Bisphenol AF Is a Full Agonist for the Estrogen Receptor ER[alpha] but a Highly Specific Antagonist for ER[Beta]

Abstract

Bisphenol AF has been acknowledged to be useful for the production of CF3-containing polymers with improved chemical, thermal, and mechanical properties. Because of the lack of adequate toxicity data, bisphenol AF has been nominated for comprehensive toxicological characterization.

We aimed to determine the relative preference of bisphenol AF for the human nuclear estrogenic receptors ERalpha and ERbeta and the bisphenol A-specific estrogen-related receptor ERRgamma, and to clarify structural characteristics of receptors that influence bisphenol AF binding.

We examined receptor-binding activities of bisphenol AF relative to [3H]17beta-estradiol (for ERalpha and ERbeta) and [3H]bisphenol A (for ERRgamma). Functional luciferase reporter gene assays were performed to assess receptor activation in HeLa cells.

We found that bisphenol AF strongly and selectively binds to ERs over ERRgamma. Furthermore, bisphenol AF receptor-binding activity was three times stronger for ERbeta [IC50 (median inhibitory concentration) = 18.9 nM] than for ERalpha. When examined using a reporter gene assay, bisphenol AF was a full agonist for ERalpha. In contrast, it was almost completely inactive in stimulating the basal constitutive activity of ERbeta. Surprisingly, bisphenol AF acted as a distinct and strong antagonist against the activity of the endogenous ERbeta agonist 17beta-estradiol.

Our results suggest that bisphenol AF could function as an endocrine-disrupting chemical by acting as an agonist or antagonist to perturb physiological processes mediated through ERalpha and/or ERbeta.

Headnote

BACKGROUND: Bisphenol AF has been acknowledged to be useful for the production of CFr containing polymers with improved chemical, ihermal. and mechanical properties. Because of the lack of adequate toxicity data, bisphenol AF has been nominated for comprehensive toxicological characterization.

OBJECTIVES: We aimed to determine the relative preference of bisphenol AF for the human nuclear estrogenic receptors ERα and ERβ and the bisphenol A-specific estrogen-related receptor ERRγ, and to clarify structural characteristics of receptors that influence bisphenol AF binding.

METHODS: We examined receptor-binding activities of bisphenol AF relative to [^sup 3^H]17β-estradiol (for ERα and ERβ) and pHJbisphenol A (for ERRγ). Functional luciferase reporter gene assays were performed to assess receptor activation in HeEa cells.

RESULTS: We found that bisphenol AE strongly and selectively binds to ERs over ERRγ. Furthermore, bisphenol AF receptor-binding activity was three times stronger for ERβ [IC^sub 50^ (median inhibitory concentration) = 18.9 nM] than for ERα. When examined using a reporter gene assay, bisphenol AF was a full agonist for ERα. In contrast, it was almost completely inactive in stimulating die basal constitutive activity of ERβ. Surprisingly, bisphenol AF acted as a distinct and strong antagonist against the activity of the endogenous ERβ agonist 17β-estradiol.

CONCLUSION: Our results suggest that bisphenol AF could (unction as an endocrine-disrupting chemical bv acting as an agonist or antagonist to perturb physiological processes mediated through ERα and/or ERβ.

KEY WORDS: bisphenol A, bisphenol AF, endocrine disruptor, estrogen receptors, receptor antagonist, receptor binding. Environ Health Perspect 118:1267-1272 (2010). doi: 10.1289/ehp.0901819 [Online 28 April 2010]

Bisphenol AF (also referred to as hexafiuorobisphenol A) is a homolog of bisphenol A (BPA) (Figure 1). Bisphenol AF has a symmetrical chemical structure of HO-C^sub 6^H^sub 4^C(CF^sub 3^)^sub 2^-C^sub 6^H^sub 4^-OH and is designated as 1 . K 1 .3,3.3-hexafluoro-2,2-bis(4-hydroxyphenyl)propane by IUPAC (International Union of Pure and Applied Chemistry) nomenclature. Bisphenol AF-containing polymers such as polycarbonate copolymers, polyimides, polyamides, and polyesters arc used in high-temperature composites, electronic materials, and gas-permeable membranes. Bisphenol AF is also used in many other specialty polymer applications, including plastic optical fibers and waveguides. Although industrial production of bisphenol AF seems to be increasing considerably, no data are available on annual production or concentrations of bisphenol AF in environmental substrates.

