* National Center for Environmental Assessment, U.S. Environmental Protection Agency (8623-D), Ariel Rios Building, 1200 Pennsylvania Avenue NW, Washington, DC 20460;
Department of Biostatistics, Virginia Commonwealth University, Richmond, Virginia 23298;
Laboratories for Reproductive Biology, Department of Pediatrics, and Department of Biochemistry and Biophysics, University of North Carolina, Chapel Hill, North Carolina 27599; and
Drug Development Toxicology, Pharmacia Corporation (7226300228), 7000 Portage Road, Kalamazoo, Michigan 49001
Received March 25, 2002; accepted June 27, 2002
![]() |
ABSTRACT |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Key Words: endocrine disruptor; androgen antagonist; antiandrogen; vinclozolin; critical windows; male development; response-surface modeling; testosterone; androgen receptor.
![]() |
INTRODUCTION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
In vivo dose-response data for vinclozolin male rat developmental effects have revealed at least two critical windows of exposure during prenatal and prepubertal development. Vinclozolin exposure to male rats via the pregnant dam during the time of sexual differentiation, gestation day 14 to postnatal day 3, at 100 mg/kg/day resulted in a female-like anogenital distance, nipple retention, cleft phallus with hypospadias, suprainguinal ectopic scrota/testes, vaginal pouch, epididymal granulomas, and small to absent sex accessory glands (Gray et al., 1994). Statistically significant effects on anogenital distance were observed after treatment with vinclozolin doses as low as 12.5 mg/kg/day using the same treatment regimen (Gray et al., 1999
). Prenatal vinclozolin treatment studies suggest that the critical window for exposure is GD 1419 (Wolf et al., 2000
). Additional antiandrogenic effects have been observed following prepubertal exposure: vinclozolin (100 mg/kg/day) treatment beginning on postnatal day (PND) 22 (day of weaning) led to delayed pubertal onset (as measured by preputial separation), decreased weight of the epididymis, ventral prostate, and seminal vesicles as well as increased serum testosterone, 5
-androstanediol, and luteinizing hormone in the male rat (Monosson et al., 1999
). In vitro and in vivo results predict that the observed developmental effects are a consequence of inhibited or blocked expression of androgen-dependent genes in androgen-responsive tissues (Kelce et al., 1995).
The two critical windows for vinclozolin susceptibility likely correspond to the time of androgen-dependent male sexual development and differentiation events. Androgen action is predicted to be affected by the concentrations of androgens, AR, sex hormone binding globulin and accessory proteins required for transcriptional activity. T and AR levels vary throughout development in humans and rodents (reviewed by Euling and Kimmel, 2001). The free T concentration ([free T]) ranges during the critical windows for external genitalia differentiation, urogenital tract differentiation, and puberty are shown in Table 1
. Compared to 0.170.73 nM free T in adult males (0.050.21 ng/ml; Fisher, 1998
), [free T] during these three critical windows is lower (Table 1
) on average possibly leading to an increased susceptibility of the fetus to effects of androgen antagonists compared to the adult. While less is known about human AR concentrations and tissue expression levels during critical windows of male development, rodent data indicate that AR is detected in urogenital tissues coincident with urogenital tract differentiation and the presence of relatively high fetal androgen levels (Bentvelsen et al., 1995
; Cooke et al., 1991
).
|
This assay allowed for the manipulation of the concentrations of two developmental stage components, AR and DHT, and for direct measurement of gene expression changes resulting from effects on androgen-AR complex transcriptional activation (Fig. 1A). This transient cotransfection model system performed in cultured cells has been used previously to characterize the antagonistic effects of vinclozolin (Wong et al., 1995
) and other AR antagonists (Kemppainen et al., 1999
) and reflects the in vivo AR activity in terms of high affinity induction of gene transcription by testosterone and DHT and its modulation by coactivator proteins (He et al., 1999
, 2001
). Dose-response data were generated at different concentration combinations of M2, AR, and DHT and statistical modeling of the dose-response data was performed to address the four questions. Response-surface modeling of the dose-response data was used to determine whether each component had an additive, synergistic, or antagonistic effect on inhibition of transcription by M2 and to compare the shapes of the response surfaces at each DHT concentration. Statistical analysis (parameter estimation and associated hypothesis testing) was used to determine whether the relationship of AR concentration to the M2 response changed at different DHT concentrations.
