Ex Vivo and in Vitro Testis and Ovary Explants

TOHCOLOGICAL SCIENCES 46, 61-74 (1998)

ARTICLE NO. TX98256O

Ex Vivo and in Vitro Testis and Ovary Explants: Utility for IdentifyingSteroid Biosynthesis Inhibitors and Comparison to a

Tier I Screening Battery1

Stephanie S. Powlin,2 Jon C. Cook,3 Stephen Novak, and John C. O'Connor

DuPont Haskell Laboratory for Toxicology and Industrial Medicine, P.O. Box 50, Newark, Delaware 19714

Received April 23, 1998; accepted August 7, 1998

Ex Vivo and in Vitro Testis and Ovary Explants: Utility forIdentifying Steroid Biosynthesis Inhibitors and Comparison to aTier I Screening Battery. Powlin, S. S., Cook, J. C, Novak, S., andO'Connor, J. C. (1998). Toxicol. Set 46, 61-74.

Testis and ovary explants have been proposed as in vitro screensfor identifying potential inhibitors of steroid biosynthesis. Thegoals of the current study were to optimize the conditions of thetwo assays, to characterize these assays using several compoundswith well-defined endocrine activity, and to compare the responsesfrom the explant assays with an in vivo male battery currentlyundergoing validation using the Crl:CD BR rat in order to eval-uate their utility as test systems for screening unknown com-pounds for possible steroid biosynthesis inhibition activity. Therewere two components to the testis/ovary assays: ex vivo and invitro. The ex vivo component used testes/ovaries from animalsdosed with the test compounds in vivo, and the in vitro componentused testes/ovaries from control animals. For the testis assays,decapsulated testis explants (50 mg) were placed into glass scin-tillation vials, ±1.0 IU/ml hCG for 3 h in a shaking water bath(34°C). Following the incubation period, medium was removed,centrifuged, and frozen until assayed for hormone concentrations.A similar procedure was used for the ovary explant assay exceptthat each ovary was incubated separately. The testis explants wereevaluated using the following compounds: ketoconazole (KETO),a testosterone biosynthesis inhibitor; aminoglutethimide (AG)(only in vitro) and anastrozole (ANA), aromatase inhibitors; fin-asteride (FIN), a 5a-reductase inhibitor; 17p-estradiol (17^E2),an estrogen receptor agonist; flutamide (FLUT), an androgenreceptor antagonist; ICI-182,780 (ICI), an estrogen receptor an-tagonist; haloperidol (HALO), a D2 receptor antagonist; and re-serpine (RES), a dopamine depletor. In the ovary assay, AG (onlyin vitro), ANA, ICI, and HALO (only in vitro) were evaluated.Addition of fetal calf serum to the medium allowed measurementof estradiol (E2) in the testis assay, but production was not inhib-ited by ANA or AG. In the ovary explant assay, only AG wasidentified as inhibiting E2 production in vitro. Hence, both thetestis and ovary explant assays appear to have limited utility for

' This study was funded partially by the Chemical Manufacturers Associa-tion and the Chlorine Chemistry Council, 1300 Wilson Boulevard, Arlington,Virginia, 22209.

2 Supported by a DuPont postdoctoral fellowship.3 To whom reprint requests should be addressed.

detecting aromatase inhibitors. Screening of these nine diverseendocrine-active compounds resulted in all of them being identi-fied as altering the endocrine system when assessed by ex vivo andin vitro testis explants. Using only the in vitro assessment with thecriteria of steroid biosynthesis inhibition, four of nine compoundswere correctly identified in the testis explant assay (17/J-E2,KETO, FLUT, and HALO). The predictability of both the in vitroand ex vivo ovary assay was 50%, suggesting a 50% false positiveor negative rate with unknown compounds. However, of the sevencompounds assessed to date (17/J-E2, ICI, ANA, KETO, FLUT,HALO, and RES), all were correctly identified using an in vivomale battery, which also has the capability to detect other endo-crine activities. Therefore, the testis and ovary explant assaywould not be necessary if one were using an in vivo male battery,since this screen would identify steroid biosynthesis inhibitors andwould also identify several other endocrine activities. Because ofthe difficulties in assessing cytotoxicity and the high false positive/negative rates, the ovary and testis explant assays are not useful asroutine screening procedures for detecting steroid biosynthesisinhibitors; however, they may have utility in confirming in vivo

findings. O 1998 Soday of Torfcotoor.

Key Words: explants; testis; ovary; steroid biosynthesis inhibi-tion; in vifro; ex vivo; rat.

Endocrine-active compounds (EACs) have been implicatedin producing developmental abnormalities (hypospadias andcryptorchidism in humans and masculinization/feminization inwildlife species), reproductive deficits (decreased spermcounts), and cancer (mammary gland, testis, and prostate inhumans; liver and thyroid in wildlife) (Ankley et al., 1997;Colbom et al., 1993; Birnbaum, 1994; IEH, 1995; Kavlock etal., 1996). Bonafide examples of EACs producing adverseeffects in humans (e.g., diethylstilbestrol) (Herbst et al., 1971;Gill et al., 1979) and wildlife (e.g., DDT) (Bitman and Cecil,1970) have been documented, which lends credibility to thishypothesis. However, in most cases, a link between EACs andadverse effects in humans, fish, and wildlife remains to beestablished (Safe, 1995; Crisp et al., 1997). In response to theheightened awareness around this issue, Congress passed dieFood Quality Protection Act of 1996 and the Safe Drinking

1096-6O8O/98 $25.00Copyright C 1998 by the Society of Toxicology.All right! of reproduction in any form reserved

Dow

nloaded from https://academ

ic.oup.com/toxsci/article/46/1/61/1684159 by guest on 10 Septem

ber 2022

62 POWLIN ET AL.

Water Act of 1996, which require the United States Environ-mental Protection Agency (USEPA) to implement screeningand testing strategies to identify EACs. The USEPA respondedby forming a multistakeholder committee, the Endocrine Dis-ruptor Screening and Testing Advisory Committee (EDSTAC).The purpose of EDSTAC is to recommend to the USEPA ascreening and testing strategy for detecting EACs. The imple-mentation of testing is Congressionally mandated to begin inAugust 1999. EDSTAC has focused its efforts on developing ascreening and testing strategy that can identify compounds thatare androgenic/antiandrogenic, estrogenic/antiestrogenic, orsteroid biosynthesis inhibitors, as well as compounds that alterthyroid hormone function.

The Organisation for Economic Co-operation and Develop-ment (OECD) has commissioned a comprehensive survey oftest methods for identifying sex-hormone-disrupting chemicalsthat is currently undergoing comment by the member countries(Holmes et al., 1998). In addition, several different tieredtesting approaches have been proposed for detecting EACs(Reel et al, 1996; Shelby et al, 1996; Ashby et al., 1997;Carney et al., 1997; Cook et al., 1997; Gray et al., 1997).Common to all these approaches is the combination of in vivoand in vitro assays, recognizing that there are limitations toeach approach (reviewed in Gray et al., 1997). In vivo assaysallow integration of ADME principles (adsorption, distribu-tion, metabolism, and excretion) and homeostasis, which areimportant with highly integrated systems such as the endocrinesystem, while in vitro approaches are useful for clarifying thesite and the mechanism of action. It is clear that no single invitro or in vivo assay will be able to stand alone as an accurate,sensitive, and cost-effective means to screen and test all po-tential EACs; however, individual assays may be valuabletools for answering specific questions. For example, testis andovary explant cultures have been proposed as simple testsystems for identifying EACs that alter steroid biosynthesiseither through up- or downregulation or via direct inhibition(Gray et al., 1997). The testis explants are used primarily toidentify compounds which inhibit testosterone biosynthesis,while ovary explants are used primarily to identify aromataseinhibitors (Laskey and Berman, 1993; Berman and Laskey,1993; Laskey et al., 1994; Piasek and Laskey, 1994; Gray etal., 1995; Laskey et al, 1995). As proposed, both the testis andovary explant assays include in vitro and ex vivo (in vivodosing followed by in vitro assessment of function) assess-ments. The in vitro assessment can be used to identify com-pounds that directly inhibit steroid biosynthesis. The ex vivoassessment can be used to identify compounds that up- ordownregulate the enzymes of the steroid biosynthetic pathway.

The perceived value of the testis explant culture is primarilybased on data generated using isolated Leydig cells (e.g.,Klinefelter and Kelce, 1996). However, to date, the testisexplant has been evaluated with only ethane dimethanesulfon-ate (EDS) (Laskey et al, 1994; Gray et al, 1995). In contrast,the ovary explant culture has been evaluated with aminoglute-

thimide (AG) (Berman and Laskey, 1993; Laskey et al, 1995),cadmium (Piasek and Laskey, 1994) and di(2-ethylhexy-l)phthalate (DEHP) (Berman and Laskey, 1993; Laskey andBerman, 1993) using ovaries from cycling rats and with cad-mium using ovaries from pregnant rats (Piasek and Laskey,1994). Because of the limited number of compounds evaluated,as well as the different assay conditions used, the goals of thecurrent study were to optimize the conditions of the two assays,to characterize these assays using several compounds withwell-defined endocrine activity, and to compare the responsesfrom the explant assays with an in vivo Tier I male batterycurrently undergoing validation (Cook et al, 1997; O'Connoret al, 1998a,b) in order to evaluate their utility as test systemsfor screening unknown compounds for possible steroid biosyn-thesis inhibition activity. The rationale for the study design ofthe Tier I battery undergoing validation has been previouslydescribed (Cook et al, 1997). The Tier I male battery isdesigned to identify several different endocrine activities usinga comprehensive hormonal assessment coupled with organweight measurements.

