Evaluation of a Tier I Screening Battery for Detecting Endocrine-Active Compounds (EACs) Using the Positive Controls Testosterone, Coumestrol, Progesterone, and RU486

John C. O'Connor*,1, Leonard G. Davis*, Steven R. Frame* and Jon C. Cook{dagger}

* DuPont Haskell Laboratory for Toxicology and Industrial Medicine, P.O. Box 50, Elkton Road, Newark, Delaware, 19714; and {dagger} Pfizer, Inc., Central Research, Building 274, Eastern Point Road, Groton, Connecticut 06340-8014

Received July 18, 1999; accepted December 3, 1999


    ABSTRACT
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
After previously examining 12 compounds with known endocrine activities, we have now evaluated 4 additional compounds in a Tier I screening battery for detecting endocrine-active compounds (EACs): a weak estrogen receptor (ER) agonist (coumestrol; COUM), an androgen receptor (AR) agonist (testosterone; TEST), a progesterone receptor (PR) agonist (progesterone; PROG), and a PR antagonist (mifepristone; RU486). The Tier I battery incorporates 2 short-term in vivo tests (5-day ovariectomized female battery; 15-day intact male battery) and an in vitro yeast transactivation system (YTS). The Tier I battery is designed to identify compounds that have the potential to act as agonists or antagonists to the estrogen, androgen, progesterone, or dopamine receptors; steroid biosynthesis inhibitors (aromatase, 5{alpha}-reductase, and testosterone biosynthesis); or compounds that alter thyroid function. In addition to the Tier I battery, a 15-day dietary restriction experiment was performed using male rats to assess confounding due to treatment-related decreases in body weight. In the Tier I female battery, TEST administration increased uterine weight, uterine stromal cell proliferation, and altered hormonal concentrations (increased serum testosterone [T] and prolactin [PRL]; and decreased serum FSH and LH). In the male battery, TEST increased accessory sex gland weights, altered hormonal concentrations (increased serum T, dihydrotestosterone [DHT], estradiol [E2], and PRL; decreased serum FSH and LH), and produced microscopic changes of the testis (Leydig cell atrophy and spermatid retention). In the YTS, TEST activated gene transcription in the yeast containing the AR or PR. In the female battery, COUM administration increased uterine weight, uterine stromal cell proliferation, and uterine epithelial cell height, and increased serum PRL concentrations. In the male battery, COUM altered hormonal concentrations (decreased serum T, DHT, E2; increased serum PRL) and, in the YTS, COUM activated gene transcription in the yeast containing the ER. In the female battery, PROG administration increased uterine weight, uterine stromal cell proliferation, and uterine epithelial cell height and altered hormonal concentrations (increased serum progesterone and decreased serum FSH and LH). In the male battery, PROG decreased epididymis and accessory sex gland weights, altered hormonal concentrations (decreased serum T, PRL, FSH, and LH; increased serum progesterone and E2), and produced microscopic changes of the testis (Leydig cell atrophy). In the YTS, PROG activated gene transcription in the yeast containing the AR or PR. In the female battery, RU486 administration increased uterine weight and decreased uterine stromal cell proliferation. In the male battery, RU486 decreased epididymis and accessory sex gland weights and increased serum FSH and LH concentrations. In the YTS, RU486 activated gene transcription in the yeast containing the ER, AR, or PR. Dietary restriction data demonstrate that confounding due to decrements in body weight are not observed when body weight decrements are 10% or less in the Tier I male battery. In addition, minimal confounding is observed at body decrements of 15% (relative liver weight, T3, and T4). Hence, compounds can be evaluated in this Tier I at levels that produce a 10% decrease in body weight without confounding of the selected endpoints. Using the responses obtained for all the endpoints in the Tier I battery, a distinct "fingerprint" was produced for each type of endocrine activity against which compounds with unknown activity can be compared. These data demonstrate that the described Tier I battery is useful for identifying EACs and they extend the compounds evaluated to 16.

Key Words: screening; Tier I battery; rats; hormone concentrations; organ weights; endocrine-active compounds (EACs); testosterone; coumestrol; progesterone; RU486.


    INTRODUCTION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Under the current legislative requirements of the Food Quality Protection Act of 1996 and the Safe Drinking Water Act of 1996, the United States Environmental Protection Agency (U.S. EPA) was mandated to implement a testing strategy to screen chemicals and pesticides for endocrine activity, starting in August 1999. In response, the U.S. EPA established the Endocrine Disruptor Screening and Testing Advisory Committee (EDSTAC, 1998Go). EDSTAC's final recommendation was to implement a screening strategy to detect compounds that are agonists or antagonists to the estrogen or androgen receptors or steroid biosynthesis inhibitors, or that alter thyroid function (EDSTAC, 1998Go).

As part of our research program to develop effective methods to screen for endocrine-active compounds (EACs), we have initiated a validation of a tiered testing scheme using 16 model EACs (Cook et al., 1997bGo; O'Connor et al., 1998aGo,bGo, 1999aGo, O'Connor et al., bGo,cGo). The rationale for this approach has been previously described and was based on 10 years of experience in identifying weak EACs (Cook et al., 1997bGo). Our Tier I testing scheme incorporates 2 short-term in vivo tests (5-day ovariectomized female battery; 15-day intact male battery) and an in vitro yeast transactivation system (YTS) for identifying compounds that alter endocrine homeostasis. The Tier I battery that we are currently validating is designed to identify a broader spectrum of EACs than was proposed by EDSTAC. Our Tier I battery is designed to identify compounds that have the potential to act as agonists or antagonists to the estrogen, androgen, progesterone, or dopamine receptors, steroid biosynthesis inhibitors (aromatase, 5{alpha}-reductase, and testosterone biosynthesis), or compounds that alter thyroid function.

We have previously used 12 compounds with known endocrine activities to examine the usefulness of the Tier I screening battery for identifying EACs with various endocrine activities (O'Connor et al., 1998aGo,bGo, 1999aGo, bGo,cGo). In the current study, 4 additional EACs with diverse endocrine activities have been examined using this Tier I screening battery. In order to characterize this screening battery's ability to detect a variety of endocrine activities, we evaluated a weak estrogen-receptor (ER) agonist, coumestrol (COUM) (Markaverich et al., 1995Go), an androgen-receptor (AR) agonist, testosterone (TEST) (Wilson, 1990Go), a progesterone-receptor (PR) agonist, progesterone (PROG) (Graham and Clarke, 1997Go), and a PR antagonist, mifepristone (RU486) (Philibert et al., 1984Go). Characterization of the responses obtained in Tier I for each of these compounds will test the hypothesis that distinct "fingerprints" of endocrine activity can be detected for model EACs.


    MATERIALS AND METHODS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Test materials.
The following materials were obtained from the Sigma Chemical Company (St. Louis, MO): ammonium sulfate, 5-bromo-2'-deoxyuridine (BrdU), 2-mercaptoethanol, o-nitrophenyl ß-o-galactopyranoside (ONPG), PROG, sodium bicarbonate, sodium dodecyl sulfate (SDS), sodium phosphate, and TEST. All other materials were obtained from the following manufacturers: COUM, Acros Chemicals, a division of Fisher Scientific (Pittsburgh, PA); RU486, SiniWest Holdings, Inc. (San Diego, CA); Certified Rodent Diet #5002, PMI®Feeds, Inc. (St. Louis, MO); osmotic minipumps model 2ML1, Alza Corporation (Palo Alto, CA); methylcellulose and potassium chloride, Fisher Scientific (Springfield, NJ); dextrose, J.T. Baker, Inc. (Phillipsburg, NJ); oxalyticase, Enzogenetics, Inc. (Corvallis, OR); luteinizing hormone (LH; catalog #RPA.552), prolactin (PRL; catalog #RPA.553), thyroid stimulating hormone (TSH; catalog #RPA.554), and follicle stimulating hormone (FSH; catalog #RPA.550) radioimmunoassay (RIA) kits, Amersham Corp. (Arlington Heights, IL); testosterone (T; catalog #TKTT5), estradiol (E2; catalog #KE2D5), tri-iodothyronine (T3; catalog #TKT35), and thyroxine (T4; catalog #TKT45) RIA kits, Diagnostic Products Corp. (Los Angeles, CA); rT3 RIA kits (catalog #10834), Polymedco Inc. (Cortlandt Manor, NY); and, dihydrotestosterone (DHT; catalog #DSL-9600) RIA kit, Diagnostic Systems Laboratories (Webster, TX).

Tier I Battery Studies
Test species.
Male and female Sprague-Dawley (Crl:CD®(SD)IGS BR) rats were acquired from Charles River Laboratories, Inc. (Raleigh, NC). Malerats were approximately 63 days old upon arrival. Female rats, approximately 42 days old, were ovariectomized on the day of shipment (41 days old). Upon arrival, rats were housed in stainless steel, wire-mesh cages suspended above cage boards and were fed irradiated PMI®Feeds, Inc., Certified Rodent Diet #5002 and provided with tap water (United Water Delaware) ad libitum. Animal rooms were maintained on a 12-h light/dark cycle (fluorescent light), a temperature of 23 ± 2°C, and a relative humidity of 50% ± 10%.