![]() |
MATERIALS AND METHODS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Twelve separate experiments were conducted using various concentration combination designs for AR vector, M2, and DHT concentrations. The M2 concentrations ([M2]) that were tested ranged from 50 nM to 1 µM. The DHT concentrations ([DHT]) ranged from 0.001 to 1 nM; the range included the maximal dose response range of luciferase expression in the in vitro assay with the maximum response occurring at approximately 0.1 nM DHT, the equilibrium binding constant (Kd) for the androgen receptor. The response at 0.001 nM DHT was at the limit of detection and at 1 nM DHT sometimes showed a decrease in luciferase response, likely reflecting a limiting concentration of transcription cofactors. The AR expression vector concentration ([AR vector]) ranged from 10 to 200 ng/dish. For each experiment, duplicate data points were performed for each three-way dose combination. For the data analysis and modeling, dose-response data from all 12 experiments was included, with the exception that response data for AR vector at 200 ng/dish were excluded. Of the 12 experiments, four included 25 ng/dish AR vector; 0, 50, 100, 200, 400, or 600 nM M2; and 0.01, 0.1, or 1 nM of DHT. Three experiments used either 10 or 50 ng/dish of AR vector; 0, 50, 100, 250, or 500 nM M2; and 0.01, 0.1, or 1 nM DHT. The remaining 5 experiments used designs with concentrations that were different but overlapped with each other and those in the 7 experiments (above); as a group, these included 0.001, 0.01, 0.1, and 1 nM DHT; 0, 10, 50, 100, 200, 500, and 1000 nM M2; and 10, 20, 25, 50,100, and 200 ng/dish AR vector. The amount of AR expression vector DNA of 1 µg or less was shown to be directly proportional to the level of expressed AR protein, using the method of calcium phosphate precipitation for DNA transfection and by densitometric analysis of protein on immunoblots (Choong et al., 1996). AR expression vector concentration was therefore referred to as AR concentration.
The luciferase activity response data were modeled as the fraction of mean control to normalize the response to the mean response of the positive controls (i.e., controls with DHT and without M2). The fraction of mean control was calculated as (mean luciferase response for replicates #1 and 2)/(DHT positive control luciferase response). Data are presented in this manner to remove interexperiment variability in the luciferase response, allowing for data from different experiments to be analyzed together. The sample means and variances for each concentration group were calculated.
Statistical modeling and hypothesis testing.
To determine the relationships among the three components of the system, [DHT], [AR], and [M2], on the luciferase response and to account for the intra- and interexperiment variability, a multivariate nonlinear model with associated hypothesis testing was performed. A modeling scheme was based on fitting the observed sample means across replicate values (Seber and Wild, 1989). Thus, the sample mean for each concentration group was the average of two values. The assumption was made that the interexperiment variability (variances across experimental groups) was constant. The model accounted for the inter- and intra-experiment variability. However, it allowed for the assumption that intra-experiment observations were correlated. The correlation structure assumed that the intra-experiment variability was compound symmetric, which assumes a nonzero common correlation for within-experiment observations. Hypothesis testing used a general linear hypothesis with an appropriate F test.
The definition of additivity used is given by Berenbaum (1985) and is based on the classical isobolograms for the combination of two chemicals (Loewe, 1953; Loewe and Muischnek, 1926
). For a combination of c chemicals, let Xi represent the concentration of the ith component alone that yields a fixed response, y0, and let xi represent the concentration of the ith component in combination with the c agents that yields the same response. According to this definition of additivity, if the substances combine with zero interaction, then
![]() | (1) |
The left side of Equation 1 is equal to 1 for an additive relationship (either a linear addition or subtraction of response) between the chemical components and the response. The left side of Equation 1
is less than 1 for a synergistic relationship between the components measured and the response. The left side of Equation 1
is greater than 1 for an antagonistic relationship between the components and the response.
Analysis using a Gompertz Nonlinear Model.
