An investigation of the endocrine-disruptive effects of bisphenol a in human and rat fetal testes

Abstract

Few studies have been undertaken to assess the possible effects of bisphenol A (BPA) on the reproductive hormone balance in animals or humans with often contradictory results. We investigated possible direct endocrine disruption by BPA of the fetal testes of 2 rat strains (14.5-17.5 days post-coitum) and humans (8-12 gestational weeks) and under different culture conditions. BPA concentrations of 10(-8)M and 10(-5)M for 72 h reduced testosterone production by the Sprague-Dawley fetal rat testes, while only 10-5M suppressed it in the Wistar strain. The suppressive effects at 10-5M were seen as early as 24h and 48 h in both strains. BPA at 10(-7)-10(-5)M for 72 h suppressed the levels of fetal rat Leydig cell insulin-like factor 3 (INSL3). BPA exposure at 10(-8)M, 10(-7)M, and 10(-5)M for 72 h inhibited testosterone production in fetal human testes. For the lowest doses, the effects observed occurred only when no gonadotrophin was added to the culture media and were associated with a poorly preserved testicular morphology. We concluded that (i) BPA can display anti-androgenic effects both in rat and human fetal testes; (ii) it is essential to ascertain that the divergent effects of endocrine disruptors between species in vitro do not result from the culture conditions used, and/or the rodent strain selected; (iii) the optimization of each in vitro assay for a given species should be a major objective rather than the search of an hypothetical trans-species consensual model-system, as the organization of the testis is intrinsically different between mammalian species; (iv) due to the uncertainty existing on the internal exposure of the human fetal testis to BPA, and the insufficient number of epidemiological studies on the endocrine disruptive effects of BPA, caution should be taken in the extrapolation of our present results to the human reproductive health after fetal exposure to BPA.

Fig 1. Time and dose-dependent effects of BPA (10^-9M-10^-5M) on the rat fetal testis.

A) Basal testosterone production by the rat fetal testis: dose-dependent effects of BPA (10^-9M-10^-5M) on basal testosterone levels. Rat strain (Sprague-Dawley versus Wistar) and vehicle in which BPA solutions were tested and prepared (DMSO versus ethanol) were studied in order to determine their influence on the responses to BPA exposures. B) Dose-dependent effects of BPA (10^-9M-10^-5M) on the basal INSL3 production by the Sprague-Dawley fetal rat testis: BPA was diluted in DMSO and the media collected after 72hr of culture were assayed. Values are mean +/- SEM of 12–24 testes from at least 3 independent experiments. Dose-responses were analyzed for significance with one-way ANOVA, Holm-Sidak post-hoc test. The effects of vehicle and strain were analyzed with two-way ANOVA. *p <0.05, ** p <0.01, *** p <0.001.

Fig 2. BPA effects at 10^-5M on the testicular histology and testicular cells.

A) A treatment with 10^-5M of BPA did not affect the morphology and the testis organization of the fetal rat. B) Effect of 10^-5M of BPA on the total number of gonocytes, Sertoli cells and Leydig cells. Values are mean ± SEM of 4–7 fetuses. Responses to BPA were measured by comparing one control testis (DMSO-treated) with the contralateral testis cultured in medium containing the tested factor. C) Quantitative analysis of BrdU incorporation into gonocytes and Sertoli cells after 72 h of culture measured as the percentage of BrdU-positive gonocytes or Sertoli cells in at least 1000 cells (n = 4 fetuses). D) Apoptosis was detected using a cleaved-caspase 3 staining, caspase-3 being cleaved in the cells undergoing apoptosis. Caspase 3-positive cells were counted on the whole testis with regard to their tubular or interstitial localization (n = 5 fetuses). *p < 0.05 by Wilcoxon signed rank tests on paired data.

Fig 3. Effects of BPA on testosterone production after 72h of culture of 7–12 GW human fetal testis explants.

A) Dose-dependent effects of BPA—diluted in DMSO or ethanol—with hLH on testosterone production by human fetal testis explants (RTP; RTP, %Ctrl): BPA was diluted in DMSO or ethanol and the media collected after 72 hr of culture were assayed. Results are expressed as normalized production of testosterone of treated samples as the percentage of that of the respective untreated first day of culture sample (RTP) (top) and as the percentage of that of the respective untreated first day of culture sample and control (RTP, %Ctrl) (bottom). Values are mean +/- SEM of testosterone from the respective untreated first day of culture basal sample and control. The number of testes (n) is indicated below the graphic for each condition. Dose-responses were analyzed for significance with a Wilcoxon test. The effects of the vehicle were analyzed with two-way ANOVA. *p<0.05. B) Testosterone production after culture of human fetal testis in the presence of BPA and hCG, hLH or no gonadotrophin (-gonado) supplementing the medium (RTP). n = 7–13 testes for the hCG group, n = 3–20 testes for hLH group and n = 9–21 in the absence of gonadotrophin. *p<0.05, ***p<0.001 (two-way ANOVA followed by a Holm-Sidak test or a Wilcoxon test to compare matched samples). C) Testosterone production represented as a fold change from the respective first day of culture sample and control (RTP, %Ctrl). Results are expressed as normalized production of testosterone of treated samples as the percentage of that of the respective untreated first day of culture sample (RTP) and control (RTP, %Ctrl). n = 7–13 testes for the hCG group, n = 3–20 testes for hLH group and n = 9–21 in the absence of gonadotrophin. *p<0.05, ***p<0.001 (two-way ANOVA followed by a Holm-Sidak test or a Wilcoxon test to compare matched samples).

Fig 4. INSL3 production after culture of 8–12 GW human fetal testis explants in absence (Ctrl) or presence of 10^-8M-10^-5M.

A) INSL3 production after culture of human fetal testis explants in the presence of BPA and hCG, hLH or no gonadotrophin (-gonado) supplementing the medium (RIP). INSL3 concentrations after culture of 8–12 GW human fetal testis in the presence of ethanol (Ctrl) or from 10^-8–10^-5 M BPA. Results are expressed as normalized production of INSL3 of treated samples as the percentage of that of the respective untreated first day of culture sample (RIP). n = 3–9 for hLH group, n = 3–11 testes for hCG group and n = 6–9 for the group without gonadotrophin. Values are means +/- SEM. *p<0.05, **p<0.01, ***p<0.001 (two-way ANOVA, followed by a Holm-Sidak test or a Wilcoxon test to compare matched sample). B) INSL3 production represented as normalized production of INSL3 of treated samples as the percentage of that of the respective untreated first day of culture sample and control (RIP, %Ctrl). INSL3 concentrations after culture of 8–12 GW human fetal testis explants in the presence of ethanol (Ctrl) or from 10^-8–10^-5 M BPA. Results are expressed as normalized production of INSL3 of treated samples as the percentage of that of the respective untreated first day of culture sample and control (RIP, %Ctrl). n = 3–9 for hLH group, n = 3–11 for hCG group and n = 6–9 for the group without gonadotrophin. Values are expressed as least squares means +/- SE. *p<0.05, **p<0.01, ***p<0.001 (two-way ANOVA, followed by a Holm-Sidak test or a Wilcoxon test to compare matched sample).

Fig 5. Representative immunostaining of CYP11A1 of human fetal testis explants 11–12 GW.

The 3,3-diaminobenzidine tetrahydrochloride staining appears dark brown in all panels, and sections were counterstained with hematoxylin. Testis cords could be easily delineated in all treated explants except in the sections of explants that have been exposed to BPA 10^-5M without gonadotrophin, in which the boundaries of the testicular cords appeared somewhat blunted. Scale bar corresponds to 100 μm.