After a quarantine period of ~1 week, rats that displayed adequate weight gain and freedom from clinical signs were divided by computerized, stratified randomization into 5 groups of 20 females and 5 groups of 15 males, so that there would be no statistically significant differences among group body-weight means. Within each group of 20 female rats, 14 animals were designated for biochemical/hormonal evaluation and 6 rats were designated for cell proliferation/morphometric evaluation. All male rats were designated for biochemical/hormonal evaluation.

Study design.
All rats were weighed daily and cage-side examinations were performed to detect moribund or dead rats. At each weighing, rats were individually handled and examined for abnormal behavior or appearance. All test compounds were prepared in 0.25% methylcellulose vehicle and administered by intraperitoneal injection at approximately 0900 h daily. The intraperitoneal route was selected to enhance the sensitivity of the assay (i.e., reduce false negative responses) and to facilitate potency comparisons by eliminating absorption as a variable (Cook et al., 1997bGo). Because there will be several routes of potential exposure of humans (e.g. dermal, oral, or inhalation), and unique exposure routes of wildlife (e.g., gills, eggs), the intraperitoneal route was selected to maximize sensitivity, realizing that subsequent studies would be required to determine whether sufficient exposure could occur via the most relevant route. Male rats were dosed for 15 days and euthanized on the morning of test day +15. Female rats were dosed for 4 days and euthanized on the morning of test day +5. The following compounds were used in the male and female in vivo battery: TEST (0.5, 2, 10, and 20 mg/kg/day), COUM (0.1, 0.5, 1, and 2.5 mg/kg/day), PROG (1, 10, 50, and 100 mg/kg/day), and RU486 (1, 10, 50, and 100 mg/kg/day). Doses were selected in order to obtain the maximal pharmacologic effect for each compound and/or not exceed the maximum tolerated dose (MTD), defined as a 10% difference in final body weight from the ad libitum control group, as determined in range-finder studies. The dose volume was 2.0-ml/kg body weight.

In addition, a dietary restriction experiment was performed using male rats to determine which endpoints were body weight-dependent, in order to evaluate potential confounding from treatment-related decreases in body weight or body weight gain. A dietary restriction experiment was previously conducted for the in vivo female battery (O'Connor et al., 1996Go). Male rats were given a range of PMI®Feeds, Inc., Certified Rodent Diet #5002 meal in order to produce different levels of weight loss. The levels of diet restriction were ad libitum control, or, in g feed/day, 22, 19, 16, or 13. Levels of diet restriction were based on calculated food consumption from a range-finder study. Daily food consumption levels for the dietary restriction experiment were targeted to achieve decreases in body weight to 95, 90, 85, and 80% of the ad libitum control. The endpoints included organ weights, serum hormone concentrations, and histopathology of the testis, epididymis, and thyroid gland.

Cell proliferation.
Six female rats from each group were designated for cell proliferation analyses. The day prior to study start (test day –1), rats were anesthetized using isoflurane and implanted (subcutaneously) with Alzet osmotic pumps loaded with 20 mg/ml BrdU dissolved in 0.5 N sodium bicarbonate buffer. On test day +5, the rats were sacrificed and the uteri were trimmed, affixed to dental wax, fixed for 2 h in Bouin's solution, routinely processed, paraffin embedded, sectioned, and stained for immunohistochemical analysis of BrdU incorporation into DNA, or sectioned and evaluated for epithelial cell height. Uterine stromal cell proliferation and uterine epithelial cell height evaluation were performed as previously described (O'Connor et al., 1996Go).

Estrous conversion.
Rats assigned to the biochemical/hormonal subset were evaluated for the stage of estrous by vaginal cytology. Vaginal washes were collected once daily by repeated pipetting of 75 µl of 0.9% sterile saline into the vagina. Slides were air dried, stained by the Wright-Geimsa method, and evaluated for conversion out of diestrus (Davis, 1993Go).

Pathological evaluations (female rats).
Between 0700 and 1000 h on the morning of test day +5 (approximately 24 h after the last administered dose), rats were anesthetized using CO2 and euthanized by exsanguination. Blood was collected from the descending vena cava and serum prepared for hormonal analyses. The presence of fluid in the uterine horns was recorded as a gross observation. The thyroid glands from the biochemical/hormonal subset were placed in formalin fixative and examined microscopically. Uteri from the cell proliferation subset were dissected, weighed, and fixed in Bouin's solution. Processed tissues were embedded in paraffin, sectioned at 5 µm, and affixed to slides for staining. Uterine stumps from all rats were saved in 10% neutral buffered formalin and processed to confirm the absence of ovarian tissue by the pathologist. If ovarian tissue was detected, animals were excluded from analyses.

Pathological evaluations (male rats).
On the morning of test day +15, rats were injected with the final dose of test compound approximately 2 h prior to sacrifice. Between 0700 and 1000 h, rats were anesthetized using CO2 and euthanized by exsanguination. Blood was collected from the descending vena cava and serum prepared for hormonal analyses. The liver, testes, epididymis, prostate, seminal vesicles, and accessory sex gland (ASG) unit were weighed and organ weights calculated relative to body weight. Weights for the testis, epididymides, and seminal vesicles were paired weights. The epididymis and thyroid gland were placed in formalin fixative, the testes were placed in Bouin's fixative, and all were examined microscopically.

Hormonal measurements.
Blood was collected at the time of euthanization from all animals. Serum was prepared and stored between –65°C and –85°C until analyzed for serum hormone concentrations. For males, serum T, E2, DHT, LH, FSH, PRL, TSH, T3, and T4 concentrations were measured by commercially available RIA kits. For the females, serum LH, FSH, and PRL concentrations were measured by commercially available RIA kits.

In Vitro Yeast Studies
Binding of TEST, COUM, PROG, and RU486 to the human ER, AR, and PR was examined using a yeast reporter gene system as previously described (Cook et al., 1997bGo; O'Connor et al., 1998bGo). Receptor-specific compounds were previously used to validate this system (O'Connor et al., 1998bGo); these included E2, DHT, and progesterone as the primary ligand for the 3 receptor systems, respectively. Briefly, stable yeast (Saccharomyces cerevisiae) transformants from single-plated colonies were grown overnight at 30°C with orbital shaking at 300 rpm in a selective medium containing yeast nitrogen base without amino acids plus ammonium sulfate and dextrose.

Assays were started when the cells reached an OD600nm greater than 1.00. One hundred microliters of cell suspension, diluted to achieve an OD600nm of 1.00, were placed into each well of a 96-well plate and 100 µl selective medium was added. Two microliters of each test compound, prepared in methanol, was then added to each well to develop a dose-response assay in triplicate. These cultures were incubated for 3 h at 30°C with rocking.

After incubation, cells were read at OD600nm to evaluate the effect of the compounds on cell growth before adding 100 µl of assay buffer (60 mM Na2HPO4, 40 mM NaH2PO4, 10 mM KCl, 1 mM MgSO4, 2 mg/ml ONPG, 50 mM 2-mercaptoethanol, 0.1% SDS, and 200 U/µl oxalyticase) to each well to assay for ß-galactosidase induction. The change in concentration of o-nitrophenol, the yellow product that results from ß-galactosidase activity cleaving the galactopyranoside, after a period of one h, was determined by reading each well at OD410nm. Nonspecific colorimetric distortions (i.e., debris) were evaluated by also reading each well at OD570nm.

For competitive inhibition testing of the test compounds, approximately a one-half maximal concentration of E2 for the YTS containing the ER, DHT for the YTS containing the AR, or PROG for the YTS containing the PR was included in each well of the dose response for each test compound, respectively. The samples were then treated as above except that the percent inhibition was calculated for each test compound.

Statistical Analyses
Mean final body weights and organ weights were analyzed by a one-way analysis of variance (ANOVA). When the corresponding F test for differences among test-group means was significant, pairwise comparisons between test and control groups were made with Dunnett's test. Bartlett's test for homogeneity of variances was performed and, when significant (p < 0.005), was followed by nonparametric procedures (Dunn's test). Serum hormone concentrations, cell proliferation indices, and uterine morphometry measurements were analyzed using Jonckheere's test for trend. If a significant dose-response trend was detected, data from the top-dose group was excluded and the test repeated until no significant trend was detected. Uterine fluid imbibition and estrous conversion data were analyzed by Fisher's test. Except for Bartlett's test, all other significance was judged at p < 0.05. EC50 values were determined using a 4-parameter logistic function (Origin 4.1, MicroCal Software, Inc., Northhampton, MA). EC50 values represent the dose at which 50% of the maximal response occurred.


    RESULTS
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
In Vivo Female Battery
Final body and organ weights, uterine fluid imbibition, and estrous conversion (Table 1Go).
Mean final body weights and relative liver weights were not affected by TEST, COUM, PROG, or RU486 administration at the dosages tested. Absolute uterine weight was numerically increased in rats treated with 10 mg/kg/day and significantly increased at 20 mg/kg/day TEST (134 and 217% of control, respectively). Absolute uterine weight was increased in a dose-dependent manner in rats treated with COUM, and was significantly increased at 0.5, 1.0, and 2.5 mg/kg/day, with the greatest increase at 2.5 mg/kg/day (218% of control). Absolute uterine weight was significantly increased in rats treated with 50 or 100 mg/kg/day PROG, with the greatest increase at 100 mg/kg/day (131% of control). Absolute uterine weight was significantly increased in rats treated with 100 mg/kg/day RU486 (113% of control). TEST administration resulted in one rat in the 20 mg/kg/day group (incidence of 7%) having uterine fluid imbibition without the concomitant estrous conversion. Uterine fluid imbibition or estrous conversion was not observed in rats treated with COUM, PROG, or RU486.