The following EACs were used to evaluate the testis assay:an estrogen receptor agonist (17/3-estradiol, 17/3-E2) (Clarkand Markaverich, 1983), an estrogen receptor antagonist (ICI-182,780, ICI) (Wakeling et al, 1991), an androgen receptorantagonist (flutamide, FLUT) (Neri et al, 1972; Simard et al,1986), a steroid biosynthesis inhibitor (ketoconazole, KETO)(Feldman, 1986), a selective aromatase inhibitor (anastrozole,ANA) (Dukes et al, 1996), a mixed aromatase/C^-side chaincleavage inhibitor (AG) (only in vitro) (Salhanick, 1982), aselective 5a-reductase inhibitor (finasteride, FIN) (Rittmaster,1994), a D2 receptor antagonist (haloperidol, HALO) (Good-man and Gillman, 1996), and a dopamine depletor (reserpine,RES) (Goodman and Gillman, 1996). In the ovary assay, AG(only in vitro), ANA, ICI, and HALO (only in vitro) wereevaluated.

MATERIALS AND METHODS

Test maltrials. The following materials were purchased from SigmaChemical Company (St. Louis, MO): 8-bromoadenosine 3':5' cyclic mono-phosphate (8-bromo-cAMP) (Sigma No. B-7880), dimethyl sulfoxide(DMSO), 17/3-E2, fetal calf serum (FCS) (Sigma No. F-2442), FIN, FLUT,follicle-stimulating hormone (FSH) (Sigma No. F-4021), HALO, Medium 199(Ml99), soybean trypsin inhibitor (Sigma No. T-9003), and trypan blue dye.The remaining materials were obtained from the following manufacturers: AG,Research Biochemicals International (Natick, MA); irradiated Certified RodentDiet No. 5002, PMI Feeds, Inc. (St. Louis, MO); methylcellulose, FisherScientific (Springfield, NT); KETO, ICN Biomedical, Inc. (Aurora, OH);M-199 and RPMI 1640 medium, Gibco BRL Products (Grand Island, NY);human chorionic gonadotropin (hCG), Wyeth-Ayerst (Philadelphia, PA); andtestosterone (T), estradiol (E2), and progesterone (P4) radioimmunoassay(RIA) kits, Diagnostic Products Corp. (Los Angeles, CA). ICI and ANA weregenerously donated by Dr. B. M. Vose (Macclesfield, UK) and Dr. J. A.Schwartz (Wilmington, DE), respectively, of Zeneca Pharmaceuticals.

Test species. Male and time-pregnant female (gd 5) Crl:CD BR rats wereacquired from Charles River Laboratories, Inc. (Raleigh, NC), and wereapproximately 63 days of age upon arrival. After a quarantine period of

Dow



ber 2022

TESTIS/OVARY EXPLANTS FOR IDENTIFYING STEROID BIOSYNTHESIS INHIBITORS 63

approximately 1 week, male rats designated for Tier I and explant studies thatdisplayed adequate weight gain and freedom from clinical signs were dividedby computerized, stratified randomization into five groups of 15 males so thatthere were no statistically significant differences among group body weightmeans. Males designated for Leydig cell isolation (n = 12) were also 63 daysof age upon arrival and were quarantined for 1 week. Due to study designrequirements, pregnant females did not undergo normal quarantine proceduresand upon arrival were randomly assigned to study groups so that there were nostatistically significant differences among group body weight means (5 rats pergroup, 5 groups).

All rats were individually housed in stainless steel, wire-mesh cages sus-pended above cage boards and were fed irradiated PMI Feeds, Inc., CertifiedRodent Diet No. 5002 and provided with tap water (United Water Delaware)ad libitum. Animal rooms were maintained on a 12-h light/dark cycle (fluo-rescent light), a temperature of 23 ± 2°C, and a relatively humidity of 50 ±10%.

Study design. All rats were weighed daily and cage-side examinationswere performed to detect moribund or dead rats. At each weighing, rats wereindividually handled and examined for abnormal behavior or appearance. Eachtest compound was prepared in 0.25% methylcellulose and administered byintraperitoneal injection (2.0 ml/kg body wt) at approximately 0900 h daily.Male rats were dosed for 15 days and euthanized on the morning of test day 15.Pregnant females were dosed for 3 days (gd 6-8) and euthanized on themorning of gd 8. On the day of sacrifice, male and female rats were euthanizedapproximately 2 h after the last administered dose.

The following dosages were used in the ex vivo experiments with male rats:KETO (25, 50, 75, and 100 mg/kg/day), ANA (0.1, 0.5, 0.75, and 1.5mg/kg/day), FIN (1, 5, 10, and 25 mg/kg/day), 17/3-E2 (0.001,0.0025,0.0075,and 0.05 mg/kg/day), FLUT (0.25, 1, 5, and 20 mg/kg/day), HALO (0.1, 0.25,0.5, and 1 mg/kg/day), ICI (0.05, 0.25, 0.5, and 1 mg/kg/day), and RES (0.1,0.5, 0.75, and 1 mg/kg). The following dosages were used in ex vivo experi-ments with pregnant female rats: ICI (0.05, 0.25, 0.5, and 1 mg/kg/day) andANA (1, 5, 7.5, and 10 mg/kg/day). Doses were selected in order to obtain themaximal pharmacologic effect for each compound and/or not exceed themaximum tolerated dose as determined in range-finder studies.

Concentrations of compounds used in vitro were selected based on limits ofsolubility (ANA, AG, FIN, HALO, and RES), knowledge of receptor activity(170-E2, ICI, and FLUT), or previous work (KETO) (Biegel et al., 1995).

Pathology. Male rats were anesthetized using carbon dioxide and eutha-nized via exsanguination on test day 15 and were evaluated for gross obser-vations of toxicity and organ weights Giver, testes, prostate, seminal vesicles,and accessory sex gland unit). Serum from blood collected at the time ofsacrifice was used to measure hormone concentrations: T, E2, dihydrotestos-terone (DHT), FSH, leutinizing hormone (LH), prolactin (PRL), thyroid-stimulating hormone (TSH), thyroxine (T4), and triiodothyronine (T3). Themethods and data are described elsewhere (O'Connor et at, 1998a) and in theaccompanying paper (O'Connor et al., 1998b).

Pregnant female rats were anesthetized using carbon dioxide and euthanizedvia exsanguination on gd 8. At sacrifice, the number of fetuses in each uterinehorn was counted, and both ovaries were removed and cleared of fat.

Testis explant assay. Ex vivo and in vitro analyses of T and E2 productionwere conducted on five animals from each treatment group. For the ex vivocomponent, testis explants from animals dosed with 17/3-E2, ICL FLUT,KETO, FIN, ANA, HALO, or RES were cultured with medium containinghCG (1 IU/ml). hCG binds to the LH receptor on Leydig cells to stimulate Tproduction. This stimulation confirms the viability of explants. For the in vitrocomponent, testis explants from untreated animals were cultured with andwithout hCG (1 IU/ml) and several concentrations of the test compounds.Methods were modified from Laskey and co-workers (1994) to facilitateincorporation into an ongoing validation of a Tier I male screening battery(O'Connor et al., 1998a,b). Specifically, the amount of testicular parenchymawas reduced from 300-400 mg to 50 mg so the in vitro experiments could beconducted using the control rats from the Tier I male battery. Other measure-ments (organ weights and serum hormone levels) were made using the same

control animals (O'Connor et al, 1998a,b). One testis per animal was re-moved, placed in a labeled cassette, moistened with cold physiological saline,placed on ice, and covered with saline-moistened paper towels. The testes weresubsequently decapsulated and approximately 50 mg of testicular parenchymawas placed into a 20-ml glass scintillation vial containing 5 ml of culturemedium (RPMI-1640 medium, 10% FCS, 50 ^tg/ml soybean trypsin inhibitor)and the appropriate concentrations of the test compounds. The final solvent(DMSO) concentration in the culture media did not exceed 0.2% (v/v). Finally,10 td of hCG (500 IU/ml stock solution) (or 10 /xl of distilled water fornon-hCG stimulated controls) was added, and the vials were capped, brieflyvortexed, and incubated vertically for 3 h at 34°C under vigorous shaking (175rpm). At the end of the incubation period, 1.4 ml of culture medium wasremoved and centrifuged at 14,000g for 5 min (4°C) to pellet all remainingtesticular parenchyma. The resulting supernatant was frozen between —65 and-85°C until analyzed for T and E2 levels.

Leydig cell isolation and culture. Leydig cells from male Crl:CD BR ratswere isolated according to the methods of Biegel et al. (1993). Briefly, thetestes were removed from the rats and perfused with a solution of collagenase(1.0 mg/ml) m M-199 medium. The tunicate was removed, and the testes wereplaced in a digestion medium and shaken in a 37°C water bath for 15 min. Thetubules were removed by centrifugation and filtration, and the remaining cellswere loaded onto a Percoll gradient and centrifuged at 800g for 30 min. Thefraction containing Leydig cells was removed and washed with M-199 mediumto remove excess Percoll. Cells were placed in 24-well culture plates at aconcentration of 50,000-70,000 cells/well and allowed to attach for 1 h. RPMI1640 medium (with 10% FCS and 1% antibiotic/antimycotic solution) (500 jxlper well) containing 4 IU/ml hCG and five concentrations of each test com-pound (0 (±hCG), 0.1, 1, 10, and 100 iiM (all + hCG)) was then added to thewells for 3 h. Culture medium was then removed from the wells and frozenbetween —65 and -85°C until analyzed for steroid concentrations.

