Prenatal Androgen Exposure Leads to Alterations in Gene and Protein ...

REPRODUCTION-DEVELOPMENT

Prenatal Androgen Exposure Leads to Alterations in Gene and Protein Expression in the Ovine Fetal Ovary Kirsten Hogg, Alan S. McNeilly, and W. Colin Duncan Division of Reproductive and Developmental Sciences (K.H., W.C.D.), The University of Edinburgh, and Medical Research Council Human Reproductive Sciences Unit (A.S.M.), Centre for Reproductive Biology, The Queens Medical Research Institute, Edinburgh EH16 4SA, United Kingdom

Exposure of a female fetus to increased androgens in utero results in an adult phenotype reminiscent of polycystic ovary syndrome. We investigated whether prenatal androgens could directly alter the structure and function of the fetal ovary. We examined fetal ovarian cell proliferation, germ cell volume, and the expression of steroid receptors and steroidogenic enzymes. In addition, we studied the inhibitors of differentiation (Ids) and the SLIT/Roundabout developmental pathways. Female fetuses were collected from ewes treated with 100 mg testosterone propionate (TP) or vehicle control (C), twice weekly from d 60 to 70 (C ⫽ 3, TP ⫽ 6) or d 90 (C ⫽ 6, TP ⫽ 8). Female fetuses were also collected at d 70 after a single injection of TP (20 mg) or vehicle C into the fetal flank at d 60 (C ⫽ 4, TP ⫽ 8). Prenatal androgenization had no effect on fetal ovarian morphology, cell proliferation, or germ cell volume. However, there was a reduction in the expression of StAR, CYP11A, CYP17, and LHR at d 90 of gestation. There was also an increase in Id1 immunostaining at d 90 and an increase in Id3 immunostaining at d 70. Direct injection of TP into the fetus down-regulated ovarian CYP11A, estrogen receptor ␣ and ␤ mRNA, and ROBO1 and up-regulated CYP19, androgen receptor immunostaining, and Id3 mRNA and protein. Although at d 90 prenatal androgenization does not result in structural changes of the fetal ovary, there are functional changes that may impact on ovarian development. TP has direct actions on the fetal ovary, and these may contribute to the adult ovarian phenotype in the ovine model of polycystic ovary syndrome. (Endocrinology 152: 2048 –2059, 2011)

T

he developing fetus is vulnerable to changes in the hormonal and nutritional environment that can manipulate highly plastic processes and predispose many adult pathophysiologies (1). Dysregulation of fetal programming can lead to an increased stress response as well as a propensity to develop cardiovascular, metabolic, and reproductive disorders (2– 4). Polycystic ovary syndrome (PCOS) is common, affecting 5–10% of women of reproductive age (5) and may have fetal origins (6). Evidence from studies in nonhuman primates, supplemented by experiments in rodents and sheep, reveal that increased in utero exposure to androgens leads to perturbed menstrual and estrous cycles, polyfollicular ovaries, hyperandrogenism, increased LH secretion, hyperinsulinemia, and increased adiposity in the adult (7–11). Because these traits are reminiscent of PCOS, the androgen-

exposed female fetus can inform us about the developmental programming of various features of PCOS. The effects of maternal androgen administration on the reproductive phenotype of female adult offspring have been well documented; however, these are likely to be secondary to an early direct effect on one or more fetal tissues. The adult consequences of increased prenatal androgens include LHdependent ovarian hyperandrogenism and hyperinsulinemia, both of which can cause polycystic ovaries (PCO) (12, 13). We therefore hypothesized that the adult PCO phenotype is secondary to fetally programmed hormonal and/or metabolic changes, rather than a direct effect on the developing ovary. Thus, we aimed to assess the affect of in utero androgenization on the fetal ovary at the time of follicle formation.

ISSN Print 0013-7227 ISSN Online 1945-7170 Printed in U.S.A. Copyright © 2011 by The Endocrine Society doi: 10.1210/en.2010-1219 Received October 20, 2010. Accepted January 24, 2011. First Published Online February 15, 2011

Abbreviations: AR, Androgen receptor; C, control; DHT, dihydrotestosterone; ER, estrogen receptor; FSHR, follicle-stimulating hormone receptor; Id, inhibitor of differentiation; LHR, LH receptor; PCO, polycystic ovaries; PCOS, polycystic ovary syndrome; qRT, quantitative real-time; ROBO, Roundabout; StAR, steroidogenic acute regulatory; TP, testosterone propionate.

2048

endo.endojournals.org

Endocrinology, May 2011, 152(5):2048 –2059


In fetal life, germ cells migrate into the developing ovary and multiply within oogonial clusters (14). These clusters break down, and the germ cells associate with somatic cells to form primordial follicles (15). Concomitantly, there is a marked reduction in germ cell number through apoptosis (16). In the sheep, these key stages of ovarian development peak between d 55 and 90 of gestation (15). We used two investigative approaches to study the effects of androgens on the developing ovary. The first study assessed the d 90 fetal ovary after maternal androgenization using a standard protocol of twice weekly testosterone propionate (TP) injections from d 60. The second study assessed the d 70 ovary using the same treatment protocol. The placenta is however able to metabolize androgens to estrogens, and therefore, variable amounts of testosterone, as well as estradiol, will reach the fetus. We therefore developed a novel direct fetal androgenization model where the fetus was injected with TP at d 60 of gestation to avoid placental metabolism. We investigated the effect of androgenization on cell proliferation and germ cell volume, by immunolocalization of proliferative (Ki67) and germ cell (OCT3/4) specific markers. However, we also hypothesized that if androgens had a direct effect on the ovary, they would alter the expression of genes involved in pathways associated with ovarian function or development. We therefore examined four key candidate pathways: 1) the expression of genomic receptors for testosterone and estradiol within the fetal ovary (17); 2) the steroidogenic potential of the fetal ovary, by quantifying the expression of key genes involved in ovarian steroid synthesis (18); and 3) the Roundabout (ROBO)/SLIT pathway that has important roles during mammalian development (19). SLIT1–3 ligands are secreted glycoproteins that bind the ROBO1– 4 receptors (19). They can regulate cell migration by acting as a repulsive cue (20) and promote apoptosis and tissue remodeling (21). In the ovine fetal ovary, the expression of SLIT2, SLIT3, ROBO1, ROBO2, and ROBO4 peaks at around d 70, coinciding with the onset of follicle formation (22). Importantly, these genes have been shown to be regulated by gonadotropins and steroids in the female reproductive tissues (19, 23, 24). Finally, we examined 4) the expression and localization of the inhibitors of differentiation (Ids). The Ids (1– 4) are basic helixloop-helix transcription factors that negatively regulate transcription. They are crucial to the timing of tissue-specific gene expression during embryogenesis, by interaction with DNA-binding domains (25) to control growth and differentiation (26). They are differentially expressed during folliculogenesis in sheep (27), where they can be regulated by members of the TGF␤ superfamily that are involved in ovarian development (27, 28). Here, we describe the effects of maternal exposure to androgens on the developing ovine fetal ovary at d 90 of


2049

gestation. We also quantify the effects on the fetal ovary at d 70 of gestation, where we assess the effect of maternal androgenization and direct fetal exposure to increased testosterone. We report differential functional alterations that could be early determinants of an adult ovarian phenotype.