In 2008, the U.S. National Institute of Environmental Health Sciences nominated bisphenol AF for comprehensive toxicological characterization based on the lack of adequate toxicity data [National Toxicology Program (NTP) 2008a]. In this nomination report, the NTP noted concern regarding potential exposure of the general population to bisphenol AF. Structural dissimilarities between bisphenol AF and BPA are determined by the presence of a rrifluoromethyl Cl:^) or methy (CH^sub 3^) group, respectively. Ihe potential toxicity of bisphenol AF is of concern in part because its CF^sub 3^ group is much more electronegative (and potentially reactive) than is the CH^sub 3^ group of BPA.

Various "'low-dose effects" of BPA have recently been reported in vivo for reproductive organ tissues in mice and rats. For example, in utero exposures to very low levels of BPA have been shown to increase the size and weight of the fetal mouse prostate (Gupta 2000; Nagel et al. 1997), and low-dose exposures have also been reported to decrease daily sperm production and fertility in male mice (Gupta 2000; vom Saal et ai. 1998). Many lines of evidence have recently indicated that low doses of BPA affect the central nervous system as well (vom Saal and Welshons 2005; Welshons et al. 2003, 2006). All of these lowdose effects of BPA have been attributed to effects on steroid hormone receptors such as estrogen receptor (FR) and androgen receptor (AR) (Welshons et al. 2003; Xu et al. 2005). In the report by the NTP (2008b) on the potential for BPA exposure to affect human reproduction or development, "some concern" was indicated as the level of concern tor potential effects on the brain, behavior, and the prostate gland.

BPA exhibits extremely weak binding activitv for ER and AR. Based on the idea that BPA may interact wirb nuclear receptors (NRs) other than ER and AR, we screened a series ot NRs and eventually discovered estrogenrelated receptor γ (ERRγ) as the BPA target receptor (Takayanagi et al. 2006). BPA binds to FRRγ very strongly |dissociarion constant (K^sub d^) = 5.5 nM] with high constitutive basal activity (Liu et al. 2007; Okada et al. 2008; Takayanagi et al. 2006). Strong binding of BPA to ERRγ was further demonstrated by direct X-ray erystallpgraphic analysis of this complex (Matsushima et al. 2007, 2008). Moreover, using real-time PCR (polymerase chain reaction), we recently demonstrated that human ERRγ mRNA is expressed abundantly in the placenta, prostate, and fetal brain (Takeda et al. 2009).

Our efforts to explore the target receptor of BPA suggested rhar it is essential to examine endocrine chemicals for interactions with all 48 human NRs. We previously reported that bisphenol AF binds to ERα more strongly than does BPA, and that the receptor selectivity of bisphenol AF is seven times higher for ERα than for FRRγ (Okada et al. 2008). There are two subtypes of estrogen receptors, ERα and ERß. with distinctly different physiological distributions and functions. Because effects of a number of chemicals have been reporred to differ berween ERα and ERß (Harris et al. 2003; Manas et al. 2004), it is iniportanr to examine the effects of bisphenol AF on both ERs. In the present study, we evaluated the binding activity and functional biological activity of bisphenol AF for ERß and found that bisphenol AF is a potent ligand thar functions as an antagonist on ERß.

Materials and Methods

Test compounds. We obtained 17β-estradiol (CAS no. 50-28-2; 98.9%) from Research Biochemicals International (Natick, MA, USA), and BPA (CAS no. 80-05-7; purity 99%) and bisphenol AF (CAS no. 1478-61-1; purity 99%) from Tokyo Kasei Kogyo Co. Ltd. (Tokyo, Japan). 4-Hydroxytamoxifen (4-OHT; CAS no. 68047-06-3; purity 98%) and 2,2-bis(p-hydroxyphenyl)- 1 , 1 , 1-trichloroethane (HPTE) were obtained from SigmaAldrich Inc. (St. Louis, MO, USA).

Preparation of glutathione S-transferase(GST)-fused NR ligand-binding domain (LBD) protein. cDNA clones of ERa and ERß were purchased from OriGene Technologies, Inc. (Rockville, MD, USA). GST-fused receptor LBDs expressed in Escherichia coli BL21CX (GST-ERα-LBD, GST-ERß-LBD, and GST-ERRγ-LBD) were purified on an affinity column of glutathione-Sepharose 4B (GE Healthcare BioSciences Co., Piscataway, NJ, USA) followed by gel filtration on a Sephadex G-10 column (15 ? 10 mm; GE Healthcare BioSciences).