Trypan blue dye exclusion was used to determine cell viability and wasperformed as follows: 500 /xl of the diluted Leydig cell preparation wascombined with 500 jd of 4% trypan blue dye and allowed to incubate for 15min. Cells were then counted using a hemacytometer and the percentage ofviable cells (total unstained cells divided by total cells X 100) was calculated.At the conclusion of the 3-h incubation, cell viability was determined byremoving the culture medium from the wells and replacing it with 500 /xl 4%trypan blue dye. Cells were incubated for 15 min, the dye was removed, andsix fields were counted using a phase-contrast microscope. The percentage ofviable cells was calculated as described above.

Ovary explant assay. Ex vivo and in vitro analyses of E2 and P4 produc-tion were conducted on five animals from each group. Methods were modifiedfrom Laskey and coworkers (1995). For the ex vivo component, ovaries fromtreated pregnant rats (gd 8) were cultured without fleft ovary) and with (rightovary) hCG stimulation (1 IU/ml). For the in vitro component, ovaries fromuntreated pregnant rats (gd 8) were cultured with medium containing severalconcentrations of the test compounds and cultured without (left ovary) andwith (right ovary) hCG stimulation (1 IU/ml). Ovaries were removed, clearedof fat, placed in labeled cassettes, moistened with physiological saline, andplaced on ice until weighed. After weighing, each ovary was placed into a1.5-ml microfuge tube containing 1 ml of culture medium (RPMI 1640medium, 10% FCS, 50 /xg/ml soybean trypsin inhibitor) containing the appro-priate concentration of test compound Ovaries were thoroughly minced withscissors into approximately <l-mm3 pieces, and microfuge tubes were vor-texed and incubated horizontally for 3 h at 36°C under vigorous shaking (175rpm). At the end of the incubation period, tubes were centrifuged at 14,000gfor 5 min (4°C). The resulting supernatant was frozen between - 6 5 and-85°C until analyzed for E2 and P4 levels.

Spectral binding. In order to determine if HAL and RES inhibit T pro-duction via inhibition of cytochrome P450 activity, spectral binding assayswere performed according to Lewis (1986). Liver microsomes from pheno-barbital-treated rats were diluted to a concentration of approximately 1 mgprotein/ml buffer. Compounds were solubilized in methanol at concentrations

Dow



ber 2022

64 POWLIN ET AL.

M 70-

60-

I "•3 40-

2 30-2

i M-0 -

—c—(-)hCG—•—(+) hCG

V

• 4 _ _ _ _ ^ - ^ :

^ ^

"3 loo-

(-)hCG-(+)hCG

2 3 4

Incubation Time (hr)

2 3 4

Incubation Time (hr)

FIG. 1. Testis explant assay: Steroid production time-course experiment Tune course of T (A) and E2 (B) production by 50-mg testis explants from controlrats incubated in the presence (closed circles) or absence (open circles) of 0.1 IU/ml hCG. Culture medium was removed from the scintillation vials at theappropriate time and analyzed for T and E2 concentrations (n = 2). T levels from hCG-stimulated explants were statistically different from the non-hCG-stimulated explants at all time points evaluated (*p £ 0.05).

of 32 mM (HAL) and 0.7 mM (RES) and incubated with microsomes, and theabsorbance from 350 to 500 nm was measured.

Statistical analyses. All data are expressed as means ± standard error. Alldata were analyzed by one-way analysis of variance (ANOVA) except hor-monal data. Hormone levels were analyzed using Jonckheere's test for trend.If a significant dose-response trend was detected, data from the top dose groupwas excluded and the test repeated until no significant trend was detected.Significance was judged a t p £ 0.05.

RESULTS

Testis Explant Assay

Optimization. Experiments were conducted to optimize theculture conditions for the testis explant assay. The parametersexamined were incubation time, incubation vessel, amount oftissue, FCS ( + / - ) , hCG concentration, type of culture me-dium, bovine serum albumin (+/—), and addition of soybeantrypsin inhibitor. A time-course experiment (1, 2, 3, 4, and 5 h)was conducted to assess the linearity of T and E2 production.T production was linear through 4 h (Fig. 1A), while E2production was linear through 5 h (Fig. IB). A 3-h incubationtime was used in subsequent experiments to insure that steroidproduction was linear over time. As expected, hCG stimulatedT production, but not E2 production (Fig. 1). An hCG dose-response experiment (0, 0.01, 0.1, 1.0, and 10 IU/ml) wasconducted to determine the concentration of hCG that pro-duced maximal T production. Maximal T production was ob-tained at a concentration of 1.0 IU/ml hCG (Fig. 2); thisconcentration was used in all subsequent experiments. Similarto Fig. 1, E2 production was not affected by increasing con-centrations of hCG (data not shown).

Addition of FCS to the culture medium was necessary forthe measurement of E2 (Table 1). To ensure that there were nomeasurable amounts of T and E2 in the culture medum, theculture medum containing FCS was assayed for T and E2; none

was detected (data not shown). Inclusion of FCS in the mediumdid not affect T production (Table 1). Other investigators haveused quartered testis explants (300-400 mg) in their experi-ments (Gray et ai, 1995; Laskey et al, 1994). In order toincorporate this assay into our ongoing Tier I battery, it wasnecessary to reduce the size of the testis explants; therefore,steroid production was compared using tissue amounts of 50and 200 mg. T and E2 production was not significantly differ-ent between 50- and 200-mg explants (Table 1); hence, 50 mgof testicular parenchyma were used in all subsequent experi-ments. Surprisingly, the incubation vessel was a critical factorfor the testis explant assay. hCG-stimulated explants culturedin 20-ml glass scintillation vials produced approximately 9

ii

\*.

\A.

1J).

O«.

0J.

OJO.

10° 10J 10*

hCG Concentration (IU/ml)10"

FIG. 2. Testis explant assay; hCG dose-response experiment. Dose re-sponse of hCG stimulation of T production by 50-mg testis explants fromcontrol rats incubated in the presence (closed circles) or absence (open circle)of hCG. Explants were incubated with hCG for 3 h before collection ofmedium for measurement of T concentrations (n = 3).

Dow



ber 2022


TABLE 1Testis Explant Assay: Summary of Optimization Parameters

Conditions

Fetal calf serum

(+)(-)

Tissue amount50 mg200 mg

Incubation vesselScintillation vialGlass tube

N

22

33

22

Testosterone

(-)hCG

0.8 ±0.18"0.9 ± 0.01

0.3 ±0.130.1 ±0.01

0.8 ± 0.180.2 ± 0.03

production (ng/mg/h)

(+)hCG

2.7 ± 0.33*3.2 ± 0.20*

0.9 ±0.150.4 ± 0.08

2.7 ± 0.33*0.3 ±0.01**

Estradiol

(-)hCG

3.2 ± 1.260

1.7 ±0.032.0 ± 0.25

3.2 ± 1.260.8 ± 0.06

production (pg/mg/h)

(+)hCG

2.5 ± 0.270

2.2 ± 0.152.3 ± 0.23

2.5 ± 0.270.7 ± 0.25**

1 Means ± standard error.* Significantly different from (-) hCG." Significantly different from scintillation vial.

times more T and 3.6 times more E2 than explants cultured in16 X 100-mm borosilicate glass tubes (Table 1).

Based on the above experiments, the optimized assay con-ditions (which allow maximal hCG stimulation of T produc-tion) used for the compound evaluation experiments were asfollows: RPMI1640 medium without phenol red, 10% FCS, 50jig/ml medium soybean trypsin inhibitor, 50 mg testis, 1.0 IUhCG/ml medium, and a 3-h incubation time. The incubationmedium (M-199 versus RPMI 1640), inclusion of bovine se-rum albumin, or the addition of soybean trypsin inhibitor is notcritical to steroid production (data not shown). RPMI 1640medium was selected to be consistent with the Leydig cellisolation procedure used in this laboratory (Biegel et ai, 1993).

Compound evaluation. Following the optimization of thetestis explant assay, effects of the test compounds on steroidbiosynthesis were evaluated both ex vivo and in vitro. Resultsfrom the ex vivo experiments can be found in Table 2. 17/3-E2(0.0075 and 0.05 mg/kg/day) inhibited T production (48 and10% of control, respectively). ANA (0.75 and 1.5 mg/kg/day)inhibited E2 production (63 and 74% of control, respectively)and had no effect on T production. FLUT (20 mg/kg/day)increased T and E2 production (217 and 121% of control,respectively). RES (0.75 and 1 mg/kg/day) decreased T pro-duction (60 and 23% of control, respectively) and E2 produc-tion (64 and 40% of control, respectively). ICI, KETO, FIN,and HALO had no effect on either T or E2 production ex vivo.

In nine in vitro experiments with testis explants, hCG-stimulated T production was increased from 1.6- to 5.4-foldwhen compared to unstimulated explants; the mean fold in-crease in hCG-stimulated T production was 4.0 ± 0.5 (Table3). The responses observed in the in vitro experiments (Table3) differed from those observed in the ex vivo evaluation (Table2). ICI had no effect on T production, but increased E2 pro-duction at 0.1, 1, and 10 /xM (183, 391, and 878% of control,respectively). ANA inhibited T production at 100 JIM (57% of

control), but surprisingly had no effect on E2 production. AGdecreased T production at 100 and 1000 /xM (46 and 16% ofcontrol, respectively) and had no effect on E2 production.KETO decreased T production at 1 and 10 /xM (45 and 9% ofcontrol, respectively), but had no effect on E2 production. FINdecreased T and E2 production at 10 and 100 /xM (68 and 16%and 73 and 16% of control, respectively). HALO inhibited Tproduction in a dose-dependent manner with the greatest de-crease at 100 /xM (28% of control), but did not affect E2production. 17/3-E2, FLUT, and RES had no effect on T or E2production in vitro.