Materials and Methods Reagents All reagents and chemicals were obtained from Sigma-Aldrich (Poole, UK), unless otherwise stated.

Animal treatments and tissue collection After regulatory and local ethical committee approval, Scottish Greyface ewes were synchronized before mating so that the gestational time points could be accurately calculated, and two breeding seasons were used to collect two treatment cohorts. The singletons (14%), twins (66%), and triplets (20%) were evenly divided between treatment and control (C) groups. In the first cohort, pregnant ewes were treated im twice weekly with 1 ml vehicle C or 100 mg TP (AMS Biotechnology Ltd,, Abingdon, UK), in vegetable oil, from d 60 of gestation. Mothers were euthanized and female fetuses collected at d 90 (C, n ⫽ 6; TP, n ⫽ 8). In the second cohort, the same regimen was used and female fetuses collected at d 70 (C, n ⫽ 3; TP, n ⫽ 6). In an additional treatment group, mothers were anesthetized at d 60 and fetuses injected into the flank once with 0.2 ml vehicle C or 20 mg TP using a 20G Quincke spinal needle (BD Biosciences, Oxford, UK) under ultrasound guidance. Animals were killed on d 70 and female fetuses collected (C, n ⫽ 4; TP, n ⫽ 8). Both fetal ovaries were dissected, and one was stored in Bouins fixative for 24 h before transfer to 70% ethanol and embedded in paraffin wax, whereas the other was snap frozen and stored at ⫺80 C. Hormonal analysis of fetal serum was performed by standard in-house RIA, where the overall coefficient of variation for testosterone and 17␤-estradiol was less than 10%, as described previously (29). Testosterone (pg/ml) was significantly increased in d 70 (C, 44.9 ⫾ 9.8; T, 297.5 ⫾ 141.7; P ⬍ 0.05) and d 90 (C, 41.6 ⫾ 5.0; T, 145.9 ⫾ 22.5; P ⬍ 0.001) indirectly exposed fetuses to levels comparable with age-matched C male sheep fetuses from these cohorts (d 70, 171.3 ⫾ 52.1; d 90, 143.1 ⫾ 35). Estradiol (pg/ml) was also significantly increased in indirectly TP-exposed d 90 fetuses (C, 4.6 ⫾ 1.0; T, 14.0 ⫾ 3.5; P ⬍ 0.01). Testosterone measurements in directly TP-treated d 70 fetuses were substantially increased (C, 52.3 ⫾ 13.4; T, 4533.5 ⫾ 2199.3; P ⬍ 0.01), whereas estradiol was not altered by this direct TP injection treatment and comparable with d 70 indirect treatment Cs.

Immunohistochemistry Immunohistochemistry was performed on fetal ovaries using OCT3/4, Ki67, Id1, Id2, Id3, Id4, androgen receptor (AR), estrogen receptor (ER)␣, and ER␤ antibodies (Table 1). Tissue sections (5 ␮m) were dewaxed and rehydrated before antigen retrieval by pressure cooking [0.01 M sodium citrate buffer (pH 6.0)] for 5 min. Sections were washed twice for 5 min in PBS, incubated in 3% H2O2 for 10 min, followed by two further 5-min PBS washes. Subsequently, tissue was blocked with 20%

2050

Hogg et al.

TABLE 1.

Effect of Prenatal Androgens on the Developing Ovary


Lists primary antibodies and respective concentrations used for immunohistochemistry

Antigen

Antibody clone/source

Dilution

Secondary antibody

AR ER␣ ER␤1 Id1 Id2 Id3 Id4 Ki67 OCT3/4

Polyclonal rabbit (N20; sc-816) (Santa Cruz Biotechnology, Inc., Santa Cruz, CA) Monoclonal mouse (VP-E614; clone 6F11) (Vector Laboratories Ltd.) Monoclonal mouse (MCA1974S; clone PPG5/10) (Serotec, Oxford, UK) Polyclonal rabbit (C20; sc-488, BP sc-488 P) (Santa Cruz Biotechnology, Inc.) Polyclonal rabbit (C20; sc489, BP sc-489 P) (Santa Cruz Biotechnology, Inc.) Polyclonal rabbit (C20; sc490, BP sc-490 P) (Santa Cruz Biotechnology, Inc.) Polyclonal rabbit (L20; sc491) BP (sc-491 P) (Santa Cruz Biotechnology, Inc.) Monoclonal mouse (NCL-Ki67-MMI) (Novocastra Laboratories, Newcastle, UK) Polyclonal goat (C20; sc-8629) (Santa Cruz Biotechnology, Inc.)

1:150 1:20 1:20 1:750 1:800 1:200 1:500 1:200 1:200

GARB GAMB GAMB GARB GARB GARB GARB GAMB RAGB

Secondary antibodies applied were goat-antirabbit biotinylated (GARB), goat-antimouse biotinylated (GAMB), or rabbit-antigoat biotinylated (RAGB) at a dilution of 1:500.

normal goat serum/5% BSA/PBS, for 1 h, with the exception of OCT3/4, where normal rabbit serum was used. Primary antibodies were diluted in the appropriate blocking solution and applied to tissue overnight at 4 C. Sections were then washed for 5 min with PBS ⫹ 1% Tween twice before incubation with a biotinylated goat antirabbit IgG or goat antimouse IgG secondary antibody (Dako, Glostrup, Denmark) at a 1:500 dilution. After two further 5-min washes in PBS ⫹ 1% Tween, tissue was incubated with Vectastain ABC Elite tertiary complex (PK-1600 Series; Vector Laboratories, Peterborough, UK) for 1 h, and washed twice in PBS for 5 min. In the case of ER antibodies, a further amplification step was used after incubation with the tertiary compound, using a tyramide signal amplification kit according to the manufacturer’s instructions (PerkinElmer, Cambridge, UK), and all protocol washes were performed with PBS. Staining was visualized colorimetri-

TABLE 2.

Stereological analysis of fetal ovaries Maternal exposed d 90 fetal ovaries were stained for Ki67 and OCT3/4 as described above and subjected to a stereological assessment to quantify proliferation index or germ cell volume, respectively, in C vs. TP-treated animals. Stereology was undertaken using Image-Pro Plus 6.2 software combined with a Stereology 5.0 plug-in (MagWorldwide, Wokingham, UK). Each fixed ovary was sectioned through its entirety (5␮m sections),

Lists forward and reverse primer sequences for qRT-PCR and respective amplicon size

Gene (accession no.) AR (XM_001253942) Forward Reverse CYP11A (NM_001093789) Forward Reverse CYP17 (NM_001009483) Forward Reverse CYP19 (NM_001123000) Forward Reverse ER␣ (NM_001001443) Forward Reverse ER&␤ (NM_001009737) Forward Reverse FSHR (NM_001009289) Forward Reverse GAPDH (NM_001034034) Forward Reverse HSD3B1 (NM_001135932) Forward Reverse Id1 (NM_001097568) Forward Reverse

bp, Base pairs.

cally by application of 3,3⬘-diaminobenzidine (Dako) for 30 sec. Tissue was rinsed in distilled water, dehydrated, counterstained with hematoxylin, and mounted. For negative Cs, primary antibodies were incubated with blocking peptide, before application. In the absence of a specific blocking peptide, negative Cs consisted of incubation with a nonspecific rabbit or mouse immunoglobulins and omission of the primary antibody.

