Estrogenic effects of environmental chemicals: an interspecies comparison

The development of various in vitro screening methods has led to identification of novel estrogenic chemicals of natural and anthropogenic origin. In this study, the (anti)estrogenic potential of several environmental chemicals were compared in an array of in vitro test systems comprising: (i) competitive binding to estrogen receptors derived from the human breast cancer cell line MCF-7 (hER) and rainbow trout (Oncorhynchus mykiss) (rtER), (ii) a proliferation assay with MCF-7 cells (E-SCREEN), and iii) induction of vitellogenin (rtVtg) in isolated rainbow trout hepatocytes. The results showed substantial differences in assay sensitivity for potent estrogens like 17beta-estradiol, diethylstilbestrol and zearalenone (ranking order of sensitivity: E-SCREEN > hER approximately rtER approximately rtVtg). Chemicals like 4-n-nonylphenol and bisphenol A had higher relative binding affinity to the hER, whereas 4-t-butylphenol and 4-n-butylphenol showed highest affinity to the rtER. Zearalenone and the novel estrogen 4-t-butylhexanol displayed a considerable higher relative potency in the E-SCREEN than the rtVtg assay, whereas alkylphenols and the novel estrogen mimic 4-t-butyl-nitrobenzene were most potent in fish cells. Correlation analysis of data from the test systems suggest that interspecies differences is largely due to inter-assay variation of the ER-dependent cellular responses, whereas binding to the ER are fairly similar in the two species tested.     

Limited species differences in estrogen receptor alpha-medicated reporter gene transactivation by xenoestrogens

Estrogen receptors (ERs) play an important role in estrogen function. However, it is well known that there are species differences in amino acid sequences of the ligand binding domains. Here, we report on the analysis of species differences in ER-dependent transactivation with some chemicals using reporter gene assays. Full-length ER cDNAs from human, rat, chicken, alligator (Caiman), whiptail lizard, African clawed frog and rainbow trout were prepared from hepatic mRNA by the RT-PCR method and inserted into expression plasmids. Both expression and reporter plasmids were transiently transfected into HeLa cells, and then the estrogenic effects of chemicals were analyzed in terms of induction of luciferase activity. No species differences in transactivation were found among human, rat, chicken, alligator, whiptail lizard and African clawed frog ERs. However, thermo-dependent alteration in susceptibility to 17-beta-estradiol was observed with the rainbow trout ER because of thermo-dependence of estrogen binding.     

Nestorone: a progestin with a unique pharmacological profile

Nestorone(R) (Nestorone 16-methylene-17alpha-acetoxy-19-norpregn-4-ene-3,20-dione), formerly referred to as ST 1435, is a potent progestin when given parenterally via sustained release formulations. The pharmacological profile of Nestorone was compared with that of levonorgestrel and 3-keto-desogestrel by steroid receptor binding studies and by in vivo bioassays in rats and rabbits. 3-Keto-desogestrel showed the highest binding affinity to progesterone receptors (PR) followed by Nestorone, levonorgestrel, and progesterone. The binding affinity of Nestorone to androgen receptors (AR) was 500- to 600-fold less than that of testosterone. However, both levonorgestrel and 3-keto-desogestrel showed significant binding (40 to 70% of testosterone) to AR. None of the progestins bound to estrogen receptors (ER). The progestational activity of Nestorone, levonorgestrel, and progesterone was compared using McPhail index in immature rabbits and pregnancy maintenance and ovulation inhibition tests in rats after subcutaneous (s.c.) administration. In all three tests, Nestorone was the most potent progestin. The progestational activity of Nestorone was also compared after oral and s.c. administration in rabbits. The potency of Nestorone was over 100-fold higher upon s.c. administration than via the oral route. The androgenic activity of progestins, based on the stimulation of ventral prostate (androgenic target) and levator ani (anabolic target) growth in castrated immature rats, showed good correlation with their binding affinity to AR. Nestorone showed no androgenic or anabolic activity. Nestorone did not bind to sex hormone binding globulin (SHBG), whereas both levonorgestrel and 3-keto-desogestrel showed significant binding to SHBG. The estrogenic/antiestrogenic activity of Nestorone was investigated in immature ovariectomized rats. In contrast to estradiol and levonorgestrel, Nestorone showed no uterotropic activity in ovariectomized rats. Despite significant binding to glucocorticoid receptors (GR), Nestorone showed no glucocorticoid activity in vivo. It is concluded that a strong progestational activity, combined with lack of androgenic, estrogenic, and glucocorticoid-like activities, confer special advantages to Nestorone for use in contraception and hormone replacement therapy.     

