Obstacles to the prediction of estrogenicity from chemical structure: assay-mediated metabolic transformation and the apparent promiscuous nature of the estrogen receptor

Endocrine disrupting potency of organic pollutant mixtures isolated from commercial fish oil evaluated in yeast-based bioassays

The aim of this work was to evaluate the activity of xenobiotic mixtures containing persistent organic pollutants isolated from commercial fish oil samples against sex hormone receptors, including estrogen and androgen. The applied bioassay was based on transgenic yeast strains. The mixtures were extracted from the samples using the semi-permeable membrane dialysis technique and analyzed with gas chromatography/ion trap mass spectrometry. It turned out that mixtures of chemicals isolated from fish oil may interact with human steroid sex hormone receptors in various ways: the tested samples showed both estrogenic and anti-androgenic activity. Calculated 17β-estradiol equivalents for the tested samples ranged between 0.003 and 0.073 pg g-1 (fat). Anti-androgenic activity expressed as the flutamide equivalent concentration was in the 18.58-216.21 ng g-1 (fat) range. Polychlorinated biphenyls and various DDT metabolites were the main fish oil pollutants influencing the receptors. Additivity and/or synergy between chemicals was observed in the ER/AR mediated response.

Optimized yeast-based in vitro bioassay for determination of estrogenic and androgenic activity of hydroxylated / methoxylated metabolites of BDEs / CBs and related lipophilic organic pollutants

Persistent organic pollutants (POPs) are known to show endocrine disrupting (ED) activity, including interactions with hormone receptors. The aim of this work was to develop a bioassay applicable for evaluation of ED potency of highly lipophilic metabolites of POPs. To that end, a yeast-based bio-assay protocol was used. Estrogenic / androgenic activity of some native brominated biphenyl ethers (BDEs) / chlorinated biphenyls (CBs), and their hydroxylated / methoxylated metabolites was assessed. Since data (including potency compared to reference native hormones) obtained using different protocols vary, the possibility that yeast transforms POPs into some more potent compounds was first checked; it seems that no such transformation is important from the test applicability standpoint. The developed method was sensitive with EC50 values 6.5*10^-11 M and 4.5*10^-9 M calculated for E2 and DHT, respectively. Both CBs and BDEs show weak estrogenic activity negatively correlated with the degree of their halogenation, but their metabolites are significantly more potent xenohormones. 4-OH-2,2',4',6'-TeCB was the most potent estrogen receptor (ER) agonist among all tested compounds; its activity was only 1,000 times lower than that of native E2.

A Demonstration of the Uncertainty in Predicting the Estrogenic Activity of Individual Chemicals and Mixtures From an In Vitro Estrogen Receptor Transcriptional Activation Assay (T47D-KBluc) to the In Vivo Uterotrophic Assay Using Oral Exposure

In vitro estrogen receptor assays are valuable tools for identifying environmental samples and chemicals that display estrogenic activity. However, in vitro potency cannot necessarily be extrapolated to estimates of in vivo potency because in vitro assays are currently unable to fully account for absorption, distribution, metabolism, and excretion. To explore this issue, we calculated relative potency factors (RPF), using 17α-ethinyl estradiol (EE2) as the reference compound, for several chemicals and mixtures in the T47D-KBluc estrogen receptor transactivation assay. In vitro RPFs were used to predict rat oral uterotrophic assay responses for these chemicals and mixtures. EE2, 17β-estradiol (E2), benzyl-butyl phthalate (BBP), bisphenol-A (BPA), bisphenol-AF (BPAF), bisphenol-C (BPC), bisphenol-S (BPS), and methoxychlor (MET) were tested individually, while BPS + MET, BPAF + MET, and BPAF + BPC + BPS + EE2 + MET were tested as equipotent mixtures. In vivo ED50 values for BPA, BPAF, and BPC were accurately predicted using in vitro data; however, E2 was less potent than predicted, BBP was a false positive, and BPS and MET were 76.6 and 368.3-fold more active in vivo than predicted from the in vitro potency, respectively. Further, mixture ED50 values were more accurately predicted by the dose addition model using individual chemical in vivo uterotrophic data (0.7-1.5-fold difference from observed) than in vitro data (1.4-86.8-fold). Overall, these data illustrate the potential for both underestimating and overestimating in vivo potency from predictions made with in vitro data for compounds that undergo substantial disposition following oral administration. Accounting for aspects of toxicokinetics, notably metabolism, in in vitro models will be necessary for accurate in vitro-to-in vivo extrapolations.