Nucleotide sequence (5ⴕ-3ⴕ)

Product size (bp) 233

GCCCATCTTTCTGAATGTCC CAAACACCATAAGCCCCATC 172 CAACGTCCCTCCAGAACTGT CAGGAGGCAGTAGAGGATGC 215 AGACATATTCCCTGCGCTGA GCAGCTTTGAATCCTGCTCT 152 AATCCAGCACTCTGGAAAGC ACGTCCACATAGCCCAAGTC 187 GAATCTGCCAAGGAGACTCG CCTGACAGCTCTTCCTCCTG 208 GAGGCCTCCATGATGATGTC GGTCTGGAGCAAAGATGAGC 196 TAAGCACTTGCCAGCTGTTC CTCATCGAGTTGGGTTCCAT 229 GGCGTGAACCACGAGAAGTATAA AAGCAGGGATGATGTTCTGG 200 GGAGACATTCTGGATGAGCAG TCTATGGTGCTGGTGTGGA 151 TCTGGGATCTGGAGTTGGAG ATACGATCGTCCGCTGGAA

Gene (accession no.) Id2 (NM_001034231) Forward Reverse Id3 (NM_001014950) Forward Reverse Id4 (NM_001546) Forward Reverse LHR (NM_214449) Forward Reverse ROBO1 (NM_133631) Forward Reverse ROBO2 (NM_001128929) Forward Reverse ROBO4 (NM_019055) Forward Reverse SLIT2 (NM_004787) Forward Reverse SLIT3 (NM_003062) Forward Reverse StAR (NM_001009243) Forward Reverse

Nucleotide sequence (5ⴕ-3ⴕ)

Product size (bp) 185

CATCTTGGACTTGCAGATCG AGAGAGCTTTGCTGTCATTTG 160 ACTCAGCTTAGCCAGGTGGA TTTGGTCGTTGGAGATCACA 132 TCACTGCGCTCAACACCGACC TTCCCCCTCCCTCTCTAGTGCTCCTG 199 TCCGAAAGCTTCCAGATGTT GAAATCAGCGTTGTCCCATT 161 TTGAATTCAGGAGCAACTCC ATTAGCTGCCCTCACAAGGC 181 CTGAGAATCGGGTTGGAAAA AGGTTCTGGCTGCCTTCTTT 164 TGTGAGGCCAGCAACCGGCT GTGGGCTCTGGGTGGCCCCA 173 TGGAGGTGTCCTCTGTGATG TTATCCTTTCCCCTCGACAA 166 GGAGCCTTCACCCAGTACAA ACACCAGCCCATCAAACAGT 194 GCATCCTCAAAGACCAGGAG CTTGACACTGGGGTTCCACT



2051

and 10 evenly spaced tissue sections were studied for each animal. Each tissue section was analyzed using the Leitz DMRB microscope (⫻60 objective) with a Prior Pro-Scan automatic stage attached (Prior Scientific Instruments Ltd., Cambridge, UK) for tiling purposes. Due to relatively few Ki67 positive cells present in the d 90 ovary, all positive cells were counted per section. The proliferation index was calculated relative to the average area of each counted section for that individual. Germ cell volume was measured by selecting crosses that overlapped positive OCT3/4 cells in at least five randomly placed grids/fields per ovarian section. Germ cell volume was calculated by dividing the number of crosses selected by the number of crosses per field (432) and multiplied by 100 to give percentage values.

Semiquantitative analysis of fetal ovaries Fetal ovaries underwent immunohistochemistry for Id1, Id2, Id3, Id4, AR, ER␣, and ER␤ as described above and were microscopically assessed to determine changes in protein expression between C and TP-treated animals. Staining intensity was classified as weak, moderate, or intense to give a measure of expression, and fetal ovarian sections were subsequently categorized in a randomized blinded assessment by two investigators with reference to staining exemplars. Variation in staining assessment between observers was less than 5%.

Quantitative real-time (qRT)-PCR RNA was extracted from whole fetal ovaries using the QIAGEN RNeasy Micro kit (QIAGEN Ltd., West Sussex, UK) as per the manufacturer’s instructions and stored at ⫺80 C. RNA concentration and purity were measured using a NanoDrop 1000 spectrophotometer (Thermo Fisher Scientific, Loughborough, UK), and cDNA was synthesized from 200 ng total RNA using the High Capacity cDNA RT kit (Applied Biosystems, Foster City, CA) and stored at ⫺20 C. Forward and reverse primers were designed for amplification of target genes by qRT-PCR (Table 2), using Primer3 Input version 0.4 online software (30). Conventional PCR and PCR-product DNA sequencing was carried out to validate and confirm the authenticity of the gene product in the sheep. Primer efficiency and suitability for SYBR Green qRT-PCR was tested by generating standard curves (cDNA diluted 1:2, 1:4, 1:8, and 1:16) in qRT-PCRs. A 10 ␮l final reaction volume was prepared with PowerSYBR Green PCR Master Mix (Applied Biosystems), primer pairs (0.5 ␮M), cDNA (1 ␮l), and nucleasefree water. The qRT-PCR cycling program consisted of a denaturing step (95 C for 10 min), combined annealing and extension step (95 C for 15 sec, 60 C for 1 min), repeated 40 times, and a final dissociation step (95, 60, and 95 C for 15 sec each). Reactions were carried out in duplicate. Negative Cs were included per gene and consisted of a reaction using cDNA minus reverse transcriptase and a reaction where cDNA was substituted with nuclease-free water. The expression of the unknown target gene relative to GAPDH as an internal C was quantified using the ⌬Ct method.

Statistical analysis Statistical analyses were performed with GraphPad Prism version 4.0 (GraphPad Software, Inc., San Diego, CA) using an unpaired Student’s t test to compare the means of two groups or a ␹2 test for proportional data. P values of less than 0.05 were regarded as significant.

FIG. 1. Immunohistochemistry for Ki67 (A) and OCT3/4 (C) with respective stereological quantification of ovarian cell proliferation (B) and germ cell volume (D) in the maternal testosterone propionate (TP)exposed d 90 fetal ovary. Scale bars, 50 ␮m. Student’s t test analyses revealed no significance differences (B and D).

