Robert M. Blair,*,1 Hong Fang,* William S. Branham,* Bruce S. Hass,* Stacey L. Dial,* Carrie L. Moland,* Weida Tong,† Leming Shi,† Roger Perkins,† and Daniel M. Sheehan*


*Division of Genetic and Reproductive Toxicology, National Center for Toxicological Research, Jefferson, Arkansas 72079; and †R.O.W. Sciences, Jefferson, Arkansas 72079.


Received July 19, 1999; accepted November 4, 1999.


1 To whom correspondence should be addressed at Division of Genetic and Reproductive Toxicology, National Center for Toxicological Research, 3900 NCTR Road, Jefferson, AR 72079. Fax: (870) 543-7682; E-mail: rblair@nctr.fda.gov.


Key Words: estrogen receptor competitive-binding assay; relative binding affinity; estrogens; antiestrogens; alkylphenols; organochlorines; pesticides; parabens; phthalates.


Summary

We have utilized a validated (standardized) estrogen receptor (ER) competitive-binding assay to determine the ER affinity for a large, structurally diverse group of chemicals. Uteri from ovariectomized Sprague-Dawley rats were the ER source for the competitive-binding assay. Initially, test chemicals were screened at high concentrations to determine whether a chemical competed with [3H]-estradiol for the ER. Test chemicals that exhibited affinity for the ER in the first tier were subsequently assayed using a wide range of concentrations to characterize the binding curve and to determine each chemical’s 3H and relative binding affinity (RBA) values. Overall, we assayed 188 chemicals, covering a 1 X 106-fold range of RBAs from several different chemical or use categories, including steroidal estrogens, synthetic estrogens, antiestrogens, other miscellaneous steroids, alkylphenols, diphenyl derivatives, organochlorines, pesticides, alkylhydroxybenzoate preservatives (parabens), phthalates, benzophenone compounds, and a number of other miscellaneous chemicals. Of the 188 chemicals tested, 100 bound to the ER while 88 were non-binders. Included in the 100 chemicals that bound to the ER were 4-benzyloxyphenol, 2,4-dihydroxybenzophenone, and 2,2'-methylenebis(4-chlorophenol), compounds that have not been shown previously to bind the ER. It was also evident that certain structural features, such as an overall ring structure, were important for ER binding. The current study provides the most structurally diverse ER RBA data set with the widest range of RBA values published to date.


Introduction


A variety of synthetic chemicals has been released into the environment, some in large quantities, during the last few decades. Rapidly increasing scientific evidence suggests that many of these chemicals, structures of which cross a wide range, can interfere with normal, hormonally regulated biological processes to adversely affect development and/or reproductive function in wildlife, experimental animals, and humans (see Colborn et al., 1993; Danzo, 1998; Daston et al., 1997; Kavlock et al, 1996; Sonnenschein and Soto, 1998; Toppari et al., 1996 for recent reviews). These environmental contaminants are able to alter the normal functioning of the endocrine and reproductive systems by mimicking or inhibiting endogenous hormone action, modulating the production of endogenous hormones, or altering hormone receptor populations (Sonnenschein and Soto, 1998). Due to the ability of these types of chemicals to interfere with endocrine systems, they have been labeled as “endocrine disruptors.” A major mechanism of endocrine disruption is binding of a xenochemical to the ER. Other mechanisms of endocrine disruption, besides receptor-mediated events, may include such mechanisms as inhibition or stimulation of hormone metabolism, actions involved in the regulation of various neural centers or the pituitary, or alterations in serum hormone-binding proteins. Endocrine disruptors are comprised of numerous types of chemicals, which can be categorized by usage (herbicides, fungicides, insecticides) or chemical structure (polychlorinated biphenyls (PCBs), dioxins, organochlorines and alkylphenols) (Colborn et al., 1993; Toppari et al., 1996), and by regulatory authority (U.S. Food and Drug Administration, U.S. Environmental Protection Agency), and their specific statutory authority.


In 1996, due to the increasing concern regarding the adverse health effects of endocrine disruptors, the U.S. Congress passed the Food Quality Protection Act (FQPA) and amended the Safe Drinking Water Act (SDWA). These laws required the EPA to develop and implement a screening strategy to assess the risk associated with estrogenic endocrine disruptors. Subsequently, the Administrator of the EPA determined that androgens and thyroid-active chemicals should also be included and that chemicals active in the screens should be rigorously tested in animal studies. In response to the passage of these laws, the EPA formed the Endocrine Disruptor Screening and Testing Advisory Committee (EDSTAC), which was given the task of designing a screening and testing program to assess potential endocrine disrupting chemicals. It was recommended by EDSTAC that screening and testing should be accomplished in 3 steps, to include (1) initial priority setting, (2) Tier 1 screening, and (3) Tier 2 testing. One prioritization method currently being assessed involves high throughput assays of receptor-dependent responses. A second method currently being examined for the priority-setting process is a quantitative structure-activity relationship (QSAR) analysis (Tong et al., 1997; Waller et al., 1996). This type of analysis correlates a chemical’s structural characteristics with biological activity and the models derived from these relationships can be used to predict the activity of untested environmental contaminants. Such models could augment, and possibly replace, high throughput screening for priority setting. According to EDSTAC recommendations, Tier 1 screening will consist of a number of in vivo and in vitro assays. Included in the list of suggested in vitro assays is the ER competitive-binding assay, which assesses a chemical’s ability to bind to the ER. This assay can be used to rapidly ascertain whether an environmental contaminant is capable of acting through the same binding mechanism as endogenous estradiol. Relative binding-affinity data can be used for (1) priority setting using models developed from the RBA values and (2) Tier 1 screening using the methods/results described in this study. The final EDSTAC report can be found on the internet at http://www.epa.gov/ opptintr/opptendo/finalrpt.htm.


It has been demonstrated that a number of the environmental contaminants are able to function in a manner similar to estradiol (Colborn etal., 1993;Daston etal., 1997; Roy etal., 1997; Sonnenschein and Soto, 1998; Toppari et al., 1996). These include such compounds as bisphenol A (Krishnan et al., 1993; Nagel et al., 1997; Perez et al., 1998), the alkylphenols (Nimrod and Benson, 1996; Soto et al., 1991; White et al., 1994), PCBs (Bergeron et al, 1994; Bitman et al., 1970; Korach et al, 1988), kepone (Gellert, 1978; Hammond et al., 1979;), and the parabens (Routledge et al., 1998). When the ER is bound by its endogenous hormone, subsequent activation of the ER results in conformational changes, protein interactions, and gene transcription (Beekman et al, 1993; Parker et al., 1993; Tsai and O’Malley, 1994). Therefore, xenoestrogen-induced alterations in normal endocrine function can result in adverse effects at the cellular level (Roy et al., 1997).


The majority of QSAR models developed to date are based on the biological activity of small groups of compounds with similar activity and structural features. However, the structural diversity of estrogenic chemicals is very broad. For the current study, chemicals were selected such that 1) a large data set was generated, 2) a wide diversity of structural features was represented and 3) a wide range of biological activity was measured. To our knowledge, the results presented here represent the largest and most diverse data set publicly available for chemicals binding to the ER. These data can be utilized to develop a highly robust 3D-QSAR model, as well as separate chemometric models (the development and implementation of which will be presented in a separate manuscript). In addition, these data are also useful for comparing ER-binding results from a large number of chemicals to small data sets using different assay conditions.



