Using an ovine model of polycystic ovary syndrome (PCOS), (pregnant ewes injected with testosterone propionate (TP) (100 mg twice weekly) from day (d)62 to d102 of d147 gestation (maternal injection – MI-TP)), we previously reported female offspring with normal glucose tolerance but hyperinsulinemia. We therefore examined insulin signalling and pancreatic morphology in these offspring using quantitative (Q) RT-PCR and western blotting. In addition the fetal pancreatic responses to MI-TP, and androgenic and estrogenic contributions to such responses (direct fetal injection (FI) of TP (20 mg) or diethylstilbestrol (DES) (20 mg) at d62 and d82 gestation) were assessed at d90 gestation. Fetal plasma was assayed for insulin, testosterone and estradiol, pancreatic tissue was cultured, and expression of key β-cell developmental genes was assessed by QRT-PCR. In female d62MI-TP offspring insulin signalling was unaltered but there was a pancreatic phenotype with increased numbers of β-cells (P<0.05). The fetal pancreas expressed androgen receptors in islets and genes involved in β-cell development and function (PDX1, IGF1R, INSR and INS) were up-regulated in female fetuses after d62MI-TP treatment (P<0.05–0.01). In addition the d62MI-TP pancreas showed increased insulin secretion under euglycaemic conditions (P<0.05) in vitro. The same effects were not seen in the male fetal pancreas or when MI-TP was started at d30, before the male programming window. As d62MI-TP increased both fetal plasma testosterone (P<0.05) and estradiol concentrations (P<0.05) we assessed the relative contribution of androgens and estrogens. FI-TP (commencing d62) (not FI-DES treatment) caused elevated basal insulin secretion in vitro and the genes altered by d62MI-TP treatment were similarly altered by FI-TP but not FI-DES. In conclusion, androgen over-exposure alters fetal pancreatic development and β-cell numbers in offspring. These data suggest that that there may be a primary pancreatic phenotype in models of PCOS, and that there may be a distinct male and female pancreas.

Prenatal androgenization induces a polycystic ovary syndrome-like phenotype in adult female offspring, which is associated with alterations that can be detected in the fetal ovary, suggesting gestational origins of this condition. We therefore investigated whether increased prenatal androgen exposure also altered testicular development using ovine animal models. Biweekly maternal testosterone propionate (TP; 100 mg) from day 62 to day 70/day 90 of gestation altered male developmental trajectory. In male fetuses serum LH was decreased (P < .01), and testicular STAR, CYP11, and CYP17 abundance were reduced. Coincident with this, basal testicular T synthesis was decreased in vitro (P < .001). Leydig cell distribution was severely perturbed in all testes prenatally exposed to TP (P < .001). To examine the contribution of estrogens, fetuses were injected with TP (20 mg), the potent estrogen agonist, diethylstilbestrol (DES; 20 mg), or vehicle control at day 62 and day 82 and assessed at day 90. The effects of fetal (direct) TP treatment, but not DES, paralleled maternal (indirect) TP exposure, supporting a direct androgen effect. Cessation of maternal androgenization at day 102 returned Leydig cell distribution to normal but increased basal T output, at day 112, demonstrating Leydig cell developmental plasticity. Earlier maternal androgen exposure from day 30 similarly influenced Leydig cell development at day 90 but additionally affected the expression of Sertoli and germ cell markers. We show in this study that increased prenatal androgen exposure alters development and function of Leydig cells at a time when androgen production is paramount for male development. This supports the concept that gestational antecedents associated with polycystic ovary syndrome may have effects on the male fetus.