Leydig Cell Cultures

HALO, ICI, and ANA were examined using Leydig cellcultures. The concentrations of each test compound used were0 (±hCG), 0.1, 1, 10, and 100 ;xM (all + hCG). Similar to thetestis explants, HALO decreased T production in a dose-dependent manner (Table 4). Cytotoxicity was noted only atthe 100 /xM concentration (Control, 95.0% viable cells, 100/xM; HALO, 70.6%). ICI did not produce cytotoxicity and,similar to the testis explants, increased E2 production in adose-dependent manner (Table 4). ANA produced cytotoxicityat the three highest concentrations (Control, 98.5% viable cells;Treated, 60.0-75.2%), yet did not affect E2 or T production(Table 4).

Ovary Explant Assay

Optimization. One goal of these studies was to make theovary explant conditions as similar as possible to the optimizedtestis explant conditions to facilitate target organ toxicity com-parisons. Initial experiments compared hormone productionfrom ovaries obtained on different gestation days, left versusright fetus numbers, ovary weights, and hormone production,

Dow



ber 2022

66 POWLIN ET AL.

TABLE 2Testis Explant Assay: Ex Vivo Steroid Biosynthesis Experiments

Compound Dosage (mg/kg/day) Testosterone (ng/mg/h) Estradiol (pg/mg/h)

ICI-182,780

Anastrozole

Ketoconazole

Finasteride

Flutamide

Haloperidol

Reserpine

0 (+hCG)0.0010.00250.00750.0500(+hCG)0.050.250.51.00(+hCG)0.10.50.751.50(+hCG)2550751000 (+hCG)1510250(+hCG)0.2515200(+hCG)0.10.250.51.00(+hCG)0.10.50.751.0

0.31 :0.20:0.31 :0.15:0.03:1.04:1.11 :1.21 :1.57:1.06:0.90:0.66:0.77:0.69:0.66:1.29:0.79:1.04:1.09:0.79:1.13:0.67:0.71 :0.85:0.80:1.00:1.18:0.85 :1.72:2.17:0.80:0.56:0.90:0.66:0.88:0.60:0.74:0.71 :0.36:0.14:

0.028"0.0220.0580.032*0.007*0.2080.1930.2040.4200.2150.0980.0860.0680.0970.1100.2230.1230.2490.1290.2180.1380.1050.0840.0850.0840.0830.1190.0760.2150.309*0.1020.0800.1530.0730.0710.0570.1260.0830.096*0.028*

NDNDNDNDND

2.66 ± 0.2092.75 ± 0.1902.28 ± 0.1832.55 ± 0.1773.24 ± 0.2921.09 ±0.0950.98 ± 0.1020.85 ±0.1180.69 ± 0.067*0.81 ±0.107*2.85 ± 0.5762.79 ± 0.6402.92 ± 0.4872.90 ± 0.4712.13 ± 0.2031.46 + 0.1511.51 ±0.1681.57 ±0.1471.49 ±0.1941.74 ±0.2311.65 ±0.2531.63 ±0.1011.62 ±0.0801.53 ±0.1141.99 ±0.101*2.00 ± 0.1021.90 ±0.1321.65 ± 0.1461.90 ±0.1211.88 ±0.1361.68 ±0.1271.27 ±0.1501.42 ±0.1791.08 ±0.132*0.67 ±0.110*

Note. ND, not determined.° Means ± standard error, n = 5.* Significantly different (p £ 0.05) from control by Jonckheere's test for trend.

as well as culture vessels and stimulant (hCG, FSH, and8-bromo-cAMP).

Hormone production was evaluated on gd 8, 12, and 17. Thehighest production of E2 (Fig. 3A) and P4 (Fig. 3B) was seenon gd 8. AG (1000 /iM) inhibited E2 production on gd 8 (21%of control) and gd 12 (44% of control), but not on gd 17. P4production was inhibited by AG on on gd 8, 12, and 17 (3, 3,and 14% of control, respectively). In all subsequent experi-ments, ovaries from rats on gd 8 were evaluated. There were nodifferences in the number of fetuses in the left and right uterinehorns, in left and right ovary weights, or in left and right ovary

hormone production (Table 5). Based on these data, the leftovary was used as the unstimulated control and the right ovarywas stimulated with hCG. Hence, these conditions allowed asimilar experimental study design to the testis explants, where(+/—) hCG stimulation was evaluated using the same animal.

In examining the effect of incubation vessel on ovary ex-plant hormone production, P4 production was similar betweenmicrofuge tubes and scintillation vials, but E2 production wasalmost six-fold higher in microfuge tubes (25.3 ± 5.73 versus4.4 ± 0.63 pg/mg/h). Microfuge tubes were chosen as theincubation vessel for subsequent studies.

Dow



ber 2022


TABLE 3Testis Explant Assay: In Vitro Steroid Biosynthesis Experiments

Compound Dosage Testosterone (ng/mg/h) Estradiol (pg/mg/h)

17/3-Estradiol

ICI-182,780

Anastrozole

Aminoglutethimide

Ketoconazole

Finasteride

Flutamide

Haloperidol

Reserpine

0 (-hCG)0(+bCG)0.010.11100( -hCG)0(+hCG)0.010.11100 (-hCG)0(+hCG)0.11101000( -hCG)0(+hCG)11010010000 (-hCG)0 (+hCG)0.010.11100 ( - h C G )0(+hCG)0.11101000 ( - h C G )0 (+hCG)0.010.11100 (-hCG)0(+hCG)0.11101000(-hCG)0 (+hCG)0.1110100

0.20:0.31 :0.34:0.43 :0.30:0.42 :0.33 :1.04:0.80:1.14:0.67:0.77 :0.20:0.90:0.91 :0.60:0.72 :0.51 :0.19:0.96:0.90:0.84:0.44:0.15 :0.24:1.29:1.20:1.11 :0.58:0.11 :0.21 :1.13:1.29:1.11 :0.77 :0.18 :0.20:1.00:0.93 :0.97 :0.98 :1.10:0.22:0.80:0.44 :0.36:0.29:0.22:0.24:0.60:0.58:0.73:0.76:0.65:

: 0.029°: 0.028: 0.043: 0.047: 0.032: 0.071: 0.027: 0.208: 0.069: 0.309: 0.120: 0.100: 0.017: 0.098: 0.111: 0.100: 0.080: 0.028*: 0.032: 0.164: 0.207: 0.256: 0.107*: 0.042*: 0.043: 0.223: 0.208: 0.192: 0.065*: 0.026*: 0.011: 0.138: 0.161: 0.158: 0.116*: 0.036*: 0.025: 0.083: 0.093: 0.145: 0.1000.1470.0280.1020.090*0.049*0.055*0.021*0.0340.0570.1040.1270.0950.063

NDNDNDNDNDND

2.20 ± 0.1992.66 ± 0.2092.96 ± 0.2214.02 ± 0.210*8.61 ± 0.690*

23.36 ± 2.208*0.87 ± 0.1001.09 ±0.0951.04 ±0.1171.00 + 0.1351.16 ± 0.1470.95 ± 0.0911.06 ± 0.0941.60 ± 0.2001.40 ± 0.0981.53 ±0.1581.61 ± 0.1321.25 ± 0.1331.97 ± 0.3322.85 ± 0.5762.46 ± 0.4612.08 ± 0.2142.35 ± 0.4132.74 ± 0.7631.05 ±0.1831.46 + 0.1511.38 ±0.1621.31 ±0.1941.06 ±0.130*0.23 ± 0.060*1.35 ± 0.0811.65 ±0.2531.55 ±0.1271.41 ± 0.1261.78 + 0.1121.41 ±0.1101.74 ±0.1592.00 ±0.1022.08 + 0.1561.89 + 0.0882.21 +0.1531.71 ±0.1691.23 ± 0.2091.68 ±0.1271.78 ±0.3131.29 ±0.1671.71 ±0.2131.55 ± 0.189

Note. ND, not determined." Means ± standard error, n = 5.• Significantly different (p s 0.05) from (+) hCG control by Jonckheere's test for trend.

Dow



ber 2022

68 POWLIN ET AL.

TABLE 4In Vitro Leydig Cell Steroid Hormone Production

Concentration Testosterone Estradiol % ViableCompound (JJM) (pg/cell) (fg/cell) cells

Haloperidol

Anastrozole

ICI-182,780

0 ( - hCG)0 (+ hCG)0.11101000 ( - h C G )0(+hCG)0.11101000 ( - hCG)0 (+ hCG)0.1110100

0.203 :0.986 :0.783 :0.633 :0.611 :0.455 :0.200:1.574:1.068:1.172 :1.135:1.144 :0.146:1.053 :1.098:0.852 :0.776 :0.850:

: 0.004°: 0.043: 0.012*: 0.007*: 0.051*: 0.012*: 0.010: 0.059: 0.073: 0.117: 0.076: 0.047: 0.008: 0.136: 0.145: 0.129: 0.113: 0.103

0.583 :0.757 :0.551 :0.676 :0.818:1.275 :0.466:0.634:0.631 :0.635 :0.848:0.841 :0.373 :0.466:0.695 :1.567 :8.520 :

31.579 :

: 0.056: 0.023: 0.012: 0.034: 0.0510.127*0.0130.0300.0790.0770.158

: 0.146: 0.014: 0.045: 0.037: 0.101*: 0.191*: 0.766*

95.092.195.592.596.270.698.596.987.374.860.075.296.397.293.496.596.894.1

" Means ± standard error, n = 4.* Significantly different (p ^ 0.05) from control by Jonckheere's test for

trend.

hCG, FSH, and 8-bromo-cAMP were examined for theirability to stimulate E2 and P4 production. hCG produced lessthan two-fold increases in E2 production and approximatelytwo- to threefold increased in P4 production (Tables 6 and 7).To determine if hCG produced the maximal stimulation ofsteroid production, FSH and 8-bromo-cAMP were also exam-ined. FSH, which normally stimulates granulosa cells to syn-thesize E2, did not affect E2 production, but did produce adose-dependent increase in P4 production (data not shown).8-Bromo-cAMP, at concentrations up to 10 mM, did not in-

TABLE 5Ovary Explant Assay: Comparison of Left vs Right FetusNumber, Ovary Weight, and Steroid Hormone Production

End point N Left Right

Number of fetuses 20 6.30 ± 0.35" 6.35 ± 0.33Ovary weight (mg) 20 33.33 ± 0.94 33.66 ± 1.04Estradiol (pg/mg/h) 10 48.29 ± 5.54 50.49 ± 7.62Progesterone (ng/mg/h) 10 9.39 ± 0.88 8.98 ± 1.07

Note. Comparisons performed on gd 8. No endpoints were significantlydifferent between left and right.