Materials and methods


Chemicals

Trizma base, Trizma hydrochloride, glycerol, EDTA, dithiothreitol, and sodium azide were obtained from Sigma (St. Louis, MO). The [2, 4, 6, 7, 16, 17-3H]-E2 (141 Ci/mmol) used in the competitive-binding assay was obtained from Dupont-New England Nuclear (Boston, MA). Hydroxylapatite was obtained from Bio-Rad Laboratories (Hercules, CA) and the UltimaGold scintillation cocktail was purchased from the Packard Instrument Company (Meriden, CT). The source and purity (when available) for each of the competing test compounds is provided in tabular form with the results from the competitive-binding assays.


Uterine cytosol (estrogen receptor) preparation

Adult (retired breeders; 244.87 ± 18.27 days of age), non-pregnant Sprague-Dawley rats (n = 188; NCTR:SDN, an outbred albino rat stock originating from CRL:CD [SD]BR rats, obtained in 1979 and maintained as a closed colony) were maintained in a controlled environment (23°C and 50% humidity) on a 12-h light/dark cycle (lights on at 0600 h). Animals received Purina rat chow and filtered tap water ad libitum. Females (a mean of 14 rats per cytosol batch) were ovariectomized a minimum of 10 days prior to receptor preparation. After sacrifice by CO2 asphyxiation, uteri were excised, trimmed of excess fat and mesentery, weighed and placed in ice-cold TEDG buffer (10 mM Tris, 1.5 mM EDTA, 10 mM dithiothreitol, 10% glycerol, pH 7.4). The pooled uteri were placed in fresh, ice-cold TEDG buffer at a concentration of 1.0 g of tissue/10 ml buffer. After pre-cooling (4°C) the homogenization probe, uterine tissue was homogenized at 4°C with a Polytron homogenizer (Brinkman Instruments, Westbury, NY) using 5-s bursts. The resulting homogenate was transferred to pre-cooled (4°C) ultracentrifuge tubes and centrifuged at 105,000 X g for 60 min at 4°C. After centrifugation, the ER-rich supernatant was decanted into 15-ml conical tubes (—3 ml/aliquot; —12 aliquots/average cytosol batch), and was stored at -70°C until used in competition assays.


Estrogen receptor (ER) competitive-binding assays

[3H]-E2 (10 μl; 1 X 10-9 M final assay concentration) was incubated with 10 μl of increasing concentrations of radioinert competitor, 50 μl of uterine cytosol preparation and 230 μl of 50 mM Tris buffer (pH 7.4) in duplicate tubes. Reaction-mixture tubes were placed in a drum roller (Glas-Col, Terre Haute, IN) and incubated at 4°C for 20 h. Following the incubation period, 750 μl of a cold 60% hydroxylapatite (HAP) slurry (made in 50 mM Tris, pH 7.4) was added to each tube to separate the bound ligand from the free ligand. These tubes were incubated in an icewater bath for 20 min and vortexed for 10 s at 5-min intervals. Tubes were subsequently centrifuged (600 X g) at 4°C for 5 min. The supernatant was discarded and the resulting HAP pellet was resuspended in 2.0 ml of cold 50 mM Tris buffer and vortexed and centrifuged as above. After 3 washes, the supernatant was discarded and 2.0 ml of cold (4°C) 100% ethanol was added to each tube to extract the radiolabeled E2 from the HAP. Tubes were incubated on ice for 15 min and vortexed at 5-min intervals. Following the ethanol incubation, the tubes were centrifuged (600 X g) at 4°C for 10 min. The resulting supernatant was decanted into vials containing 10 ml of scintillation cocktail. Radioactivity was measured on a Packard Tri-Carb 1600TR Liquid Scintillation Analyzer (Packard Instrument Company, Meriden, CT). In addition to the radioinert competitors, each assay included a zero tube (no competitor added; represented total binding of [3H]-E2; averaged approximately 15,000 dpm) and an E2 standard curve (1 X 10-7, 1 X 10-8, 1 X 10-9, 3.33 X 10-10, 1 X 10-10, and 3.33 X 10-11 M concentrations) for quality control purposes. The 1 X 10-7 M E2 tube contained a 100-fold molar excess ofradioinert E2 compared to [3H]-E2 and thus represented non-specific binding (NSB; averaged approximately 800 dpm). Radioactivity counts (dpm) of the NSB tubes were subtracted from all tubes prior to calculation of percent [3H]-E2 bound. Data for each competitor and the E2 standard curve were plotted as percent [3H]-E2 bound versus molar concentration, and the 3H (50% inhibition of [3H]-E2 binding) for each competitor determined. The RBA for each competitor was calculated by dividing the IC5Q of E2 by the IC5Q of the competitor and was expressed as a percent (E2 = 100). Details and validation of the ER competitive-binding assay will be published in a separate manuscript (unpublished).


Test chemicals were dissolved in 100% ethanol at the highest concentration possible. Stock solutions were then subsequently diluted in ethanol for analysis in the ER competitive-binding assay. Due to the large number of chemicals tested, the ER competitive-binding assays were set up in a tiered design. Unless known to bind to the ER, test chemicals were initially run at only 2 high concentrations spanning 3 log concentrations (Tier 1). If a test compound exhibited binding to the ER, then a second assay (Tier 2) was run using a wide range of concentrations, ranging generally from 1 X 10-4 to 1 X 10-9 M in 10-fold increments, though this varied depending upon the competitor. If necessary, a Tier 3 assay (consisting of one-half log molar concentrations which bracketed the 3H observed in the Tier 2 assay) was run to more accurately determine a competitor’s 3H. In the Tier 1 assay, approximately 36 chemicals could be assayed in replicate per cytosol batch, while only 18 chemicals per cytosol batch could be assayed in replicate in either the Tier 2 or Tier 3 assay. In the final analysis, it required one rat per chemical assayed in replicate. Chemicals which failed to bind the ER were designated as “non-binders.” All assays were replicated a minimum of 2 times and 3H values of positive chemicals are the means of the replicates.



Results


Mean 3Hs and RBAs for the 188 chemicals tested (100 binders and 88 non-binders) in the ER competitive-binding assay are presented in Tables 1-12. Of the 100 active test chemicals, 26 were strong binders (log RBA > 0), 33 were moderate binders (log RBA between 0 and -2), and 41 were weak binders (log RBA < -2). For purposes of clarity and convenience, test compounds were grouped in tables according to chemical or use classifications. Figure 1 shows representative ER-binding curves over the range of concentrations used for various chemicals in the ER competitive-binding assay.



Figure 1. Representative estrogen receptor binding curves. This figure demonstrates the variety of chemicals, based on both structure and affinity for the ER, assayed in the ER competitive-binding assay. It also clearly demonstrates the parallelism in the linear portion of the curve between different test competitor curves.


As expected, all of the selected steroidal estrogens (Table 1), synthetic estrogens (Table 2) and antiestrogens (Table 3) showed affinity for the ER. With the exception of one chemical in each of these 3 classes, all these chemicals exhibited moderate to strong binding affinity. This was especially evident for the synthetic estrogens, which exhibited strong affinity for the ER in 13 of the 16 chemicals tested. Of the synthetic estrogens, diethylstilbestrol (DES), meso-hexestrol, and ethinyl estradiol (EE2) exhibited greater affinity for the ER than E2. The only other chemical analyzed that exhibited greater affinity than E2 for the ER was the antiestrogen 4-hydroxytamoxifen.