A rule-based expert system for chemical prioritization using effects-based chemical categories

A rule-based expert system (ES) was developed to predict chemical binding to the estrogen receptor (ER) patterned on the research approaches championed by Gilman Veith to whom this article and journal issue are dedicated. The ERES was built to be mechanistically transparent and meet the needs of a specific application, i.e. predict for all chemicals within two well-defined inventories (industrial chemicals used as pesticide inerts and antimicrobial pesticides). These chemicals all lack structural features associated with high affinity binders and thus any binding should be low affinity. Similar to the high-quality fathead minnow database upon which Veith QSARs were built, the ERES was derived from what has been termed gold standard data, systematically collected in assays optimized to detect even low affinity binding and maximizing confidence in the negatives determinations. The resultant logic-based decision tree ERES, determined to be a robust model, contains seven major nodes with multiple effects-based chemicals categories within each. Predicted results are presented in the context of empirical data within local chemical structural groups facilitating informed decision-making. Even using optimized detection assays, the ERES applied to two inventories of >600 chemicals resulted in only ~5% of the chemicals predicted to bind ER.

Development of molecular docking-based binding energy to predict the joint effect of BPA and its analogs

A general proposal for predicting the joint effect of endocrine disrupting chemicals by examining binding energy models was developed in this study. 2,2-bis(4-hydroxyphenyl)propane (BPA) and 11 of its analogs were chosen, and the estrogenic activity of each compound was measured by determining its EC50 value using a recombinant gene yeast assay. Binding energies (BEs) were calculated using Surflex-Docking software. The analysis of the relationship between EC50 values and BEs showed that there is a linear correlation between the BEs and EC50 values. Furthermore, the analysis of the given binary and quaternary mixtures of BPA and three of its analogs showed that the joint effects of the mixtures were affected by the proportions of the chemicals in each mixture and their relative binding energy. The correlation between the joint effects of mixtures and the binding energy of the individual compounds has been described using one formula, which can be used to predict the joint effects of other mixtures.

Molecular modelling and QSAR analysis of the estrogenic activity of terpenoids isolated from Ferula plants

The relationship between chemical structure and estrogenic activity in a series of terpenoid esters with aromatic and aliphatic acid substituents isolated from Ferula plants, was studied. The fragments of the terpenoid structure that are potentially responsible for estrogenic activity were revealed. A quantitative structure-estrogenic activity study has been carried out using the QSAR approach with use of data derived from quantum-chemical calculations as well as data generated from three-dimensional structures of terpenoids. A number of molecular descriptors was obtained from the density functional theory (DFT) at the B3LYP/6-31G(d, p) level of calculation. Comparative analysis of the quantum-chemical computational data was also performed to confirm hypothesis concerning importance of the distance between the oxygen of alcohol hydroxyl group and the functional group in the para-position of the benzene ring (the hydroxyl or methoxy group). Use of the Genetic Algorithm in the QSAR analysis allowed the structural and physicochemical parameters of the terpenoids responsible for estrogenic activity to be determined. A significant QSAR model was obtained with an r(2) value of 0.892. The resulting model showed a reliable dependence of estrogenic activity of the terpenoids on such parameters as molecular shape, number of phenolic groups, surface polarity and the energy of the highest occupied molecular orbital.

Experimental parameters affecting sensitivity and specificity of a yeast assay for estrogenic compounds: results of an interlaboratory validation exercise

In vitro assays are considered as the first step in a tiered approach to compound screening for hormonal activity. Although many new assays have been developed in recent years, little attention has been paid towards assay validation. Our objective was to identify critical experimental parameters in a yeast estrogen screen (YES) that affect its sensitivity and specificity. We investigated the role of incubation time, solvent type, yeast inoculum growth stage and concentration on the outcome of the YES. Compounds tested included new and established agonists, antagonists and negative controls, and results were evaluated according to prefixed statistical criteria. In addition, we assessed the assay's performance in a blind interlaboratory validation exercise (IVE). An incubation time of five days was necessary to positively identify the estrogenic properties of all agonists tested, when dissolved in DMSO. Longer incubation times were required when using an ethanol protocol. Similar estrogenic activity was reported for benzyl butyl phthalate, bisphenol-A, methoxychlor, permethrin and genistein in the IVE. One out of the three laboratories did not classify alpha,beta-endosulfan, dissolved in DMSO, as an estrogen. The same was true for 4,4'-DDE and lindane, dissolved in ethanol, a result that might be attributable to an inappropriate yeast start concentration and/or growth stage. These validation experiments show that under appropriate experimental conditions the YES yields sensitive, specific and reliable results. Therefore it fulfills the requirements as a first step screening assay to evaluate the capacity of chemicals to interact with the estrogen receptor.