" Means ± standard error, n = 5.

crease E2 production and increased P4 production to the samelevel as 1.0 IU/ml hCG (data not shown). Neither factor ap-peared to be better than hCG in terms of hormone stimulation;thus, hCG was used in the remaining experiments.

Compound evaluation. hCG stimulated P4 production to asimilar extent in both ex vivo and in vitro experiments (approx-imately 2.4 ± 0.13-fold) (Tables 6 and 7). Ex vivo, ICI had noeffect on P4 production (Table 6). Similar to what was seen inthe testis explants, ICI increased E2 production in a dose-dependent manner, with the greatest increase in the 1 mg/kg/day dose group (246% of control) (Table 6). ANA inhibited P4production in a dose-dependent manner, with the greatest de-crease in the 10 mg/kg/day dose group (31% of control), andinhibited E2 production at 10 mg/kg/day (32% of control)(Table 6).

In vitro, ICI increased E2 production at 10 /xM (273% ofcontrol) and decreased P4 production at 1 and 10 /xM (44 and64% of control) (Table 7). Surprisingly, ANA did not affect E2or P4 production (Table 7). AG inhibited E2 production at 10and 100 /iM (23 and 16% of control, respectively) and alsodecreased P4 production at 10 and 100 /xM (69 and 21% of

I I (-) Aminoglutethimidc! Aminoglutethimide

Gestation Dav

1 ) l-l Aminoglutethimidef^^ 1+) Aminoglutethimide

Gestation Dav

FIG. 3. Ovary explant assay: Steroid production on different gestation days. Comparison of E2 (A) and P4 (B) production on gd 8,12, and 17. Ovary explantswere incubated for 3 h with 1 IU/ml hCG in the presence (solid bars) or absence (open bars) of 1000 jxM AG (gd 8, n = 4; gd 12, n = 6, gd 17, n = 6) (*p s0.05 when compared to explants incubated in the absence of AG).

Dow



ber 2022


TABLE 6Ovary Explant Assay: Ex Vivo Steroid Biosynthesis Experiments

Compound

IQ-182,780

Anastrozole

Dosage(mg/kg/day)

0(-hCG)0 (+hCG)0.050.250.51.00(-hCG)0(+hCG)1.05.07.510.0

Estradiol(pg/mg/h)

17.3 ± 2.08"16.2 ± 2.7224.9 ± 5.9125.9 ± 3.6427.5 ± 2.63*39.8 + 4.91*17.2 + 2.1320.6 ± 3.1732.2 ± 4.5528.0 ± 3.1614.0 ± 0.986.5 ± 1.08*

Progesterone(ng/mg/h)

3.3 ± 0.546.9 ± 1.098.9 ± 1.289.0 ± 1.627.7 ± 0.479.0:4.1 :

10.9;5.5:5.3i3.3;

t 1.00t 0.45t 1.24t 0.63*t 0.87*t 0.39*

3.4 ± 0.42*

" Means ± standard error, n = 5.* Significantly different (p S 0.05) from (+) hCG control by Jonckheerc's

test for trend.

control, respectively). HALO did not affect either E2 or P4production (Table 7).

Spectral Binding Assays

The profile for a Type I spectra is a peak at 390 nm and atrough at 420 nm, while the profile for a Type II spectra is apeak at 430 nm and a trough at 390 nm. Both HALO (peak, 415nm; trough, 385 nm) and RES (peak, 415 nm; trough, 390 nm)exhibited Type II binding spectra, indicating that both HALOand RES bind to the heme iron of cytochrome P450 to inhibitenzyme activity (data not shown).

DISCUSSION

Testis and ovary explant assays have been proposed as invitro screens to primarily identify potential steroid biosynthesisinhibitors. The testis explant assay is designed primarily toidentify T biosynthesis inhibitors, while the ovary explantassay is designed to detect aromatase inhibitors (Gray et al,1997). The goals of the current study were to optimize theassay procedures, to evaluate whether the testis explant assaycould also detect aromatase inhibitors, and to validate thesesystems using several compounds with well-defined endocrineactivity in order to evaluate their utility as test systems forscreening unknown compounds for possible steroid biosynthe-sis inhibition activity.

Optimization

For the testis explant assay, the following parameters wereevaluated: incubation time, incubation vessel, amount of tissue,incubation medium, FCS (+/—), and hCG concentration. Ofthese six parameters, the most critical component was theincubation vessel. To obtain consistent hCG stimulation of T

biosynthesis, 20 ml scintillation vials must be used rather thanglass test tubes (16 X 100 mm). The size of the incubationvessel may have been critical because the 20-ml scintillationvial allows greater physical dispersion of the explants. The20-ml scintillation vial allows approximately two times asmuch dispersion as the 16 X 100-mm test tubes. Inclusion ofFCS into culture medium was necessary for measurement ofE2 by RIA, presumably by the presence of serum bindingproteins. Measurement of medium E2 levels from testis ex-plants is beneficial since it may facilitate detection of aro-matase inhibitors using testis explants and therefore eliminatethe need for the ovary assay. Consistent with previous work(Biegel et al., 1995), hCG does not stimulate E2 production.The inability to enhance estradiol biosynthesis is presumablydue to the substrate (i.e., testosterone) not being rate limitingbased on the apparent Km for aromatase of 40 nM (Hutchisonet al., 1997). To incorporate the testis assay into our ongoingTier I battery, it was necessary to use smaller explants thanpreviously described in the literature (quartered testis) (Laskeyet al, 1994; Gray et al., 1995). The 50-mg testis explants hadT production (0.9 ng/mg/h) that was approximately twofoldgreater than quartered testis (0.4-0.5 ng/mg/h) (Laskey et al,1994; Gray et al., 1995), demonstrating that smaller explantscan be used.

TABLE 7Ovary Explant Assay: In Vitro Steroid Biosynthesis Experiments

DosageCompound QiM)

ICI-182,780 0 (-hCG)0 (+hCG)0.010.11.010.0

Anastrozole 0 (-hCG)0 (+hCG)0.1110100

Aminoglutethimide 0 (-hCG)0 (+hCG)0.1110100

Haloperidol 0 (-hCG)0 (+hCG)0.1110100

Estradiol(pg/mg/h)

18.7 :24.7 :18.8 :24.3 :30.5 :67.5 :4.8 ;9.5 ;3.6;3.4:3.5;3.1 ;

14.7:21.4:23.6:

9 .3 :5.0;3.5;

14.0;23.4 :19.0;20.5 :18.8:17.6:

t 1.86°t 1.88t3.52t3.35t 3.41t 6.25*t 1.93t3.57t0.59t 0.43t0.52i 0.41t 1.10t6.93t 3.64t0.67t 1.15*t 0.28*t2.97t4.87:3 .U:3.92b6.27b2.9O

Progesterone(ng/mg/h)

4.7 :10.7:9.0:8.4:4.7:6.9:3.5:9 . 1 :6.9:8.5:7.8:5 .3:3.6:6 . 1 :7.2:5.7:4 .2 :1.3:2.7:4 .2 :7.4:6.2:5.7:4 .0 :

t 1.49tO.83tO.83t 1.26t 0.62*t 1.27*;0.37t 1.30i 1.18i 1.81t 0.65:0.96= 0.75:0.73t 0.75t 1.04:0.84*: 0.08*:0.12:0.78:0.69:0.14:0.55:0.33

" Means ± standard error, n = 5.* Significantly different (p =s 0.05) from (+) hCG control by Jonckheere's

test for trend.

Dow



ber 2022

70 POWLIN ET AL.

TABLE 8Testis Explant Assay: Summary of Expected and Observed Responses for Steroid Hormone Production

Compound

170-EstradiolICI-182,780AnastrozoleAminoglutethimideKetoconazoleFinasterideFlutamideHaloperidolReserpine

Testosteroneproduction

Expected

1—

ND—

r—i

Observed

1——ND——

t—1

Ex vivo explant

Estradiolproduction

Expected Observed

ND ND— —

- 1ND ND— —

—

t Ti i

Assay

predictiveof activity?

YesYesYes"NDYesYesYesYesYes

Testosteroneproduction

Expected Observed

— —

- Ii 14 1

- 1—

- i— —

In vitro explant

Estradiolproduction

Expected Observed

ND ND

- ti -1 -

— —- I

—— —

Assay

predictiveof activity?