Table 1. IC50s and Relative Binding Affinities (RBA) for Steroidal Estrogens



Table 2. IC50s and Relative Binding Affinities (RBA) for Synthetic Estrogens



Table 3. IC50s and Relative Binding Affinities (RBA) for Type I“ and Type II* Antiestrogens



A number of miscellaneous steroids were also analyzed in the ER competitive-binding assay (Table 4). The majority of these chemicals was inactive and did not bind to the ER. However, norethynodrel and 5α-androstane-3β,17β-diol exhibited moderate binding while 5α-androstane-3α,17β-diol showed weak binding. Unexpectedly, both of these latter chemicals exhibited U-shaped ER-binding curves in which the percent [3H]-E2 bound increased with increasing concentration of test compound after reaching a nadir below 50% (Fig. 2). Nonetheless, the initial portions of these curves were parallel to the E2 standard curve. Of the 188 chemicals assayed, 22 (20 binders and 2 non-binders) demonstrated a U-shaped binding curve (identified individually in Tables 1-12). In 5 of these 22 chemicals, an increase in percent [3H]-E2 bound was evident for the 2-3 highest concentrations tested. In the remaining 17 chemicals, an increase in the percent [3H]-E2 binding was evident only at the highest concentration tested. The majority (15) of these 22 chemicals consisted of steroidal estrogens, synthetic estrogens, antiestrogens, or miscellaneous steroids.


Table 4. IC50s and Relative Binding Affinities (RBA) for Other Miscellaneous Steroids




Figure 2. U-shaped binding curves were evident for a few of the chemicals assayed in the ER competitive-binding assay. As described in the results, this may be due to an increased dissociation rate of the ER-E2 complex in the presence of high concentrations of competitors.


Alkylphenolic compounds (Table 5) were a major group of chemicals assayed. All but 3 of the alkylphenols exhibited binding to the ER with their binding affinity ranging from moderate to weak. Of concern was the possibility that different sources or different lots of an individual chemical might bind to the ER with substantially different affinities. Therefore, we tested this possibility on 4-nonylphenol, a compound known to be a mixture of isomers. Nonylphenol was chosen due to its importance in National Toxicology Program studies (Chapin et al, 1999) and the fact that it is easily obtained from several different sources. We obtained 4-nonylphenol from Aldrich, Fluka, Schenectady International, and Lancaster. In addition, we tested 2 different lots of 4-nonylphenol from Aldrich and 2 different lots from Fluka. The 3H values for 4-nonylphenol from Aldrich, Fluka, and Schenectady International ranged from 2.4 X 10-6 to 4.7 X 10-6, thus demonstrating very little difference across sources. Also, there were no substantial differences in ER-binding affinities between different lots of 4-nonylphenol from the same source. The one source of 4-nonylphenol which exhibited a significantly lower affinity for the ER compared to the other sources was the Lancaster (3H = 2.8 X 10-5); however, this was 4-n-nonylphenol, an analytical standard, which is a pure, unbranched isomer unlike the other nonylphenol compounds which are a mixture of para-isomers.


Table 5. IC50s and Relative Binding Affinities (RBA) for Alkylphenolic Compounds



It also appeared that length of a chemical’s side chain influenced the ER-binding affinity of the alkylphenols. In general, the longer the side chain, the greater the binding affinity for the ER, such that nonylphenol > octylphenol > heptyloxyphenol > amylphenol > butylphenol > ethylphenol. However, there appears to be a limit to the number of side-chain carbons that increases binding to the ER since 4-dodecylphenol (12 carbons) exhibited a lower RBA than 4-nonylphenol (9 carbons).


Table 6 provides the binding affinities of another major group of test chemicals, the diphenyl derivatives. Of the diphenyl methane derivatives (bisphenol As), only bisphenol B was moderately active at binding the ER, while the remaining chemicals in this group were weak binders or inactive at the concentrations tested. Three of the 5 diphenyl ethanes were moderate ER binders, while the biphenyl compounds were either weak binders or non-binders.


Table 6. IC50s and Relative Binding Affinities (RBA) for Diphenyl Derivatives



In this study, we tested 3 groups of organochlorines (Table 7): DDT isomers, methoxychlor and it’s derivatives, and polychlorinated biphenyls (PCBs). Only dihydroxymethoxychlor olefin showed strong affinity for the ER. The remaining chemicals exhibited either moderate or weak RBAs or were inactive at the concentrations tested. Only one of the DDT isomers, o,p'-DDT, bound to the ER and did so with weak affinity. The majority of the methoxychlor derivatives (5 of 7) and PCBs (6 of 9) competed for the ER. A difference in ER binding between


Table 7. IC50S and Relative Binding Affinities (RBA) for Organochlorines



methoxychlor with different levels of purity was evident. The 95% methoxychlor exhibited weak affinity for the ER, while the 99% methoxychlor did not compete for the ER. It has been demonstrated previously that a phenolic contaminant in methoxychlor preparations is estrogenic and that this contaminant may be dihydroxymethoxychlor (HPTE), a metabolite of methoxychlor (Bulger et al, 1978a,b; Cummings, 1997). Since the pesticides (the DDT and methoxychlor isomers) within this group of chemicals exhibited affinity for the ER, it was important to determine whether other pesticides could also compete. Of the other pesticides tested (Table 8), only kepone bound with moderate affinity to the ER. None of the remaining pesticides analyzed in the ER competitive-binding assay exhibited any activity at the concentrations tested.


Table 8. IC50s and Relative Binding Affinities (RBA) for Other Pesticides



Table 9 shows the RBAs for several paraben compounds. All of the parabens examined in this study competed for the ER. Of the 7 chemicals analyzed, one bound the ER with moderate affinity while the rest exhibited weak binding. The parabens, like the alkylphenolic compounds, demonstrated a positive correlation between binding affinity and chain length.


Table 9. IC50s and Relative Binding Affinities (RBA) for Alkyl Hydroxy Benzoate Preservatives (Parabens)



The chemicals with the longer side chains (2-ethylhexyl, heptyl, and benzyl 4-hydroxybenzoates) showed greater affinity for the ER compared to the parabens with shorter side chains (butyl, propyl, ethyl, and methyl 4-hydroxybenzoates).


Under the conditions of our ER competitive-binding assay, none of the 8 phthalate compounds exhibited an 3H (Table 10). However, 2 of these chemicals, benzylbutyl phthalate and BIS(2-ethylhexyl) phthalate, did compete for the ER (Fig. 3a).


Table 10. IC50s and Relative Binding Affinities (RBA) for Phthalates




Figure 3. Estrogen receptor binding affinity curves for the phthalate compounds. A depicts the binding curves for the entire phthalate group using our standard ER competitive-binding assay. None of the phthalates competed strongly for the ER; however, benzylbutyl and BIS(2-ethylhexyl) phthalate showed slight competition for the ER. B shows curves from a single equilibrium test with benzylbutyl phthalate. Two concentrations of benzylbutyl phthalate were assayed in our ER competitive-binding assay with extended incubation periods. Extending the incubation period did not increase affinity for the ER. C demonstrates phthalate binding in an alternative assay procedure in which the 20-h, 4°C incubation was replaced by a 30-min, 30°C incubation period followed by a 30-min, 4°C incubation period. Replacing the normal incubation conditions with a short-term, high-temperature incubation resulted in 3 of the 4 phthalates tested binding to the ER with very weak affinity.


