|Home | About | Journals | Submit | Contact Us | Français|
Targeted disruption of the inhibin α gene (Inha−/−) in mice results in an ovarian phenotype of granulosa cell tumors that renders the animals infertile. Little is known about the reproductive defects prior to tumor development. Here, we report novel data on early follicle dynamics in Inha−/− mice, which demonstrate that inhibin α has important consequences upon follicle development. Morphological changes in both germ and somatic cells were evident in postnatal day 12 ovaries, with Inha−/− mice exhibiting numerous multilayered follicles that were far more advanced than those observed in age-matched controls. These changes were accompanied by alterations in follicle dynamics such that Inha−/− ovaries had fewer follicles in the resting pool and more committed in the growth phase. Absence of inhibin α resulted in advanced follicular maturation as marked by premature loss of anti-Müllerian hormone (AMH) in secondary follicles. Additionally, gene expression analysis revealed changes in factors known to be vital for oocyte and follicle development. Together, these data provide key evidence to suggest that regulation of the inhibin/activin system is essential for early folliculogenesis in the prepubertal mouse ovary.
Inhibins, activins and other related dimeric members of the transforming growth factor β (TGFβ) family are renowned for their pleiotropic roles in various developmental and physiological processes. Development of the ovarian follicle (folliculogenesis) is a complex event orchestrated by intraovarian growth factors, pituitary gonadotropins, and steroidogenic hormones (Matzuk et al., 2002). The elucidation of many of the specific roles of locally-derived growth factors during folliculogenesis has been revealed through the analysis of transgenic mouse models (Matzuk and Lamb, 2008; Myers and Pangas, 2009; Pangas and Matzuk, 2008), which have provided key tools in dissecting the molecular pathways involved in follicular growth and development. Particular examples of targeted deletions of members belonging to the TGFβ family include anti-Müllerian hormone (AMH), growth differentiation factor 9 (GDF9) and activin; all of which show alterations in folliculogenesis (Dong et al., 1996; Durlinger et al., 1999; Pangas et al., 2007). Unlike the aforementioned growth factors, the targeted disruption of the inhibin α gene (Inhα−/−) results in granulosa cell tumors, rendering the mice infertile and ultimately resulting in a cachexia-related death (Matzuk et al., 1994; Matzuk et al., 1992).
Initially discovered for their opposing abilities to suppress pituitary follicle stimulating hormone (FSH), activins and inhibins have also been shown to have diverse endocrine, paracrine and autocrine functions in both gonadal and extragonadal systems (Chang et al., 2002; Pangas and Matzuk, 2008). These two members of the TGFβ family are well characterized for their mutually antagonistic properties; in part, this is a consequence of their related protein structure. The generation of an α subunit (encoded by the Inha gene) combined with a β subunit (encoded by the Inhba or Inhbb gene) makes inhibins (α:βA, α:βB), whilst homodimers (βA;βA, βB;βB) or heterodimers (βA;βB) of the β subunits make activin. The major gonadal sites of inhibin synthesis are the Sertoli cells in males and the granulosa cells in females. Reports thus far on the Inha−/− mouse model confirm that there are important roles for inhibin in normal follicle recruitment and gonadal tumor development (Matzuk et al., 1996; Wu et al., 2004). Because of the cancer phenotype, much of the published data on the Inha−/− mouse has focused upon the role of inhibin as a tumor-suppressor (Andreu-Vieyra et al., 2007; Brown et al., 2000; Burns et al., 2003; Cipriano et al., 2001; Kumar et al., 1996; Li et al., 2007; Nagaraja et al., 2008) rather than its role in fertility.
Although many studies have shown activins and inhibins to be critical regulators of gonadal function, little is known about their specific roles during the early stages of folliculogenesis. Emerging evidence in both mouse and human systems suggests that the inhibin/activin system is important for both cell proliferation (germ and somatic) and follicular assembly (Bristol-Gould et al., 2006; Martins da Silva et al., 2004). We therefore hypothesized that loss of inhibin in the Inha−/− mouse, or alternatively, the gain of unopposed activins, would affect early postnatal folliculogenesis, as the homeostatic relationship between these related dimeric glycoproteins is disrupted. Our findings demonstrate that ovaries from Inha−/− mice display a precocious follicular development phenotype that is evident by the first week of life and markedly apparent in the subsequent weeks of postnatal life. Inha−/− ovaries exhibit alterations in follicular dynamics with evidence of early follicle recruitment and strikingly advanced follicular growth. Analysis of candidate genes revealed the alterations in intraovarian factors notorious for playing key roles in oocyte and follicle development. Additionally, a marked defect in cell-to-cell communication between the germ and somatic cell types further contributes to the perturbed ovarian phenotype in Inha−/− mice. Our findings highlight the importance of the inhibin/activin system during early folliculogenesis and illustrate how a refined balance of growth factors is required for proper follicle growth and development.