Results Maternal prenatal androgenization does not alter d 90 ovine ovary morphology/structure To investigate the effect of midgestation androgen exposure (d 60 –90) on the ovine fetal ovary, a morphological and immunohistochemical assessment was undertaken. The gross morphological structure of the d 90 ovary was not altered as a consequence of maternal androgen treatment when assessed microscopically. Cell proliferation was quantified across the ovary by immunohistochemistry for Ki67 (Fig. 1A). Stereological measurements indicated no change in the number of proliferative cells (Fig. 1B). Furthermore, germ cell-specific OCT3/4 immunohistochemistry (Fig. 1C) revealed no variation in germ cell volume for C vs. maternal TP-exposed ovaries, after quantitative stereological evaluation (Fig. 1D). Maternal prenatal androgenization affects d 90 ovarian steroidogenic gene expression The d 90 fetal ovary expresses receptors to androgens and estrogens. Maternal androgenization had no effect on the expression of mRNA for genomic AR (Fig. 2A), ER␣ (Fig. 2B), or ER␤ (Fig. 2C). Functional changes in response to increased in utero androgens were however observed in the d 90 fetal ovary, resulting in gene expression that could significantly impact ovarian steroidogenic po-

2052

Hogg et al.



trol of cell proliferation, death, and differentiation in the d 90 fetal ovary. We first examined the expression of the SLITs and ROBOs that are expressed during follicle formation in the developing ovine ovary. There were no differences in the expression of SLIT2 or SLIT3 ligands or ROBO1, ROBO2, or ROBO4 receptors after maternal androgenization (data not shown). All four Id proteins (Id1– 4) were expressed in the d 90 ovine fetal ovary (Fig. 3, A–D). Immunohistochemistry for Id1 revealed protein localization confined mostly to the germ cells (Fig. 3D). In contrast, Id2 and Id3 protein (Fig. 3, A and B, respectively) were strongly expressed in the germ cells and somatic cell streams, and Id3 staining was present in pregranulosa cells adjacent to germ cells (Fig. 3B). Id4 protein expression was present in some but not all somatic cells and pregranulosa cells and localized to the germ cells (Fig. 3C). Although no changes were observed for Id2, Id3, and Id4 mRNA or protein expression after maternal exposure to androgens (data not shown), staining intensity for Id1 was increased in C vs. TP-exposed d 90 ovaries in both germ and somatic cell types (P ⬍ 0.01; Fig. FIG. 2. Effect of maternal TP exposure on AR, ER␣, ER␤, StAR, CYP11A, HSD3B1, CYP17, 3E), although no variation existed at CYP19, and LHR (A–I) gene expression in the d 90 fetal ovary. Values are mean ⫾ SEM relative to the internal housekeeping gene GAPDH. A Student’s t test was performed to detect the transcript level (Fig. 3F). Id1 protein significant alterations in gene expression, where *, P ⬍ 0.05 and **, P ⬍ 0.01. was present in the ovarian stromal cell streams and more strongly expressed in tential. Quantitative RT-PCR detected reductions in the the germ cell cytoplasmic compartment in TP-exposed d expression of steroidogenic genes, including steroidogenic 90 ovaries compared with Cs (Fig. 3D). acute regulatory (StAR) protein (P ⬍ 0.01; Fig. 2D), CYP11A (P450scc) (P ⬍ 0.05; Fig. 2E), and CYP17 (P ⬍ 0.01; Fig. 2G), but not HSD3B1 (Fig. 2F) nor CYP19 (Fig. The effects of maternal and direct 2H). Interestingly, LH receptor (LHR) (Fig. 2I) and folli- androgenization on d 70 ovarian steroidogenic cle-stimulating hormone receptor (FSHR) (data not gene expression The d 70 fetal ovary also expresses receptors for anshown), crucial components in stimulating the steroidodrogen and estrogens. Although there were no differences genic pathway in ovarian steroidogenic cells in the adult ovary, were also expressed in the d 90 ovine fetal ovary, in the expression of AR mRNA (data not shown), immuand in the case of LHR, it was significantly down-regu- nohistochemistry for the AR protein revealed variations in the level of receptor present (Fig. 4A). Staining intensity lated after maternal androgen exposure (P ⬍ 0.05). for AR was increased in direct fetal TP-exposed d 70 ovaries (P ⬍ 0.05; Fig. 4B) but not significantly altered in the Maternal prenatal androgen exposure and other maternally exposed d 70 ovaries (Fig. 4C). In contrast, developmental regulatory molecules in the d 90 direct fetal exposure to TP resulted in a down-regulation ovary We assessed the effect of maternal androgenization on of the expression of both ER␣ (P ⬍ 0.001; Fig. 4D) and the expression of candidate molecules involved in the con- ER␤ (P ⬍ 0.05; Fig. 4F), whereas maternal androgeniza-



2053

respectively). Direct fetal androgenization reduced the expression of ROBO1 (P ⬍ 0.05; Fig. 5L), whereas maternal androgenization had no effect on its expression (Fig. 5I). There were no trends or differences in the expression of ROBO2 or ROBO4 after both direct and indirect androgenization (data not shown). Id genes were also examined in d 70 direct fetal and indirect maternal TP-exposed ovaries. Both mRNA and protein expression for Id3 changed after in utero androgen exposure (Fig. 6), whereas no changes were observed for Id1, Id2, or Id4 (data not shown). Id3 transcript was increased after direct fetal TP exposure (P ⬍ 0.05; Fig. 6A), although this trend was not significant after maternal TP exposure (Fig. 6B). Id3 protein, however, was up-regulated in maternal TP-exposed d 70 ovaries, with stronger expression observed particularly in the stromal compartment (Fig. 6C), after both direct fetal treatments (P ⬍ 0.05; Fig. 6D) and indirect maternal treatments (P ⬍ 0.01; Fig. 6E).

Discussion One key feature of PCOS is the development of PCO characterized by an increased number of small peripheral preantral follicles, an enlarged ovarian volume, and thickened stroma (5). It is not clear if the PCO phenotype represents fundamentally different ovaries or if it is an ovarian response to an altered hormonal or metabolic environment. Exogenous or endogenous androgens can make a normal ovary polycystic in appearance (13), as can hyperinsulinemia (12) and increased LH concentrations (31). These hormonal and metabolic changes are present in women with PCOS (5, 32, 33) and are seen in the ovine model of PCOS (34, 35), so it is possible that PCO occurs in response to the postnatal environment. Because prenatal androgenization in an ovine model leads to both an adult and prepubertal PCO phenotype, we investigated whether this was reflected in changes at the time of follicle formation in fetal life. In women, it has been suggested that because PCO contain more growing follicles, without a reduction in primordial follicles, they may contain a larger pool of follicles (36, 37). Given that the follicle pool is set in fetal life, we investigated the d 90 fetus after maternal androgenization. At d 90 of gestation, around the time of follicle formation, maternal androgenization had no effects on cell prolifer-

FIG. 3. Immunohistochemistry for Id2, Id3, Id4, and Id1 (A–D) in the d 90 ovine fetal ovary. Id1 protein expression is altered by maternal TP exposure (D and E); however, mRNA is not altered (F). Insets represent negative Cs. Scale bars, 50 ␮m. A Student’s t test (F) or a ␹2 statistical test (E) was performed, where **, P ⬍ 0.01.