View this table:
[in this window]
[in a new window]
 
TABLE 1 In Vivo Female Battery: Final Body Weight (bw), Absolute Uterine Weight, Uterine Fluid Imbibition, and Estrous Conversion
 
Uterine stromal cell proliferation and epithelial cell height (Table 2Go).
TEST and COUM administration caused dose-dependent increases in uterine stromal cell proliferation that were significantly increased at 10 and 20 mg/kg/day for TEST, and were significantly increased at all dosages for COUM, with the greatest increases at the highest dosages (1057 and 697% of control for TEST and COUM, respectively). PROG administration increased uterine stromal cell proliferation at dosages of 10, 50, and 100 mg/kg/day, with statistically significant increases at 50 and 100 mg/kg/day, and the greatest increase was at the highest dosage (1689% of control). RU486 administration produced a statistically significant decrease in uterine stromal cell proliferation at the highest dosage (64% of control).


View this table:
[in this window]
[in a new window]
 
TABLE 2 In Vivo Female Battery: Uterine Stromal Cell Proliferation and Uterine Epithelial Cell Height
 
Uterine epithelial cell height was increased in a dose-dependent manner and was significantly increased at all dosages in rats treated with COUM, with the greatest increase at the highest dosage (148% of control). PROG administration caused a statistically significant increase in uterine epithelial cell height at dosages of 10 mg/kg/day and higher, with the greatest increase at the highest dosage (181% of control). Uterine epithelial cell height was unchanged in rats treated with TEST or RU486 at the dosages tested.

Serum hormone concentrations (Table 3Go).
Serum was collected for hormonal analyses approximately 24 h after the last administered dose. TEST administration produced a dose-dependent increase in serum PRL concentrations that was significantly increased at 10 and 20 mg/kg/day, with the greatest increase at the highest dosage (379% of control). Serum FSH concentrations were significantly decreased at the highest dosage and serum LH concentrations were significantly decreased at 10 and 20 mg/kg/day, with the greatest decreases at the highest dosages (81% and 36% of control for FSH and LH, respectively). Serum T concentrations, undetectable in the control group, were significantly increased at 20 mg/kg/day to 2.3 ng/ml.


View this table:
[in this window]
[in a new window]
 
TABLE 3 In Vivo Female Battery: Serum Hormone Concentrations
 
COUM administration caused a statistically significant increase in serum PRL concentrations at 2.5 mg/kg/day (190% of control). Serum FSH concentrations were significantly increased at 1 and 2.5 mg/kg/day, with the greatest increase at 1 mg/kg/day (122% of control). Serum LH concentrations were not affected by COUM administration at dosages as high as 2.5 mg/kg/day.

PROG administration caused a dose-dependent decrease in serum FSH concentrations that was significantly decreased at all dosages, with the greatest decrease at the highest dosage (73% of control). Serum LH concentrations were significantly decreased at 100 mg/kg/day (79% of control). Serum P4 concentrations were significantly increased at 50 and 100 mg/kg/day, with the greatest increase at the highest dosage (223% of control). Serum PRL concentrations were not affected by PROG administration at dosages as high as 100 mg/kg/day. In contrast, RU486 administration did not affect serum PRL, FSH, or LH concentrations at dosages as high as 100 mg/kg/day.

Histopathology (data not shown).
The thyroid gland was examined microscopically to identify compounds that target the thyroid gland. No microscopic changes were observed in the thyroid glands from animals treated with TEST, COUM, PROG, or RU486.

In Vivo Male Battery: Dietary Restriction Experiment
Final body weights (Table 4Go).
Daily food-consumption levels for the dietary restriction experiment were targeted to achieve decreases in body weight to 95, 90, 85, and 80% of the ad libitum control. Final body weights were significantly decreased at all levels of dietary restriction, achieving final body weights of 90, 85, 79, and 74% of the ad libitum control at 22, 19, 16, and 13 grams of feed/day, respectively.


View this table:
[in this window]
[in a new window]
 
TABLE 4 In Vivo Male Battery: Effect of Dietary Restriction on Final Body and Organ Weights
 
Absolute organ weights (Table 4Go).
Absolute liver weights were decreased in a dose-dependent manner and were significantly decreased in all treatment groups, with the greatest decrease (61% of control) at a dietary restriction level that achieved a final body weight of 74% of the ad libitum control. Absolute epididymis and prostate weights were significantly decreased when final body weights were decreased to 74% of the ad libitum control (93 and 74% of control for epididymides and prostate weights, respectively). Absolute ASG unit weight and seminal vesicle weight were significantly decreased when final body weights were decreased to <= 85% of the ad libitum control, with the greatest decreases (70 and 69% of control for ASG and seminal vesicle weights, respectively) at a dietary restriction level that achieved a final body weight of 74% of the ad libitum control. Absolute testis weights were unaffected by dietary restriction at levels that resulted in a final body weight decrement of up to a 26% when compared to the ad libitum control.

Relative organ weights (Table 4Go).
Relative liver weights were decreased in a dose-dependent manner and were significantly decreased in all treatment groups, with the greatest decrease (82% of control) at a dietary restriction level that achieved a final body weight of 74% of the ad libitum control. Relative testis and epididymis weights were increased in a dose-dependent manner and were significantly increased at all levels of dietary restriction, with the greatest increases (132 and 125% of control for testis and epididymis weights, respectively) at a dietary restriction level that achieved a final body weight of 74% of the ad libitum control. Relative ASG weight, and the individual component weights of the ASG (relative seminal vesicles and prostate) were unaffected by dietary restriction at levels that resulted in a final body weight decrement of up to a 26% when compared to the ad libitum control.

Histopathology (data not shown).
No microscopic changes were observed in the testes, epididymides, or thyroid (O'Connor et al., 1999cGo) by dietary restriction at levels that resulted in a final body weight decrement of up to 26% when compared to the ad libitum control.

Serum hormone concentrations (Table 5Go).
Serum T and DHT concentrations were statistically significantly decreased when final body weights were <= 79% of the ad libitum control, with the greatest decreases (26 and 37% of control for T and DHT, respectively) when final body weights were 74% of the ad libitum control. Serum PRL concentrations were numerically decreased (considered biologically significant) when final body weights were <= 79% of the ad libitum control, and were statistically significantly decreased (56% of control) when final body weights were 74% of the ad libitum control. Serum TSH concentrations were numerically decreased (considered biologically significant) when final body weights were <= 79% of the ad libitum control, and were statistically significantly decreased (62% of control) when final body weight were 74% of the ad libitum control. Serum T3 and T4 concentrations were significantly decreased when final body weights were <= 85% of the ad libitum control, with the greatest decreases (75 and 72% of control for T3 and T4, respectively) at a dietary restriction level that achieved a final body weight of 74% of the ad libitum control. Serum FSH and LH concentrations were unaffected by dietary restriction at levels that resulted in a final body weight decrement of up to a 26% when compared to the ad libitum control. Serum E2 concentrations were unaffected by dietary restriction at levels that resulted in a final body weight decrement of up to a 21% when compared to the ad libitum control. Serum E2 concentrations were not measured at the maximum level of dietary restriction due to a lack of serum.


View this table:
[in this window]
[in a new window]
 
TABLE 5 In Vivo Male Battery: Effect of Dietary Restriction on Serum Hormone Concentrations
 
In Vivo Male Battery
Final body and liver weights (Table 6Go).
Statistically significant decreases in mean final body weights were observed in rats treated with 100 mg/kg/day PROG (91% of control) and in rats treated with 50 or 100 mg/kg/day RU486 (93 and 90% of control, respectively), and were unchanged in rats treated with TEST or COUM. Relative liver weights were significantly decreased in rats treated with 20 mg/kg/day TEST (89% of control), and significantly increased in rats treated with 100 mg/kg/day PROG or 100 mg/kg/day RU486 (107 and 110% of control, respectively). Relative liver weights were unchanged in rats treated with COUM at the dosages tested.


View this table:
[in this window]
[in a new window]
 
TABLE 6 In Vivo Male Battery: Final Body Weights and Organ Weights
 
Reproductive organ weights (Table 6Go).
Absolute testis weights were not affected by treatment with TEST, COUM, PROG, or RU486. Absolute epididymis weights were decreased in a dose-dependent manner by PROG treatment and were significantly decreased in rats treated with 50 or 100 mg/kg/day, with the greatest decreases at the highest dosage (83% of control). RU486 administration caused a statistically significant decrease in epididymis weight at 100 mg/kg/day (89% of control). Absolute epididymis weights were not affected by treatment with TEST or COUM at the dosages tested.