At a concentration of 1 X 10-3 M, benzylbutyl phthalate and BIS(2-ethylhexyl phthalate) exhibited 66% and 78% [3H]-E2 bound, respectively. Since the phthalates are apparently such weak binders, it is possible that they had not reached equilibrium during the 20-h, 0°C incubation period of the ER competitive-binding assay. Therefore, an assay was conducted that examined benzylbutyl phthalate binding at 4°C at twice-daily intervals for 3 days (Fig. 3b). The results of this assay demonstrated that extending the incubation time had no effect on binding of benzylbutyl phthalate to the ER. As a final test, we conducted an assay in which the 20-h, 4°C incubation period was replaced by a 30-min, 30°C incubation followed by a 30-min, 4°C incubation (Fig. 3c). Under these assay conditions, ER binding of benzylbutyl phthalate (3H = 7.2 X 10-5 M), diethyl phthalate (3H = 5.0 X 10-3 M), and dimethyl phthalate (3H = 9.9 X 10-3 M) was achieved, with benzylbutyl and dibutyl phthalate exhibiting U-shaped binding curves.


The remaining test chemicals consisted of benzophenone compounds (Table 11) and several miscellaneous classes of chemicals (Table 12). Only 2 of the 5 benzophenone compounds competed for the ER. Of the remaining 39 chemicals assayed (Table 12), 3 showed moderate affinity and 4 exhibited weak affinity for the ER, while the other 32 were inactive at the concentrations tested.


Table 11. IC50s and Relative Binding Affinities (RBA) for Benzophenone Compounds



Table 12. IC50s and Relative Binding Affinities (RBA) for Miscellaneous Compounds



There were also a number of test chemicals that competed for the ER, but did so with such a weak affinity that an 3H was not attainable (Table 13). While the chemicals that exhibited this slight affinity for the ER were structurally diverse, including 4 organochlorines and 2 phthalates, they were also environmentally important.


Table 13. Test Compounds that Competed for the ER but Did Not Attain a Measurable 3H




Discussion


To our knowledge, this is the first paper to report the ER-binding affinities of such a large number (188) of structurally diverse chemicals assayed under identical conditions. Other published results (Andersen et al, 1999; Bolger et al, 1998; Kuiper et al., 1997, 1998; Perez et al., 1998; Shelby et al., 1996; Waller et al., 1996) report the ER RBAs for smaller and less structurally diverse data sets.


In the current study, the rats utilized as the source of the ER were retired breeders. Although this is not a common practice, ER levels in ovariectomized, retired breeders are comparable to both immature, intact animals (Clark et al., 1978; van Doom et al., 1982) and adult, ovariectomized rats (Medlock et al., 1991). Similar binding results were evident between the current study and a previous study using immature rats (Perez et al., 1999), further indicating that the age of the rats used as an ER source is not critical. While ER levels are similar between these rats of different ages, the use of retired breeders has three distinct advantages. First, retired breeders have greater uterine weights, which results in more total ER available for use in binding studies. Second, due to the larger uterine weights, fewer animals are required. Lastly, the use of retired breeders is essentially an animal-sparing process, since the ability to use these animals, which would be disposed of under normal situations, allows us to conduct these studies without having to purchase and subsequently sacrifice new animals.


The reason for the U-shaped binding curves observed in the present study remains unclear. However, it has been shown that high concentrations of steroids and antiestrogens can markedly accelerate the dissociation rate of the ER-[3H]-E2 complex (Borgna and Ladrech, 1982). Chemicals in the present study that exhibited this response did so at concentrations greater than 1 X 10-5 M, with the majority of them being at concentrations greater than 1 X 10-4 M. Therefore, the U-shaped binding curves presented here are most likely due to the high doses assayed and the kinetics of the binding assay, and as such they are probably not associated with the U-shaped doseresponse curves observed at low doses in vivo for some chemicals (vom Saal et al, 1997).


Of the variety of chemical classes analyzed in the current study, two were of particular interest. These were the phthalates and the parabens. The phthalic acid esters (phthalates) are widely used as plasticizing agents; however, since they are not covalently bound within the plastic, the phthalates can be released into the environment (Autian, 1973; Giam et al, 1978; Thomas and Thomas, 1984). Since phthalates are capable of altering reproductive function in rats (Ema et al, 1998; Wine et al., 1997) and since estrogens are a primary regulator of female reproduction, it is possible that phthalates are working through the ER. Previous studies have suggested that di-nbutyl phthalate, benzylbutyl phthalate and BIS(2-ethylhexyl) phthalate are weak binders or weakly estrogenic in a variety of systems (Bolger et al., 1998; Coldham et al, 1997; Jobling et al., 1995; Nakai et al, 1999; Soto et al, 1995; Waller et al, 1996; Zacharewski et al, 1998). However, we were unable to determine 3H values for any of the phthalate compounds analyzed in our ER competitive-binding assay. Nonetheless, two of the phthalates, benzylbutyl phthalate and BIS(2-ethylhexyl) phthalate, competed with E2 for the ER. Utilizing rainbow trout ER, Jobling et al. (1995) was able to establish an 3H for benzylbutyl phthalate as well as binding for BIS(2ethylhexyl) and di-n-butyl phthalate. However, the binding curves were not parallel with that of E2, suggesting that the phthalates may be working through or influenced by an alternative mechanism. Other studies have also reported 3H values for benzylbutyl phthalate and di-n-butyl phthalate, using a human recombinant ER (hrER) (Bolger et al, 1998; Nakai et al, 1999) and rat uterine cytosol (Zacharewski et al, 1998). Zacharewski and colleagues (1998) also demonstrated that dihexyl phthalate competed with [3H]-E2 for the ER, though not with strong enough affinity to attain an 3H value. However, these studies utilized one-day ER competitive-binding assays, which included either a 30-min, 30°C (Zacharewski et al, 1998) or a 60-min, room temperature (Bolger et al, 1998; Nakai et al, 1999) incubation period. Replacing the standard 20-h, 4°C incubation with a 30-minute, 30°C incubation followed by a 30-min, 4°C incubation increased the ability of the phthalates to bind the ER. Waller and co-workers (1996) reported pKis for both benzylbutyl phthalate and di-n-butyl phthalate using an 18-h, 4°C incubation in their ER competitive-binding assay. However, these authors measured binding to a mouse uterine cytosol ER preparation and differences in binding affinities may be related to variation in species sensitivity. When comparing the data presented here to data collected from competitive-binding assays that utilized a shortterm incubation period at room temperature or above (Arcaro et al, 1998; Bolger et al, 1998; Klotz et al., 1996), it was evident that differences in RBA values were relatively small. In general, RBA values for low affinity xenobiotics were slightly higher in the short term, high temperature assays compared to the assay utilized in the current study, while high affinity chemicals exhibited similar binding affinities. This suggests that differences in assay conditions can lead to the observed differences in binding affinities and that high-temperature assays might be somewhat more sensitive to certain chemicals. However, high-temperature assays run the risk of ER degradation during the assay.