YesNoNoNoYesNoYesYes*No'

Note, i , decrease; t . increase; —, no change; ND, not determined." Although no change in estradiol production was expected, the decrease is consistent with the activity of ANA; thus, the assay was considered predictive.* Although no change in testosterone production was expected, the decreases are supported because HAL exhibited a Type II binding spectra; thus, the assay

was considered predictive.c Although no change in testosterone production was expected, the ability of RES to produce Type II binding spectra suggests that testosterone production

should have been decreased; thus, the assay was considered not predictive.

The goal of the ovary explant assay optimization was tooptimize the conditions for detecting aromatase inhibitors andalso to make the assay as similar as possible to the testisexplant assay in order to facilitate target-organ toxicity com-parisons. In the current studies, ovaries from pregnant rats werechosen because steroid production is more predictable on agiven day of gestation and not subject to the fluctuations seenin cycling female rats (Piasek and Laskey, 1994; Gray et al,1997). Gestation day 8 was chosen since there is a higher levelof E2 production early in gestation, and this will facilitate thedetection of compounds that inhibit E2 production. By elimi-nating the need for monitoring of estrous cycle, the variability,complexity, and cost of this assay were reduced. One of thelimitations of using ovaries from time-pregnant rats is thatestradiol production is lower than in rats sacrificed in proestrus.Similar to the testis assay, the incubation vessel was importantfor optimal steroid (E2) production. For the ovary explantassay, the use of 1.5-ml microfuge tubes lying horizontalproduced the greatest E2 production, again, probably due to thegreater physical dispersion than was obtained with 20-ml scin-tillation vials. Because left and right ovary steroid productionwas found to be equivalent, an experimental design similar tothe testis explant assay was used, where the left ovary wasunstimulated and the right ovary was stimulated with hCG.Using this experimental design, side-by-side (+/—) hCG stim-ulation of ovaries produced a 2.4-fold increase in P4 produc-tion, whereas other investigators did not observe stimulation ofP4 production (Piasek and Laskey, 1994). In all likelihood, thisdifference is due to a different experimental design where thenon-hCG incubation was conducted in the first hour, and

compared to subsequent hCG stimulation in the second, third,and fourth hours.

Compound Evaluation

Responses of the testis and ovary explants to each testcompound were compared to the responses observed in theTier I male battery (O'Connor et al, 1998a,b), as well as thepublished literature, in order to determine the predictability ofthe testis and ovary assays for identifying steroid biosynthesisinhibitors. Predictability was determined by consistency of theresponses seen with the explants versus the well-characterizedactivity of the endocrine controls. Summaries of the predictedand observed responses can be found in Tables 8 and 9.

17/3-E2 (estrogen receptor agonist) produced a decrease in Tproduction in ex vivo testis explants, but not in vitro. The exvivo decrease in T production is most likely due to a combi-nation of Leydig cell atrophy (O'Connor et al., 1998a) anddownregulation of the C]7a-hydroxylase enzyme (Nishihara etal, 1988). Using isolated Leydig cells, 17/3-E2 does not pro-duce a decrease in T production in vitro until 3 days of culture(Tsai-Morris et al., 1986; Nishihara et al, 1988); hence,17/3-E2 would not be expected to produce a decrease in Tproduction in vitro under the current study design, which usesa 3-h incubation. In the in vivo Tier I male battery, 17/3-E2produced a dose-dependent decrease in serum T, DHT, and LHlevels (O'Connor et al, 1998a), consistent with the ex vivofindings.

ICI was predicted to be a negative control in the testis andovary explant assays because it had no effect in the in vivo Tier

Dow



ber 2022

TESHS/OVARY EXPLANTS FOR IDENTIFYING STEROID BIOSYNTHESIS INHIBITORS 71

TABLE 9Ovary Explant Assay: Predicted Responses for Steroid Hormone Production

Compound

ICI-182,780AnastrozoleAminoglutethimideHaloperidol

Estradiolproduction

Expected Observed

- t- 1ND NDND ND

Ex vivo explant

Progesteroneproduction

Expected Observed

- IND NDND ND

Assaypredictive

of activity?

NoYes"NDND

Estradiolproduction

Expected Observed

- Ti -i i

In vitro explant

Progesteroneproduction

Expected Observed

|

1 -1 1

Assaypredictive

of activity?

NoNoYesYes

Note, i , decrease; f, increase; —, no change; ND, not determined." Although no change in hormone production was expected, the decreases are consistent with the activity of ANA; thus, the assay was considered predictive.

I battery (O'Connor et al, 1998b) and since its endocrineactivity is characterized as a pure estrogen receptor antagonist(Wakeling et al, 1991). ICI did not affect ex vivo steroidproduction, yet it unexpectedly increased E2 production bytestis explants in vitro (8.8-fold compared to the hCG control).When isolated Leydig cells were cultured with identical con-centrations of ICI, E2 production was similarly increased in adose-dependent manner. Furthermore, an increase in E2 pro-duction was seen in both ex vivo and in vitro ovary explants.This paradoxical response is not consistent with in vivo find-ings (Wakeling et al, 1991; O'Connor et al, 1998b), and themechanism is unclear. One possibility is that ICI may interferewith the E2 RIA; however, no interference was detected whenmedium containing ICI was examined in the E2 RIA by usingICI-spiked calibrators with known concentrations of E2 (datanot shown). The implication is that the effective concentrationsof ICI were sufficient to elicit an agonistic response through alocal feedback loop that stimulates E2 production.

ANA is a potent and highly selective aromatase inhibitorwith no intrinsic hormonal activities (Dukes et al, 1996). Itwas expected that ANA would inhibit E2 production and haveno effect on T production. This pattern of responses was seenin the Tier I male battery (O'Connor et al, 1998b) and in theex vivo testis explant assay. However, in vitro T productionwas inhibited and there was no effect on E2 production. Theinability to detect ANA as an aromatase inhibitor in the in vitrotestis explant assay probably reflects the low aromatase activitywithin the testis. Furthermore, when isolated Leydig cells wereexposed to ANA, T and E2 production were not affected. ANAhas not been previously reported to inhibit T production, al-though it has been shown to inhibit other cytochrome P450

isozymes in vitro when tested at relatively high concentrations(PDR, 1996). In the ovary explant assay, ANA decreased bothE2 and P4 production ex vivo but not in vitro. The decrease inE2 production observed ex vivo was unexpected since the exvivo assessment is not designed to detect compounds thatdirectly inhibit enzyme activity but is designed to identifycompounds that up- or downregulate the enzymes of the ste-

roidogenic pathway or compounds that cause atrophy or hy-perplasia of Leydig, theca, or granulosa cells. The reason E2production was not affected in vitro in the testis or ovary assayis unclear. Repeated administration of aromatase inhibitors hasbeen shown to decrease estrogen levels by inducing estrogenmetabolism (Purba et al, 1994), which may be another meansby which ANA decrease E2 ex vivo, but not in vitro. Theseresults suggest diat both the in vitro testis and ovary explantassays are limited in their ability to detect aromatase inhibitors.

KETO is a T biosynthesis inhibitor (reviewed in Feldman,1986). In the testis explant assay, KETO did not affect T or E2production ex vivo. As expected, in vitro, KETO decreased Tproduction and had no effect on E2 production. The inability ofKETO to decrease T production ex vivo has been previouslyshown (Kan et al, 1985; Pont et al, 1982). The discrepancybetween the in vitro and ex vivo results is attributed to thedilution of residual KETO when the testis from treated rats isplaced into the culture medium, a finding consistent withenzyme inhibitors that act by directly binding to inhibit thesteroidogenic pathway. In the Tier I battery, KETO loweredserum T and DHT, increased LH and FSH, and had no effecton PRL (O'Connor et al, 1998b). Hence, there was concor-dance between the testis explants, the in vivo Tier I malebattery, and the published literature.

AG is a mixed aromatase/C^-side chain cleavage inhibitor(Salhanick, 1982). In the in vitro testis explant assay, AGdecreased T but not E2 production, suggesting that only theC27-side chain cleavage enzyme is being inhibited. As ex-pected based on published literature (Steinetz et al, 1985), AGdecreased both E2 and P4 production in the in vitro ovaryexplant assay. The inability to detect AG and ANA as aro-matase inhibitors in the in vitro testis explant assay probablyreflects the low aromatase activity within the testis.

In the testis explant assay, FIN, a 5a-reductase inhibitor, hadno effect on hormone production ex vivo, but decreased T andE2 production in vitro. These data are not predicted based onthe pharmacologic activity (Rittmaster, 1994). The reason forthe decrease is unclear and may be due to cytotoxicity. As-

Dow



ber 2022

72 POWLIN ET AL.

sessment of FIN in the Tier I male battery revealed decreasesin serum DHT levels and increases in LH and PRL levels,which is consistent with its mode of action (O'Connor et al,1998b). Therefore, the data from the testis explant assay werenot consistent with the expected results based on the data fromthe Tier I male battery (O'Connor et al, 1998b) or the pub-lished literature.

As expected, FLUT, an androgen receptor antagonist, in-creased T production in the ex vivo testis explant assay, but notin vitro. In the Tier I male battery, FLUT increased serum T,DHT, E2, FSH, and LH levels and produced diffuse Leydigcell hyperplasia (O'Connor et al, 1998b). Hence, the increasein T production in the ex vivo testis explant assay is attributedto the Leydig cell hyperplasia and increased steroidogeniccapacity due to hypertrophy as a result of the increased serumLH levels (Ewing and Zirkin, 1983; Wing et al, 1984; Mendis-Handagama et al, 1988; Cook et al, 1993). These data dem-onstrate concordance between the testis explants, the in vivoTier I male battery, and the published literature (Neri et al,1972; Simard et al, 1986; O'Connor et al, 1998b).