Relative ASG (prostate, seminal vesicles with fluid, and coagulating glands) weights were numerically increased at all dosages of TEST and were significantly statistically increased at 0.5, 10, and 20 mg/kg/day, with the greatest increase at the highest dosage (189% of control). In addition, individual component weights of the ASG were also increased by TEST administration. For example, relative seminal vesicle weights were numerically increased at all dosages of TEST, and were significantly increased at 10 and 20 mg/kg/day, with the greatest increase at 20 mg/kg/day (197% of control). Similarly, relative prostate weights were numerically increased at all dosages of TEST, and were significantly increased at 0.5, 10, and 20 mg/kg/day, with the greatest increase at 20 mg/kg/day (171% of control). Relative ASG weights were decreased in a dose-dependent manner by PROG treatment and were statistically significantly decreased at 10, 50, and 100 mg/kg/day, with the greatest decrease at the highest dosage (50% of control). In addition, individual component weights of the ASG were also decreased by PROG administration. Relative seminal vesicle weights were decreased in a dose-dependent manner by PROG administration and were statistically significantly decreased at 10, 50, and 100 mg/kg/day, with the greatest decrease at the highest dosage (39% of control). Relative prostate weights were statistically significantly decreased at 50 and 100 mg/kg/day PROG, with the greatest decrease at 50 mg/kg/day (72% of control). Relative ASG weight was significantly decreased by RU486 treatment at 100 mg/kg/day (74% of control). In addition, relative seminal vesicle weights were significantly decreased at 100 mg/kg/day by RU486 administration (73% of control). Relative prostate weights were not affected by RU486 treatment at the dosages tested. Neither the relative ASG unit weight nor the individual component weights of the ASG were affected by COUM treatment at the dosages tested.

Histopathology of the reproductive organs.
Minimal to moderate atrophy of Leydig cells was present in all groups treated with TEST. Low incidences of minimal retention of late-step spermatids were also present in the 10- and 20-mg/kg/day TEST groups (incidence of 3/15 and 6/15 for the 10 and 20 mg/kg/day groups, respectively) (Fig. 1Go). Atrophied Leydig cells had small, hyperchromic nuclei and scant, pale-staining cytoplasm. Spermatid retention was characterized by adluminal or basilar accumulation of late-step elongated spermatids, most commonly in stages IX–XII tubules. No microscopic changes were observed in the epididymides of rats treated with TEST.



View larger version (135K):
[in this window]
[in a new window]
 
FIG. 1. Testes from a control rat (1a) and a rat treated with 20 mg/kg/day testosterone (1b). The interstitium of the testosterone-treated rats is atrophied (asterisks), and late-step spermatids are retained at various levels within the seminiferous epithelium (arrows). HE; x250.

 
Minimal Leydig cell atrophy was present in 6/15 and 13/15 male rats administered 50 or 100 mg/kg/day PROG, respectively. Leydig cell atrophy was morphologically similar, although less severe, than that noted in rats treated with TEST. No microscopic changes were observed in the epididymides of rats treated with PROG or in the testes or epididymides of rats treated with COUM or RU486.

Reproductive hormone concentrations (Table 7Go).
Serum was collected for hormonal analyses approximately 2 h after the last administered dose. TEST, an AR agonist, caused a statistically significant increase in serum T and DHT concentrations at all dosages, with the greatest increases at the highest dosage (2978 and 4300% of control, respectively). Serum E2 concentrations were numerically increased at 10 and were statistically significantly increased at 20 mg/kg/day (227% of control). Serum PRL concentrations were numerically increased at all dosages and were significantly increased at 20 mg/kg/day, with the greatest increase at the highest dosage (152% of control). Serum FSH and LH concentrations were significantly decreased at all dosages, with the greatest decreases at 20 mg/kg/day for FSH and 10 mg/kg/day for LH (69 and 46% of control for FSH and LH, respectively).


View this table:
[in this window]
[in a new window]
 
TABLE 7 In Vivo Male Battery: Reproductive Hormone Concentrations
 
COUM, a weak ER agonist, caused a statistically significant decrease in serum T and DHT concentrations at 1 and 2.5 mg/kg/day, with the greatest decreases at 2.5 mg/kg/day for T (36% of control) and at 1 mg/kg/day for DHT (47% of control). Serum E2 concentrations were significantly decreased at 0.5, 1, and 2.5 mg/kg/day, with the greatest decrease at 0.5 mg/kg/day (78% of control). Serum PRL concentrations were significantly increased at the highest dosage (265% of control). Serum FSH and LH concentrations were not affected by COUM treatment at dosages as high as 2.5 mg/kg/day.

PROG, a PR agonist, caused a statistically significant decrease in serum T concentrations at 100 mg/kg/day (45% of control). Serum E2 concentrations were numerically increased at all dose levels and were significantly increased at 10, 50, and 100 mg/kg/day, with the greatest increase at the highest dosage (190% of control). Serum DHT concentrations were significantly increased at 100 mg/kg/day (152% of control). Serum PRL concentrations were numerically decreased at all dosages and were significantly decreased at the 10-, 50-, and 100-mg/kg/day dosages, with the greatest decrease at 50 mg/kg/day (40% of control). Serum FSH and LH concentrations were significantly decreased at 50 and 100 mg/kg/day, with the greatest decreases at the highest dosage (62 and 53% of control for FSH and LH, respectively). Serum P4 concentrations were increased in a dose-dependent manner and were statistically significantly increased at all dosages, with the greatest increase at 100 mg/kg/day (39,750% of control).

RU486, a PR antagonist, caused statistically significant increases in serum FSH concentrations at 10, 50, and 100 mg/kg/day, with the greatest increase at the highest dosage (122% of control). Serum LH concentrations were significantly increased at 50 and 100 mg/kg/day, with the greatest increase at the highest dosage (143% of control). Serum T, E2, DHT, and PRL concentrations were not affected by RU486 treatment at dosages as high as 100 mg/kg/day.

Thyroid hormone concentrations (Table 8Go) and histopathology.
Serum thyroid hormone analyses (TSH, T3, and T4) were performed in order to evaluate the Tier I male battery for its ability to detect compounds that alter thyroid hormone homeostasis. TEST administration caused a statistically significant decrease in serum T3 concentrations at all dosages, with the greatest decrease at the highest dosage (62% of control). Serum T4 concentrations were numerically decreased at all dosages and were significantly decreased at 2, 10, and 20 mg/kg/day, with an equivalent decrease at 10 and 20 mg/kg/day (0% of control). Serum TSH concentrations were not affected by TEST administration at the dosages tested. No microscopic changes were observed in the thyroid glands from animals treated with TEST.


View this table:
[in this window]
[in a new window]
 
TABLE 8 In Vivo Male Battery: Thyroid Hormone Concentrations
 
COUM administration caused a statistically significant decrease in serum T4 concentrations at 1 and 2.5 mg/kg/day, with the greatest decrease at the highest dosage (80% of control). Serum TSH and T3 concentrations were not affected by COUM administration at the dosages tested. No microscopic changes were observed in the thyroid glands from animals treated with COUM.

PROG administration caused a dose-dependent decrease in serum TSH concentrations that was statistically significantly at 50 and 100 mg/kg/day, with the greatest decrease at 100 mg/kg/day (66% of control). Serum T3 and T4 concentrations were not affected by PROG administration at the dosages tested. No microscopic changes were observed in the thyroid glands from animals treated with PROG.

RU486 administration caused a statistically significant increase in TSH at 100 mg/kg/day (156% of control), and a statistically significant decrease in T3 and T4 concentrations at 10, 50, and 100 mg/kg/day, with the greatest decrease at the highest dosage (55% and 61% of control for T3 and T4, respectively). Rats given 50 or 100 mg/kg/day RU486 had low incidences (3/15 and 5/15, respectively) of minimal to mild depletion of colloid in the thyroid gland. Colloid depletion was characterized by pale staining or absence of colloid in thyroid follicles. In some affected rats, equivocal hypertrophy of follicular epithelium was also present but was difficult to definitively distinguish from controls.

In Vitro YTS
In the YTS, yeast containing constructs that express human ER, AR, or PR were used to evaluate TEST, COUM, PROG, and RU486 for their ability to bind to the receptor and to activate a response element driving an inducible reporter (ß-galactosidase). During the testing of each compound, a positive control (E2 for ER, DHT for AR, and PROG for PR) was included in each experiment, which served as a reference point for each day's maximum response when comparing test substances across assays, and all results were calculated as a percentage of the maximal response. No compound-related effects on growth were detected with any of the test compounds in our 3-h assay.

In the YTS containing the ER (Fig. 2AGo), COUM and RU486 activated transcription, while TEST and PROG did not. In the YTS containing the ER, 17ß-estradiol, the endogenous ER ligand used for determination of the maximal response, had an EC50 value of 1.0 x 10–9 M. COUM, a weaker ER agonist than 17ß-estradiol, activated transcription via ER binding, with an EC50 value of 2.0 x 10–6 M, and RU486, a PR antagonist, had an EC50 value of 8.7 x 10–3 M. In the YTS containing the AR (Fig. 2BGo), DHT, an endogenous AR ligand used for determination of the maximal response, had an EC50 value of 2.4 x 10–9 M. In the YTS containing the AR, TEST, PROG, and RU486 activated transcription with EC50 values of 1.2 x 10–8 M, 5.2 x 10–6 M, and 2.1 x 10–3 M, respectively. In the YTS containing the PR (Fig. 2CGo), PROG, the endogenous PR ligand used for determination of the maximal response and one of the test compounds, had an EC50 value of 1.6 x 10–8 M. In the YTS containing the PR, TEST and RU486 also activated transcription with EC50 values of 3.3 x 10–6 M and 8.2 x 10–3 M, respectively. COUM did not activate transcription of the AR or PR. In the YTS competition assays, none of the test compounds was active in the ER competition assay. Only RU486 was active in that assay, although it did not inhibit the DHT response but rather, increased it, with an EC50 value of 3.1 x 10–3 M (data not shown).