It has been demonstrated in numerous studies that parabens compounds produce a variety of toxic and physiological effects. Chang and Voelkel (1986) reported that both methyl and propyl parabens are present in a commercial preparation of naloxone (Narcan®) at high concentrations (1.8 mg/ml of methyl [1.2 X 10-2 M] and 0.2 mg/ml of propyl [1.1 X 10-3 M] parabens). They also demonstrated that the vasodilatory effects of Narcan are due entirely to the parabens. Various paraben compounds have also been shown to cause severe damage to rat hepatocytes (Nakagawa et al., 1998; Sugihara, et al., 1997) and reduced ciliary beat frequency in rat trachea (Jian and Po, 1993a,b). Nakagawa et al. (1998) indicated that hepatocyte toxicity was correlated with alkyl side-chain length such that butyl and iso-butyl parabens were more toxic than propyl and iso-propyl parabens, which in turn were more toxic than ethyl and methyl parabens. This correlates well with the binding affinity data in the current study, which demonstrated increasing RBA values with increasing alkyl-side chain length. Despite these obvious toxic effects, parabens compounds are present in a wide variety of products including cosmetics (Rastogi et al., 1995), pharmaceutical products (Chang and Voelkel, 1986; Pompy et al, 1991; van Faassen et al, 1990), cigarettes (Castano, et al, 1988), and honey bee royal jelly (Ishiwata et al, 1995). In fact, in 1981, the Food and Drug Administration reported that four parabens compounds (methyl, ethyl, propyl, and butyl parabens) were used as preservatives in more than 13,200 formulations, including most cosmetic products, due to their antimicrobial properties (Elder, 1984). Although the antimicrobial activity of the parabens compounds is positively correlated with alkyl-chain length, their water solubility is negatively correlated. Therefore, the short-chain parabens compounds are generally used in formulations (Dal Pozzo and Pastori, 1996). Nonetheless, we and others (Routledge et al., 1998) have demonstrated that even the smaller parabens compounds are weak ER competitors. Cytotoxicity (Nakagawa et al, 1998) and antimicrobial activity (Dal Pozzo and Pastori, 1996) are not the only characteristics associated with alkyl side-chain length. Percutaneous absorption (Lee and Kim, 1994; Twist and Zatz, 1986) and transactivation of the ER (Routledge et al, 1998) also increase with alkyl side-chain length. Thus, due to their inherent estrogenicity and their wide range of applications, it is apparent that the parabens compounds pose a potential hazard as endocrine disruptors.


The determination of ER-binding affinities in a group of chemicals with such a wide variety of structural diversity is also beneficial because it increases one’s knowledge of chemicals which are inadvertently estrogenic. For example, to our knowledge, 4-benzyloxyphenol, 2,4-dihydroxybenzophenone, and 2,2'-methylenebis(4-chlorophenol) have not previously been demonstrated to bind the ER. The 4-benzyloxyphenol (commonly called monobenzone) is a therapeutic agent used as a depigmenting compound. It has been widely used in the treatment of vitiligo, a skin disease which effects 1-3% of the population and is characterized by depigmentation of areas of the skin (Kenney, 1988). This chemical has also been examined as an alternative branding method in beef cattle (Schwartzkopf et al, 1994). Benzoresorcinol (2,4-dihydroxybenzophenone) is commonly used as a plastic additive (Spyropoulos 1998). It is also a metabolite of benzophenone-3, an absorber of ultraviolet light used extensively as a sunscreen and color fastener (Okereke et al, 1993, 1994). Dermal treatment of Sprague-Dawley rats with benzophenone-3 resulted in detection of it’s metabolites, including 2,4-dihydroxybenzophenone, in plasma five min after administration (Okereke et al, 1994). This metabolite was detected in most tissues within six h after administration, with the highest concentration observed in the liver (Okereke et al, 1993). 2,2'-Methylenebis(4chlorophenol), commonly called dichlorophen, is used as an agricultural fungicide, a germicide in soaps and shampoos, and therapeutically as an anthelmintic.


Even absent the computational models, certain structural features important for ER binding can be discerned. A ring structure is of primary importance to ER-binding affinity. Although chemicals with a ring structure may or may not bind the ER, chemicals lacking a ring structure apparently will not bind the ER. In the present study, we were specifically attempting to identify potentially estrogenic chemicals so the vast majority of chemicals tested possessed a ring structure. Of the 6 chemicals that did not possess ring structures, none exhibited affinity for the ER. With the exception of 4 chemicals (kepone, norethynodrel, 5α-androstane-3α, 17β-diol, and 5α-androstane-3β,17β-diol), all test chemicals that actively bound the ER contained benzene rings. In addition to a ring structure, all the actively competing test chemicals, with the exception oftriphenylethylene, o,p'-DDT, and 2,4-dichlorobiphenyl, contained an oxygen atom on the ring. It is well known that a phenolic ring is important for estrogenicity (Anstead et al., 1997) and it has been suggested that the hydroxyl group on the phenol ring acts as both a hydrogen bond donor and acceptor in the ER-binding site (Duax and Weeks, 1980). Substitution of the hydroxyl hydrogen atom by an alkyl group such as a methyl group significantly decreases the chemical’s affinity for the ER. For example, diethylstilbestrol dimethyl ether (RBA = 0.056) exhibited a greatly lowered affinity for the ER when compared to the non-methylated diethylstilbestrol (RBA = 399.56). In addition, methoxychlor, which contains 2 methyl groups, showed a much lower ER-binding affinity (RBA = 0.001) than dihydroxymethoxychlor (RBA = 0.253), which contains no methyl groups. However, not all substitutions result in significant suppression of ER-binding affinity. Some of the antiestrogens (tamoxifen, clomiphene, nafoxidine, and toremifene) remain strong binders despite the lack of a phenolic ring. In general, the results indicate that chemicals with 2 ring structures separated by 2 carbon atoms (steroidal and synthetic estrogens and diphenyl ethanes) have higher RBAs compared to chemicals with a single ring structure (alkylphenols, phthalates and parabens) or 2 rings separated by one carbon atom (bisphenol A derivatives and benzophenone compounds). Interestingly, ER binding was even observed for very small chemicals such as ethylphenol and 2-chloro-4-methylphenol, further confirming the importance of the phenolic ring for binding. Crystallographic data of the ER complex with a number of chemicals, including E2, DES, raloxifene and 4-hydroxytamoxifen, demonstrated that the critical spacing of hydrophobic and hydrogen bond interactions play a major role in determining binding affinity (Brzozowski et al, 1997; Shiau et al, 1998). Although hydrophobic interactions and hydrogen bonding are important for high-affinity ligands, they are clearly less important for low-affinity ligands (unpublished data).