Experimental Inhatm1Zuk/tm1Zuk (herein called Inha−/−) mice were maintained on a C57BL/6J;129S5/SvEvBrd mixed genetic background and kept in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Inha−/− mice were generated as previously described (Matzuk et al., 1992) and genotyped by PCR analysis of genomic tail DNA using the primers E2-2 (5'-GGTCTCCTGCGGCTTTGCGC-3'), HPRT (5'-GGATATGCCCTTGACTATAATG-3'), and Intron (5'- CCTGGGTGGAGCAGGATATGG-3') to determine whether the mouse was wildtype, heterozygous or homozygous null at the inhibin α locus.
Mice were anesthetized by isoflurane inhalation (Abbott Laboratories; Abbott Park, IL) and euthanized by decapitation. Ovaries were collected from mice postnatal day 3, 6 and 12 days, with the day of birth considered as postnatal day 0. Ovarian tisssue was either fixed in Bouin's solution (Sigma; St. Louis, MO) or stored in RNAlater (Ambion; Austin, TX) at −80°C for subsequent use.
Tissue processing and embedding were performed by the Department of Pathology Core Facility (Baylor College of Medicine) using standard techniques. Ovaries were serially sectioned into 5μm sections with approximately 5 sections per slide. Sections were stained with periodic acid Schiff's reagent and hematoxylin. For follicle counting, 3–5 ovaries were assessed per genotype for each age group and every 5th section was counted. Follicles were classified as follows; primordial, an oocyte surrounded by squamous pregranulosa cells; primary, an oocyte surrounded by predominantly cuboidal granulosa cells; secondary, an oocyte surrounded by 2–4 layers of granulosa cells, or an oocyte surrounded by more than 4 layers of granulosa cells without an antrum. To avoid double counting secondary follicles, only oocytes with a visible nucleus were counted. Data were expressed as percentage of follicles type per total follicles. For oocyte and follicle measurements, the shortest and longest diameter of both the oocyte and follicle was taken and the mean was recorded. Oocytes were grouped in 10μm increments based upon their size. Graphical representation of oocyte size vs. mean follicle diameter was plotted with SEM for follicle diameters on the Y axis.
Immunohistochemistry was performed using the Vectorstain ABC method (Vector Laboratories; Burlingame, CA) as previously described (Pangas et al., 2007). Goat polyclonal anti-AMH (1:1500 dilution, Santa Cruz Biotechnology; Santa Cruz, CA) was used on at least 3 different sections from 5 animals of each genotype. Immunoreactivity was visualized by diaminobenzidine (Vector Laboratories) and counterstained with hematoxylin.
Blood was recovered from postnatal day 3, 6 and 12 day old animals. Due to limited volume of blood in younger mice, pooled serum (for all age groups) from 3–5 animals was used, and 6 independent pools were assayed. Blood was collected and serum separated by centrifugation in Microcontainer tubes (Sarstedt; Nümbrecht, Germany) and stored at −20°C until use. FSH serum measurements were performed by The University of Virginia Ligand Core Facility (Specialized Cooperative Centers Program in Reproduction Research NICHD/NIH U54HD28934). The mouse FSH RIA has a sensitivity of 2ng/ml and an average intraassay coefficient of 10.1% and an interassay coefficient variation of 13.3%.