Radioligand binding assays for saturation binding. We conducted the saturation binding assays for ERa and ERβ essentially as reported by Nakai et al. (1999) using tritium-labeled ligand [3H]17β-estradiol (5.96 TBq/mmoí; GE Healthcare UK Ltd., Buckinghamshire, UK). Receptor protein GST-ERa-LBD or GST-ERβ-LBD (0.3 nM) was incubated with increasing concentrations of [^sup 3^H] 1 7β-estradiol (0.1-30 nM) in a final volume of 100 pL binding buffer (10 mM Tris, 1 mM EDTA, 1 mM EGTA, 1 mM sodium vanadate(V), 0.5 mM phenylmethylsulfonyl fluoride, 0.2 mM leupeptin, 10% glycerol; pH 7.4). Nonspecific binding was determined in a parallel set of incubations that included 10 pM nonradiolabeled 17β-cstradiol. After incubation for 2 hr at 200C, free radioligand was removed by incubation with 0.4% dextrancoated charcoal (Sigma-Aldrich Inc.) in phosphate-buffered saline (PBS; pH 7.4) for 10 min on ice and then centrifugea1 for 10 min at 1 5,000 rpm.

We performed the saturation binding assay for ERRγ as reported previously (Okada et al. (2008) using [^sup 3^H]BPA (5.05 TBq/mmol; Moravek Biochemicals, Brea, CA, USA). Specific binding of tritium-labeled ligand was calculated by subtracting the nonspecific binding from the total binding. Receptor proteins that were expressed and purified were evaluated in a saturation binding assay to estimate K^sub d^ and receptor density (B^sub max^), and only good-quality preparations with appropriate K^sub d^ and B^sub max^ were used for competitive receptor-binding assays.

Radioligand binding assays for competitive binding. Bisphenol AF, BPA, 17β-estradiol, and 4-OHT were dissolved in 0.3% DMSO in 1% bovine serum albumin (BSA; a blocker of nonspecific adsorption to the reaction vessels). HPTE was tested as a reference compound that acted as an ERa agonist and an ERβ antagonist. These chemicals were examined for their ability to inhibit the binding of [3H]17β-estradiol (5 nM in final) to GSTERa-LBD (26 ng) and GST-ERβ-LBD (26 ng). The reaction mixtures were incubated overnight at 4°C, and free radioligand was removed with 1% dextran-coated charcoal by filtration. Radioactivity was determined on a liquid scintillation counter (TopCount NXT; PerkinElmer Life Sciences Japan, Tokyo, Japan). We calculated the half-maximal inhibitor)' concentrations (IQ0) for 17β-estradiol from dose-response curves obtained using the nonlinear analysis program ALLFIT (DeLean et al. 1978). Each assay was performed in duplicate and repeated at least five times. For reconfirmation, we also performed the binding assay for ERRy using [3H]BPA (5 nM final concentration) and GST-ERRy-LBD (26 ng).

Luciferase reporter gene assay. HeLa cells were maintained in Eagle's minimum essential medium (MEM; Nissui, Tokyo, Japan) in the presence of 10% (vollvoi) fetal bovine serum at 37°C. For luciferase assays, HeLa cells were seeded at 5 × 10^sup 5^ cells per 6-cm dish for 24 hr and then transfected with 4 pg reporter gene (pGL3/3xERE) and 3 pg of ERa or ERβ expression plasm id (pcDNA3/ERs) by Lipofectamine Plus reagent (Invittogen Japan, Tokyo, Japan) according to the manufacturer's protocol. Approximately 24 hr after transfection, ceils were harvested and plated into 96-well plates at 5 × 10^sup 4^ cells/well. The cells were then treated with varying doses of chemicals diluted with 1% BSA/PBS (voF/vol). After 24 hr, luciferase activity was measured with the appropriate reagent using a Luciferase Assay System (Promcga, Madison, WI, USA) according to the manufacturer's instructions. Light emissions were measured using a Wallace 1420 ARVOsx mulrilabel counter (PerkinElmer). Cells treated with 1% BSA/PBS were used as a vehicle control. Each assay was performed in triplicate and repeated at least three times. The assay for ERRy was carried out as previously reponed (Okada et al. 2008).

To measure the antagonistic activity of bisphenol AF for ERβ, we examined four concentrations (0.01, 0.1, 1.0. and 10 pM) of bisphenol AF for a serial concentration of 17β-estradiol (10^sup -12^ to 10^sup -5^ M in the final solution). Also, a serial concentration of bisphenol AF (10^sup -12^ to 10^sup -3^ M in the final solution) was assayed in the presence of 10 or 100 nM concentrations of I 7β-estradiol, which normally elicit full activation ot ERβ.

Results

Strong binding activity of bisphenol AF to ERβ receptor. We selected receptor protein preparations suitable for the competitive receptor-binding assay based on Scatchard plot analyses of saturation-binding assays. Receptor populations with the appropriate dissociation constant K^sub d^) and receptor density (B^sub max^) were used for each radioligand receptor-binding assay. Because ali of the NRs are secreted protein preparations, observed B^sub max^ values were comparable with those calculated from their molecular weight.