View larger version (15K):
[in this window]
[in a new window]
 
FIG. 2. Activity of 17ß-estradiol, testosterone, dihydrotestosterone, coumestrol, progesterone, and RU486 in the YTS. Yeast cells expressing estrogen (A), androgen (B), or progesterone (C) receptor were exposed to increasing concentrations of the test compounds, incubated for 3 h at 30°C, and assayed for ß-galactosidase activity after a 1h incubation with the colorimetric substrate. The percent of the maximum response on the day of testing (determined by 17ß-estradiol for the ER, dihydrotestosterone for the AR, or progesterone for the PR) for indicated doses of each test compound are shown.

 

    DISCUSSION
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
In this report, we have summarized the results of our ongoing Tier I screening battery using 4 well-characterized EACs. We have previously described the results for 12 compounds using this Tier I battery (Cook et al., 1997bGo; O'Connor et al., 1998aGo,bGo, 1999aGo, O'Connor et al., bGo,cGo). The 2 main goals of this validation are to test the hypothesis that distinct "fingerprints" of endocrine activities can be identified for each EAC, and to evaluate which endpoints should be included in a final Tier I-type screen. In order to accomplish these goals, the responses observed for each of the positive controls were compared to the expected responses based on the published literature.

In Vivo Male Battery: Dietary Restriction Experiment
To investigate potential confounding secondary-to-body-weight decrements, a dietary restriction experiment was conducted in male rats. A similar experiment has been previously performed for the in vivo female battery (O'Connor et al., 1996Go) and for the thyroid endpoints evaluated in the in vivo male battery (O'Connor et al., 1999cGo). In the male dietary restriction experiment, liver weights were affected by dietary restriction at levels that resulted in final body weight decrements of 90% of ad libitum control or greater. It is generally accepted that liver weight is body weight-dependent, and that expression on a relative-to-body-weight basis will correct for body weight decrements (Feron et al., 1973Go). Surprisingly, both absolute and relative liver weights were affected by dietary restriction, although relative liver weight did correct for most of the body-weight decrement. These data demonstrate that body-weight decrements cannot be totally corrected by expression on a relative-to-body-weight basis. For the reproductive organ weights, the data from the dietary-restriction experiment illustrate that testis and epididymis weights should be analyzed on an absolute-weight basis, while ASG and the individual component weights of the ASG should be analyzed on a relative (to body weight)-weight basis. The only exception to this generalization was a slight decrease in absolute epididymis weight by dietary restriction at levels that resulted in final body-weight decrements of 74% of ad libitum control or greater. However, by targeting an MTD where body-weight decrements are no greater than 10% of control, confounding of absolute epididymis weights and relative liver weights would be minimized, although statistically significant effects on relative liver weight may still be observed.

Dietary restriction did not affect the serum hormone concentrations of E2, FSH, or LH. In contrast, serum T, DHT, PRL, TSH, T3, and T4 were attenuated by dietary restriction at levels that resulted in final body weight decrements of <= 85% of ad libitum control, depending on the hormone (Table 5Go). Clearly, body weight decrements can confound interpretation of compound-related hormonal effects, either preventing the detection of weak EACs or incorrectly concluding an agent is an EAC. The decrease in TSH and thyroid hormones is physiologically consistent with lowering metabolism during conditions of weight loss. In the dietary restriction experiment using the in vivo female battery (O'Connor et al., 1996Go), the only endpoint affected by dietary restriction levels that resulted in final body-weight decrements up to 80% of ad libitum control was uterine progesterone receptor concentrations. By targeting an MTD where body weight decrements are no greater than 10% of control, body weight-dependent decrements in the selected endpoints in the in vivo male (Tables 4 and 5GoGo) or female battery (O'Connor et al., 1996Go) would not occur; hence, the potential of confounding will be minimized when screening for EACs.

Development of a "Fingerprint" for the Tier I Battery for Four Distinct Endocrine Activities
The pattern of the responses observed with TEST, an AR agonist, was characteristic of the responses expected using the described Tier I screening battery. In the in vivo female battery, several of the uterotrophic endpoints were affected. As expected, based on previously published data (Jones and Edgren, 1973Go), absolute uterine weight (Table 1Go) and uterine stromal cell proliferation (Table 2Go) were increased. A slight increase (1 of 14 animals) in the incidence of uterine fluid imbibition was also observed, although this was considered not to be compound-related. Estrous conversion is a more sensitive marker for estrogenicity than uterine fluid imbibition (O'Connor et al., 1996Go), and there were no changes in that endpoint. For example, bisphenol A and methoxychlor, 2 weak ER agonists, increased the incidence of uterine fluid imbibition to 1/9 and 10/10, respectively, and increased the incidence of estrous conversion to 4/9 and 10/10, respectively. Control animals had incidence of 0/10 for both endpoints (Cook et al., 1997bGo). Uterine epithelial cell height, a less specific and less sensitive endpoint than uterine stromal cell proliferation for detecting estrogen receptor agonists (O'Connor et al., 1996Go; 1998aGo), was not affected by TEST administration at the dosages tested. The hormonal pattern was also characteristic of an AR agonist. Serum PRL and T concentrations were increased and serum FSH and LH concentrations were decreased (Table 3Go). The increase in serum T was a direct result of compound administration with the endogenous androgen TEST, the identical ligand to serum T, which is being measured in the RIA. YTS containing the ER did not detect TEST as a ligand, which is inconsistent with other published data (Gaido et al., 1997Go; Nicholson et al., 1978Go; O'Connor et al., 1998bGo; Pelissero et al., 1993Go; Zava and McGuire, 1978Go). The estrogen-like effects observed with TEST in the in vivo female battery were either the result of the binding of T and DHT to the ER and/or local aromatization of T to E2 since serum levels of E2 were not increased following TEST administration. Based on the data from the in vivo female battery, one cannot distinguish between an AR agonist such as TEST and a weak estrogen receptor agonist such as COUM (see below). However, by examining TEST in the integrated Tier I battery consisting of the in vivo female and male batteries and the in vitro YTS, it is possible to distinguish between androgenic versus estrogenic compounds such as COUM (discussed below).

In the in vivo male battery, TEST increased weights of the ASG unit, seminal vesicle, and prostate (Table 6Go), and produced characteristic hormonal alterations (Table 7Go) (i.e., increased T, E2, DHT, PRL, and decreased FSH and LH). As a result of TEST administration, and similar to the in vivo female battery, serum T concentrations are increased. Secondary to the increase in T, serum concentrations of DHT (the major metabolite of T) and E2 (the aromatization product of T) also increase (Ascoli and Segaloff, 1990Go; DeKretser et al., 1995Go; Wilson, 1990Go). The increases in serum DHT and E2 would not be expected with androgenic xenobiotics, as they would not be substrates for the enzymes 5{alpha}-reductase and aromatase. T and DHT bind to the AR, resulting in increased intracellular androgenic stimulus. This androgenic stimulus drives the increased weights observed for the androgen-dependent tissues of the ASG (Ascoli and Segaloff, 1990Go; DeKretser et al., 1995Go; Wilson, 1990Go). Centrally, there is a high level of androgenic feedback. As a result of the androgenic feedback mechanism, gonadotropin release from the anterior pituitary is attenuated in order to decrease T production from the Leydig cells (Ascoli and Segaloff, 1990Go; DeKretser et al., 1995Go; Wilson, 1990Go). Because TEST is being exogenously administered, serum concentrations of T, DHT, and E2 remain elevated. With xenobiotic AR agonists, androgenic feedback centrally would result in decreased T, DHT, FSH, and LH, and possibly E2 (Ascoli and Segaloff, 1990Go; DeKretser et al., 1995Go; Wilson, 1990Go). The Leydig cell atrophy observed microscopically in male rats treated with TEST is consistent with the decreases in serum gonadotropins.

In the in vitro YTS, TEST activated transcription in the yeast containing the AR and PR, and was negative in the yeast containing the ER (Figs. 2A–2CGo). Binding to the AR was expected, since T is the endogenous ligand for the AR and binding to the PR has been previously documented (Gaido et al., 1997Go). The lack of transcriptional activation by TEST in the YTS containing the ER was unexpected, because the published literature illustrates that TEST binds to the ER (Gaido et al., 1997Go; Nicholson et al., 1978Go; Pelissero et al., 1993Go; Zava and McGuire, 1978Go). The reason for this discrepancy is unknown.