A novel carborane analog, BE360, with a carbon-containing polyhedral boron-cluster is a new selective estrogen receptor modulator for bone

Carboranes are a class of carbon-containing polyhedral boron-cluster compounds with globular geometry and hydrophobic surface that interact with hormone receptors. Estrogen deficiency results in marked bone loss due to increased osteoclastic bone resorption in females, but estrogen replacement therapy is not generally used for postmenopausal osteoporosis due to the risk of uterine cancer. We synthesized a novel carborane compound BE360 to clarify its anti-osteoporosis activity. BE360 showed a high binding affinity to estrogen receptors (ER), ERα and ERβ. In ovariectomized (OVX) mice, femoral bone volume was markedly reduced and BE360 dose-dependently restored bone loss in OVX mice. However, BE360 did not exhibit any estrogenic activity in the uterus. BE360 also restored bone loss in orchidectomized mice without androgenic action in the sex organs. Therefore, BE360 is a novel selective estrogen receptor modulator (SERM) that may offer a new therapy option for osteoporosis.     

Estrogenic activity of procymidone in rainbow trout (Oncorhynchus mykiss) hepatocytes: a possible mechanism of action

It is known that procymidone modifies sexual differentiation in vivo and in vitro, and that it induces vitellogenin (Vtg) synthesis in primary cultured rainbow trout hepatocytes. The aim of this study was to evaluate the mechanism underlying this latter in vitro estrogenic action. The cells were treated for 24 h with procymidone 150 microM (with 17beta-estradiol [E2] 20 microM as a positive control) combined with an estrogen receptor (ER) antagonist (tamoxifen 20 microM or ICI 182,780 1 microM) or, given the drug toxic action on the production of reactive oxygen species (ROS), a free radical scavenger (alpha-tocopherol 30 microM). The results from ELISA experiments provided evidence that procymidone Vtg-induction is inhibited by ER antagonists and by alpha-tocopherol suggesting that both ER and ROS are involved in this effect. The ROS detection revealed that the treatment with alpha-tocopherol and tamoxifen completely prevented ROS induction by procymidone, that was not inhibited by ICI 182,780. In exploring the mechanism mediating these events and its timing, we found that procymidone induced mitogen-activated protein kinase (MAPK) at 30 and 60 min, and that this effect was blocked by co-treatment with alpha-tocopherol. In summary, the results of the study clearly support the idea that the estrogenic activity of procymidone in primary cultured trout hepatocytes is mediated by ROS production, and that this activity is similar to that of the ligand-independent ER activation involving MAPK.     

The EPI bioassay identifies natural compounds with estrogenic activity that are potent inhibitors of androgenic pathways in human prostate stromal and epithelial cells