Since the data from the ER competitive-binding assays are being utilized for the development of a computational QSAR model (Shi et al, unpublished), a large number of test chemicals, which provided a wide range of structural diversity with approximately similar numbers of binders in each decade of RBA values, was intentionally selected. These included chemicals that were both ER binders and non-binders, and were environmentally and commercially relevant. Specifically, the data for chemicals that do bind are being used to develop a robust comparative molecular field analysis (CoMFA) QSAR model, which can be used to predict the ER-binding affinity for specific chemicals of interest, including some 8,000 FDAregulated chemicals. This allows one to identify new avenues for future research. Due to the large number of chemicals tested and their structural diversity, this data set will also be useful in the development of other computational predictive models. In addition to the CoMFA model, other QSAR models (classical, hologram, and pharacophore-based models), as well as a number of classification models (e.g., K-Nearest Neighbor and Self Independent Modeling of Chemical Analogy), which qualitatively predict whether a chemical will bind or not bind, are currently being explored. These models, individually or in a complementary sequence, may allow more complicated biological systems to be modeled. The strengths and weaknesses of various QSAR models/programs (Hansch, CoMFA, HQSAR, etc.) have been previously presented (Tong et al, 1997, 1998). The published ER crystal structure (Brzozowski et al., 1997; Shiau et al, 1998) is also being utilized, in combination with the current data set, to examine a “uniform docking” approach, which will enable one to identify the binding conformation of a diverse group of chemicals. Using an array of molecular modeling approaches, common structural features that are related to biological activity can be identified. Once identified, these structural features can be used to rapidly identify untested, potentially estrogenic chemicals.


In summary, the current study has analyzed a large number of structurally diverse chemicals for their ability to bind to the rat uterine ER. Compared to other published data sets, this is the largest and most diverse group of chemicals tested to date using a single assay. By examining such a large group of chemicals, we were able to identify several chemicals not previously known to bind the ER. Relative binding affinities of these diverse chemicals were often highly correlated with specific structural features. In particular, an overall ring structure was important to a chemical’s ability to bind the ER; the presence of a phenolic ring was indicative of a chemical’s affinity for the ER; and alterations in hydroxylation and alkyl side-chain length also affected a chemical’s RBA. These data are useful in a number of areas. For example, the RBA data are being utilized as a training set for the construction of a robust CoMFA QSAR model as well as other QSAR and chemometric models; these models can be useful as priority setting tools in the EDSTAC process. The large RBA data set presented here can also be used for comparisons with other ER-binding results generated under varying assay conditions.



Acknowlegments


We gratefully acknowledge the U.S. Department of Energy and the U.S. Food and Drug Administration (FDA) for postdoctoral support through the Oak Ridge Institute for Science and Education. This work was partially supported by FDA’s Office of Women’s Health and the Chemical Manufacturer’s Association.



References


Andersen, H. R., Andersson, A-M., Arnold, S. F., Autrup, H., Barfoed, M., Beresford, N. A., Bjerregaard, P., Christiansen, L. B., Gissel, B., Hummel, R., Jorgensen, E. B., Korsgaard, B., Le Guevel, R., Leffers, H., McLachlan, J. A., Moller, A., Nielsen, J. B., Olea, N., Oles-Karasko, A., Pakdel, F., Pedersen, K. L., Perez, P., Skakkeboek, N. E., Sonnenschein, C., Soto, A. M., Sumpter, J. P., Thorpe, S. M., and Grandjean, P. (1999). Comparison of short-term estrogenicity tests for identification of hormone-disrupting chemicals. Environ. Health Perspect. 107(Suppl 1), 89-108. 

Anstead, G. M., Carlson, K. E., and Katzenellenbogen, J. A. (1997). The estradiol pharmacophore: Ligand structure-estrogen receptor binding affinity relationships and a model for the receptor binding site. Steroids 62, 268-303. 

Arcaro, K. F., Vakharia, D. D., Yang, Y., and Gierthy, J. F. (1998). Lack of synergy of mixtures of weakly estrogenic hydroxylated polychlorinated biphenyls and pesticides. Environ. Health Perspect. 106(Suppl 4), 1041-1046. Autian, J. (1973). Toxicity and health threats of phthalate esters: review of the literature. Environ Health Perspect. 4, 3-26. 

Beekman, J. M., Allan, G. F., Tsai, S. Y., Tsai, M. J., and O’Malley, B. W. (1993). Transcriptional activation by the estrogen receptor requires a conformational change in the ligand binding domain. Mol. Endocrinol. 7, 1266-1274. 

Bergeron, J. M., Crews, D., and McLachlan, J. A. (1994). PCBs as environmental estrogens: turtle sex determination as a biomarker of environmental contamination. Environ. Health Perspect. 102, 780-781. 

Bitman, J., and Cecil, H. C. (1970). Estrogenic activity of DDT analogs and polychlorinated biphenyls. J. Agric. Food Chem. 18, 1108-1112. 

Bolger, R., Wiese, T. E., Ervin, K., Nestich, S., and Checovich, W. (1998). Rapid screening of environmental chemicals for estrogen receptor binding capacity. Environ. Health Perspect. 106, 551-557. 

Borgna, J. L., and Ladrech, S. (1982). The dissociation rate of estrogen receptor-ligand complexes is increased by high concentrations of steroids and antiestrogens. Mol. Cell. Endocrinol. 27, 1-15. 

Brzozowski, A. M., Pike, A. C., Dauter, Z., Hubbard, R. E., Bonn, T., Engstrom, O., Ohman, L., Greene, G. L., Gustafsson, J. A., and Carlquist, M. (1997). Molecular basis of agonism and antagonism in the oestrogen receptor. Nature 389, 753-758. 

Bulger, W. H., Muccitelli, R. M., and Kupfer, D. (1978a). Studies on the in vivo and in vitro estrogenic activities of methoxychlor and its metabolites. Role of hepatic mono-oxygenase in methoxychlor activation. Biochem. Pharmacol. 27, 2417-2423. 

Bulger, W. H., Muccitelli, R. M., and Kupfer, D. (1978b). Interactions of methoxychlor, methoxychlor base-soluble contaminant, and 2,2-BIS(p-hydroxyphenyl)-1,1,1-trichloroethane with rat uterine estrogen receptor. J. Toxicol. Environ. Health 4, 881-893. 

Castano, J. I., Palacio, F. J., and Vargas, L. R. (1988). Liquid chromatographic determination of propyl paraben in cigarette filler. J. Assoc. Anal. Chem. 71, 1115-1117. 

Chang, S.-W., and Vaelkel, N. F. (1986). Actions of opiate agonists, naloxone, and paraben preservatives in the rat lung circulation. Proc. Soc. Exp. Biol. Med. 181, 404-410. 

Chapin, R. E., Delaney, J., Wang, Y., Lanning, L., Davis, B., Collins, B., Mintz, N., and Wolfe, G. (1999). The effects of 4-nonylphenol in rats: A multigeneration reproduction study. Toxicol. Sci.. 52, 80-91. 

Clark, J. H., Hardin, J. W., Upchurch, S. and Eriksson, H. (1978). Heterogeneity of estrogen binding sites in the cytosol of the rat uterus. J. Biol. Chem. 253, 7630-7634. 

Colborn, T., vom Saal, F. S., and Soto, A. M. (1993). Developmental effects of endocrine-disrupting chemicals in wildlife and humans. Environ. Health Perspect. 101, 378-384. 

Coldham, N. G., Dave, M., Sivapathasundaram, S., McDonnell, D. P., Conner, C. , and Sauer, M. J. (1997). Evaluation of a recombinant yeast cell-estrogen screening assay. Environ. Health Perspect. 105, 734-742. 

Cummings, A. M. (1997). Methoxychlor as a model for environmental estrogens. Crit. Rev. Toxicol. 27, 367-379. 