Evaluating the interactions of vertebrate receptors with persistent pollutants and antifouling pesticides using recombinant yeast assays

The development of in vitro methods for screening potentially harmful biological activities of new compounds is an extremely important way to increase not only their intrinsic environmental safety, but also the public perception of the safety standards associated with them. In this work we use two yeast systems to test the ability of different chemicals to bind and activate two vertebrate receptors which are intimately related to adverse biological effects of pollution in exposed fauna: the estrogen receptor (ER) and the aryl hydrocarbon receptor (AhR). The panel of compounds analysed here includes well-known pollutants, like PCBs, pp'-DDT and hexachlorobenzene, together with the less-known, emerging putative pollutants, such as Sea-Nine, Irgarol and diuron. Results show the ability of some of these compounds to interact with one or both receptors, provide hints about the relationship between structure and activity, and suggest mechanistic explanations for the biological activities already described in whole-animal experiments. In addition, we show that AhR may have an intrinsic ligand promiscuity comparable to that of ER, a feature not fully appreciated in the past due to the technical difficulties involved with testing highly lipophilic substances in yeast-based assays.

Receiver operating characteristic analysis for environmental diagnosis. A potential application to endocrine disruptor screening: in vitro estrogenicity bioassays

The realization that certain chemicals are able to disrupt hormonal systems in humans and wildlife has challenged the way we assess risk from chemicals and led national and international agenciesto devise programsto screen chemicals for endocrine-disrupting properties. Chemicals capable of mimicking sex hormones, such as estrogens and androgens, have received the most attention, and although not yet validated, in vitro techniques to test for such properties are well developed. Receiver operating characteristic (ROC) analysis has been successfully used in the biomedical and military fields for several decades to assess the accuracy of diagnostic tests in terms of both their sensitivity and specificity. This approach is applied here to demonstrate its potential to assess how well in vitro bioassays can predict estrogenicity in vivo. Despite the limited availability of suitable data, the ROC curves obtained indicate that these bioassays are effective diagnostic tests. The potential sources of false positives and false negatives are identified and potential applications to endocrine disruptor screening programs discussed.

Analytical methodology for the identification of estrogenic contaminants in fish bile

Effluents from wastewater treatment works (WwTWs) contain estrogenic contaminants that can cause feminised responses in fish. In order to assess the identity of estrogenic contaminants taken up by fish exposed to effluents, an analytical method was developed to detect estrogenic substances in fish bile, where many xenobiotics are excreted and concentrated. Estrogenic metabolites in bile were deconjugated using enzymatic hydrolysis and the estrogenic activity was determined using a yeast estrogen receptor transcription screen (YES). Hydrolysed samples were concentrated by solid-phase extraction (SPE) prior to fractionation by reversed-phase high-performance liquid chromatography (HPLC). Active HPLC fractions were detected by YES assay and analysed by gas chromatography-mass spectrometry (GC-MS) after trimethylsilylation. The method was validated using bile samples from immature female rainbow trout, which had been exposed to either tap water or an undiluted estrogenic effluent for 10 days. Hydrolysis of bile from effluent-exposed fish was complete within 16 h add most of the estrogenic activity in the bile was released by 3-glucuronidase rather than sulfatase or 3-glucosidase treatment. The estrogenic activity of hydrolysed bile from effluent-exposed fish ranged between 530 and 1440 ng E2eq/mL and was 17-48-fold greater than the activity of bile from reference fish exposed to tap water. The estrogenic activity of bile samples decreased with time in storage (at-70 degrees C by 7% per month). The recovery of estrogenic activity from SPE was 96 +/- 7% (mean +/- SD), from HPLC fractionation 87 +/- 7% and for the whole method 81 +/- 7% (n = 7). 17beta-Estradiol, estrone, 17alpha-ethinylestradiol, nonylphenol and short-chain nonylphenol polyethoxylates were all identified from GC-MS analysis of active HPLC fractions of bile from effluent-exposed trout, whereas only 17beta-estradiol was detected in bile from fish exposed to tap water. There were also several other minor estrogenic components, at present unidentified, in bile of effluent-exposed fish. The work shows that fractionation of fish bile is a useful approach to identifying mixtures of estrogenic contaminants taken up by fish from WwTW effluents and has the potential for application in the detection of other endocrine disrupting chemicals in fish tissues.