A nonlinear model was chosen for these data as a sigmoid-shaped relationship was expected for dose-response data. The form of the model was based on a Gompertz model parameterized as follows:
![]() | (2) |
For these data, / was assumed known and was specified at two values that were above the observed responses (2 and 5) for comparison. Here, the parameter /
was fixed as the response-surface did not plateau within the experimental region, and the maximum response was not observed. To account for the intra-experimental variability of the luciferase response in this assay, within-experiment observations could be correlated. Observations across experiments were assumed to be independent. The response means from duplicates of each dose combination and variances for each group were calculated. The model given in Equation 2
was fit to the sample means. The variance of each group was assumed to be constant across concentrations. In addition, the correlation of observations within an experiment was assumed to be constant. Thus, a compound symmetric correlation structure was assumed for (yij, yij', where j
j'). The method of maximum likelihood was used to estimate unknown model parameters using the MIXNLIN macro in SAS version 6.12 (SAS, 1999).
The DHT concentrations tested were 0.001, 0.01, 0.1, and 1 nM. As the model given in Equation 2 is based on the arithmetic concentration scale and not the log scale, the concentrations of DHT do not uniformly span the experimental region. One concern was that the responses at 1 nM DHT may overly influence the model fit and associated inference. Therefore, to address the effect of the fit of DHT concentrations, we considered a model that allowed for a response surface associated with AR and M2 at each fixed DHT level. More specifically, x'ß in Equation 2
was parameterized as follows:
![]() | (3) |
![]() |
RESULTS |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
|
|
|
|
Effect of [M2], [AR] and their interaction at fixed levels of DHT.
To determine whether [M2] and [AR] had an additive effect on the response, the fitted four-dimensional relationships are described as three-dimensional surfaces at fixed DHT concentrations (Figs. 3B, 3E, 3H, and 3K). Following the work of Carter et al. (1988) and Gennings (2000), the detection of an M2*AR interaction is equivalent to finding an isobologram with a curvilinear relationship different from the line of additivity. If an M2*AR interaction was observed, then the null hypothesis of additivity was rejected.
At 0.001 nM DHT, the parameter associated with the effect of [M2] was positive but not significant (p = 0.081, Table 2) and the parameter associated with [AR] was negative but not significant (p = 0.459, Table 2
). These results indicate that the effect of [AR] and [M2] did not significantly alter the response at 0.001 nM DHT. M2*AR was negative and borderline significant at 0.001 nM DHT (p = 0.057, Table 2
), indicating that increasing concentration combinations of AR and M2 decrease the luciferase response to DHT. All contour lines bow below the line of additivity (Fig. 3C
) indicating a consistently synergistic relationship between M2 and AR at 0.001 nM DHT.
At 0.01 nM DHT, the parameters associated with the effect of [M2] and [AR] on the response were negative and significant (M2, p < 0.001; AR, p = 0.025, Table 2), suggesting that [M2] and [AR] exhibited an inhibitory effect on the response. The M2*AR interaction was positive and significant (p = 0.033, Table 2
) suggesting that M2 and AR have an antagonistic interaction. This result indicates that the response to M2 inhibition of luciferase activity decreases as [AR] increases. While some contour lines bow up and others bow down (Fig. 3F
), all contours are above their corresponding line of additivity and thus are consistent with an antagonistic relationship between M2 and AR at 0.01 nM DHT.
At 0.1 nM DHT, the parameter associated with the effect of [M2] was negative and significant (p < 0.001, Table 2), indicating that [M2] exhibited an inhibitory effect on the response. The parameter associated with the effect of [AR] was negative and not significant at 0.1 nM DHT (p = 0.965, Table 2
), indicating that the effect of [AR] on the response to DHT did not significantly alter the response to M2. The M2*AR was positive and significant (p < 0.001, Table 2
), indicating that the inhibiting effect of [M2] on luciferase response diminished with an increase in [AR] (Figs. 3H and 3I
). This result suggests an antagonistic relationship between [AR] and [M2]. The negative slope of M2 can be visualized in Figure 3H
at the edge of the surface corresponding to 10 ng/dish AR. The edge at 100 ng/dish AR demonstrates a less steep, negative slope and this change in slope is due to the M2*AR. The corresponding contour plot (Fig. 3I
) shows all response contours bow above the line of additivity, indicating a consistently antagonistic M2*AR relationship at 0.1 nM DHT.