Therefore, the expected pattern in the Tier I battery for a xenobiotic AR agonist would involve estrogen agonist-like responses in the in vivo female battery (i.e., increased uterine weight and uterine stromal cell proliferation) (Table 9Go) and increases in ASG weight and hormonal alterations (decreased T, DHT, E2, LH, and FSH) in the in vivo male battery (Table 10Go). While increases in serum PRL were observed in both the male and female battery after TEST administration, this increase was either a result of binding of TEST to the ER, and/or possible tissue aromatization of TEST to E2. Therefore, xenobiotic AR agonists would not be expected to increase serum PRL levels if these elevations are due primarily to aromatization, since xenobiotics would not be substrates for aromatase. Microscopically, atrophy of Leydig cells would be expected following treatment with an AR agonist, although effects on the seminiferous epithelium are less predictable. Administration of AR agonist has produced dose-dependant effects ranging from enhancement of spermatogenesis to degeneration and exfoliation of germ cells (Bansal and Davies, 1986Go; Flickinger, 1978Go; Jezek et al., 1993Go). Direct binding would be expected by AR agonists in the AR YTS assay (or competition for binding in the YTS competition assay containing the AR) (Table 11Go). Transcriptional activation in the YTS containing the ER and PR may also be observed with some androgenic compounds, depending upon the transcriptional system used.


View this table:
[in this window]
[in a new window]
 
TABLE 9 In Vivo Female Battery: Expected "Fingerprint" for Different Endocrine Activities Using a 5-Day Treatment Scenario
 

View this table:
[in this window]
[in a new window]
 
TABLE 10 In Vivo Male Battery: Expected "Fingerprint" for Different Endocrine Activities Using a 15-Day Treatment Scenario
 

View this table:
[in this window]
[in a new window]
 
TABLE 11 In Vitro YTS: Expected "Fingerprint" for Different Endocrine Activities
 
COUM, a weak ER agonist, produced the expected characteristic responses in the Tier I screening battery. In the in vivo female battery, several of the uterotrophic endpoints were increased: absolute uterine weight (Table 1Go), uterine stromal cell proliferation (Table 2Go), and uterine epithelial cell height (Table 2Go). Uterine fluid imbibition and estrous conversion were not affected by COUM administration at the dosages tested, illustrating the limitations of these endpoints for detecting weak estrogen receptor agonists. The observations for COUM are consistent with other weak estrogen receptor agonists such as estriol, bisphenol A, and methoxychlor (Cook et al., 1997bGo; O'Connor et al., 1996Go). Clearly, of the endpoints evaluated in the current in vivo female battery, uterine stromal cell proliferation appears to be the most sensitive marker for estrogen receptor agonists, where COUM administration resulted in a 7-fold increase in uterine stromal cell proliferation. This conclusion is also consistent with the results for 3 other weak ER agonists, estriol, bisphenol A, and methoxychlor, which produced a 12-fold, 11-fold, and 14-fold increase in uterine stromal cell proliferation, respectively (Cook et al., 1997bGo). However, the hormonal pattern was not completely typical for an ER agonist (Table 3Go). Consistent with other ER agonists (O'Connor et al., 1996Go, 1998bGo), serum PRL concentrations were increased along with serum FSH concentrations, which was an unexpected response. This increase in FSH was judged spurious, as a similar response was not observed in the in vivo male battery. This example illustrates the utility of redundant endpoints across the components of a tier battery. Estrogen agonists typically result in negative feedback to the hypothalamus and pituitary, resulting in decreased release of gonadotropins (Wilson, 1990Go). For that reason, it was expected that both FSH and LH would be decreased by COUM administration. The lack of a response in LH is attributed to the low potency of COUM based on the responses seen with E2 and estriol, which are high and low potency ER agonists, respectively (O'Connor et al., 1996Go). Another important point is that the responses of weak ER agonists were similar, independent of whether they were endogenous (estriol) or xenobiotic (COUM, bisphenol A, methoxychlor) in nature (Cook et al., 1997bGo; O'Connor et al., 1996Go).

In the in vivo male battery, organ weights and testicular histology were unaffected by COUM administration at the dosages tested (Table 6Go). This is in contrast to the changes that occurred in these parameters after administration of the ER agonist 17ß-estradiol (O'Connor et al., 1998bGo), and reflects the lower potency of COUM. Hormonal changes in responses to COUM included decreased T, E2, and DHT, and increased PRL, which was also typical for an ER agonist (Table 7Go) (O'Connor et al., 1998bGo). However, another reflection of the low potency of COUM, serum FSH and LH concentrations were unaffected by COUM administration at the dosages tested.

In the in vitro YTS, COUM activated transcription in the yeast containing the ER, and was negative in the yeast containing the AR or PR (Figs. 2A–2CGo). Binding to the AR and PR have been shown with 17ß-estradiol, a very potent estrogen receptor agonist (Gaido et al., 1997Go; O'Connor et al., 1998bGo); however, this has not been previously shown for weaker estrogen receptor agonists such as COUM. The lack of transcriptional activation by COUM in the YTS containing the AR and PR also reflects the lower binding affinity of COUM for the AR and PR, and is consistent with other weak ER agonists such as bisphenol A, methoxychlor, and estriol (data not shown).

Therefore, the expected pattern in the Tier I battery for a weak ER agonist would be increased uterine weight and uterine stromal cell proliferation in the in vivo female battery (Table 9Go), hormonal alterations (decreased T, DHT, and E2) in the in vivo male battery (Table 10Go), and direct binding in the ER YTS assay (or competition for binding in the YTS competition assay containing the ER) (Table 11Go). More potent ER agonists may also decrease androgen-dependent tissue weights in the in vivo male battery, and in both in vivo batteries, decreased serum LH and FSH and increased PRL concentrations may be observed. (O'Connor et al., 1996Go, 1998bGo).

PROG was used as the model for a PR agonist, and it produced the expected characteristic responses in the Tier I screening battery. In the in vivo female battery, several of the uterotrophic endpoints were affected. As expected based on previously published data (Jones and Edgren, 1973Go), absolute uterine weight (Table 1Go) and uterine stromal cell proliferation (Table 2Go) were increased. In addition, uterine epithelial cell height was increased by PROG administration. While these endpoints are sensitive for detecting ER agonists, these data for PROG, as well as those from TEST, RU486 (see below), and other steroidal compounds (Jones and Edgren, 1973Go), demonstrate that they lack specificity (Jones and Edgren, 1973Go; O'Connor et al., 1996Go). The hormonal pattern observed with PROG in the in vivo female battery was also characteristic of that of a PR agonist. Serum FSH and LH concentrations were decreased, and serum P4 concentrations were increased (Table 3Go). The increase in serum P4 was a result of exogenous administration of PROG, and would not be expected for xenobiotic progestins. The serum LH and FSH concentrations were decreased as a consequence of the normal physiological feedback inhibition of gonadotropins by P4 at the anterior pituitary (Hsueh and Billig, 1995Go).

In the in vivo male battery, responses to PROG included decreased epididymis, ASG unit, seminal vesicle, and prostate weights (Table 6Go). This pattern of organ-weight changes were similar to those observed for the estrogen receptor agonist 17ß-estradiol (O'Connor et al., 1998bGo), the androgen receptor antagonist flutamide (O'Connor et al., 1998aGo), and the testosterone biosynthesis inhibitor ketoconazole (O'Connor et al., 1998aGo). In order to distinguish between these different modes of action, hormonal and YTS data were required. The effects of PROG on the physiological feedback mechanism were apparent in the serum hormone alterations (Table 7Go). PROG administration decreased serum T, FSH, and LH, and increased serum E2 and P4. The increase in serum E2 concentration was attributed to the normal steroidogenic conversion of P4 to E2 (Hsueh and Billig, 1995Go) and would not be expected to occur with xenobiotics since they would not be substrates for these enzymes. Similar to the ovariectomized female battery, elevated P4, as well as E2, feedback to inhibit gonadotropin release from the anterior pituitary (Ascoli and Segaloff, 1990Go; Hsueh and Billig, 1995Go); hence, the decrease in serum LH and FSH concentrations. It is unclear why PRL concentrations decrease following P4 administration, but it may also be due to a centrally-mediated feedback mechanism. Secondary to the decreased LH and FSH concentrations, serum concentrations of T were decreased as a result of the attenuated LH-stimulation of the Leydig cells (Ascoli and Segaloff, 1990Go). The minimal increase in DHT concentrations in the 100 mg/kg/day PROG group was judged to be spurious, because changes in T and DHT have been observed to occur in tandem with several EACs (O'Connor et al., 1998aGo,bGo; 1999bGo). Microscopically, atrophy of Leydig cells in rats treated with PROG was consistent with the decreased gonadotropins and has been observed following exposure to other progestins (Chaudhary et al., 1990Go). It is not known if direct effects on Leydig cells may also contribute to PROG-associated decreases in T.

In the in vitro YTS, PROG activated transcription in the yeast containing the AR and PR, and was negative in the yeast containing the ER (Figs. 1A–1CGo). Binding to the PR was expected since PROG is the endogenous ligand for the PR, and binding to the AR has been previously documented (Gaido et al., 1997Go). The lack of PROG binding to the ER was consistent with the scientific literature (Gaido et al., 1997Go).