The impact of variable selection on the modelling of oestrogenicity

Many oestrogenic chemicals exert their activity via specific interactions with the oestrogen receptor (ER). The objective of the present study was to identify significant descriptors associated with the ER binding affinities of a large and diverse set of compounds to drive quantitative structure-activity relationships (QSARs). To this end, a variety of statistical methods were employed for variable selection. These included stepwise regression and partial least squares (PLS) analyses, as well as a non-linear recursive partitioning method (Formal Inference-based Recursive Modelling). A total of 157 molecular descriptors including quantum mechanical, graph theoretical, indicator variables and log P were used in the study. Furthermore, cluster analysis of variables was performed to identify groups of descriptors representing similar molecular features. Hierarchical PLS analyses were performed, where the scores of the significant components of either PLS or principle component analysis (PCA), performed separately on each cluster, were used as the variables for the top model. This reduced the number of the variables representing the larger clusters, leading to a similar number of descriptors for each distinct molecular feature. The results showed that the most important molecular properties for stronger ER binding affinity are molecular size and shape, the presence of a phenol moiety as well as other aromatic groups, hydrophobicity and presence of double bonds. The best PLS model obtained, in terms of predictive ability, was a hierarchical PLS model. However, a rigorous validation study showed that the MLR model using descriptors selected by stepwise regression has greater predictive power than the PLS models.

Examination of the in vitro (anti)estrogenic, (anti)androgenic and (anti)dioxin-like activities of tetralin, indane and isochroman derivatives using receptor-specific bioassays

Molecules derived from tetralin, indane and isochroman are often used in the synthesis of fragrance materials. The two polycyclic musk fragrances AHTN (6-acetyl-1,1,2,4,4,7-hexamethyltetralin), HHCB (1,2,4,6,7,8-hexahydro-4,6,6,7,8,8-hexamethylcyclopenta-gamma-2-benzopyran) and ADBI (4-acetyl-1,1-dimethyl-6-tert-butylindane) are derived from tetralin, isochroman and indane, respectively. In previous studies, AHTN and HHCB have been shown to antagonize estrogen receptors (ERs), both in vitro and in vivo. Here, we used two newly developed reporter gene assays, to examine the agonistic and antagonistic properties of several indane, tetralin and isochroman derivatives towards the human androgen receptor (AR) and aryl hydrocarbon receptor (AhR). Additionally, we also assessed (anti)estrogenicity of these compounds. A number of compounds showed weak estrogenic activity towards the human ER alpha. Several compounds showed (anti)estrogenic effects, starting at a concentration of 0.1 microM. Surprisingly, almost all compounds were found to be AR antagonists, starting at 0.1 microM. None of the compounds tested, showed either agonism or antagonism towards the AhR. Non-specific effects via crosstalk of the AhR and the ER or AR can therefore be ruled out. As far as we are aware, molecules derived from indane, tetralin and isochroman showing direct interaction with the ER and AR have not been reported previously.

Structure--activity relationships of the estrogenic sesquiterpene ester ferutinin. Modification of the terpenoid core

Esterification of p-hydroxybenzoic acid, a very weak estrogenic compound, with the daucane alcohol jaeschkeanadiol (1b) leads to a spectacular magnification of the estrogenic activity. To identify the structural elements responsible for this effect, the terpenoid core of jaeschkeanadiol p-hydroxybenzoate (ferutinin, 1a) was modified, capitalizing on the presence of two functionalities, the monoacylated, hydrogen-bonded 1,3-diol system and the double bond. The hydrogen bonding, while possibly useful, was not critical for activity, while hydrogenation and cyclopropanation of the double bond were tolerated. Conversely, oxidative modifications of the double bond that placed a hydroxyl on the alpha-face of the molecule proved detrimental. Taken together, these observations identified the substitution at C-8/C-9 as critical for activity.