At 1 nM DHT, the parameters associated with the effect of [M2] and [AR] were negative but not significant (M2, p = 0.131; AR, p = 0.749, Table 2) indicating that neither the effect of [AR] nor [M2] significantly altered the response to DHT. The parameter associated with M2*AR did not show a significant effect on the fitted response surface (p = 0.674, Table 2
; Figs. 3K and 3L
), indicating that their interaction cannot be determined with the modeled data at 1 nM DHT.
![]() |
DISCUSSION |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
The relevance of the DHT concentrations tested in vitro to physiological androgen levels was explored by identifying androgen concentrations during critical windows for human male urogenital tract, external genitalia, and pubertal development in the literature. Free androgen concentrations in vivo were considered comparable to the DHT concentrations used in this study because serum binding proteins (i.e., sex hormone binding globulin) were not present during the in vitro assay. DHT concentrations were compared to [free T] since [free T] but not [free DHT] during these critical windows of human development were identified in the literature, and responses to T and DHT in the AR-mediated luciferase reporter gene assay were similar (Kemppainen et al., 1999). At stages overlapping the critical window for male urogenital tract differentiation, the [free T] measurements in cord blood range from
0.050.5 nM (Table 1
; Abramovich and Rowe, 1973
; Diez dAux and Murphy, 1974). Within the critical window for external genitalia differentiation, the [free T] measurements in cord blood ranged from
0.10.5 nM (Table 1
; Diez dAux and Murphy, 1974). During Tanner stage I, a stage just before puberty onset that may correspond to the critical window for pubertal development, the [free T] measurements ranged from
0.0010.2 nM. Thus, three of the four tested DHT concentrations, 0.001, 0.01, and 0.1 nM, fell within the expected physiological [free T] range during some androgen-sensitive developmental intervals while 1 nM DHT was above the in vivo [free T] during human development.
At 0.01 and 0.1 nM DHT, the model is most consistent with the data and is the most robust since the greatest number of data points and the most complete coverage of three-way dose combinations were tested (Fig. 3). The shapes of the response surfaces were similar indicating that the relationships among the three components were similar at these two DHT concentrations. [M2] exhibited a significant inhibitory effect on the response that is consistent with the mechanism of action of M2 as an androgen antagonist. As predicted by the mode of action for vinclozolin, the inhibitory effect of M2 decreased as [AR] increased at these two DHT concentrations. The additivity tests for [M2] and [AR] indicated that their relationship was significant and antagonistic for the response since the null hypothesis of additivity was rejected. This finding is consistent with the biology of receptor binding, i.e., as [AR] increases, the luciferase response would be expected to increase and, as [M2] increases, the luciferase response would be expected to decrease. As expected, at 0.1 nM and 0.01 nM DHT the response contours of the contour plots (Figs. 3F and 3I
) suggest a consistently antagonistic relationship between AR and M2 across all dose combinations. At 0.01 nM DHT, [AR] had a significant inhibitory effect on the response, which is inconsistent with the fact that DHT has a higher binding affinity for AR than M2, i.e., within the [AR] range tested, as [AR] increases a greater proportion of DHT-AR compared to M2-AR complexes would be expected to form. Comparable in vivo [AR] during human development was not identified in the literature. Further data on [AR] and distribution during development in humans and rodents are needed to understand comparable in vivo AR levels and their effect on the response to vinclozolin exposure.
The models at 0.001 and 1 nM DHT were less reliable than at either 0.01 or 0.1 nM DHT since they were based on fewer data points and less coverage of the arithmetic dose intervals. The statistics did not reveal significant patterns about the relationship between [M2] and [AR] at 0.001 and 1 nM DHT. The response-surface shapes at 0.001 and 1 nM DHT were significantly different from those at 0.1 nM DHT, but the results of M2*AR were not significant (1 nM DHT) or were borderline significant (0.001 nM DHT). The differences in response-surface shapes may be due to constraints of the in vitro system, and therefore, may not inform the in vivo developmental scenario. For example, at 1 nM DHT, the decreased luciferase response may reflect a limiting concentration of transcription cofactors (i.e., titration of transcription cofactors at saturating [DHT]; "squelching," E.M. Wilson, unpublished results). The DHT level at 0.001 nM is often too low to achieve a robust transcriptional response in the assay (data not shown). A high [AR] was found to be inhibitory and may also be due to titration of transcription cofactors by the AR. Alternatively, changes in the M2*AR relationship to the response at the highest and lowest DHT levels may indicate that at lower DHT levels and in the presence of excess M2 and limited DHT, M2 may antagonize DHT-AR more effectively. Certain EDCs have been observed to behave as an antagonist at relatively high concentrations and as an agonist at very high concentrations (Kemppainen and Wilson, 1996; Maness et al., 1998
; Wong et al., 1995
).
