated estrogenic activities of BPA, BPAF, and BPS in three different human cancer cells. Our preliminary data suggests that BPS is comparable with BPA and BPAF for ER-mediated reporter activity and gene expression changes, but further analysis is underway.

Much of the BPA, BPAF, and BPS work has been done in vitro, but in vivo rodent studies examine the effects of BPA at different stages of development.⁴ As the hormonal milieu differs during development, the timing of exposure to environmental toxicants is key to understanding increased susceptibility to EDCs. Studies have begun to examine the fetal basis of adult disease in rodents exposed to BPA,² but studies for BPAF and BPS need to be conducted. Human data for BPA is controversial due to the ubiquitous levels of BPA found in hospital laboratory plastics, but very few, if any, human studies are noted for BPAF and BPS.¹⁸ Of need for the studies involving BPA analogs are well-controlled studies using high and low doses at critical windows of development to examine the developmental basis of disease that results after a lag between the time of exposure and the manifestation of a disorder.^2,19 Even with the ubiquitous nature of these EDCs, much is still needed to fully elucidate the complexity of these synthetic chemicals.

Some EDCs naturally occur in foods derived from plant origin (especially soy-based foods). The best known examples are phytoestrogens such as daidzein and genistein.²⁰ Since this group of EDCs structurally resembles endogenous estrogens, they compete with endogenous estrogens for binding to ERs.²¹ Many research reports show that genistein influences lipid homeostasis and insulin resistance, counteracts inflammatory cytokines, and possesses antidiuretic properties.²² With similar chemical structures, both treatment with daidzein and genistein exhibited strong ERE-mediated estrogenic activities via ERα and ERβ in our studies.⁷ Interestingly, this group of EDCs has stronger binding affinity to ERβ than ERα.²¹ By interacting with ERs, the phytoestrogens exert estrogenic effects in humans; however, according to the dose or the exposure duration, this group of EDCs can also act as antagonists.²³ Genistein is known to have health benefits; it offers protection against cardiovascular diseases, attenuates post-menopausal problems, and helps prevent osteoporosis, and is implicated in prostate cancer prevention.^24,25 However, the dose of phytoestrogens used in these studies is higher than the concentrations found in the plasma of people who regularly consume soy.²⁴ Studies that use doses of daidzein and genistein that are representative of levels found in the human population need to be undertaken.²⁴

In summary, our studies demonstrate the mechanistic importance of chemical structure similarities and cell type/promoter specificity when evaluating the potential activities of EDCs. The synthetic EDCs such as BPA and its analogs mainly affect ERα and can act as both agonists and antagonists in a dose-dependent manner. The natural EDCs (daidzein and genistein) are stronger estrogens due to their estrogenic activations for both ERα and ERβ. Most importantly, the dosage determines whether a specific endocrine disrupting compound is beneficial or toxic to the human population and one mechanism cannot explain the complexity of these compounds actions.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

10.4161/endo.27197