BPA was a very weak ligand for FRa (IC^sub 50^ = L030 nM) based on its ability to inhibit [^sup 3^H]17β-estradiol binding (Figure 2A, Table 1), as we previously reported (Okada et al. 2008). In the present study, we confirmed that BPA is also a very weak ligand for ERβ (IC^sub 50^ = 900 nM; Figure 2B, fable 1), indicating comparable interactions of BPA with ERa and ERβ despite the subtle structural differences between these ERs. In contrast, bisphenol AF was 20 times more potent than BPA as a ligand for ERa (IC^sub 50^ = 53.4 nM; Figure 2A, Table 1) and was approximately 48 times more potent for ERβ (IC^sub 50^ = 18.9 nM; Figure 2B, Table 1). This high binding activity for ERβ suggests that the binding pocket of ERβ possesses specific structural elements that interacr much more favorably with the CF^ groups of bisphenol AF than with the CH3 groups of BPA. We also assayed HPTE, an analog of BPA and bisphenol AF with the CCI3 group. HPTE was almost cquipotent to bisphenol AF in the assays for both ERα and ERβ (Table 1), but approximately 10 times more potent than bisphenol AF for ERRγ.

Receptor-binding selectivity of bisphenol AF and BPA. We used the IC^sub 50^ values shown in Table 1 (from the competitive receprorbinding assay for nuclear ERα, ERβ, and ERRγ) ro estimate receptor selectivity ratios lor BPA and bisphenol AF (Table 2). The results indicate that BPA is exclusively selective for ERRγ, being 90-100 times more active for ERRγ than for ERα or ERβ. In contrast, bisphenol AF receptor binding is much more selective for ERα and ERβ than for ERRγ (6.70 times more selective for ERα than for ERRγ and 18.94 times more selective for ERβ than for ERRγ; Table 2). Bisphenol AF binding is also about three times more potent for ERβ than for ERα.

Differential effects of bisphenol AF in the reporter gene assay. We next examined reporter gene activity after bisphenol AF exposure in HeLa cells transiently cotransfecred with an ERα or ERβ expression plasmid and an estrogen-response element (ERE)-lucif erase reporter plasmid. Bisphenol AF fully activated ERα (increasing activity to - 7 times the baseline level) in a dose-dependent manner at concentrations of 10"'° to 1O-"1 M (Figure 3A). The halt-maximal effective concentration (ECs0) of bisphenol AF was 58.7 nM.

When we compared potencies for ERα activation vetsus ERα binding to determine receptor activation potency [expressed as EQo (juVD/ICso (nM)], we found a clear discrepane)' between 17β-estradiol and bisphenol AF. As shown in Table 3, we estimated the receptor activation potency for 1 7β-estradiol to be 0.085 (0.075 nM/0.88 nM based on values from Figure 3A and Table 1, respectively). In contrast, the receptor activation potency of bisphenol AF [1.099 (58.7 nM/ 53.4 nM)] was approximately i3 times greater than char of 17β-estradiol (Table 3). This means that the concentration of 17β-estradiol required to stimulate a 50% response is about 13 times lower than the concentration required to occupy 50% of receptors, whereas the concentration of bisphenol AF required to stimulate a 50% response is about the same as that required to occupy 50% of receptors. This suggests that the receptor conformation induced by bisphenol AF is not as conducive to receptor activation as that induced by 17β-estradiol when measured in HeLa cells.

BPA was an extremely weak activator of both ERα (EC^sub 50^ = 317 nM) and ERβ (EC^sub 50^ = 693 nM) based on the luciferase reporter gene assay. The receptor activation potencies of'BPA for ERα (0.308) and ERβ (0.770) were 3.6 and 18.8 times greater than the receptor activation potencies of l7β-estradiol for ERα and ERβ, respectively (Table 3). These suggests that, compared with 17β-estradiol, the concentration of BPA required to stimulate a 50% response is much higher than the concentration required to occupy 50% of receptors. In addition, as shown in Figure 3B, BPA exhibited a reduced ability to bring about full activation of ERβ (3.5 times greater activity relative to baseline in response to BPA vs. an increase to 6 times the baseline level in response to 17β-estradiol). This difference in efficacy indicates that BPA does not have the same ability as 17β-estradiol to induce activation conformation when measured in HcLa cells on this promoter.

Antagonist activity of bisphenol AF on ERβ. For ERβ, bisphenol AF was almost completely inactive, with very little increase in activity even at 10 µM, the highest concentration tested (Figure 3B). Based on the strong receptor-binding activity of bisphenol AF for ERβ (IC^sub 50^ = 18.9 nM; Table D, we expected that bisphenol AF would also have a high receptor activation potency for ERβ. This unexpected inactivity in the reporter gene assay suggests that bisphenol AF binding disrupts the ERβ-LBD activation conformation, in which the α-helix 12 (H 12) of the receptor is normally positioned to recruit the coactivator protein conformation (Brzozowski et al. 1997: Ruff et al. 2000).