The results suggest that the response to vinclozolin is altered by changes in androgen and AR concentration, both of which are known to vary during development. If the relationship between DHT and AR concentrations and their effect on M2 inhibition pertain in vivo, susceptibility to vinclozolin exposure could be affected by developmental stage as well as individual differences in AR or DHT concentrations. The model predicts that susceptibility to an equivalent vinclozolin exposure would be expected to be greatest when DHT and AR concentrations are relatively low but sufficiently high to induce a transcriptional response (i.e., in the absence of AR-dependent transcription, inhibition cannot be observed). However, the free androgen concentration cannot be the only factor affecting sensitivity and response to an androgen antagonist. In the human, the free androgen concentration is lower before puberty than it is during male sexual differentiation in utero (Table 1) in the rat. It was this period of male sexual development in the fetal rat that was shown to be most susceptible to androgen antagonist effects. Significant effects on anogenital distance at birth, the most sensitive endpoint of sexual differentiation in the developing rat embryo, were observed after treatment at doses as low as 3.125 mg/kg/day (Gray et al., 1999
). In contrast, effects on the most sensitive endpoint of puberty, the timing of preputial separation, were not observed at 10 or 30 mg/kg/day but were observed at 100 mg/kg/day (Monosson et al., 1999
).
Decreased sensitivity to antiandrogens prior to puberty versus the time of sexual differentiation in the developing male fetus may result from differences in feedback regulation at the hypothalamic-pituitary-gonadal axis. In the prepubertal and adult male, androgen feedback inhibits luteinizing hormone (LH) production by the anterior pituitary. Exposure to an antiandrogen such as vinclozolin would inhibit androgen-induced downregulation of LH. The subsequent LH-induced rise in testosterone synthesis by the Leydig cells of the testis leads to elevated androgen levels, which compete with antiandrogens for binding to the AR (Monosson et al., 1999). The compensatory increase in testosterone thereby minimizes the sensitivity of the prepubertal rat to the effects of antiandrogen exposure. In contrast, in the fetal rat and human male, the developing hypothalamic-pituitary-gonadal axis feedback regulation is not fully functional (Forest, 1985
; Huhtaniemi et al., 1981a
,b
, 1995
; Quigley, 2002
; Warren et al., 1987
). LH production by the fetal male pituitary is suppressed by placental estrogen and fetal testicular androgen production is stimulated in the human by human chorionic gonadotropin (hCG) from the placenta. Because fetal androgens do not regulate LH production, there may be no compensatory rise in testosterone production when challenged with an antiandrogen. The absence of hypothalamic-pituitary-gonadal feedback regulation in the fetus may contribute to the increased sensitivity to antiandrogen exposure during male sexual differentiation.
Developmental stage information on the concentration of components other than androgen and androgen receptor concentrations is needed to further understand the effect of developmental stage on response to an antiandrogen. One other important component is sex hormone binding globulin (SHBG), the mean serum concentration of which decreases from six months to 14 years of age, concomitant with an increase in non-SHBG-bound T (Belgorosky and Rivarola, 1987). While [T] is relatively low during the prepubertal stage, [SHBG] is also low, presumably leading to an increase in unbound T. Additionally, serum SHBG decreased from prepuberty (Tanner stage I) to puberty onset (Tanner stage II), suggesting that androgen-dependent developmental events correspond to a relatively high unbound androgen concentration that is available to bind to AR (Kim et al., 1999). Data on SHBG concentrations throughout development would be useful to define its effect on the response to an antiandrogen.