Dal Pozzo, A., and Pastori, N. (1996). Percutaneous absorption of parabens from cosmetic forumlation. Int. J. Cosmetic Sci. 18, 57-66. 

Danzo, B. J. (1998). The effects of environmental hormones on reproduction. Cell. Mol Life Sci. 54, 1249-1264. 

Daston, G. P., Gooch, J. W., Breslin, W. J., Shuey, D. L., Nikiforov, A. I., Fico, T. A., and Gorsuch, J. W. (1997). Environmental estrogens and reproductive health: A discussion ofthe human and environmental data. Reprod. Toxicol. 11, 465-481. 

Duax, W. L., and Weeks, C. M. (1980). Molecular basis of estrogenicity: x-ray crystallographic studies. In Estrogens in the Environment, (J. A. McLachlan, Ed.), pp. 11-31. Elsevier, North Holland, NY. 

Elder, R. L. (1984). Final report on the safety assessment of methylparaben, ethylparaben, propylparaben, and butylparaben. J. Am. Coll. Toxicol. 3, 147-209. 

Ema, M., Miyawaki, E., and Kawashima, K. (1998). Reproductive effects of butyl benzyl phthalate in pregnant and pseudopregnant rats. Reprod. Toxicol. 12, 127-132. 

Gellert, R. J. (1978). Kepone, mirex, dieldrin, and aldrin: Estrogenic activity and the induction of persistent vaginal estrus and anovulation in rats following neonatal treatment. Environ. Res. 16, 131-138. 

Giam, C. S., Chan, H. S., Neff, G. S., and Atlas, E. L. (1978). Phthalate ester plasticizers: A new class of marine pollutant. Science 199, 419-421. 

Hammond, B., Katzenellenbogen, B. S., Krauthammer, N., and McConnell, J. (1979). Estrogenic activity of the insecticide chlordecone (Kepone) and interaction with the uterine estrogen receptors. Proc. Natl. Acad. Sci. USA 76, 6641-6645. 

Ishiwata, H., Takeda, Y., Yamada, T., Watanabe, Y., Hosagai, T., Ito, S., Sakurai, H., Aoki, G., and Ushiama, N. (1995). Determination and confirmation ofmethyl-p-hydroxybenzoate in royal jelly and other foods produced by the honey bee. Food Addit. Contam. 12, 281-285. 

Jian, L., and Po, A. L. W. (1993a). Kinetic evaluation of the ciliotoxicity of methyland propy-p-hydroxybenzoates using factorial experiments. J. Pharm. Pharmacol. 45, 98 -101. 

Jian, L., and Po, A. L. W. (1993b). Ciliotoxicity of methyland propyl-phydroxybenzoates: A dose-response and surface response study. J. Pharm. Pharmacol. 45, 925-927. 

Jobling, S., Reynolds, T., White, R., Parker, M. G., and Sumpter, J. P. (1995). 

A variety of environmentally persistent chemicals, including some phthalate plasticizers, are weakly estrogenic. Environ. HealthPerspect. 103, 582-587. 

Kavlock, R. J., Daston, G. P., DeRosa, C., Fenner-Crisp, P., Gray, L. E., Kaattari, S., Lucier, G., Luster, M., Mac, M. J., Maczka, C., Miller, R., Moore, J., Rolland, R., Scott, G., Sheehan, D. M., Sinks, T., and Tilson, H. A. (1996). Research needs for the risk assessment of health and environmental effects of endocrine disruptors: A report of the U.S. EPAsponsored workshop. Environ. Health Perspect. 104(Suppl 4), 715-740. 

Kenney, J. A., Jr. (1988). Vitiligo. Dermatological Clinics 6, 425-434. 

Klotz, D. M., Beckman, B. S., Hill, S. M., McLachlan, J. A., Walters, M. R., and Arnold, S. F. (1996). Identification of environmental chemicals with estrogenic activity using a combination of in vitro assays. Environ. Health Perspect. 104, 1084-1089. 

Korach, K. S., Sarver, P., Chae, K., McLachlan, J. A., and McKinney, J. D. (1987). Estrogen receptor-binding activity of polychlorinated hydroxybiphenyls: Conformationally restricted structural probes. Mol. Pharmacol. 33, 120 -126. 

Krishnan, A. V., Stathis, P., Permuth, S. F., Tokes, L., and Feldman, D. (1993). Bisphenol-A: An estrogenic substance is released from polycarbonate flasks during autoclaving. Endocrinology 132, 2279-2286. 

Kuiper, G. G. J. M., Carlsson, B., Grandien, K., Enmark, E., Haggblad, J., Nilsson, S., and Gustafsson, J-A. (1997). Comparison of the ligand binding specificity and transcript tissue distribution of estrogen receptors a and 0. Endocrinology 138, 863-870. 

Kuiper, G. G. J. M., Lemmen, J. G., Carlsson, B., Corton, J. C., Safe, S. H., van der Saag, P. T., van der Burg, B., and Gustafsson, J.-A. (1998). Interaction of estrogenic chemicals and phytoestrogens with estrogen receptor 0. Endocrinology 139, 4252-4263. 

Lee, C. H., and Kim, H. J. (1994). A study on the absorption mechanisms of drugs through biomembranes. Arch. Pharm. Res. 17, 182-189. 

Medlock, K. L., Lyttle, C. R., Kelepouris, N., Newman, E. D., and Sheehan, D. M. (1991). Estradiol down-regulation of the rat uterine estrogen receptor. Proc. Soc. Exp. Biol. Med. 196, 293-300. 

Nagel, S. C., vom Saal, F. S., Thayer, K. A., Dhar, M. G., Boechler, M., and Welshons, W. V. (1997). Relative binding affinity-serum modified access (RBA-SMA) assay predicts the relative in vivo activity ofthe xenoestrogens bisphenol A and octylphenol. Environ. Health Perspect. 105, 70-76. 

Nakagawa, Y., and Moldeus, P. (1998). Mechanism of p-hydroxybenzoate ester-induced mitochondrial dysfunction and cytotoxicity in isolated rat hepatocytes. Biochem. Pharmacol. 55, 1907-1914. 

Nakai, M., Tabira, Y., Asai, D., Yakabe, Y., Shimyozu, T., Noguchi, M., Takatsuki, M., and Shimohigashi, Y. (1999). Binding characteristics of dialkyl phthalates for the estrogen receptor. Biochem. Biophys. Res. Commun. 254, 311-314. 

Nimrod, A. C., and Benson, W. H. (1996). Environmental estrogenic effects of alkylphenol ethoxylates. Crit. Rev. Toxicol. 26, 335-364. 

Okereke, C. S., Abdel-Rahman, M. S., and Friedman, M. A. (1994). Disposition of benzophenone-3 after dermal administration in male rats. Toxicol. Lett. 73, 113-122. 

Okereke, C. S., Kodry, A. M., Abdel-Rahman, M. S., Davis, R. A., and Friedman, M. A. (1993). Metabolism of benzophenone-3 in rats. Drug Metab. Dispos. 21, 788-791. 

Parker, M. G., Arbuckle, N., Dauvois, S., Danielian, P., and White, R. (1993). Structure and function of the estrogen receptor. Ann. NY Acad. Sci. 684, 119-126. 