The reactive stromal phenotype is an important factor for prostate cancer progression and may be a new target for treatment and prevention. A new high efficiency preclinical protocol, the EPI bioassay, reflects the interaction of endocrine, paracrine and immune, (EPI) factors on induced androgen metabolism in human prostate reactive stroma. The bioassay is based on co-culturing human primary prostate stromal cells and LAPC-4 prostatic adenocarcinoma cells in a downscaled format of 96-well-plates for testing multiple doses of multiple target compounds. Metabolism of dehydroepiandrosterone (DHEA) with or without TGFβ1-induced stimulation (D+T) of the reactive stroma phenotype was assessed by increased testosterone in the media and PSA production of the epithelial prostate cancer cells. Using the non-metabolizable androgen R1881, effects from direct androgen action were distinguished from stromal androgen production from DHEA. Stromal cell androgenic bioactivity was confirmed using conditioned media from D+T-treated stromal cell monocultures in an androgen-inducible AR screening assay. We further showed that both agonists to estrogen receptor (ER), DPN (ERβ) and PPT (ERα), as well as estrogenic natural compounds including soy isoflavones attenuated D+T-induced PSA production. Studies with the pure ER agonists showed that activating either ERα or ERβ could inhibit both D+T-mediated and R1881-mediated PSA production with the D+T effect being more pronounced. In conclusion, natural compounds with estrogenic activity and pure ER agonists are very potent inhibitors of stromal conversion of DHEA to androgenic metabolites. More studies are needed to characterize the mechanisms involved in estrogenic modulation of the endocrine-immune-paracrine balance of the prostate microenvironment.     

Hormonal properties of norethisterone, 7alpha-methyl-norethisterone and their derivatives

Norethisterone (NET) is a progestagenic compound with very weak androgenicity and estrogenicity. These low androgenic and estrogenic activities may be attributed to NET itself or induced by metabolites of NET. In order to improve the bioactivity of NET, the effects of a 7alpha-methyl substitution were studied. Thus this study has two objectives: first the comparison between biological activities of NET and 7alpha-methyl-NET (MeNET), and second the biological activity of tentative metabolites of NET and those of MeNET. The metabolites consist of a 3-keto-, 3alpha- or 3beta-hydroxy-group located next to a carbon 4 to 5 double bond (Delta(4)) or a 5alpha-hydrogen atom. The 7alpha-methyl substitution was of special interest as it prevents 5alpha-reduction. The biological activities of NET, MeNET and their potential metabolites were assessed by in vitro binding, transactivation and proliferation assays on progesterone (PR), androgen (AR), estrogen (ER) and glucocorticoid (GR) receptors and by in vivo progestagenic McPhail, androgenic Hershberger, estrogenic Allen-Doisy tests and combined estrogenic and progestagenic ovulation inhibition tests. NET is a compound with five- to eight-fold weaker PR binding and transactivation activities than the reference compound Org 2058 (100%) and two-fold stronger than progesterone. Binding and transactivation activities of NET for AR (DHT=100%) are 3.2 and 1.1%, respectively, for ER none (E2=100%) and for GR below 1% (DEX=100%). MeNET is 1.5- to two-fold less progestagenic and ten- to 20-fold more androgenic than NET, while it does not show activity for ER and GR. The relative binding affinity of 5alpha-NET was seven-fold lower for PR and 1.5-fold higher for AR than for NET, while in transactivation assays 5alpha-NET was only active at levels below 1% for all tested receptors. 3beta-Hydroxy-(5alpha-reduced)-metabolites showed clear ER binding and transactivation activities, while 3alpha-hydroxy-(5alpha-reduced)-metabolites did hardly possess these characteristics. These hydroxy metabolites did not bind or activate other receptors. Substitution of 7alpha-methyl to NET metabolites led to similar characteristics, but with higher activities for AR and ER and weaker activity for PR. The outcome of in vivo tests showed a remarkable effect for MeNET. Progestagenic activity in rabbits appeared for NET equipotent to or eight-fold higher than for MeNET, after subcutaneous or oral treatment, respectively. On the other hand, MeNET showed in rats a ten-fold higher androgenicity and eight-fold higher estrogenicity than NET. Ovulation inhibition was induced at very low oral or subcutaneous dose levels, being 120- or ten-fold lower than for NET, respectively. The estrogenicity can also be induced by 3alpha- or 3beta-hydroxy metabolites of MeNET, which are 15 or even more than 40-fold stronger than those of NET, respectively. In conclusion, after the introduction of a 7alpha-methyl substituent to NET an increased estrogenicity and androgenicity and a reduced progestagenic activity was found. The in vivo estrogenicity is mainly due to 3beta-hydroxy-MeNET and to a lesser extent to 3alpha-hydroxy-MeNET, while the androgenicity and progestagenicity are most likely caused by MeNET itself. Since the 7alpha-methyl substituent inhibits 5alpha-reductase, 5alpha-reduced MeNET metabolites can be excluded from biological activities. As MeNET is a very effective ovulation inhibitor, due to its mixed progestagenic and estrogenic profile, a further reduction of androgenicity of MeNET may yield new contraceptives with an attractive profile for contraception.     