tion did not alter their expression (Fig. 4, E and G, respectively). There were no differences in the expression of StAR, CYP11A (Fig 5A), HSD3B1, CYP17 (Fig. 5B), CYP19 (Fig. 5C), LHR, or FSHR in the d 70 fetal ovary after maternal androgenization. Unlike the effect of maternal androgen exposure on the d 70 ovine fetal ovary, direct fetal androgenization resulted in alterations to the potential ovarian steroidogenic capacity at d 70. Although StAR, HSD3B1, CYP17 (Fig. 5E), LHR, and FSHR were not changed, there was a decrease in CYP11A (P ⬍ 0.05; Fig. 5D) and an increase in CYP19 expression (P ⬍ 0.05; Fig. 5F). The effect of maternal and direct androgenization on other developmental regulatory molecules in the d 70 ovary The ROBO and SLIT genes are known to peak in the ovine fetal ovary during midgestation. No significant changes were identified for SLIT2 (Fig. 5G) or SLIT3 (Fig. 5H) mRNA expression due to maternal androgenization, nor after direct fetal androgenization (Fig. 5, J and K,

2054

Hogg et al.


FIG. 4. Immunohistochemical expression of ovarian AR in a directly treated d 70 fetus (A; inset represents negative C), and quantification of the effect of direct and indirect treatments (B and C). Analysis of direct and indirect treatment effects on ER␣ (D and E) and ER␤ (F and G) gene expression. A ␹2 statistical test (B and C) or a Student’s t test were performed, where *, P ⬍ 0.05 and ***, P ⬍ 0.001 (D–G).

ation or germ cell volume. Smith et al. (38) used a similar ovine model of prenatal androgenization from gestational d 30 to 90 and also found no alteration in oocyte number at d 90; however, by d 140 of gestation, there was a significant increase in the number of growing follicles. Because germ cell proliferation is complete by d 90, it is unlikely that there is more germ cell proliferation or follicles formed. It is possible that reduced germ cell loss occurs at later stages of gestation, leading to an increased


number of follicles at birth. However, because the PCO phenotype develops when maternal administration of androgens stops at d 90 (34, 39), any effect at later gestational stages may be secondary to effects on other fetal reproductive or metabolic tissues programmed by androgenization before d 90. The midgestation sheep fetal ovary has the capacity to be steroidogenically active, expressing a number of enzymatic components of the steroid biosynthesis pathway, including StAR, CYP11A, HSD3B1, CYP17, and CYP19 (18). Products of these pathways, including progesterone, androstenedione, testosterone, and estradiol, are readily detectable in ovine fetal ovarian tissue homogenates by d 75 (18). The role of fetal ovarian steroid production is not clear, but there are ARs and ERs in multiple tissues, including the ovary (17). We showed a significant reduction of StAR, CYP11A, and CYP17 transcripts in the d 90 maternal TP-exposed ovary. StAR and CYP11A are rate limiting steps in the steroidogenic pathway (40), therefore these intrinsic alterations after increased exposure to androgens could have a significant impact on the steroidogenic capability of the fetal ovary. In addition, maternal prenatal androgenization also led to a decrease in d 90 ovarian CYP17 that encodes the enzyme 17␣-hydroxylase, responsible for androgen production (reviewed in Ref. 41). Furthermore, the ability of the d 90 fetal ovary to respond to circulating endocrine steroidogenic molecules was impaired, because LHR expression was significantly reduced. Although the consequence of these alterations is not clear, these changes are an indication that the steroidogenic function of the fetal ovary at d 90 is altered by maternal androgenization. As the midgestation fetal ovary matures, cells differentiate, and this is associated with a reduction in germ cell number. We therefore investigated whether the Id genes are expressed in the fetal ovary and assessed whether fetal androgenization could augment their expression to delay some differentiation pathways to potentially alter ovarian cell growth and development. We have shown for the first time that all four Ids are differentially expressed in germ cells and somatic cells of the fetal ovary at the time of follicle formation. Although maternal androgen treatment had no effect on Id expression at the mRNA level, we found a striking difference in the immunostaining for Id1 in the d 90 fetal ovary. In the fetal ovaries of C animals, moderate Id1 staining could be detected in the germ cells; however, after maternal androgenization, there was intense staining of the germ and somatic cells. Whether this is a direct or indirect effect of androgen is not known, and it is also not clear what functional effect this change would have. Considering its role in other tissues, Id1 may facilitate suppression of subsets of genes required for cell-spe-



2055

should be noted that that study, and most previous studies of the ovine adult phenotype after prenatal androgenization, commenced maternal androgen treatment at d 30 (10, 35, 39, 44, 45). However, it is clear that starting treatment on d 60 still results in a reproductive phenotype with reduced estrous behavior (46), as well as hormonal abnormalities, including increased LH pulse frequencies (34). However, the nonhuman primate model is much more characterized and shows a similar effect on menstrual cycles, serum gonadotropins and androgens, and importantly PCO morphology, when the androgens are started at the early or later stage (8, 47– 49). One of the benefits of starting androgenization at d 60 is that the female offspring do not develop external male genitalia and, like women with PCOS, have more normal female genitalia (46). Starting at d 60 also facilitates androgenization of the fetus using direct injection. We were therefore able to study the effects of androgenization on the d 70 fetal ovary using both indirect maternal treatment and a novel model where TP was directly injected into the fetus. The placenta acts as a metabolic maternalfetal barrier to TP, because it has the capacity to aromatize androgens into estrogens. Therefore, some effects observed in offspring after maternal androgenization may be mediated by alterations in fetal estradiol exposure. Indeed, although elements of the PCOS phenotype were seen in sheep exposed FIG. 5. Effect of TP exposure on ovarian CYP11A, CYP17, CYP19, SLIT2, SLIT3, and ROBO1 to the nonaromatizable androgen dihygene expression in the indirect maternal (A–C and G–I, respectively) and direct fetal-treated drotestosterone (DHT), there was not (D–F and J–L, respectively) d 70 fetus. A Student’s t test was carried out, where *, P ⬍ 0.05 and ***, P ⬍ 0.001. the clear alterations in ovarian morphology observed in offspring from TPexposed mothers (45), although adult cific development and apoptosis. The Ids are generally gonadotropin profiles may still be perturbed (50, 51). We most highly expressed at times when proliferation is high injected TP into fetuses at d 60 and studied the ovaries at and are decreased in terminally differentiated cells (42, 43). With this in mind, it is possible that the up-regulation d 70, when the circulating testosterone concentrations of Id1 protein reflects further functional changes in the were still markedly elevated. The fetal ovary expresses fetal ovary and may delay gene expression necessary for aromatase, and we do not know if the intraovarian estradiol concentrations are increased after fetal injection of ovarian developmental and maturational processes. It remains possible that an increase in Id1 at d 90 could TP. However, there was no difference in the circulating be a precursor to the increased number of growing follicles estradiol concentration, unlike fetuses exposed to materpresent in the d 140 ovary, reported previously (38). It nal TP administration. We were therefore able to compare

2056

Hogg et al.