Perez, P., Pulgar, R., Olea-Serrano, F., Villalobos, M., Rivas, M., Metzler, M., Pedraza, V., and Olea, N. (1998). The estrogenicity of bisphenol A-related diphenylalkanes with various substituents at the central carbon and the hydroxy groups. Environ. Health Perspect. 106, 167-174. 

Pompy, L., Karlin, A., Capuono, C. M., Cottrell, J. E., and Hartung, J. (1991). Paraben preservatives do not increase intracranial pressure in cats. Anesthesiology 75, 669-672. 

Rastogi, S. C., Schouten, A., De Kruijf, N., and Weijland, J. W. (1995). Contents of methyl-, ethyl-, propyl-, and benzylparaben in cosmetic products. Contact Dermatitis 32, 28-30. 

Routledge, E. J., Parker, J., Odum, J., Ashby, J., and Sumpter, J. P. (1998). Some alkyl hydroxy benzoate preservatives (parabens) are estrogenic. Toxicol. Applied Pharm. 153, 12-19. 

Roy, D., Palangat, M., Chen, C.-W., Thomas, R. D., Colerangle, J., Atkinson, A., and Yan, Z-J. (1997). Biochemical and molecular changes at the cellular level in response to exposure to environmental estrogen-like chemicals. J. Toxicol. Environ. Health 50, 1-29. 

Schwartzkopf, K. S., Stookey, J. M., Hull, P. R., and Clark, E. G. (1994). Screening of depigmenting compounds for the development of an alternate method of branding beef cattle. J. Anim. Sci. 72, 1393-1398. 

Shelby, M. D., Newbold, R. R., Tully, D. B., Chae, K., and Davis, V. L. (1996). Assessing environmental chemicals for estrogenicity using a combination of in vitro and in vivo assays. Environ. Health Perspect. 104, 1296-1300. 

Shiau, A. K., Barstad, D., Loria, P. M., Cheng, L., Kushner, P. J., Agard, D. A., and Greene, G. L. (1998). The structural basis of estrogen receptor/coactivator recognition and the antagonism of this interaction by tamoxifen. Cell 95, 927-937. 

Sonnenschein, C., and Soto, A. M. (1998). An updated review of environmental estrogen and androgen mimics and antagonists. J. Steroid Biochem. Mol. Biol. 65, 143-150. 

Soto, A. M., Justicia, H., Wray, J. W., and Sonnenschein, C. (1991). p-Nonylphenol: an estrogenic xenobiotic released from “modified” polystyrene. Environ. Health Perspect. 92, 167-173. 

Soto, A. M., Sonnenschein, C., Chung, K. L., Fernandez, M. F., Olea, N., and Serrano, F. O. (1995). The E-SCREEN assay as a tool to identify estrogens: an update on estrogenic environmental pollutants. Environ. Health Perspect. 103(Suppl 7), 113-122. 

Spyropoulos, D. V. (1998). Stability testing of the plastics additives 2,4dihydroxybenzophenone, 2,2'-dihydroxy-4-methoxybenzophenone, 2-hydroxy-4-n-octyloxybenzophenone, 2-ethoxy-2-ethyloxanilide (tinuvin 312) and 2,2'-methylenebis(4-methyl-t-tert-butylphenol) in aqueous and fatty food simulants. Food Addit. Contam. 15, 362-369. 

Sugihara, N., Shimomichi, K., and Furrino, K. (1997). Cytotoxicity of food preservatives in cultured rat hepatocytes loaded with linolenic acid. Toxicology 120, 29-36. 

Thomas, J. A., and Thomas, M. J. (1984). Biological effects of the (di-2ethylhexyl) phthalate and other phthalic esters. CRC Crit. Rev. Toxicol. 13, 283-317. 

Tong, W., Perkins, R., Strelitz, R., Collantes, E. R., Keenan, S., Welsh, W. J., Branham, W. S., and Sheehan, D. M. (1997). Quantitative-structure activity relationships (QSARs) for estrogen binding to the estrogen receptor: Predictions across species. Environ. Health Perspect. 105, 1116-1124. 

Tong, W., Lowis, D. R., Perkins, R., Chen, Y., Welsh, W. J., Goddette, D. W., Heritage, T. W., Sheehan, D. M. (1998). Evaluation of quantitative structure-activity relationship methods for large-scale prediction of chemicals binding to the estrogen receptor. J. Chem. Inf. Comput. Sci. 38, 669-677. 

Toppari, J., Larsen, J. C., Christiansen, P., Giwercman, A., Grandjean, P., Guillette, L. J., Jr, Jegou, B., Jensen, T. K., Jouannet, P., Keiding, N., Leffers, H., McLachlan, J. A., Meyer, O., Muller, J., Rajpert-De Meyts, 

E. R.-D., Scheike, T., Sharpe, R., Sumpter, J., and Skakkeboek, N. E.. (1996). Male reproductive health and environmental xenoestrogens. Environ. Health Perspect. 104(Suppl 4), 741-803. 

Tsai, M. J., and O’Malley, B. W. (1994). Molecular mechanisms of action of steroid/thyroid receptor superfamily members. Ann. Rev. Biochem. 63, 451486. 

Twist, J. N., and Zotz, J. L. (1986). Influence of solvents on paraben permeation through idealised skin model membranes. J. Soc. Cosmet. Chem. 37, 429-444. 

Van Doorn, L. G., Berenschot-Roozendaal, J., Poortman, J., Thijssen, J. H. H., and Schwarz, F. (1982). Binding characteristics of 5-androstene-30, 170diol and estradiol 170 to the cytoplasmic estrogen receptor of the immature rat uterus. J. Steroid Biochem. 16, 661-671. 

Van Faassen, I., Verweij-van Vught, A. M., Lomecky-Janousek, M. Z., Razenberg, P. P., and van der Veen, E. A. (1990). Preservatives in insulin preparations impair leukocyte function: In vitro study. Diabetes Care 13, 71-74. vom Saal, F. S., Timms, B. G., Montano, M. M., Palanza, P., Thayer, K. A., Nagel, S. C., Dhar, M. D., Ganjam, V. K., Parmigiani, S., and Welshons, W. V. (1997). Prostate enlargement in mice due to fetal exposure to low doses of estradiol or diethylstilbestrol and opposite effects at high doses. Proc. Natl. Acad. Sci. USA 94, 2056-2061. 

Waller, C. L., Oprea, T. I., Chae, K., Park, H.-K., Korach, K. S., Laws, S. C., Wiese, T. E., Kelce, W. R., and Gray, L. E., Jr. (1996). Ligand-based identification of environmental estrogens. Chem. Res. Toxicol. 9, 1240-1248. 

White, R., Jobling, S., Hoare, S. A., Sumpter, J. P., and Parker, M. G. (1994). Environmentally persistent alkylphenolic compounds are estrogenic. Endocrinology 135, 175-182. 

Wine, R. N., Li, L.-H., Barnes, L. H., Gulati, D. K., and Chapin, R. E. (1997). Reproductive toxicity of di-n-butylphthalate in a continuous breeding protocol in Sprague-Dawley rats. Environ. Health Perspect. 105, 102-107. 

Zacharewski, T. R., Meek, M. D., Clemons, J. H., Wu, Z. F., Fieldin, M. R., and Matthews. J. B. (1998). Examination of the in vitro and in vivo estrogenic activities of eight commercial phthalate esters. Toxicol. Sci. 46, 282-293.