Receptor profiling and endocrine interactions of tibolone

The receptor profiles and in vivo activity of tibolone, and its primary metabolites, Delta(4)-isomer, and 3alpha- and 3beta-hydroxytibolone, were studied and compared to those of structurally related compounds. The Delta(4)-isomer was the strongest binder and activator of the progesterone receptor (PR); tibolone was 10 times weaker in binding and half as potent in transactivation of PR; 3alpha- and 3beta-hydroxytibolone did not bind or activate PR. In rabbits oral tibolone produced a minor progestagenic effect in the endometrium, whereas co-administration of tibolone and the anti-estrogen ICI 164,384 unmasked tibolone's progestagenic effect. 3-Hydroxytibolones were the strongest binders and activators of the estrogen receptors (ERs), with greater affinity for ERalpha than for ERbeta. Tibolone showed weaker binding and activation of both ERs and the Delta(4)-isomer has a binding and activation activity of less than 0.1% of E2 for ERalpha or ERbeta. Tamoxifen and 4-hydroxytamoxifen showed partial ERalpha agonistic effects with a maximal response of 12% and raloxifene of 3-5%. Oral administration of 1mg tibolone to ovariectomized rats induced an estrogenic effect on vaginal epithelium. The Delta(4)-isomer was a stronger binder and activator of the androgen receptor (AR) than tibolone; both 3-hydroxytibolones did not bind or activate AR. Introducing a 7alpha-methyl group decreased progestagenic and increased androgenic activity. We conclude that the progestagenic and androgenic activities of tibolone are mediated by the Delta(4)-isomer, and the estrogenic activity, by the 3-hydroxytibolones. The estrogenic activity of the 3-hydroxytibolones masked the progestagenic activity of tibolone in rabbit endometrium. Full estrogenic response was observed in rat vaginal tissue after oral administration of tibolone.     

Structure and estrogenic activity of new lignans from Iryanthera lancifolia

Five new dibenzylbutane type lignans (1-5) were isolated from the stem bark of Iryanthera lancifolia. Their structures were determined by extensive 1D and 2D NMR spectroscopic studies and chemical evidence. Seventeen of the isolated compounds were tested for their estrogenic activities in the estrogen responsive human breast cancer cell line MCF-7 BUS using the E-Screen proliferation assay. Cell proliferation was evaluated by the SRB assay to calculate the estrogenic parameters. The majority of the compounds induced a mitogenic response. This effect, given as Relative Proliferative Effect (RPE) to reference estrogen 17beta-estradiol (E(2)), ranged between 14% and 84%.     

Monitoring Endocrine Disrupter Compounds in the Tunisian Hamdoun River using In Vitro Bioassays

Anthropogenic chemicals occurring in the environment, namely endocrine-disrupting chemicals (EDCs), have generated growing concern over their potential adverse effects on human wildlife health and ecosystem processes. This interest resulted particularly from their abilities to mimic the effect of endogenous hormones. In this study, we used stable transfected reporter cell lines to investigate the endocrine-disrupting profile of water as well as sediment samples. Samples are collected from up- and downstream of an industrial wastewater discharge point at the Hamdoun River in the vicinity of an industrial zone located at the center of Tunisia. The analysis of estrogen, androgen, and xenobiotic (pregnane X and dioxin) ligands receptors expressed by chimeric cell lines indicated that while the water and sediment samples from upstream sites have lower levels of estrogenic activity, those from downstream exhibited stronger estrogenic, aryl hydrocarbon receptor (AhR), and Pregnane X Receptor (PXR) activities. Moreover, collected samples have shown hormonal activity in terms of all tested receptors except the androgenic ones. In vitro recombinant estrogen receptor competitive binding assays revealed that while the estrogenic activities of the downstream water sample compounds had a strong affinity for estrogen receptor α (ERα), those present in the sediment samples showed a weaker one. These findings were consolidated by subsequent chemical analysis (high-performance liquid chromatography with UV detectors). Our results indicate that the water and sediment discharges at the Hamdoun River represent a major sink for EDCs from natural and industrial effluents, particularly those of the textile industry, with pernicious potential to disrupt normal endocrine functions.     