HALO, a dopamine receptor antagonist, had no effect onhormone production in the ex vivo testis explant assay or in thein vitro ovary explant assay, but inhibited T production in thein vitro testis assay. Similarly, when isolated Leydig cells wereexposed to HALO in vitro, T production was inhibited in adose-dependent manner. In the Tier I male battery, HALOincreased serum PRL, but had no other effects (unpublisheddata). It is unclear why the decrease in T production was notobserved in the ex vivo testis assay or in the Tier I battery;however, it seems possible that the dose was not high enoughto inhibit T biosynthesis based on the absence of any bodyweight effects in the HALO-treated rats. HALO has beenshown to suppress testosterone production in rats and humans,possibly via a central effect (Okonmah et al, 1986; Rinieris etal, 1990). However, the in vitro testis explant and Leydig celldata suggest that HALO is acting directly on the Leydig cell. Insupport of this, HALO produced a Type II binding spectrumwith cytochrome P450 (Lewis, 1986), confirming a direct in-hibitory mechanism. These data demonstrate agreement be-tween the testis explant assay and published literature.

RES, a dopamine depletor, inhibited both T and E2 produc-tion in the ex vivo testis explant assay. These decreases appearsecondary to Leydig cell atrophy (unpublished data). Similar toHALO, RES displayed a Type II binding spectrum with cyto-chrome P4J0, suggesting that it binds to the heme iron to inhibitactivity. However, in the in vitro testis explant assay, T and E2production were not affected, suggesting that RES does notappear to be directly inhibiting testosterone biosynthesis. Thereason for this discrepancy is unclear. In the Tier I malebattery, RES decreased serum T, DHT, E2, LH, and FSH(unpublished data). The magnitude of the decrease in LH andFSH, coupled with die testis explant data, indicates that RES isalso acting centrally. Because histopathological examination of

the testis demonstrated the presence of Leydig cell atrophy, theTier I male battery and testis explant assay data are consistent.

Conclusions

One of the goals was to determine if the testis explant assaycould be used to detect aromatase inhibitors. Addition of FCSto the medium allowed measurement of E2 production by testisexplants, which was not inhibited by ANA or AG. It seemslikely that the failure of the testis explant assay to detectaromatase inhibitors may be due to the low aromatase activitywidiLn the testis. In the ovary explant assay, only AG wasidentified as inhibiting E2 production in vitro. This was sur-prising given that the ovary has higher levels of aromatase thanthe testis, and ANA is a more specific aromatase inhibitor thanAG (Dukes et al, 1996). The reason ANA did not inhibit E2production is unclear. Under the current assay conditions, thetestis and ovary explant assays appear to have limited utility indetecting aromatase inhibitors.

A major limitation of the ovary and testis assays is thedifficulty in assessing cytotoxicity. The decreases seen in hor-mone production (ANA, AG, KETO, FIN, HALO, and RES)could possibly be due to cytotoxicity, rather than a direct effectof the compound. Viability measurements in the isolated Ley-dig cell cultures indicated that there may be some cytotoxicityassociated with exposure to ANA and high doses of HALO;however, ANA did not affect E2 production in vitro usingLeydig cell cultures or the testis explants. Standard means ofmeasuring cytotoxicity (lactate dehydrogenase production,MTT) may be confounded in these assays due to interferencefrom the high levels of fetal calf serum (10%) as well as celldamage due to the physical dispersion from the shaking duringincubation. Clearly, a reliable measure of cytotoxicity shouldbe developed before the testis and ovary explants are used forexamining unknown compounds for steroid biosynthesis inhi-bition activity.

The in vitro and ex vivo testis/ovary explant assays detecttwo different responses to EACs. The in vitro assays canpotentially identify compounds that directly inhibit steroidbiosynthesis, while the ex vivo assays indirectly identify com-pounds that produce either hyperplasia and/or atrophy of ste-roid-producing cells (Leydig, theca, and granulosa), secondaryto altered steroid production. The in vitro testis explant assaycorrectly identified four of nine compounds based on publishedliterature and the data from the Tier I in vivo battery (17/3-E2,KETO, FLUT, and HALO). For this data set, which usedcompounds with diverse endocrine activities, the predictabilityof die in vitro testis explant assay was 44%, suggesting that afalse positive or negative rate of 56% could occur with un-known compounds (Table 8). The ex vivo testis explant assaycorrectly identified eight of eight compounds. Based on theseresults, the ex vivo testis explant assessment would give alower false positive rate than a combination of an ex vivo/invitro testis explant assessment. However, the ex vivo testis

Dow



ber 2022


explant assay does not identify compounds that have the po-tential to directly inhibit steroid biosynthesis, but rather iden-tifies compounds that produce either atrophy or hyperplasia ofthe Leydig cells, albeit indirectly. The predictability of both thein vitro and ex vivo ovary assay was 50%, suggesting a 50%false positive or negative rate with unknown compounds (Ta-ble 9). It appears from these data that the ovary explant assaywill not be useful as a screening assay for identifying aro-matase inhibitors based on its inability to detect ANA. Theability of the ovary explant assay to detect other types ofsteroid biosynthesis inhibitors has not been evaluated. In con-trast, histopathology of the testis, which is included in the TierI battery, would detect all compounds that produced Leydigcell atrophy/hyperplasia and therefore would identify potentialsteroid biosynthesis inhibitors. Furthermore, of the seven com-pounds assessed to date (17/3-E2, ICI, ANA, KETO, FLUT,HALO, and RES), all were correctly identified using the TierI male battery, which also has the ability to detect otherendocrine activities (O'Connor et al., 1998a,b). Therefore, thetestis and ovary explant assay would not be necessary if onewas using an in vivo Tier I battery, since this screen wouldidentify a steroid biosynthesis inhibitors, and would also iden-tify several other endocrine responses. Based on the results ofthis study, the testis and ovary explant assays would not beuseful as routine screening procedures for detecting steroidbiosynthesis inhibitors. This conclusion is based on the highfalse positive/negative rates and the difficulties in assessingcytotoxicity. Clearly, the results of this work, coupled with ourongoing validation of a Tier I screening battery (O'Connor etal., 1998a,b), suggest that in vivo models should be used toidentify EACs, followed by confirmation in either in vitro ordefinitive tests such as a multigeneration reproduction study.

ACKNOWLEDGMENTS

We acknowledge the invaluable advice of Ms. Ann M. Mason (ChlorineChemistry Council) and Drs. James A. Barter (PPG Industries), Robert E.Chapin (NTBHS), A. Michael Kaplan (DuPont), William R. Kelce (USEPA),Roland R. Miller (Dow Chemical Company), and Ellen K. Silbergeld (Uni-versity of Maryland at Baltimore). The expert technical assistance of Bryan W.Crossley, Brian P. Shertz, and Vivian Thompson is also appreciated.

REFERENCES

Ankley, G. T., Johnson, R. D., Toth, G., Fomar, L. C , Detenbeck, N. E., andBradbury, S. P. (1997). Development of a research strategy for assessing the

ecological risk of endocrine disruptors. Rev. ToxicoL 1, 231-267.

Ashby, J., Houthoff, E., Kennedy, S. J., Stevens, J., Bars, R., Jekat, F. W.,Campbell, P., Van Miller, J., Carpanini, F. M., and Randall, G. L. P. (1997).The challenge posed by endocrine-disrupting chemicals. Environ. Health

Perspect. 105, 164-169.

Berman, E., and Laskey, J. W. (1993). Altered steroidogenesis in whole-ovary

and adrenal culture in cycling rats. Reprod. ToxicoL 7, 349-358.

Biegel, L. B., Cook, J. C, and Hum, M. E. (1993). Isolation and primaryculture of Leydig cells. In Methods in Reproductive Toxicology (R. E.Chapin and J. J. Heindel, Eds.), Vol. 3A, pp. 182-196. Academic Press,New York.

Biegel, L. B., Liu, R. C. M., Hunt, M. E., and Cook, J. C. (1995). Effects ofammonium perfluorooctanoate on Leydig cell function: In vitro, in vivo, andex vivo studies. Toxicol. Appl. Pharmacol 134, 18-25.

Bimbaum, L. S. (1994). Endocrine effects of prenatal exposure to PCB'sdioxins, and other xenobiotics: Implications for policy and future research.Environ. Health Perspect. 102, 676-679.

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

Carney, E. W., Hoberman, A. M., Farmer, D. R., Kapp, R. W., Jr., Nikiforov,A. I., Bernstein, M., Hunt, M. E., Breslin, W. J., Cagen, S. Z., and Daston,G. P. (1997). Estrogen modulation: Tiered testing for hazard evaluation.Repro. ToxicoL, in press.

Clark, J. H., and Markaverich, B. M. (1983). The agonistic and antagonisticeffects of short acting estrogens: A review. Pharmacol. Ther. 21, 429—453.

Colborn, T., Dumanoski, D., and Myers, J. P. (19%). In Our Stolen Future:Are We Threatening Our Fertility, Intelligence, and Survival? A ScientificDetective Story. Dutton Books, New York.

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

Cook, J. C , Kaplan, A. M., Davis, L. G., and O'Connor, J. C. (1997).Development of a tier I screening battery for detecting endocrine activecompounds. Regul. Toxicol. Pharmacol. 26, 60-68.

Cook, J. C , Mullin, L. S., Frame, S. R., and Biegel, L. B. (1993). Investigationof a mechanism for Leydig cell tumorigenesis by linuron in rats. Toxicol.AppL Pharmacol. 119, 195-204.