We therefore evaluated the antagonist activity of bisphenol AF against ]7β-estradiol. When we examined 17β-estradiol, an endogenous agonist ligand of ERβ, in the presence of 0.01, 0.1, 1.0, and 10 pM bisphenol AF, its activity (EC^sub 50^ = 0.075 nM) was graduallyweakened. As shown in Figure 4A, the dosedependent curves of 1 7β-estradiol shifted to the right with increasing concentrations of bisphenol AF, indicating that bisphenol AF effectively inhibits the interaction between 17β-estradiol and ERβ. When che results of Figure 4A were analyzed using a Schild plot, p/42, a measure of affinity of the antagonist for receptor, was calculated to be 7.87 from the dissociation equilibrium constant (K^sub B^= 1.35 × 10^sup -8^M).

The antagonist activity of bisphenol AF for 17β-estradiol/ERβ was further evidenced by assays in which we added serial concentrations of bisphenol AF (10^sup -12^ to 10^sup -5^ M) to a solution of 17β-estradiol maintained at a constant concentration. When 1 × IO^sup -8^ M 17β-estradiol was treated with bisphenol AF, the activity of 1 7β-estradiol was reduced in a dose-dependent manner in response to bisphenol AF concentrations ranging from 1O-10 ro 10^sup -5^ M (Figure 4B). We obtained a similar result (or 1 × 10^sup -7^ 17β-estradiol. These results demonstrate that bisphenol AF can antagonize the activity of 17β-estradiol on the ERβ receptor.

Discussion

Structural characteristics of bisphenols and ERs/ERR'f receptors. The differences in receptor selectivity between bisphenol AF and BPA are due to the CH^sub 3^ [Lef-right arrow] CF^sub 3^ substitution on the bisphenol backbone structure. Bisphenol AF is a hexafluoro derivative of BPA with the CH^sub 3^ [arrow right] CF^sub 3^ substitution on the backbone structure of 2,2-disubstituted propane CH^sub 3^-C-CH^sub 3^. BPA binds strongly to ERRγ, but bisphenol AF binds to ERRγ only weakly: we therefore judged that the binding pocket of ERRγ-LBD possesses structural elements unfavorable for interaction with the rrifluoro groups. The molecular size of CF^sub 3^ is almost the same as that of CH^sub 3^, and thus there would be no structural repulsion or steric hindrance between these groups. However, because the CF^sub 3^ group is very electron rich, the structural elements standing face to face with CF^sub 3^ must also be electron rich, resulting in their electrostatic repulsion.

In our previous study (Matsushima et al. 2007, 2008), we found that the ERRy binding sites for BPA CH^sub 3^ groups were Phe435 and Met306. Because the aromatic phenyl and S-CH^sub 3^ groups of Phe435 and Met306 are electron rich, conditions would be unfavorable for binding of bisphenol AF's electron-rich CF^sub 3^ groups. Corresponding receptor residues in ERα are Leu525 and Leu384, respectively. Apparently, there would be no electrostatic repulsion between the bisphenol AF's CF^sub 3^ groups and the Leu residues. Such a release in strucniral stress must be very favorable for receptor activity and the selectivity of bisphenol AF for ERα.

In the present study, we found bisphenol AF to be a strong ligand for both ERα and ERβ receptors, although it shows a 3 times greater preference for ERβ over ERα. A much more important finding is that bisphenol AF functions in a different way for ERα and ERβ. Bisphenol AF is a full agonist for ERα but an antagonist for ERβ. The LBDs of ERα and ERβ share a high sequence identity (59%) and similar three-dimensional structures. We observed no obvious differences between ERα and ERβ in the ERE transcriptional assays in the presence of 1 7β-estradiol.

Among the amino acid residues lining the binding pockets of ERα and ERβ, two residues differ significantly: Leu384 in a-helix 5 (H5) of ERa is replaced by Met336 in ERβ, and Met421 in loop 6-7 of ERa is replaced by Ile373 in ERβ. These two residues are most probably responsible for the discriminative affinity and reverse functional activity of bisphenol AF for ERα and ERβ. Furthermore, because bisphenol AF is an ERβ antagonist, the binding of bisphenol AF to the ERβ lígand-binding pocket must damage the ERβ-LBD activation conformation, in which the a-helix 12 (H 12) in LBD is positioned to recruit the coactivator proteins conformation (Brzozowski et al. 1997; Ruff et al. 2000). Bisphenol AF binding to LBDs of ERα and ERβ are being analyzed in light ot the crystal structures in studies in progress in our laboratory.