Mycotoxins in root extracts of American and Asian ginseng bind estrogen receptors alpha and beta

The estrogenic activity of ginseng has been the subject of conflicting reports. Cell proliferation, induction of estrogen-responsive genes, and isolated cases of adverse reactions such as postmenopausal vaginal bleeding and gynecomastia have been reported after ginseng treatment. Other studies report antiproliferative effects with no induction of estrogen-responsive genes. We developed estrogen receptor (ER) alpha and ER alpha competitive binding assays using recombinant receptors and [(3)H]-17 alpha-estradiol to detect phytoestrogens in extracts of Asian ginseng root (Panax ginseng C. A. Meyer) and American ginseng root (Panax quinquefolius L.). Root extracts contained substances that bound both receptor isoforms. These substances had a two to three times greater affinity for ER alpha. Significantly higher binding was found in methanol extracts than in hot water extracts. Subsequent analysis of the extracts revealed significant ER binding attributable to zearalenone, the estrogenic mycotoxin produced by several Fusarium species. The ER showed no binding affinity for Rb1 and Rg1, the major ginsenosides found in P. quinquefolius and P. ginseng, respectively. Thus, ginseng extraction methods, plant species tested, and mycotoxin contaminants may help to explain the disparate literature reports. The prevalence and health significance of fungal contamination in herbal products used for medicinal purposes should be further investigated.     

Phytoestrogens genistein and daidzin enhance the acetylcholinesterase activity of the rat pheochromocytoma cell line PC12 by binding to the estrogen receptor

Some compounds derived from plants have been known to possess estrogenic properties and can thus alter the physiology of higher organisms. Genistein and daidzin are examples of these phytoestrogens, which have recently been the subject of extensive research. In this study, genistein and daidzin were found to enhance the acetylcholinesterase (AChE) activity of the rat neuronal cell line PC12 at concentrations as low as 0.08 μM by binding to the estrogen receptor (ER). Results have shown that this enhancement was effectively blocked by the known estrogen receptor antagonist tamoxifen, indicating the involvement of the ER in AChE induction. That genistein and daidzin are estrogenic were confirmed in a cell proliferation assay using the human breast cancer cell line MCF7. This proliferation was also blocked by tamoxifen, again indicating the involvement of the ER. On the other hand, incubating the PC12 cells in increasing concentrations of 17 β-estradiol (E2) did not lead to enhanced AChE activity, even in the presence of genistein or daidzin. This suggests that mere binding of an estrogenic compound to the ER does not necessarily lead to enhanced AChE activity. Moreover, the effect of the phytoestrogens on AChE activity cannot be expressed in the presence of E2 since they either could not compete with the natural ligand in binding to the ER or that E2 down-regulates its own receptor. This study clearly suggests that genistein and daidzin enhance AChE activityin PC12 cells by binding to the ER; however, the actual mechanism of enhancement is not known. This revised version was published online in August 2006 with corrections to the Cover Date.     

Gel-filtration chromatographic method for determining relative binding affinities: rat hepatic estrogen receptor as an example system

We report a gel-filtration-based chromatographic method for separation of specific, nonspecific, and free radioligand in a protein receptor-ligand binding assay for the example of the estrogen receptor ERalpha. This assay affords relative binding affinities (RBAs) without the need for a separate determination of nonspecific binding. The probit method is recommended as the most satisfactory method of evaluating the data. The assay responds to both estrogen agonists and antagonists, mixtures respond additively, and the slopes of the probit plots indicate that all ligands bind to the same site on the estrogen receptor. RBAs obtained with rat and rainbow trout ERalpha were in good agreement, and also with those from other reported assays, consistent with the interspecies conservation of key regions of the ligand binding domain among estrogen receptors.     