Therefore, the expected pattern in the Tier I battery for a PR agonist would be estrogen agonist-like responses in the in vivo female battery (i.e., increased uterine weight, uterine stromal cell proliferation, and uterine epithelial cell height) (Table 9Go), decreases in epididymis and ASG weights, hormonal alterations (decreased T, DHT, LH, and FSH) and Leydig cell atrophy in the in vivo male battery (Table 10Go), and direct binding in the PR YTS assay (or competition for binding in the YTS competition assay containing the PR) (Table 11Go). Transcriptional activation in the YTS containing the AR may also be observed with some progestins. The data for PROG illustrate the importance of having redundancy of endpoints in a Tier I screening battery. For example, using only the in vivo female battery, PROG would be indistinguishable from an ER agonist such as COUM, or even an AR agonist such as TEST. Similarly, even by coupling the in vivo male battery with the in vivo female battery, it is still difficult to distinguish a PR agonist from an ER agonist such as 17ß-estradiol (O'Connor et al., 1998bGo) or COUM. The only potential differences that might be observed between a PR agonist and an ER agonist would be differences in serum PRL (ER agonists increase PRL, PR agonists do not) and serum E2 (ER agonists should decrease E2, PR agonists would increase or not alter E2). However, coupling of the in vivo batteries to an in vitro system such as the YTS facilitates the discernment of ER agonists from PR agonists.

The pattern of the responses observed with RU486, a PR antagonist, were characteristic of the responses reported in the published literature. For instance, in the in vivo female battery, absolute uterine weight (Table 1Go) was increased by RU486, consistent with previous reports (Philibert et al., 1984Go), but concomitant increases in the incidence of uterine fluid imbibition or estrous conversion were not observed. The increase in uterine weight was attributed to RU486 binding to the ER (Philibert et al., 1984Go). The lack of an effect on uterine fluid imbibition, estrous conversion, uterine epithelial cell height, and serum hormone concentrations reflects the weak binding of RU486 to the ER. Interestingly, uterine stromal cell proliferation was significantly decreased at the highest dosage yet uterine weight was increased. This paradoxical response may be due to competing signals from RU486 blocking P4 binding to the PR (i.e., attenuating uterine cell proliferation by P4) (Table 2Go) coupled with RU486 binding to the ER (i.e., stimulating uterine weight).

In the in vivo male battery, responses to RU486 were similar to those observed for the AR antagonist flutamide (O'Connor et al., 1998aGo), and reflect the partial antiandrogenic activity of RU486 (Deraedt et al., 1984Go; Philibert et al., 1984Go). RU486 administration decreased epididymis, ASG unit, and seminal vesicle weights, but did not affect the weight or microscopic appearance of the testis (Table 6Go). These antiandrogenic effects on organ weights were observed only at the highest dosage of RU486, and were smaller in magnitude than the potent antiandrogen, flutamide (O'Connor et al., 1998aGo). RU486 administration caused significant increases in serum FSH and LH, and numerical, albeit not statistically significant, increases in serum T and DHT (Table 7Go). This hormonal pattern is similar to that observed with flutamide (O'Connor et al., 1998aGo). AR antagonists such as flutamide or RU486, bind to the androgen receptor and effectively block recognition of androgens, resulting in decreased intracellular androgenic stimulus, and ultimately decreased weights for the androgen-dependent tissues (Cook et al., 1993Go; Neri et al., 1972Go; O'Connor et al., 1998aGo; Simard et al., 1986Go). Centrally, there is a lack of recognition of androgens (Cook et al., 1993Go). As a result of the decreased intracellular androgenic stimulus, gonadotropin release from the anterior pituitary increases in order to stimulate T production from the Leydig cells (Neri et al., 1972Go; Simard et al., 1986Go). In response to the increase in gonadotropins and possibly a blockade of Leydig-cell androgen receptors, serum T production increases (Cook et al., 1993Go; Neri et al., 1972Go; Simard et al., 1986Go). Secondary to the increase in T, serum concentrations of DHT (the major metabolite of T) also increase. Therefore, in the in vivo male battery, the PR antagonists produce an antiandrogenic pattern of organ weight and endocrine changes due to their ability to bind to the androgen receptor (Fig. 2BGo).

In the in vitro YTS, RU486 activated transcription in the yeast containing the ER, AR, and PR, although the activity was greatest in the YTS containing the PR (Figs. 2A–2CGo). Binding to the PR was expected since the main endocrine activity of RU486 is as a PR antagonist. The binding in the YTS containing the AR and ER illustrate that cross talk among the steroid receptors can occur via weak binding to their receptors. The AR-binding ability of RU486 further supports the AR antagonist-like effects of RU486 that were observed in the in vivo male battery.

Therefore, the expected pattern in the Tier I battery for a PR antagonist would be increased uterine weight in the in vivo female battery (Table 9Go). The effects of a PR antagonist in the in vivo male battery were reflected in a pattern similar to antiandrogens (Table 10Go). In the YTS, a PR antagonist would produce direct binding in the PR YTS assay (or competition for binding in the YTS competition assay containing the PR) (Table 11Go). In addition, transcriptional activation in the YTS containing the ER and AR may also be observed with some PR antagonists. To date, no xenobiotics have been reported in the literature as having full PR antagonist activity, possibly due to their misidentification as AR antagonists.

Utility of the Tier I Battery for Identifying Compounds That Alter Thyroid Homeostasis
In the Tier I battery, thyroid hormone analyses (in vivo male battery only) and microscopic evaluations of the thyroid gland (in vivo male and female battery) were evaluated for their utility to detect thyroid toxicants. The expectation was that TEST, COUM, and PROG would not alter thyroid function, while RU486 would, based on the results of a 6-month study where the thyroid weights were increased (Deraedt et al., 1984Go). TEST and PROG did not produce follicular cell hypertrophy/hyperplasia in long-term rodent studies (PDR, 1996).

RU486 exhibited a hormonal pattern that was consistent with a potential thyroid toxicant (increased TSH, decreased T3, T4). These changes were attributed to enhanced excretion of thyroid hormones based on the increased relative liver weights (Table 6Go). In addition, there was minimal to mild depletion of colloid in the thyroid glands from the RU486-treated animals. These data suggest that RU486 is a weak thyroid hormone-modulating compound. This conclusion was consistent with previously published data for RU486 that demonstrated increased thyroid weights and follicular epithelial cell height after 6 months of treatment (Deraedt et al., 1984Go).

Thyroid hormone concentrations were also affected by TEST, COUM, and PROG (Table 8Go), but they did not produce the characteristic hormonal pattern expected of a thyroid toxicant. For instance, TEST decreased T4 concentrations to nondetectable levels, yet did not increase TSH or produce histopathological changes in the thyroid. COUM decreased serum T4 levels at the 2 highest dosages, while PROG decreased TSH levels; neither of these increased thyroid weight or histopathological lesions in the thyroid gland. Based on the responses seen with 13 distinct EACs evaluated to date (O'Connor et al., 1998aGo,bGo, 1999aGo, cGo), thyroid toxicants should produce the following pattern of responses: a hormonal pattern of increased TSH and decreased T3 and T4 coupled with increased thyroid gland weight and/or histopathologic changes (colloid depletion and/or follicular cell hypertropy/hyperplasia). These changes may not always be statistically significant. In contrast, the absence of this characteristic pattern of changes suggests that the evaluated compound is not a thyroid toxicant, even if statistically significant hormonal changes are observed (e.g., TEST, COUM, or PROG).

Conclusions
In summary, using this integrated Tier I screening battery consisting of a 5-day ovariectomized female battery, a 15-day intact male battery, and an in vitro YTS, all 4 test compounds were identified as EACs and produced distinct fingerprints of activity. The activities that were correctly identified include the AR agonist TEST, the weak ER agonist COUM, and the PR agonist PROG. RU486 produced a pattern of responses characteristic of a weak antagonist in the in vivo models. These responses coupled with the YTS data demonstrating primary binding to the PR would allow one to distinguish RU486 as a progesterone antagonist. This example illustrates the important of redundancy across the 3 components of the Tier I battery. The completion of these 4 positive endocrine controls brings the total number of compounds examined in the integrated Tier I screening battery to 13 (O'Connor et al., 1998aGo,bGo, 1999aGo, bGo,cGo). In addition, parts of the Tier I screening battery have also been used to identify several other compounds with potential endocrine-modulating activity, including methoxychlor (a proestrogen) (Cook et al., 1997bGo), bisphenol A (a weak ER agonist) (Cook et al., 1997bGo), linuron (an AR antagonist) (Cook et al., 1993Go), atrazine and cyanazine (dopaminergic modulators) (Cook et al., 1997aGo), a placebo pellet with estrogenic activity (O'Connor et al., 1997Go), and 3 proprietary compounds (identified as being an aromatase inhibitor, mixed testosterone/aromatase inhibitor, and AR antagonist) (unpublished data). Based on this validation exercise, a distinct "fingerprint" for each of these endocrine activities was observed against which compounds with unknown activities can be compared. This validation exercise is consistent with the ICCVAM process and we believe illustrates the rigor that the EPA endocrine validation team should apply to any test before it is implemented as an endocrine screen.