FIG. 6. Expression of Id3 mRNA in the d 70 ovary of direct fetal (A) and indirect maternal (B) TP-treated animals is shown with a representative image of Id3 protein expression in the d 70 maternal C and TP-treated ovary (C; arrows indicate intense somatic cell staining, and inset represents negative C). Scale bars, 50 ␮m. Quantification of Id3 protein is illustrated in d 70 direct fetal (D) and indirect maternal (E)-treated ovaries. A Student’s t test (A and B) or a ␹2 statistical test (D and E) was performed, where *, P ⬍ 0.05 and **, P ⬍ 0.01.

the fetal ovarian effects of very high concentrations of testosterone alone with a moderate elevation of testosterone and estradiol associated with an adult ovarian phenotype. The changes in the expression of the steroidogenic enzymes observed at d 90 of gestation were not seen at d 70 of gestation after maternal androgenization. Although no change in CYP11A mRNA expression occurred at d 70 with maternal androgenization, direct fetal exposure to TP did inhibit the expression of CYP11A. This may suggest that androgens directly regulate CYP11A expression and that it occurs earlier after exposure to higher concentrations of TP. Aromatase (CYP19) was significantly increased in the direct treatment d 70 ovaries, whereas no change occurred in the maternal exposure cohort. The significance of this in the fetus is not known, nor is it clear if this is a dose effect of the testosterone or whether raised


estradiol can inhibit a testosterone-dependent increase in expression. Direct fetal TP treatment also reduced the mRNA expression of both ER␣ and ER␤, a feature not present in the maternally TP-exposed d 70 or 90 fetal ovary. This suggests that androgens can affect the ability of the ovary to respond to estradiol. Nonspecific staining in our immunohistochemistry Cs meant that we could not confirm this at a protein level, but a previous study has described expression of ER␣ and ER␤ in multiple fetal ovarian cell types, including germ cells, surface epithelium, and stromal cells (17). It is therefore possible that TP can directly affect fetal ovary somatic cell function. Although the role of ERs in the fetal ovary is not clear, while ER knockout mice are infertile (52, 53), early reproductive tract development is unaffected (54). However, exposure to the synthetic estrogen diethylstilbestrol during gestation can result in developmental abnormalities in the gonad and reproductive tract (55, 56). Unlike ER expression, AR mRNA was not altered after either maternal or direct fetal prenatal androgen treatment; however, immunolocalization of the protein revealed an up-regulation of AR in direct TP-exposed d 70 ovaries, with a comparable nonsignificant trend present in maternal TP-exposed d 70 ovaries. Administration of DHT to pregnant rats has also been shown to lead to an increase in AR protein in the urogenital tract of female offspring (57). Although the increase in AR protein may be a result of ongoing exposure to high concentrations of testosterone, the fetal ovary appears to be able to adapt to increased androgen concentrations, because no changes were found in AR in d 90 fetal ovaries. This is in contrast to findings by Ortega et al. (58) in a prenatally androgenized ovine model, where treating pregnant ewes with DHT or TP from d 30 to 90 led to small increases in AR protein in d 90 and 140 fetal ovaries. It may be that these differences represent a concentration effect or are a result of longer exposure. However, the increase in ovarian AR after direct TP injection and after maternal DHT (58) suggests an androgenic mode of up-regulation. The lack of agreement between the mRNA and protein expression levels observed in this study may be due to an early mRNA response to increased androgens or changes in mRNA stability or translation or protein turnover. Because the entire frozen ovary was extracted for mRNA, additional validation, such as Western blotting, was not possible. Our observations have revealed that functional changes occur in the fetal ovary as early as d 70 and that these are more marked after direct treatment with TP, in which there are supraphysiological concentrations of circulating androgen. Similarly, when examining the Ids, Id3 was up-regulated after direct exposure to TP at both the


mRNA and protein level. After maternal androgenization, the trend to increase fetal ovarian Id3 mRNA was not significant; however, there was an increase in Id3 immunostaining. Id3 up-regulation occurred in the oocytes and was particularly clear in the somatic cells. This is also consistent with a role for the Ids in fetal ovarian development and indicates that up-regulation of their expression by androgens could have the potential to alter future ovarian development. In each of the previous examples, with the possible exception of CYP19 expression, the effects of direct and indirect androgenization on the d 70 fetal ovary have been similar. However, there was also a significant change in the expression of ROBO1 mRNA in the fetal ovary after direct TP exposure. The SLIT/ROBO pathway can regulate cell migration and promote cell death (19); therefore, we hypothesized that they had an important role in follicle formation in the fetal ovary. Given that there are steroid response elements in the promoter region of these genes (24), and steroids can inhibit SLIT/ROBO expression in the adult female reproductive tract (19, 24), we suspected that they may be target molecules for androgens in the fetal ovary. We found that the fetal TP injection inhibited the expression of ROBO1. ROBO1 is the most abundantly expressed ROBO, and its localization to the pregranulosa cells of forming follicles suggests an involvement in primordial follicle assembly (22). Although the functional effects have not been assessed, these findings would be consistent with a direct effect of androgen in reducing the ovarian SLIT/ROBO pathway. The nature of the change in ROBO1 expression in the fetal ovary and the response to different steroid exposure in vivo is not clear. However, an alteration in expression of SLIT/ ROBO components during the time corresponding to primordial follicle formation, when a finite population of resting oocytes are produced, could have an important impact on adult reproductive success. Although this study did not examine the adult ovarian phenotype, prenatal androgen exposure at different time points is associated with changes in adult ovarian function (6, 46). We speculated that this may be secondary to the hormonal environment, such as increased LH concentrations, and the metabolic environment, with increased insulin concentrations, that such fetal androgenization causes (32, 59). The midgestation fetal ovary undergoes marked growth, development, and remodeling to form the pool of follicles that will remain throughout the reproductive lifespan, and it expresses both AR and ER (15, 17). Although the morphology of the fetal ovary at d 90 of gestation is unchanged by exposure to androgens, we have found that prenatal androgenization leads to early subtle, but nevertheless notable, functional alterations. We be-


2057

lieve that the consequence of these changes is to alter ovarian development and contribute to an ovarian phenotype in later life. Although estradiol may have a role in manipulating ovarian development, we have shown that androgens have direct effects. However, even in the presence of large concentrations of TP in the pharmacological range, the effects remain subtle. We suggest that the PCO phenotype in the adult may not entirely be due to the hormonal and metabolic environment and that there may be direct androgen effects on the developing ovary before d 90 that program future ovarian morphology.