Binding of glucocorticoid receptors to mammary chromatin acceptor sites

We have recently characterized the interaction of mouse mammary estrogen receptors (ER) with mammary chromatin acceptor sites and demonstrated that ER from estrogen resistant lactating mammary glands do not bind to chromatin. In this study we have characterized the chromatin binding of the glucocorticoid receptor from mouse mammary glands isolated from nulliparous and lactating mice in order to better understand the relationship between receptor binding to chromatin and steroidogenic sensitivity of the tissue. Mammary chromatin was linked covalently to cellulose and deproteinized sequentially by 0-8 M Gdn-HCl. Binding to intact chromatin as well as to chromatin deproteinized by Gdn-HCl was determined using partially purified [3H]dexamethasone labelled glucocorticoid-receptor complexes (GR) obtained by fractionation on DEAE-cellulose columns. The binding of [3H]GR from mammary glands of nulliparous mice to chromatin fractions from the same tissue revealed maximal binding activity (acceptor sites) on chromatin previously extracted with 5-6 M Gdn-HCl. Binding of [3H]GR was of high affinity (Kd = 0.2 nM) and saturable. A simultaneous comparison of the chromatin binding patterns for [3H]ER and [3H]GR isolated from mammary glands of nulliparous mice revealed that the chromatin subfractions obtained with 4-6 M Gdn-HCl extraction contained acceptor sites for both [3H]ER and [3H]GR; however, while the [3H]ER bound to a 4.5 M and a 5.5 M site, the [3]GR bound a 5 M and a 6 M site. Competition experiments supported the steroid receptor specificity of the chromatin acceptor sites. Thus, the 4-6 M chromatin fractions contain distinct acceptor sites for the glucocorticoid receptor and for the estrogen receptor. In addition our studies reveal that the binding patterns of [3H]GR isolated from mammary glands of nulliparous and lactating mice to their homologous chromatin is essentially similar. Thus, in contrast to estrogen receptors, glucocorticoid receptors from lactating mammary glands are able to effectively bind to chromatin acceptor sites which supports our previous suggestion that the estrogenic insensitivity of lactating mouse mammary glands may at least be in part due to the impeded interaction of ER with chromatin acceptor sites.     

Isolation and characterization of the receptor for vitellogenin from follicles of the rainbow trout,Oncorhynchus mykiss

The isolation and characterization of the receptor for vitellogenin from follicle membranes of the rainbow trout, Oncorhynchus mykiss, is described. Follicle membrane proteins subjected to SDS-polyacrylamide gel electrophoresis and subsequently to either protein staining or ligand blotting with radiolabelled vitellogenin (¹²⁵iodine-vitellogenin) demonstrated that the vitellogenin receptor has an apparent molecular mass of 200 kD (probably comprising of two 100-kD subunits) under non-reducing conditions. The vitellogenin binding sites were identified as specific receptors: binding was saturable and the binding sites were both tissue specific to follicle membranes and exhibited ligand specificity. Scatchard analyses of specific binding data revealed a single class of binding sites with a high affinity for rainbow trout vitellogenin (K _d=8.2·10^-9 mol·1^-1). Both brown trout, Salmo trutta, vitellogenin and carp, Cyprinus carpio, vitellogenin were able to displace the radiolabelled rainbow trout vitellogenin from its receptor, although they were less effective than rainbow trout vitellogenin.Abbreviations B _max maximum number of binding sites available - BSA bovine serum albumin - bt-VTG brown trout vitellogenin - c-VTG earp vitellogenin - HEPES N-2-hydroxyethylpiperazine-N-2-ethanesulfonic acid - K _d dissociatian constant - NCM nitrocellulose membranes - PMSF phenylmethylsulphonylfluoride - rt-VTG rainbow trout vitellogenin - VTG vitellogenin