Statistical modeling is useful in risk assessment, because responses can be predicted at untested dose combinations and because all concentration combinations cannot be tested. The modeling approach applied in this paper can be used to predict the effect of developmental stage on the response to exposure to an EDC or to developmental toxic agents in general. A similar modeling approach has been used for a number of mixture studies (Charles et al., 2002; Weller et al., 1999
; Gennings et al., in press). In this study, the "mixture" studied was the combination of the chemical exposure (M2) and two developmental stage components (DHT and AR). The use of an in vitro assay allowed for the manipulation of the concentrations of these two developmental stage components and consequently the effect of different dose combinations, reflecting different developmental stages, on the response to the chemical exposure could be determined by the modeling approach. Determining if the relationships among the developmental stage components and developmental toxic agent exposure change at different dose combinations is an approach that can be applied to a number of scenarios. Understanding the impact of developmental stage and its components is critical to risk assessment for children and susceptible populations. Additionally, the results of this analysis are useful to cumulative risk assessment efforts, especially for chemicals such as DHT and M2, which have the same mode of action but have opposite effects.
The model presented here suggests that developmental stage, as defined by androgen and androgen receptor concentrations can affect the response to an antiandrogen and that the relationship between [AR] and [M2] is similar and antagonistic in the 0.010.1 nM DHT range. Information on additional components of developmental stage could be incorporated into a future model to predict the response to an EDC at different developmental stages and the effect of each component on the response. Further expansion of the model could include information reflecting individual sex and racial/ethnic differences in circulating androgen levels during development and in adulthood to predict the response to an EDC.
![]() |
ACKNOWLEDGMENTS |
---|
![]() |
NOTES |
---|
1 To whom correspondence should be addressed. Fax: (202) 565-0078. E-mail: euling.susan{at}epa.gov.
![]() |
REFERENCES |
---|
![]() ![]() ![]() ![]() ![]() ![]() ![]() |
---|
Belgorsky, A., and Rivarola, M. A. (1987). Changes in serum sex hormone-binding globulin and in serum non-sex hormone-binding testosterone during prepuberty in boys. J. Steroid Biochem. 27, 291295.[ISI][Medline]
Bentvelsen, F. M., Brinkmann, A. O., van der Schoot, P., van der Linden, J. E., van der Kwast, T. H., Boersma, W. J., Shroder, F., and Nijman, J. M. (1995). Developmental pattern and regulation by androgens of androgen receptor expression in the urogenital tract of the rat. Mol. Cell. Endocrinol. 113, 245253.[ISI][Medline]
Berenbaum, M. C. (1985). The expected effect of a combination of agents: The general solution. J. Theor. Biol. 114, 413431.[ISI][Medline]
Carter, W. H., Jr., Gennings, C., Staniswalis, J. G., Campbell, E. D., and White, K. L., Jr. (1988). A statistical approach to the construction and analysis of isobolograms. J. Am. Coll. Toxicol. 7, 963973.[ISI]
Charles, G. D., Gennings, C., Zacharewski, T. R., Gollapudi, B. B., and Carney, E. W. (2002). Assessment of interactions of diverse ternary mixtures in an estrogen receptor- reporter assay. Toxicol. Appl. Pharmacol. 180, 1121.[ISI][Medline]
Choong, C. S., Quigley, C. A., French, F. S., and Wilson, E. M. (1996). A novel missense mutation in the amino-terminal domain of the human androgen receptor gene in a family with partial androgen insensitivity syndrome causes reduced efficiency of protein translation. J. Clin. Invest. 98, 14231431.
Cooke, P. S., Young, P., and Cunha, G. R. (1991). Androgen receptor expression in developing male reproductive organs. Endocrinology 128, 28672873.[Abstract]
Diez dAux, R. C., and Pearson Murphy, B. E. (1974). Androgens in the human fetus. J. Steroid Biochem. 5, 207210.[ISI][Medline]
Euling, S. Y., and Kimmel, C. A. (2001). Developmental-stage sensitivity and mode-of-action information for androgen agonists and antagonists. Sci. Total Environ. 274, 103113.[ISI][Medline]
Fisher, D. A., Ed. (1998). Endocrinology test selection and interpretation. Quest Diagnostics Manual, 2nd ed. Quest Diagnostics, Inc., San Juan Capistrano, CA.