Crisp, T. M., Clegg, E. D., Cooper, R. L., Anderson, D. G., Baetcke, K. P.,Hoffmann, J. L., Morrow, M. S., Rodier, D. J., Schaeffer, J. E., Touart,L. W., Zeeman, M. G., Patel, Y. M., and Wood, W. P. (1997). In SpecialReport on Environmental Endocrine Disruption: An Effects Assessment andAnalysis. EPA/630/R-967012.

Dukes, M., Edwards, P. N., Large, M., Smith, I. K., and Boyle, T. (1996). Thepreclinical pharmacology of "arimidex" (anastrozole; ZD1O33)—A potent,selective aromatase inhibitor. J. Steroid Biochem. Mol. Biol. 58, 439—445.

Ewing, L. L., and Zirkin, B. R. (1983). Leydig cell structure and steroidogenicfunction. Recent Prog. Horm. Res. 39, 599-635.

Feldman, D. (1986). Ketoconazole and other imidazole derivatives as inhibi-tors of steroidogenesis. Endoc. Rev. 7, 409-420.

Gill, W. B., Schumacher, F. B., Straus, F. H., and Schoenberg, H. W. (1979).Association of diethylstilbestrol exposure in utero with cryptorchidism,testicular hyperplasia and semen abnormalities. J. UroL 122, 36-39.

Goodman and Gilman (1996). In Goodman & Gilman's Pharmacologic Basisof Therapeutics (J. G. Hardman, A. G. Gilman, and L. E. Limbird, Eds.), 9thed. McGraw-Hill, New York.

Gray, L. E., Jr., Kiinefelter, G., Kelce, W., Laskey, J., Ostby, J., and Ewing, L.(1995). Hamster Leydig cells are less sensitive to ethane dimethanesulfonatewhen compared to rat Leydig cells both in vivo and in vitro. Toxicol. Appl.Pharmacol. 130, 248-256.

Gray, L. E., Jr., Kelce, W. R., Wiese, T., Tyl, R., Gaido, K., Cook, J.,Kiinefelter, G., Desaulniers, D., Wilson, E., Zacharewski, T., Waller, C ,Foster, P., Laskey, J., Reel, J., Giesy, J., Laws, S., McLachlan, J., Breslin,W., Cooper, R., Di Giulio, R., Johnson, R., Purdy, R., Mihaich, E., Safe, S.,Sonnenschein, C , Welshons, W., Miller, R., McMaster, S., and Colbom, T.(1997). Endocrine screening methods workshop report: Detection of estro-genic and androgenic hormonal and antihormonal activity for chemicals thatact via receptor or steroidogenic enzyme mechanisms. Reprod. Toxicol. 11,719-750.

Herbst, A., Ulfelder, H., and Poskanzer, D. C. (1971). Adenocarcinoma of thevagina: Association of maternal stilbestrol therapy with tumor appearance inyoung women. N. EngL J. Med. ISA, 878-881.

Holmes, P., Humfrey, C , and Scullion, M. (1998). In Appraisal of Test

Dow



ber 2022

74 POWLJN ET AL.

Methods for Sex-Hormone Disrupting Chemicals. OECD Environment Di-rectorate, Environmental Health and Safety Division, Paris.

Hutchison, J. B., Beyer, C , Hutchison, R. E., and Wozniak, A. (1997). Sexdifferences in the regulation of embryonic brain aromatase. J. SteroidBiochem. Mol. BioL 61, 315-322.

IEH (1995). In Environmental Oestrogens: Consequences to Human Healthand Wildlife. Institute for Environment and Health, Norwich.

Kan, P. B., Hirst, M. A., and Feldman, D. (1985). Inhibition of steroidogeniccytochrome P-450 enzymes in rat testis by ketoconazole and related imida-zole anti-fungal drugs. /. Steroid Biochem, 23, 1023-1029.

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 Tilsonj,H. A. (1996). Research needs for the risk assessment of health and envi-ronmental effects of endocrine disrupters: A report of the U.S. EPA-sponsored workshop. Environ. Health Perspect. 104, 715-740.

Klinefelter, G., and Kelce, W. R. (1996). Leydig cell responsiveness tohormonal and nonhormonal factors in vivo and in vitro. In 77K; Leydig Cell(A. H. Payne, M. P. Hardy, and L. D. Russell, Eds.), pp. 535-553. CacheRiver Press, Vienna.

Laskey, J. W., and Berman, E. (1993). Steroidogenic assessment using ovaryculture in cycling rats: Effects of bis(2-diethylhexyl)phthalate on ovariansteroid production. Reprod Toxicol. 7, 25-33.

Laskey, J. W., Berman, E., and Ferrell, J. M. (1995). The use of culturedovarian fragments to assess toxicant alterations in steroidogenesis in theSprague-Dawley rat Reprod Toxicol 9, 131-141.

Laskey, J. W., Klinefelter, G. R., Kelce, W. R., and Ewing, L. L. (1994).Effects of ethane dimethanesulfonate (EDS) on adult and immature rabbitLeydig cells: comparison with EDS-treated rat Leydig cells. BioL Reprod50, 1151-1160.

Lewis, D. (1986). Physical methods in the study of the active site geometry ofcytochromes P-450. Drug Metab. Rev. 17, 1-66.

Mendis-Handagama, S. L. M. C , Zirkin, B. R., and Ewing, L. L. (1988).Comparison of components of the testis interstitium with testosterone se-cretion in hamster, rat and guinea pig testes perfused in vitro. Am. J. Anat.181, 12-22.

Neri, R., Florence, K., Koziol, P., and Cleave, S. V. (1972). A biologicalprofile of a nonsteroidal antiandrogen, SCH 13521 (4'-nitro-3'-trifluorom-ethylisobutyranilide). Endocrinology 91, 427-437.

Nishihara, M., Winters, C. A., Buzko, E., Waterman, M. R., and Dufau, M. L.(1988). Hormonal regulation of rat Leydig cell cytochrome P-45017a mRNAlevels and characterization of a partial length rat P-45O17a cDNA. Biochem.Biophys. Res. Commun. 154, 151-158.

O'Connor, J. C , Frame, S. R., Biegel, L. B., Cook, J. C , and Davis, L. G.(1998a). Sensitivity of a Tier I screening battery compared to an in uteroexposure for detecting the estrogen receptor agonist 17/3-estradiol. ToxicolSet 44, 169-184.

O'Connor, J. C , Cook, J. C , Slone, T. W., Frame, S. R., and Davis, L. D.(1998b). An ongoing validation of a Tier I screening battery for detectingendocrine-active compounds (EACs). Toxicol. Sci. 46, 45-60.

Okonmah, A. D., Bradshaw, W. G., Coueyro, P., and Soliman, K. F. A. (1986).The effect of neuroleptic drugs on serum testosterone level in the male rat.Gen. Pharmacol 17, 235-238.

Physicians' Desk Reference (1996). Vol. 50, 13th ed. Medical EconomicsData, Oradell, NJ.

Piasek, M., and Laskey, J. W. (1994). Acute cadmium exposure and ovariansteroidogenesis in cycling and pregnant rats. Reprod. Toxicol 8, 495-507.

Pont, A., Williams, P. L., Azhar, S., Reitz, R. E., Bochra, C, Smith, E. R., andStevens, D. A. (1982). Ketoconazole blocks testosterone synthesis. Arch.Intern, Med. 142, 2137-2140.

Purba, H. S., King, E. J., Richert, P., and Bhatnagar, A. S. (1994). Effect ofaromatase inhibitors on estrogen 2-hydroxylase in rat liver. / SteroidBiochem. Mol. Biol 48(2/3), 215-219.

Reel, J., Lamb, J., and Neal, B. (1996). Survey and assessment of mammalianestrogen biological assays for hazard characterization. Fundam. Appl Toxi-col. 34, 288-305.

Rinieris, P., Hatzimanolis, J., Markianos, M., and Stefanis, C. (1990). Effectsof treatment with various doses of haloperidol on the pituitary-gonadal axisin male schizophrenic patients. Neuropsychobiology 22, 146-149.

Rittmaster, R. S. (1994). Finasteride. N. EngL J. Med 330, 120-125.

Safe, S. H. (1995). Environmental and dietary estrogens and human health: Isthere a problem? Environ. Health Perspect. 103, 346-351.

Shalanick, H. A. (1982). Basic studies on aminoglutethimide. Cancer Res. 42,3315s-3321s.

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

Simard, J., Luthy, I., Guay, J., Belanger, A., and Labrie, F. (1986). Charac-teristics of interaction of the antiandrogen flutamide with the androgenreceptor in various target tissues. Mol Cell. Endocrinol. 44, 261-270.

Steinetz, B. G., O'Byrne, E. M., Sawyer, W. K., Butler, M. C , Munigle, J., andSteele, R. E. (1985). Effects of aminoglutethimide on cervical dilatabilityand serum immunoreactive relaxin in pregnant rats (41989). Proc. Soc. Exp.Biol Med 178, 101-104.

Tsai-Morris, C. H., Knox, G., Luna, S., and Dufau, M. L. (1986). Acquisitionof estradiol-mediated regulatory mechanism of steroidogenesis in culturedfetal rat Leydig cell. / Biol. Chem. 261, 3471-3474.

Wakeling, A. E., Dukes, M., and Bowler, J. (1991). A potent specific pureantiestrogen with clinical potential. Cancer Res. 51, 3867-3873.

Wing, T.-Y., Ewing, L. L., and Zirkin, B. R. (1984). Effect of LH withdrawalon Leydig cell smooth endoplasmic reticulum and steroidogenic reactionswhich convert pregnenolone to testosterone. Endocrinology 115, 2290-2296.

Dow



ber 2022

Ex Vivo and in Vitro Testis and Ovary Explants

Documents