RNA was isolated using the QIAGEN (Valencia, CA) RNeasy Micro kit. RNA concentration was quantified using a NanoDrop Spectrophotometer ND-1000 (NanoDrop Technologies; Wilmington, DE). cDNA was reverse transcribed from 200ng of total RNA in a 20μl reaction using High Capacity RNA-to-cDNA Master Mix (Applied Biosystems; Foster City, CA). Real-time PCR was performed using the Applied Biosystems Prism 7500 Sequence Dectection System, Taqman Master Mix and Gene Expression Assays. The following Taqman (Applied Biosystems) Assays used for qPCR: mouse glyceraldehyde-3-phosphate dehydrogenase (Gapdh) (Pangas et al., 2007) as an endogenous control; Amh, Mm03023963_m1; Bmp15, Mm00437797_m1; Ccnb1, Mm00838401_g1; Ccnd2, Mm00438071_ml; Cyp19, Mm00484049_m1; Fshr, Mm00442819; Gdf9, Mm00433565_m1; Inhba, Mm00434338_ml; Inhbb, Mm03023992_ml; Kitl, Mm00442972_ml. All qPCR data was analyzed by the ΔΔ cycle threshold method using the ABI 7500 System Software (version 1.2.3) and normalized to the endogenous reference (Gapdh). One control sample was randomly chosen as the calibrator sample and the mean and SEM were calculated. Results were plotted using the relative expression of each target gene with each sample compared to the control.
Statistical analysis were carried out using GraphPad Prism 5 (GraphPad Software; La Jolla, CA). Two-tailed unpaired t-tests were used for single comparisons. At least 3 independent experiments were carried out at all times and P<0.05 was considered statistically significant.
During the first few weeks of life, there is significant organisation and development within the postnatal mouse ovary (Pepling, 2006). This is the time of primordial follicle formation and by postnatal day 3, the majority of oocytes are encapsulated by somatic cells (Pepling, 2006). Ovaries from WT (Fig. 1A, C, E, G) and Inha−/− (Fig. 1B, D, F, H) mice were examined at postnatal day 3, 6 and 12. At postnatal day 3, there were no obvious gross morphological differences between WT and Inhα−/− animals with both ovaries consisting predominantly of primordial follicles and some one layer follicles with growing oocytes (primary follicles) (Fig. 1A, B). By postnatal day 6, multiple 2 layer follicles (secondary) were evident in both genotypes. Although Inha−/− ovaries appeared to have more growing multilayer secondary follicles at this age (Fig. 1C, D), morphometric measurements of oocyte size vs. mean follicular diameter did not reveal any major changes (Supplementary Fig. 1). By postnatal day 12, there was a striking difference in size between the WT (Fig. E) and Inha−/− ovaries. Histological images of the ovaries taken at the same magnification showed more than a 2-fold increase in size of Inha−/− ovaries (Fig1. F). While WT littermates did not exhibit follicles with more than 2–4 layers of granulosa cells (Fig1. G), Inha−/− ovaries had numerous multilayered follicles (Fig. 1 F, H), some of which had greater than 10 layers of granulosa cells. Small antral-like cavities were also observed in many of the large multilayered follicles in Inha−/− ovaries (Fig. 1H). Such features indicated an advanced follicular phenotype in Inha−/− ovaries.
Folliculogenesis prior to the antral stage is thought to be independent of the gonadotropins (Kumar et al., 1996). There is however, a peak of FSH that is reported to occur at approximately postnatal day 10 (Weng et al., 2006). Because inhibin/activin β subunits are up regulated in Inha−/− ovaries (Table.1), we wanted to determine if the increased follicle maturation in Inha−/− mice was a consequence of increased circulating FSH. Serum FSH levels remained unchanged between WT and Inha−/− female mice at postnatal day 3 (15.11 ± 1.23 ng/ml vs 13.88 ± 1.62), postnatal day 6 (22.98 ± 1.47 vs 26.51 ± 3.47 ng/ml) and postnatal day 12 (29.98 ± 1.50 vs 27.53 ± 1.98 ng/ml) (Fig. 2 A). To assess if the advanced follicular growth was the result of increased FSH sensitivity, we also measured the mRNA levels of the FSH receptor (Fshr) and the cytochrome P450 aromatase (Cyp19) a well-known target gene of FSH action. There was a non-significant trend for Fshr expression to increase and no change in Cyp19 (FSH target gene) expression in Inha−/− ovaries (Fig. 2B,C). To assess any changes in granulosa cell proliferation, we also investigated cell cycle genes, cyclin D2 (Ccnd2) (Fig. 2D) and cyclin B1 (Ccnb1) (Table. 1). Ccnd2, a known target gene of activin or GDF9 (Ogawa et al., 2003), and activin and FSH (Ogawa et al., 2003; Park et al., 2005) was significantly increased in Inha−/− ovaries (Fig. 2D). There was no change in the mRNA levels of Ccnb1 (Table 1). Together, these data suggest that the acceleration of follicle growth in Inha−/− mice at postnatal day 12 is not secondary to increased circulating FSH or increased FSH sensitivity, but more likely a result of locally produced growth factors.