Forest, M. G. (1985). Sexual maturation of the hypothalamus: Pathophysiological aspects and clinical implications. Acta Neurochir. (Wien) 75, 2342.[ISI]
Forest, M. G., Cathiard, A. M., and Bertrand, J. A. (1973). Total and unbound testosterone levels in the newborn and in normal and hypogonadal children: Use of a sensitive radioimmunoassay for testosterone. J. Clin. Endocrinol. Metab. 36, 11321142.[ISI][Medline]
Gennings, C. (2000). On testing for drug/chemical interactions: Definitions and inference. J. Biopharm. Stat. 10, 457467.[Medline]
Gennings, C., Charles, G., Gollapudi, B., Zacharewski, T., Carney, E. (in press). Analysis of resulting data from estrogen receptor reporter gene assays. J. Agric. Biol. Environ. Stat.
George, F. W., and Wilson, J. D. (1994). Sex determination and differentiation. In The Physiology of Reproduction, 2nd ed. (E. Knobil and J.D. Neill, Eds.), pp. 328. Raven Press, New York.
Gray, L. E., Jr., Ostby, J. S., and Kelce, W. R. (1994). Developmental effects of an environmental antiandrogen: The fungicide vinclozolin alters sex differentiation of the male rat. Toxicol. Appl. Pharmacol. 129, 4652.[ISI][Medline]
Gray, L. E., Jr., Ostby, J., Monosson, E., and Kelce, W. R. (1999). Environmental antiandrogens: Low doses of the fungicide vinclozolin alter sexual differentiation in the male rat. Toxicol. Ind. Health 15, 4864.[ISI][Medline]
He, B., Bowen, N. T., Minges, J. T., and Wilson, E. M. (2001). Androgen-induced NH2- and carboxyl-terminal interactions inhibits p160 coactivator recruitment by activation function 2. J. Biol. Chem. 276, 4229342301.
He, B., Kemppainen, J. A., Voegel, J. J., Gronemeyer, H., and Wilson, E. M. (1999). Activation function 2 in the human androgen receptor ligand-binding domain mediates interdomain communication with the NH2-terminal domain. J. Biol. Chem. 274, 3721937225.
Huhtaniemi, I. (1995). Molecular aspects of the ontogeny of the pituitary-gonadal axis. Reprod. Fertil. Dev. 7, 10251035.[ISI][Medline]
Huhtaniemi, I. T., Katikineni, M., and Catt, K. J. (1981a). Regulation of luteinizing-hormone-receptors and steroidogenesis in the neonatal rat testis. Endocrinology 109, 588595.[ISI][Medline]
Huhtaniemi, I. T., Katikineni, M., Chan, V., and Catt, K. J. (1981b). Gonadotropin-induced positive regulation of testicular luteinizing hormone receptors. Endocrinology 108, 5865.[ISI][Medline]
Jost, A. (1953). Problems in fetal endocrinology: The gonadal and hypophyseal hormones. Recent Prog. Horm. Res. 8, 379418.[ISI]
Jost, A. (1972). A new look at the mechanism controlling sexual differentiation in mammals. Johns Hopkins Med. J. 130, 3853.[ISI][Medline]
Kelce, W. R., Gray, L. E., and Wilson, E. M. (1998). Antiandrogens as environmental endocrine disruptors. Reprod. Fertil. Dev. 10, 105111.[ISI][Medline]
Kelce, W. R., Lambright, C. R., Gray, L. E., Jr. and Roberts, K. P. (1997). Vinclozolin and p, p'-DDE alter androgen-dependent gene expression: In vivo confirmation of an androgen receptor-mediated mechanism. Toxicol. Appl. Pharmacol. 142, 192200.[ISI][Medline]
Kelce, W. R., Monosson, E., Gamcsik, M. P., Laws, S. C., and Gray, Jr. L. E. (1994). Environmental hormone disruptors: Evidence that vinclozolin developmental toxicity is mediated by antiandrogenic metabolites. Toxicol. Appl. Pharmacol. 126, 276285.[ISI][Medline]
Kemppainen, J. A., Langley, E., Wong, C. I., Bobseine, K., Kelce, W. R., and Wilson, E. M. (1999). Distinguishing androgen receptor agonists and antagonists: Distinct mechanisms of activation by medroxyprogesterone acetate and dihydrotestosterone. Mol. Endocrinol. 13, 440454.