Scientific issues associated with the validation of in vitro and in vivo methods for assessing endocrine disrupting chemicals

The assays required to assess the potential of chemicals to act as endocrine disrupting (ED) agents are either in place or are under current development. However, the validation and utilisation of these assays is currently being hampered by uncertainties regarding their purpose and required sensitivity, and uncertainties as to the intrinsic variability of the parameters being measured. This article discusses these several sources of uncertainty and the intrinsic variability of many of the key assay parameters. It is concluded that current uncertainties regarding the use of ED assays, and the extrapolation of rodent effects to humans, are due to the absence of an extensive agreed rodent control database for the developmental parameters under study, coupled to the established intrinsic variability of these parameters between strains/species of test animals and test protocols. Only when these factors are generally accepted, well studied and controlled for, will it be possible to employ ED assays with confidence and to relate assay data to effects likely to be seen in humans.

Assessment of estrogenic activity in some common essential oil constituents

Estrogenic responses have not only been associated with endocrine function, but also with cognitive function. Several studies have indicated that estrogen replacement therapy has favourable effects on cognition, and may have potential in the prevention and treatment of Alzheimer's disease. Thus, ligands for the estrogen receptor, that have a better efficacy and adverse-effect profile than drugs currently available, require investigation. This study was undertaken to investigate the potential estrogenic activity of a number of essential oil constituents. Initially, estrogenic activity was determined by a sensitive and specific bioassay using recombinant yeast cells expressing the human estrogen receptor. At high concentrations, estrogenic activity was detected for citral (geranial and neral), geraniol, nerol and trans-anethole, while eugenol showed anti-estrogenic activity. Molecular graphics studies were undertaken to identify the possible mechanisms for the interaction of geranial, neral, geraniol, nerol and eugenol with the ligand-binding domain of the estrogen alpha-receptor, using the computer program HyperChem. Citral, geraniol, nerol and eugenol were also able to displace [(3)H]17beta-estradiol from isolated alpha- and beta-human estrogen receptors, but none of these compounds showed estrogenic or anti-estrogenic activity in the estrogen-responsive human cell line Ishikawa Var I at levels below their cytotoxic concentrations, and none showed activity in a yeast screen for androgenic and anti-androgenic activity. The potential in-vivo estrogenic effects of citral and geraniol were examined in ovariectomized mice, but neither compound showed any ability to stimulate the characteristic estrogenic responses of uterine hypertrophy or acute increase in uterine vascular permeability. These results show that very high concentrations of some commonly used essential oil constituents appear to have the potential to interact with estrogen receptors, although the biological significance of this is uncertain.

Dietary bisphenol A prevents ovarian degeneration and bone loss in female mice lacking the aromatase gene (Cyp19 )

We previously generated mice lacking aromatase activity by targeted disruption of Cyp19 (ArKO mice), and reported phenotypes of the female mice, showing hemorrhage formation and follicular depletion in the ovary, diminution in uterine size, and bone loss. In the present study, we examined the influence of dietary bisphenol A (BPA), a monomer used for the production of polycarbonate and known to have estrogenic activity, on these phenotypes of the ArKO mice. When ArKO mice were fed chow diets supplemented with 0.1% or 1% (w/w) BPA for 5 months, they were protected from ovarian degeneration, uterine diminution and bone loss in a dose-dependent manner. Northern blot analyses of ovarian RNA of ArKO mice showed differences in the expression levels of insulin-like growth factor (IGF)-I, IGF-I receptor, growth differentiation factor 9 and bone morphogenetic protein 15 as compared with those in the ovaries of wild-type mice. The differences in the expression levels were restored by dietary BPA. In the ArKO uteri, expression of progesterone receptor and vascular endothelial growth factor mRNAs was diminished, and was restored by BPA to the levels in wild-type mice. In contrast, BPA had little effect on the ovarian, uterine and skeletal structures of wild-type mice. In conclusion, estrogenic effects of BPA on the reproductive tract as well as skeletal tissue were evident in adult female ArKO mice. These results suggest that the ArKO mouse is an animal model suitable for studying effects of estrogenic chemicals as well as estrogen in vivo.

Estrogenic activities of chemicals in diesel exhaust particles

In a previous study, we focused on estrogenic activity of the hexane extract of diesel exhaust particles (DEP). The extract of hexane was first fractionated to acidic, phenolic and neutral portions according to their chemical properties, of which the neutral fraction was fractionated by column chromatography on silica gel. The chemical structures of compounds in these fractions were then analyzed. It was found that the neutral fraction of the hexane extract of DEP contains dibenzothiophene derivatives, one of which, 4,6-dimethyldibenzothiophene, possesses estrogenic activity.