The formation of primordial follicles is central to a female's reproductive lifespan and is dependent upon the process of pregranulosa cells encapsulating a single oocyte. To determine if the number of oocytes within the Inha−/− neonatal ovary, or moreover whether the formation of primordial follicles was altered, we investigated postnatal day 3 Inha−/− mice. There was no difference in the total number of oocytes counted in Inha−/− ovaries compared with age-matched WT controls, and this was true for both primordial and primary follicle types (Fig. 3A). To further determine if the loss of inhibin αhad an effect upon follicles during the initial waves of folliculogenesis, we assessed Inha−/− and WT ovaries at postnatal day 12. The numbers of primordial, primary and secondary follicles were assessed. Secondary follicles were broken down into three categories, 2–4 layers of granulosa cells; more than 4 layers of granulosa cells and; more than 4 layers of granulosa cells with an antral cavity) (Fig. 3B). Analysis of follicle dynamics revealed that Inhα−/−mice exhibit an acceleration in the number of follicles leaving the resting pool (primordial follicles) and an increase in the number of follicles in the growing pool (primary and secondary stages). There was approximately a 30% reduction in the number of primordial follicles in Inhα−/− ovaries compared to WT littermates. Concomitantly, the reduction of primordial follicles in Inha−/− ovaries was associated with an increase of follicles at all growing stages, demonstrating that the follicles in Inhα−/− mice are developing at an accelerated rate compared to WT mice.
AMH protein expression is present in non-atretic preantral and small antral follicles of postnatal ovaries (Baarends et al., 1995; Hirobe et al., 1992). Once a follicle develops to a more advanced stage and becomes FSH sensitive, AMH protein expression is lost (Durlinger et al., 2002b). As expected, both adult (8 weeks) and prepubertal (postnatal day 12) WT ovaries showed positive staining for AMH protein in primary, secondary and early antral follicles (only present in 8 week old WT ovary) (Fig. 4A, B). Expression was not evident in primordial follicles (Fig. 4A inset) or corpora lutea (Fig. 4A). Although there was immunoreactivity for AMH in some secondary follicles in Inha−/− ovaries, staining patterns were variable with many 2–4 layered secondary or large multilayered follicles that had either totally lost expression of AMH or lost it with the exception of the cell layer closest to the oocyte (Fig. 4C,D). Similar to the WT ovary, positive staining for AMH was also evident in the early primary follicles of Inha−/− ovaries (Fig. 4E), and although AMH expression is not detected in WT primordial follicles (Durlinger et al., 2002b), the occasional primordial follicle located at periphery of the Inha−/−stained positive. We hypothesize that these follicles represent a population of primordial follicles that have more advanced granulosa cells and are probably in transition. Consistent with the early loss of AMH protein expression in Inha−/− ovaries, transcript levels of Amh were also significantly reduced in comparison to WT ovaries (Fig. 4F).
During early folliculogenesis, oocyte and granulosa cell proliferation occurs in a synchronous manner and is dependent upon locally produced growth factors such as GDF9, BMP15 and KITL. When viewed at high magnification, it became evident that many of the follicles in Inha−/− ovaries had lost this synchronous relationship. There were many small oocytes (20-30μm) contained in abnormally large follicles (Fig. 5B, C). These small oocytes enclosed in these large Inha−/− secondary follicles were comparable in size to those of WT primary follicles. Graphical representation of oocyte diameter vs. follicle diameter further reiterated the loss of synchronous oocyte-follicle growth in Inha−/− ovaries (Fig. 5D) at postnatal day 12, that remained unchanged at postnatal day 6 (Supplementary Fig. 1). Investigation into the expression of factors known to be involved in oocyte growth and development revealed that the expression of Gdf9 was significantly increased in Inha−/− ovaries whilst Bmp15 and Kitl were significantly reduced (Fig. 5D-E). These results suggest that the loss of synchronous growth between germ and somatic cells in Inha−/− follicles may be due to changes in the expression of growth factors that maintain synchronous oocyte-granulosa cell development.