Bisphenol AF as a candidate of potential endocrine disruptor. Bisphenol AF is a potent estrogen agonist for ERα and a potent estrogen antagonist for ERβ. ERα and ERβ are widely distribured throughout the body, displaying distinct but overlapping expression patterns in a variety of tissues. ERα is expressed primarily in the uterus, liver, kidneys, and heart (Couse and Korach 1999), whereas ERβ is expressed primarily in the ovaries (Couse and Korach 1999), prostate (Couse and Korach 1999), lungs (Kuiper et al. 1997), and gastrointestinal tract and bladder (Nilsson et al. 2001). Coexprc.vMon of both receptors occurs in the mammary glands (Pettersson and Gustafsson 2001), epididymis (Pau et al. 1998), thyroid (Pau et al. 1998), adrenals (Pau et al. 1998), bone (Arts et al. 1997; Brandenberger et al. 1997), and certain regions of the brain (Couse and Korach 1999). [For additional information, see Nuclear Receptor Signaling Atlas (2010).] 17β-Estradiol plays a critical role in many physiological processes in both females and males. These include normal growth, development, and cell-type-specific gene regulation in tissues of the reproductive tract, central nervous system, and skeleton (Couse and Korach 1999; Nilsson et al. 2001; Pettersson and Gustafsson 2001). Bisphenol AF is a potent binder of ERα and ERβ and thus would perturb rhesc physiological processes, perhaps providing significant adverse influences for the central and peripheral systems.

Effects of the bisphenol trihalogenated methyl group on receptor actions. Bisphenol AE is an agonist tor ERα and an antagonist for ERβ. Similar resulrs have been reported for HPTE, a bisphenolic metabolite of methoxychlor [1,1,1 -trichloro-2,2-bis(4-methoxyphenyOethane]. HPTE behaved as an ERα agonist and an ERβ antagonist with estrogenresponsive promoters in HeLa cells (Gaido et al. 1999). We confirmed these results in our assay systems as well. HPTE was a strong binder of ERα with IC^sub 50^ = 59,1 nM and of ERβ with IC^sub 50^ = 1S-1 nM (Table 1). As reported previously by Gaido et al. (1999) and Nettles et al. (2004), HPTE acts as a full agonist for ERα but a strong antagonist for ERβ. However, bisphenol AF and HPTE differ in their receptor preference for ERRy. HPTE was approximately 10 times more potent than bisphenol AF for ERRγ binding, although both chemicals were most strongly bound to ERβ (Tables I, 2). As an antagonist for ERβ, bisphenol AF IpA^sub 2^ = 7.87) was somewhat stronger than HPTE, the pA^sub 2^ of which was reported to be 7.52 (Gaido er al. 1999). However, both bisphenol AF and HPTE are significantly potent as ERβ antagonists.

Chemical structures of bisphenol AF and HPFE differ, with one of two CF^sub 3^ groups of bisphenol AF replaced by CCl^sub 3^ in H PIE, and the other by H (Figure 1). However, these compounds are similar in that both have trihalogenated methyl groups that may produce different activities for ERa and ERβ via their interactions with the ligand-binding pockets of each ER. namely. Leu384 in H5 of ERa <-> Met336 in ERβ, and Met421 in loop 6-7 of ERa ^ Ile373 in ERβ.

Metboxychlor is a chlorinated hydrocarbon pesticide structurally similar to DD 1 (dichlorodiphenyltricbloroethane) and thus is sometimes referred to as dimethoxy or methoxy DDT, It had been used to some degree as a replacement for DDT to protect crops, ornamentals, livestock, and pets against various insects, because it was believed to be metabolized more quickly than DDT, thus reducing or preventing bioaccumulation (Kapoor et al. 1970). Methoxychlor is uterotropic in the ovariectomized rat and can cause adverse developmental and reproductive effects in mice and rats (Aim et al. 1996; Cunimings 1997; Hall et al. 1997). However, HPTE is approximately 100 times more active at ERs rhan is merhoxychlor. To date, the use of methoxychlor has been banned in many countries, including the United States, Japan, and the European Union. All these issues clearly raise concerns that not only HPTE but also bisphenol AF may be a potential endocrine disruptor affecting either ERa or ERβ, or both.