Acknowledgments We thank Dr. Mick Rae for helpful discussions and Joan Docherty and the staff of the Marshall Building for excellent animal care. Address all correspondence and requests for reprints to: Kirsten Hogg, Centre for Reproductive Biology, The Queen’s Medical Research Institute, 47 Little France Crescent, Edinburgh EH16 4SA, United Kingdom. E-mail: [email protected]. This work was supported by Medical Research Council Project Grants G0500717 (to W.C.D.) and U.1296.00.05.00007.01 (to A.S.M.). K.H. has a University of Edinburgh College of Medicine and Veterinary Medicine studentship. W.C.D. is supported by a Scottish Senior Clinical fellowship from the Scottish Funding Council. Disclosure Summary: The authors have nothing to disclose.

References 1. Barker DJP 2004 The developmental origins of adult disease. J Am Coll Nutr 23:588S–595S 2. Langley-Evans SC 2001 Fetal programming of cardiovascular function through exposure to maternal undernutrition. P Nutr Soc 60: 505–513 3. Maric C 2007 Mechanisms of fetal programming of adult hypertension: role of sex hormones. Hypertension 50:605– 606 4. Rhind SM, Rae MT, Brooks AN 2001 Effects of nutrition and environmental factors on the fetal programming of the reproductive axis. Reproduction 122:205–214 5. Franks S 1995 Polycystic ovary syndrome. N Engl J Med 333:853– 861 6. Abbott DH, Barnett DK, Bruns CM, Dumesic DA 2005 Androgen excess fetal programming of female reproduction: a developmental aetiology for polycystic ovary syndrome? Hum Reprod Update 11: 357–374 7. Eisner JR, Barnett MA, Dumesic DA, Abbott DH 2002 Ovarian hyperandrogenism in adult female rhesus monkeys exposed to prenatal androgen excess. Fertil Steril 77:167–172 8. Eisner JR, Dumesic DA, Kemnitz JW, Colman RJ, Abbott DH 2003 Increased adiposity in female rhesus monkeys exposed to androgen excess during early gestation. Obesity 11:279 –286 9. Manneras L, Cajander S, Holmang A, Seleskovic Z, Lystig T, Lonn M, Stener-Victorin E 2007 A new rat model exhibiting both ovarian and metabolic characteristics of polycystic ovary syndrome. Endocrinology 148:3781–3791

2058

Hogg et al.


10. Recabarren SE, Padmanabhan V, Codner E, Lobos A, Duran C, Vidal M, Foster DL, Sir-Petermann T 2005 Postnatal developmental consequences of altered insulin sensitivity in female sheep treated prenatally with testosterone. Am J Physiol Endocrinol Metab 289: E801–E806 11. Padmanabhan V, Veiga-Lopez A, Abbott DH, Recabarren SE, Herkimer C 2010 Developmental programming: impact of prenatal testosterone excess and postnatal weight gain on insulin sensitivity index and transfer of traits to offspring of overweight females. Endocrinology 151:595– 605 12. Murray RD, Davison RM, Russell RC, Conway GS 2000 Clinical presentation of PCOS following development of an insulinoma: case report. Hum Reprod 15:86 – 88 13. Vendola KA, Zhou J, Adesanya OO, Weil SJ, Bondy CA 1998 Androgens stimulate early stages of follicular growth in the primate ovary. J Clin Invest 101:2622–2629 14. Fulton N, Martins da Silva SJ, Bayne RAL, Anderson RA 2005 Germ cell proliferation and apoptosis in the developing human ovary. J Clin Endocrinol Metab 90:4664 – 4670 15. McNatty KP, Smith P, Hudson NL, Heath DA, Tisdall DJ, O WS, Braw-Tal R 1995 Development of the sheep ovary during fetal and early neonatal life and the effect of fecundity genes. J Reprod Fertil Suppl 49:123–135 16. Qi A, Zhang ZP, Cao GF, Zhang Y 2008 Histological study of germ cells development and apoptosis in mongolian sheep fetal ovaries. Anim Reprod Sci 103:179 –186 17. Juengel JL, Heath DA, Quirke LD, McNatty KP 2006 Oestrogen receptor ␣ and ␤, androgen receptor and progesterone receptor mRNA and protein localisation within the developing ovary and in small growing follicles of sheep. Reproduction 131:81–92 18. Quirke LD, Juengel JL, Tisdall DJ, Lun S, Heath DA, McNatty KP 2001 Ontogeny of steroidogenesis in the fetal sheep gonad. Biol Reprod 65:216 –228 19. Dickinson RE, Duncan WC 2010 The SLIT-ROBO pathway: a regulator of cell function with implications for the reproductive system. Reproduction 139:697–704 20. Kidd T, Bland KS, Goodman CS 1999 Slit is the midline repellent for the robo receptor in Drosophila. Cell 96:785–794 21. Wang B, Xiao Y, Ding B-B, Zhang N, Yuan X-b, Gui L, Qian K-X, Duan S, Chen Z, Rao Y, Geng J-G 2003 Induction of tumor angiogenesis by Slit-Robo signaling and inhibition of cancer growth by blocking Robo activity. Cancer Cell 4:19 –29 22. Dickinson RE, Hryhorskyj L, Tremewan H, Hogg K, Thomson AA, McNeilly AS, Duncan WC 2010 Involvement of the SLIT/ROBO pathway in follicle development in the fetal ovary. Reproduction 139:395– 407 23. Dickinson RE, Myers M, Duncan WC 2008 Novel regulated expression of the SLIT/ROBO pathway in the ovary: possible role during luteolysis in women. Endocrinology 149:5024 –5034 24. Duncan WC, McDonald SE, Dickinson RE, Shaw JLV, Lourenco PC, Wheelhouse N, Lee KF, Critchley HOD, Horne AW 2010 Expression of the repulsive SLIT/ROBO pathway in the human endometrium and Fallopian tube. Mol Hum Reprod 16:950 –959 25. O’Toole PJ, Inoue T, Emerson L, Morrison IEG, Mackie AR, Cherry RJ, Norton JD 2003 Id proteins negatively regulate basic helix-loophelix transcription factor function by disrupting subnuclear compartmentalization. J Biol Chem 278:45770 – 45776 26. Norton JD 2000 ID helix-loop-helix proteins in cell growth, differentiation and tumorigenesis. J Cell Sci 113:3897–3905 27. Hogg K, Etherington SL, Young JM, McNeilly AS, Duncan WC 2010 Inhibitor of differentiation (Id) genes are expressed in the steroidogenic cells of the ovine ovary and are differentially regulated by members of the transforming growth factor-␤ family. Endocrinology 151:1247–1256 28. Childs AJ, Kinnell HL, Collins CS, Hogg K, Bayne RAL, Green SL, McNeilly AS, Anderson RA 2010 BMP Signalling in the human fetal ovary is developmentally-regulated and promotes primordial germ cell apoptosis. Stem Cells 28:1368 –1378