Inhibins and activins play integral roles during various stages of folliculogenesis. We describe novel data on the early follicle dynamics in the postnatal Inha−/− mouse prior to tumor development. This mouse model provides an excellent in vivo tool to investigate the physiological andmechanistic roles for inhibin during the early stages of follicle formation and their subsequent entry into the growth phase. Unfortunately, with respect to folliculogenesis, the Inha−/− model is limited to development prior to puberty because beginning at around 4 weeks and with the onset pituitary gonadotropin secretions, Inha−/− mice begin to develop sex-cord stromal tumors (Kumar et al., 1996; Matzuk and Lamb, 2002; Nagaraja et al., 2008). In the current study, we reveal that the targeted deletion of inhibin αresults in precocious follicle development, involving the disruption of oocytegranulosa cell interactions and the generation of aberrant oocyte and somatic cell gene expression patterns.
The formation of the earliest follicles, the primordials, is a critical stage in female fertility and in the mouse, occurs during the first few days of postnatal life (Pepling and Spradling, 2001). Utilizing the Inha−/− mouse model, we wanted to address questions regarding the roles of the inhibin/activin system during early follicular development because activins are reported to play important roles in the proliferation and survival of germ and somatic cells proceeding primordial follicle formation in the human (Coutts et al., 2008; Martins da Silva et al., 2004) and during primordial follicle formation in the mouse (Bristol-Gould et al., 2006). Surprisingly, we found that loss of inhibin αhad no detrimental effect upon the formation of primordial follicles. At postnatal day 3, Inha−/− ovaries had similar morphology to their WT counterparts and had no differences in numbers of primordial or primary follicles. While injection of exogenous activin increases the formation of primordial follicles (Bristol-Gould et al., 2006), our studies suggest that loss of inhibin αhas no effect upon the number of follicles in the neonatal ovary. Although the Inha−/− mouse provides a unique model that allows an in vivo insight into the effects of unopposed activin upon germ cell endowment and follicle formation, the potential roles of other growth factors should also be considered.
By postnatal day 6, the there was no apparent difference between WT and Inha−/− ovaries however, by day 12, there was a striking difference. Follicles in postnatal day 12 Inha−/− mice had progressed to very large multilayered structures, consisting of more than 10 layers of granulosa cells. Both activins and follicle-stimulating hormone (FSH) are known regulators of granulosa cell proliferation (Hillier, 2001; Miro and Hillier, 1996), with activins in particular stimulating granulosa cell proliferation in small follicles that are FSH-independent (Miro and Hillier, 1996). The pituitary-derived gonadotropins, FSH and luteinizing hormone (LH), are major regulators of mammalian gonadal function (Halpin et al., 1986; Kumar et al., 1999), and are essential modulators for further follicle development and subsequent ovulation (Kumar et al., 1997; Peters et al., 1975), once the follicle has acquired an antral cavity. Because inhibins suppress pituitary FSH, and adult mice display increased serum FSH levels comparison to WT mice (Matzuk et al., 1992), we investigated serum FSH in prepubertal animals. However, there were no discernible differences in serum FSH in prepubertal Inha−/− mice at any of the postnatal days investigated (days 3, 6 and 12). We propose that during these early developmental stages prior to puberty and before the requirement of a gonadotropin-surge, inhibin does not have an endocrine effect upon FSH production. This coupled with no significant changes in the expression of the FSH receptor or CYP19, a well-known target gene of FSH action (Hillier et al., 1994), infer that the acceleration in granulosa cell proliferation in Inha−/− prepubertal animals is not a major direct effect of FSH, but more likely a consequence of intraovarian growth factors signaling.
We report a regulatory role for inhibin in primordial follicle activation and also at the rate at which they entered into the growth phase. Because AMH is a known regulator of primordial follicle recruitment, we postulate that the premature loss of AMH in Inha−/− ovaries contributes to the advanced rate of primordial follicle recruitment at postnatal day 12. Although mice genetically deficient in Amh do not display a reduction in the number of primordial follicles at postnatal day 25 (Durlinger et al., 1999), in vitro evidence suggests that AMH can act as an inhibitory growth factor in neonatal ovaries (Durlinger et al., 2002a). These data imply that AMH is one of many regulators of primordial follicle recruitment and functional redundancies are likely to exist between many of these growth factors. Together, our data suggests that inhibins have inhibitory roles upon primordial follicle recruitment (possibly via influencing AMH signaling) and a regulatory role that controls the rate that follicles mature (by inhibiting positive growth factors such as activin and GDF9).