Conclusions

BPA binds strongly to ERRγ but very weakly to ERa and ERβ. In contrast, bisphenol AF binds very weakly to ERRγ but strongly to ERa and ERβ. These differences in receptor selectivity reflect subtle but distinct structural differences resulting from rhe CH^sub 3^ [Lef-right arrow] CF^sub 3^ substitution on the bisphenol backbone structure. The trifluoromethyl group is much more electronegative than the methyl group. These results suggest that apparently minor structural differences among chemicals and NRs may have pronounced effects on binding affinity and selectivity. Thus, the present study emphasizes the crucial importance of accurate evaluation of receptor responses to understanding interactions between endocrinedisrupting compounds and diverse human NRs. Taken together, these results clearly indicate the importance of examining the degree and ways in which bisphenol AF may influence the physiological roles of ERa and ERβ. Given that bisphenol AF and BPA function as endocrine disrupters, these chemicals would work differently via different NRs.

References

REFERENCES

Aim H, Tiemann U, Tomer H. 1996. Influence of organochlorine pesticides on development ol mouse embryos in vitro Reprod Toxicol 10(4):321-326

Arts J, Kuiper GGJM. Janssen JMMF, Gustafsson J-À, Lòwik CWGM, Pols NAP, et al. 1997. Differential expression of estrogen receptors a and 15 rnRNA during differentiation of human osteoblast SV-HFO cells. Endocrinology 138011:5067-5070.

Brandenberger AW, Tee MK, Lee JY, Chao V, Jaffe RB. 1997. Tissue distribution of estrogen receptors alpha lER-o) and beta (ER-JiI mRNA in the midgestational human fetus. J Clin Endocrinol Metab 821101:3509-3512.

Brzozowski AM, Pike AC, Dauter Z, Hubbard RE. Bonn T, Engstrom 0, et al. 1997. Molecular basis of agonism and antagonism in the oestrogen receptor Nature 389:753-758.

Couse JF, Korach KS 1999. Estrogen receptor null mice: what have we learned and where will they lead us? Endocr Rev 20(31:358-417

Cunimings AM 1997 Methoxychlor as a model for environmental estrogens. Crit Rev Toxicol 27(41:367 379

DeLean A, Munson PJ, Rodbard D. 1978. Simultaneous analysis of families of sigmoidei curves: application to bioassay, radioligand assay, and physiological dose-response curves. Am J Physiol 235(2):E97-E102.

Gaido KW, Leonard LS. Maness SC. Hall JM, McDonnell DP. Saville B, et al. 1999. Differential interaction of the methoxychlor metabolite 2,2-bis-(p-hydroxyphenyll-l,1,l-trichloroethane with estrogen receptors a and p\ Endocrinology 140(121:5746-5753.

Gupta C. 2000. Reproductive malformation of the male offspring following maternal exposure to estrogenic chemicals. Proc Soc Exp Biol Med 224121:61-68.

Hall DL, Payne LA, Putnam JM, Huet-Hudson YM. 1997 Effect of methoxychlor on implantation and embryo development in the mouse. Reprod Toxicol 1 1(151:703- 708

Harris HA, Albert LM. Leathurby Y, Malamas MS, Mewshaw RE, Miller CP, et al. 2003. Evaluation of an estrogen receptor-|5 agonist in animal models of human disease. Endocrinology 144(101:4241-4249

Kapoor IP, Metcalf RL, Nystrom RF, Sangha GK. 1970 Comparative metabolism of methoxychlor, methiochlor, and DDT in mouse, insects, and in a model ecosystem. J Agrie Food Chem 18(61:1145-1152.

Kuiper GGMJ, Carlsson B, Granthen K, Enmark E, Haggblad J, Nilsson S. et al 1997. Comparison of the ligand binding specificity and transcript tissue distribution ol estrogen receptors ? and )}. Endocrinology 138(31:863-870.

Liu X, Matsushima A, Okada H, Tokunaga T, Isozaki K, Shimohigashi Y. 2007. Receptor binding characteristic of the endocrine disruptor bisphenol A for the human nuclear estrogen-related receptor ?. Chief and corroborative hydrogen bonds of the bisphenol A phenol-hydroxyl group with Arg316 and Glu275 residues. FEBS J 274(24|:6340-6351.

Manas ES, Unwalla RJ, Xu ZB, Malamas MS, Miller CP, Harris HA, et al. 2004. Structure-based design of estrogen receptor-beta selective ligands. J Am Chem Soc 126(40). i5 i06- !51 19.

Matsushima A, Kakuta Y, Teramoto T, Koshiba T, Liu X, Okada H, et al 2007. Structural evidence for endocrine disruptor bisphenol A binding to human nuclear receptor ERRy. J Biochem 142(41:517-524.

Matsushima A, Teramoto T, Okada H, Liu X, Tokunaga T, Kakuta Y, et al 2008. ERRy tethers strongly bisphenol A and 4-cx-cumylphenol in an induced-fit manner Biochem Biophys Res Commun 373(31:408-413.