Kemppainen, J. A., and Wilson, E. M. (1996). Agonist and antagonist activities of hydroxyflutamide and Casodex relate to androgen receptor stabilization. Urology 48, 15763.[ISI][Medline]
Kim, M. R., Gupta, M. K., Travers, S. H., Rogers, D. G., Van Lente, F., and Faiman, C. (1999). Serum prostate-specific antigen, sex hormone binding globulin, and free androgen index as markers of pubertal development in boys. Clin. Endocrinol. 50, 203210.[ISI][Medline]
Loewe, S. (1953). The problem of synergism and antagonism of combined drugs. Arzneim. Forsh. 3, 285290.
Loewe, S., and Muischnek, H. (1926). Uber kombinationswirkunger: I. Mitteilung: Hiltsmittelder fragstellung. Naunyn-Schmiedebergs. Arch. Pharmacol. 114, 313326.
Maness, S. C., McDonnell, D. P., and Gaido, K. W. (1998). Inhibition of androgen receptor-dependent transcriptional activity by DDT isomers and methoxychlor in HepG2 human hepatoma cells. Toxicol. Appl. Phamacol. 151, 135142.[ISI][Medline]
Marshall, W. A., and Tanner, J. M. (1970). Variations in the pattern of pubertal changes in boys. Arch. Dis. Child 45, 1323.[ISI][Medline]
McKenna, N. J., Lanz, R. B., and OMalley, B. W. (1999). Nuclear receptor coregulators: Cellular and molecular biology. Endocr. Rev. 20, 321344.
Monosson, E., Kelce, W. R., Lambright, C., Ostby, J., and Gray, L. E., Jr. (1999). Peripubertal exposure to the antiandrogenic fungicide, vinclozolin, delays puberty, inhibits the development of androgen-dependent tissues, and alters androgen receptor function in the male rat. Toxicol. Ind. Health 15, 6579.[ISI][Medline]
Pasqualini, J. R., and Kincl, F. A., Eds. (1985). Hormones and the Fetus. Pergamon Press, New York.
Quigley, C. A. (2002). Editorial: The postnatal gonadotropin and sex steroid surge: Insights from the androgen insensitivity syndrome. J. Clin. Endocrinol. Metab. 87, 2428.
SAS Institute (1999). SAS Campus Drive, Cary, NC 27513.
Seber, G. A. F., and Wild, C. J. (1989). Nonlinear Regression. Wiley, New York.
Tomlin, C. D. S., Ed. (1997). The Pesticide Manual, 11th ed. British Crop Protection Council, Farnham, Surrey, UK.
U.S. EPA (2001). Vinclozolin; notice of use cancellations. U.S. Environmental Protection Agency. Fed. Reg. 66(163), 4413444136.
Warren, D. W., Huhtaniemi, I. T., Dufau, M. L., and Catt, K. J. (1987). Regulation of LH receptors and steroidogenesis in the fetal rat testis in vivo. Acta Endocrinol. (Copenh.) 115, 189195.[ISI][Medline]
Weller, E., Long, N., Smith, A., Williams, P., Ravi, S., Gill, J., Henessey, R., Skornik, W., Brain, J., Kimmel, C., Kimmel, G., Holmes, L., and Ryan, L. (1999). Dose-rate effects of ethylene oxide exposure on developmental toxicity. Toxicol. Sci. 50, 259270.[Abstract]
Wilson, E. M., and French, F. S. (1976). Binding properties of androgen receptors. Evidence for identical receptor in rat testis, epididymis, and prostate. J. Biol. Chem. 251, 56205629.[Abstract]
Wolf, C. J., LeBlanc, G. A., Ostby, J. S., and Gray, L. E., Jr. (2000). Characterization of the period of sensitivity of fetal male sexual development to vinclozolin. Toxicol. Sci. 55, 152161.
Wong, C., Kelce, W. R., Sar, M., and Wilson, E. M. (1995). Androgen receptor antagonist versus agonist activities of the fungicide vinclozolin relative to hydroxyflutamide. J. Biol. Chem. 270, 1999820003.