29. McNatty KP, Baird DT, Bolton A, Chambers P, Corker CS, McLean H 1976 Concentration of oestrogens and androgens in human ovarian venous plasma and follicular fluid throughout the menstrual cycle. J Endocrinol 71:77– 85 30. Rozen S, Skaletsky HJ 2000 Primer3 on the WWW for general users and for biologist programmers. In: Krawetz S, Misener S, eds. Bioinformatics methods and protocols: methods in molecular biology. Totowa, NJ: Humana Press; 365–386 31. Kero J, Poutanen M, Zhang F-P, Rahman N, McNicol AM, Nilson JH, Keri RA, Huhtaniemi IT 2000 Elevated luteinizing hormone induces expression of its receptor and promotes steroidogenesis in the adrenal cortex. J Clin Invest 105:633– 641 32. Abbott DH, Dumesic DA 2009 Fetal androgen excess provides a developmental origin for polycystic ovary syndrome. Expert Rev Obstet Gynecol 4:1–7 33. Xita N, Tsatsoulis A 2006 Review: fetal programming of polycystic ovary syndrome by androgen excess: evidence from experimental, clinical, and genetic association studies. J Clin Endocrinol Metab 91:1660 –1666 34. Savabieasfahani M, Lee JS, Herkimer C, Sharma TP, Foster DL, Padmanabhan V 2005 Fetal programming: testosterone exposure of the female sheep during midgestation disrupts the dynamics of its adult gonadotropin secretion during the periovulatory period. Biol Reprod 72:221–229 35. Sharma TP, Herkimer C, West C, Ye W, Birch R, Robinson JE, Foster DL, Padmanabhan V 2002 Fetal programming: prenatal androgen disrupts positive feedback actions of estradiol but does not affect timing of puberty in female sheep. Biol Reprod 66:924 –933 36. Franks S, Stark J, Hardy K 2008 Follicle dynamics and anovulation in polycystic ovary syndrome. Hum Reprod Update 14:367–378 37. Stubbs SA, Stark J, Dilworth SM, Franks S, Hardy K 2007 Abnormal preantral folliculogenesis in polycystic ovaries is associated with increased granulosa cell division. J Clin Endocrinol Metab 92: 4418 – 4426 38. Smith P, Steckler TL, Veiga-Lopez A, Padmanabhan V 2009 Developmental programming: differential effects of prenatal testosterone and dihydrotestosterone on follicular recruitment, depletion of follicular reserve, and ovarian morphology in sheep. Biol Reprod 80:726 –736 39. Birch RA, Padmanabhan V, Foster DL, Unsworth WP, Robinson JE 2003 Prenatal programming of reproductive neuroendocrine function: fetal androgen exposure produces progressive disruption of reproductive cycles in sheep. Endocrinology 144:1426 –1434 40. Stocco DM, Clark BJ 1996 Role of the steroidogenic acute regulatory protein (StAR) in steroidogenesis. Biochem Pharmacol 51:197–205 41. Sewer MB, Dammer EB, Jagarlapudi S 2007 Transcriptional regulation of adrenocortical steroidogenic gene expression. Drug Metab Rev 39:371–388 42. Barone MV, Pepperkok R, Peverali FA, Philipson L 1994 Id proteins control growth induction in mammalian cells. Proc Natl Acad Sci USA 91:4985– 4988 43. Hara E, Yamaguchi T, Nojima H, Ide T, Campisi J, Okayama H, Oda K 1994 Id-related genes encoding helix-loop-helix proteins are required for G1 progression and are repressed in senescent human fibroblasts. J Biol Chem 269:2139 –2145 44. Steckler T, Manikkam M, Inskeep EK, Padmanabhan V 2007 Developmental programming: follicular persistence in prenatal testosterone-treated sheep is not programmed by androgenic actions of testosterone. Endocrinology 148:3532–3540 45. Veiga-Lopez A, Astapova OI, Aizenberg EF, Lee JS, Padmanabhan V 2009 Developmental programming: contribution of prenatal androgen and estrogen to estradiol feedback systems and periovulatory hormonal dynamics in sheep. Biol Reprod 80:718 –725 46. Roberts EK, Padmanabhan V, Lee TM 2008 Differential effects of prenatal testosterone timing and duration on phenotypic and behavioral masculinization and defeminization of female sheep. Biol Reprod 79:43–50 47. Dumesic DA, Schramm RD, Bird IM, Peterson E, Paprocki AM,


48.

49.

50.

51.

52.


Zhou R, Abbott DH 2003 Reduced intrafollicular androstenedione and estradiol levels in early-treated prenatally androgenized female rhesus monkeys receiving follicle-stimulating hormone therapy for in vitro fertilization. Biol Reprod 69:1213–1219 Dumesic DA, Schramm RD, Peterson E, Paprocki AM, Zhou R, Abbott DH 2002 Impaired developmental competence of oocytes in adult prenatally androgenized female rhesus monkeys undergoing gonadotropin stimulation for in vitro fertilization. J Clin Endocrinol Metab 87:1111–1119 Eisner JR, Dumesic DA, Kemnitz JW, Abbott DH 2000 Timing of prenatal androgen excess determines differential impairment in insulin secretion and action in adult female rhesus monkeys. J Clin Endocrinol Metab 85:1206 –1210 West C, Foster DL, Evans NP, Robinson J, Padmanabhan V 2001 Intra-follicular activin availability is altered in prenatally-androgenized lambs. Mol Cell Endocrinol 185:51–59 Masek KS, Wood RI, Foster DL 1999 Prenatal dihydrotestosterone differentially masculinizes tonic and surge modes of luteinizing hormone secretion in sheep. Endocrinology 140:3459 –3466 Dupont S, Krust A, Gansmuller A, Dierich A, Chambon P, Mark M 2000 Effect of single and compound knockouts of estrogen receptors ␣ (ER␣) and ␤ (ER␤) on mouse reproductive phenotypes. Development 127:4277– 4291

2059

53. Krege JH, Hodgin JB, Couse JF, Enmark E, Warner M, Mahler JF, Sar M, Korach KS, Gustafsson J-Å, Smithies O 1998 Generation and reproductive phenotypes of mice lacking estrogen receptor ␤. Proc Natl Acad Sci USA 95:15677–15682 54. Lubahn DB, Moyer JS, Golding TS, Couse JF, Korach KS, Smithies O 1993 Alteration of reproductive function but not prenatal sexual development after insertional disruption of the mouse estrogen receptor gene. Proc Natl Acad Sci USA 90:11162–11166 55. Ma L 2009 Endocrine disruptors in female reproductive tract development and carcinogenesis. Trends Endocrin Met 20:357–363 56. Newbold RR 2004 Lessons learned from perinatal exposure to diethylstilbestrol. Toxicol Appl Pharm 199:142–150 57. Bentvelsen FM, McPhaul MJ, Wilson JD, George FW 1994 The androgen receptor of the urogenital tract of the fetal rat is regulated by androgen. Mol Cell Endocrinol 105:21–26 58. Ortega HH, Salvetti NR, Padmanabhan V 2009 Developmental programming: prenatal androgen excess disrupts ovarian steroid receptor balance. Reproduction 137:865– 877 59. Padmanabhan V, Manikkam M, Recabarren S, Foster D 2006 Prenatal testosterone excess programs reproductive and metabolic dysfunction in the female. Mol Cell Endocrinol 246:165–174

Final author versions of articles are published online within 25 days of acceptance. www.endo-society.org