Nagel SC. vom Saal FS, Thayer KA, Ohar MG, Boechler M, Welshons WV. 1997 Relative binding affinity-serum modified access IRBA-SMA) assay predicts the relative in vivo bioactivity of the xenoestrogens bisphenol A and octylphenol. Environ Health Perspect 105:70-76.

Nakai M, Tabira Y, Asai D, Yakabe Y, Shinmyozu T, Noguchi M, et al. 1999. Binding characteristics of dialkyl phthalates for the estrogen receptor. Biochem Biophys Res Commun 254(21:311-314.

Nettles KW, Sun J, Radek JT, Sheng S, Rodriguez AL, Katzenellenbogen JA, et al. 2004 Allosteric control of ligand selectivity between estrogen receptors a and |5: implications for other nuclear receptors. MoI Cell 13(31:317-327.

Nilsson S, Mäkelä S. Treuter E, Tujague M. Thomsen J, Andersson G, ei al. 200) Mechanisms of estrogen action. Physiol Rev 81(41:1535-1565.

NTP (National Toxicology Program). 2008a. Chemical Information Profile for Bisphenol AF [CAS No. 1478-61-1). Supporting Nomination for Toxicological Evaluation by the National Toxicology Program. Available: http://ntp. niehs.nih.gov/ntp/htdocs/Chem_Background/ExSumPdf/ BisphenolAF_093008_508.pdf !accessed 30 March 2010].

NTP (National Toxicology Program). 2008b. NTP-CERHR Monograph on the Potential Human Reproductive Developmental Effects of Bisphenol A. NlH Publication No. 08-5994. Available: http://cerhr.niehs.nih.gov/evals/ bisphenol/bisphenol.pdf [accessed 2 August 2010]

Nuclear Receptor Signaling Atlas. 2010. Datasets: Tissue-Specific Expression Patterns of Nuclear Receptors. Available: www. nursa.org/10.1621/datasets.02001 [accessed 2 August 2010].

Okada H, Tokunaga T, Liu X, Takayanagi S, Matsushima A, Shimohigashi Y. 2008. Direct evidence revealing structural elements essential for the high binding ability of bisphenol A to human estrogen-related receptor-γ. Environ Health Perspect 116:32-38.

Pau CY, Pau KY, Spies HG. 1998. Putative estrogen receptoi beta and alpha mRNA expression in male and female rhesus macaques. MoI Cell Endocr 146(1-21:59-68.

Pettersson K, Gustafsson J-Â. 2001 Role of estrogen receptor beta in estrogen action. Annu Rev Physiol 63:165-192.

Ruff M, Gangloff M, Wurtz JM, Moras D. 2000. Estrogen receptor transcription and transactivation: structure-function relationship in DNA- and ligand-binding domains of estrogen receptors. Breast Cancer fies 2|2):353~359.

Takayanagi S, Tokunaga T, Liu X, Okada H, Matsushima A, Shimohigashi Y. 2006. Endocrine disruptor bisphenol A strongly binds to human estrogen-related receptor y (ERRyI with high constitutive activity. Toxicol Lett 167(21:95-105.

Takeda Y, Liu X, Sumiyoshi M, Matsushima A, Shimohigashi M, Shimohigashi Y. 2009. Placenta expressing the greatest quantity of bisphenol A receptor ERRy among the human reproductive tissues: predominant expression of type-1 ERRy isoform. J Biochem 146(11:113-122.

vom Saal FS, Cooke PS, Buchanan DL, Palanza P, Thayer KA, Nagel SC, et al. 1998. A physiologically based approach to the study of bisphenol A and other estrogenic chemicals on the size of reproductive organs, daily sperm production, and behavior. Toxicol lnd Health 14:239-260.

vom Saal FS, Welshons WV. 2005. Large effects from small exposures. II. The importance of positive controls in lowdose research on bisphenol A. Environ Res 100:50-76.

Welshons WV, Nagel SC, vom Saal FS. 2006. Large effects from small exposures. III. Endocrine mechanisms mediating effects of bisphenol A at levels of human exposure. Endocrinology 147(6 suppl):56-69.

Welshons WV, Thayer KA, Judy BM, Taylor JA, vom Saal FS. 2003. Large effects from small exposures. I. Mechanisms for endocrine-disrupting chemicals with estrogenic activity. Environ Health Perspect 111:994-1006.

Xu L-C, Sun H, Chen J-F. Bian Q, Qian J. Song L, et al. 2005. Evaluation of androgen receptor transcriptional activities of bisphenol A, octylphenol and nonylphenol in vitro. Toxicology 216(2-31:197-203.