    ACKNOWLEDGMENTS
 
We gratefully acknowledge partial funding of this project by the Chemical Manufacturers Association and the Chlorine Chemistry Council, 1300 Wilson Boulevard, Arlington, Virginia. The authors would like to acknowledge the invaluable advice of Ms. Ann M. Mason (Chlorine Chemistry Council) and Drs. James A. Barter (PPG Industries), Robert E. Chapin (NIEHS), A. Michael Kaplan (DuPont), William R. Kelce (Monsanto), Ronald R. Miller (Dow Chemical Co.), and Ellen K. Silbergeld (University of Maryland at Baltimore). The authors would like to thank Vivian Thompson, Bryan Crossley, Stephen Novak, Christine Glatt, Suzanne Craven, Denise Janney, and Susan Nicastro for their technical support. Finally, the authors would also like to thank Dr. Donald P. McDonnell (Duke) for his technical expertise and generosity in providing the yeast strains for the YTS and Dr. Kevin W. Gaido (CIIT) for his suggestions.


    NOTES
 
1 To whom correspondence should be addressed. Fax: (302) 366-5003. E-mail: john.c.oconnor{at}usa.dupont.com. Back


    REFERENCES
 TOP
 ABSTRACT
 INTRODUCTION
 MATERIALS AND METHODS
 RESULTS
 DISCUSSION
 REFERENCES
 
Ascoli, M., and Segaloff, D. L. (1990). Adenohypophyseal hormones and their hypothalamic releasing factors. In Goodman-Gilman's The Pharmacological Basis of Therapeutics (A. Goodman-Gilman, Ed.), Vol. 1, pp. 1363–1382. Pergamon, New York.

Bansal, M. R., and Davies, A. G. (1986). Effect of testosterone oenanthate on spermatogenesis and serum testosterone concentration in adult mice. J. Reprod. Fert. 78, 219–224[Abstract]

Chaudhary, J., Sharma, R. K., Majumdar, S. S., Bhatti, D. P. S., Misro, M. M., and Das, R. P. (1990). Effect of STS—557 (17ß-cyanomethyl 17ß-hydroxy-estra-4,9(10)-dien-3-one) on blood hormone levels, the testis, accessory sex organs and fertility of rats. Int. J. Androl. 13, 398–407.[ISI][Medline]

Cook, J. C., Craven, S. C., Van Pelt, C. S., Arnold, S. F., Obourn, J. D., and O'Connor, J. C. (1997a). Hormonal mechanism involved in triazine-mediated rat mammary tumorigenesis. Toxicologist 36, 468.

Cook, J. C., Kaplan, M. A., Davis, L. G., and O'Connor, J. C. (1997b). Development of a Tier I screening battery for detecting endocrine active compounds (EACs). Regul. Toxicol. Pharmacol. 26, 60–68.[ISI]

Cook, J. C., Mullin, L. S., Frame, S. R. and Biegel, L. B. (1993). Investigation of a mechanism for Leydig cell tumorigenesis by linuron in rats. Toxicol. Appl. Pharmacol. 119, 195–204.[ISI][Medline]

Davis, B. (1993). Female Reproductive Toxicology. Academic Press, San Diego.

DeKretser, D. M., Risbridger, G. P., and Kerr, J. B. (1995). Basic endocrinology of the testis. In Endocrinology (L. J. DeGroot, Ed.), Vol. 3, pp. 2307–2335. W.B. Saunders, Philadelphia.

Deraedt, R., Vannier, B., and Fournex, R. (1984). Toxicological study on RU 486. In The Antiprogestin Steroid RU 486 and Human Fertility Control (E.-E. Baulieu and S. J. Segal, Eds.), Vol. 1, pp. 123–126. The Rockefeller Foundation, New York.

EDSTAC. (1998). Endocrine Disruptor Screening and Testing Advisory Committee (EDSTAC) Final Report, August 1998.

Feron, V. J., De Groot, A. P., Spanjers, M. T., and Til, H. P. (1973). An evaluation of the criterion "organ weight" under conditions of growth retardation. Food Cosmet. Toxicol. 11, 85–94.[ISI][Medline]

Flickinger, C. (1978). Effects of testosterone enanthate on the structure of the male reproductive tract of the rat. Anat. Rec. 192, 555–584.[ISI][Medline]

Gaido, K. W., Leonard, L. S., Lovell, S., Gould, J. C., Babai, D., Portier, C. J., and McDonnell, D. P. (1997). Evaluation of chemicals with endocrine-modulating activity in a yeast-based steroid-hormone-receptor gene-transcription assay. Toxicol. Appl. Pharmacol. 143, 205–212.[ISI][Medline]

Graham, J. D., and Clarke, C. L. (1997). Physiological action of progesterone in target tissues. Endocr. Rev. 18, 502–519.[Abstract/Free Full Text]

Hsueh, A. J. and Billig, H. (1995). Ovarian hormone synthesis and mechanism of action. In Endocrinology (L. J. DeGroot, Ed.), Vol. 3, pp. 2019–2030. W.B. Saunders, Philadelphia.

Jezek, D., Simunic-Banek. L., and Pezerovic-Panijan, R. (1993). Effects of high doses of testosterone propionate and testosterone enanthate on rat seminiferous tubules: A stereological and cytological study. Arch. Toxicol. 67, 131–140.[ISI][Medline]

Jones, R. C., and Edgren, R. A. (1973). The effects of various steroids on the vaginal histology in the rat. Fertil. Steril. 24, 284–291.[ISI][Medline]

Markaverich, B. M., Webb, B., Densmore, C. L., and Gregory, R. R. (1995). Effects of coumestrol on estrogen receptor function and uterine growth in ovariectomized rats. Environ. Health Perspect. 103, 574–81.[ISI][Medline]

Neri, R., Florance, K., Koziol, P., and Cleave, S. V. (1972). A biological profile of a nonsteroidal antiandrogen, SCH 13521 (4'-nitro-3'-trifluoromethylisobutyranilide). Endocrinology 91, 427-437.[ISI][Medline]

Nicholson, R. I., Davies, P., and Griffiths, K. (1978). Interaction of androgens with oestradiol-17ß receptor proteins in DMBA-induced mammary tumors: A possible oncolytic mechanism. Eur. J. Cancer 14, 439–445.[ISI][Medline]

O'Connor, J. C., Cook, J. C., Craven, S. C., Van Pelt, C. S., and Obourn, J. D. (1996). An in vivo battery for identifying endocrine modulators that are estrogenic or dopamine regulators. Fundam. Appl. Toxicol. 33, 182–195.[ISI][Medline]

O'Connor, J. C., Cook, J. C., Slone, T. W., Makovec, G. T., Frame, S. R., and Davis, L. G. (1998a). An ongoing validation of a Tier screening battery for detecting endocrine-active compounds (EACs). Toxicol. Sci. 46, 45–60.[Abstract]

O'Connor, J. C., Cook, J. C., VanPelt, C. S., Arnold, S. F., and Obourn, J. D. (1997). Weak estrogenic activity from continuous-release pellets. Reprod. Toxicol. 11, 101–106.[ISI][Medline]

O'Connor, J. C., Frame, S. R., Biegel, L. B., Cook, J. C., and Davis, L. G. (1998b). Sensitivity of a Tier I screening battery compared to an in utero exposure for detecting the estrogen receptor agonist 17ß-estradiol. Toxicol. Sci. 44, 169–184.[Abstract]

O'Connor, J. C., Frame, S. R., Davis, L. G., and Cook, J. C. (1999a). Detection of dopaminergic modulators in a Tier I screening battery for detecting endocrine-active compounds (EACs). Toxicol. Sci., submitted.

O'Connor, J. C., Frame, S. R., Davis, L. G., and Cook, J. C. (1999b). Detection of the environmental antiandrogen p,p'-DDE in CD and Long-Evans rats using a Tier I screening battery and a Hershberger assay. Toxicol. Sci. 51, 44–53.[Abstract]

O'Connor, J. C., Frame, S. R., Davis, L. G., and Cook, J. C. (1999c). Detection of thyroid toxicants in a Tier I screening battery and alterations in thyroid endpoints over 28 days of exposure. Toxicol. Sci. 51, 54–70.[Abstract]

Pelissero, C., Flouriot, G., Floucher, J. L., Bennetau, B., Donogues, J., LeGac, F., and Sumpter, J. P. (1993). Vitellogenin synthesis in cultured hepatocytes; An in vitro test for the estrogenic potency of chemicals. J. Steroid Biochem. Mol. Biol. 44, 263–272.[ISI][Medline]

Philibert, D., Moguilewsky, M., Mary, I., Lecaque, D., Tournemine, C., Secchi, J., and Deraedt, R. (1984). Pharmacological profile of RU 486 in animals. In The Antiprogestin Steroid RU 486 and Human Fertility Control (E.-E. Baulieu and S. J. Segal, Eds.), Vol. 1, pp. 49–68. The Rockefeller Foundation, New York.

Physician's Desk Reference (PDR) (1996). Vol. 50, 13th ed. Medical Economics Data, Oradell, NJ.

Simard, J., Luthy, I., Guay, J., Belanger, A., and Labrie, F. (1986). Characteristics of interaction of the antiandrogen flutamide with the androgen receptor in various target tissues. Mol. Cell. Endocrinol. 44, 261–270.[ISI][Medline]

Wilson, J. D. (1990). Androgens. In Goodman-Gilman's The Pharmacological Basis of Therapeutics (A. Goodman-Gilman, Ed.), Vol. 1, pp. 1441–1457. Pergamon, New York.

Zava, D. T., and McGuire, W. L. (1978). Androgen action through estrogen receptor in a human breast cancer-cell line. Endocrinology 103, 624–631.[ISI][Medline]