Although many follicles within Inha−/− ovaries had progressed to very large multilayeredstructures, many had abnormal interactions between their germ and somatic cell components. A striking feature of the postnatal day 12 Inha−/− ovary was the dysregulation or loss of synergy between the oocyte and granulosa cell growth dynamics in many of the growing follicles. A number of the oocytes contained in the multilayered follicles in Inha−/− ovaries had unusually small diameters for follicles of such a size; diameters that were comparable to WT primary follicles. To our knowledge, the observed follicular phenotype in the prepubertal Inha−/− mouse model is unique, and with respect to oocyte morphology displays a phenotype opposite to other observed transgenic models, whereby loss of a germ cell factor (i.e., GDF9) or loss of a somatic cell factor (i.e., FOXO3A) causes enlargement of the oocyte (Dong et al., 1996; Liu et al., 2007). Future investigations into whether these small oocytes are growth restricted (as explained below) or are simply too immature for follicles of such diameters will offer further insight into key regulators of oocyte growth.
Analysis of genes integral for oocyte-follicle dynamics demonstrated that important intraovarian growth factors such as Kitl, Bmp15 and Gdf9 were altered in Inha−/− ovaries. We found that whilst the expression of Gdf9 was increased, both Bmp15 and Kitl were significantly reduced. Although both growth factors have important roles within the ovary, only GDF9 appears to be essential for female fertility in the mouse (Dong et al., 1996; Yan et al., 2001). Signaling pathways that regulate GDF9 are unclear, and it is possible that important intrafollicular signaling dynamics exist between GDF9 and the inhibin/activin system because the removal of the inhibin αgene overcomes the block at the primary follicle stage in Gdf9−/− null mice (Wu et al., 2004). In contrast to Gdf9 expression, we found reduced transcripts of oocyte-derived Bmp15 in prepubertal ovaries. Whilst the expression patterns of BMP15 have been reported in adult mice (Dube et al., 1998), the precise role and temporal expression patterns of BMP15 in the prepubertal mouse are still unclear. This coupled with the reported differences between poly and mono-ovular species (Dong et al., 1996; Juengel et al., 2004; Juengel et al., 2002; McNatty et al., 2005; Yan et al., 2001) make the role of BMP15 hard to decipher. This may suggest that the synergistic or compensatory roles for these oocyte-derived factors are limited to the gonadotropin-dependent or at least the later stages folliculogenesis involving the development of the oocyte-cumulus cell complex (Yan et al., 2001). Subsequent investigations will further determine the relevance and potential roles for BMP15 in early mouse folliculogenesis.
Granulosa cell-derived KITL is negatively regulated by both activin (Coutts et al., 2008; Pangas et al., 2007) and oocyte-derived GDF9 (Elvin et al., 1999; Wu et al., 2004). In light of its functional importance for oocyte growth and development (Thomas and Vanderhyden, 2006), we postulate that the reduction of Kitl in Inha−/− ovaries is a likely contributor to the small oocyte phenotype. Our working model hypothesizes that in the absence of inhibin, activin and/or GDF9 suppress Kitl, resulting in the loss of key a trophic signal required for oocyte growth. Simultaneously, the adjacent granulosa cells are responding to increased activin and GDF9 as positive regulators of their own proliferation (Fig. 6). Thus, in Inha−/− ovaries, alterations in these important growth factors are likely contributors to the loss of synchronous oocyte-follicle growth. This further suggests that activin and GDF9 at times may be functionally redundant (i.e., such that they can both inhibit the production of KITL and enhance granulosa cell proliferation). Both growth factors work via the same intracellular SMAD signaling proteins [SMAD2/3 (Gilchrist et al., 2006; Kaivo-Oja et al., 2003; Pangas and Matzuk, 2008; Roh et al., 2003)], in some instances stimulate or inhibit common genes (Eramaa et al., 1995; Kaivo-Oja et al., 2003) and have recently been shown to act in a co-operative manner to regulate key components of the activin signaling pathway (Shi et al., 2009).
In summary, the prepubertal Inha−/− mouse ovary provides a unique in vivo genetic system to study the dynamics of the inhibin/activin system during early follicle development. Unlike adult Inha−/− mice, pre prepubertal Inha−/− animals do not show evidence of tumor development. They are however, subjected to perturbed folliculogenesis. We show that in the absence of inhibin, the prepubertal ovary displays precocious follicle growth and alterations in intraovarian factors that are essential for folliculogenesis (Fig. 6). We propose that loss of inhibin α results in more granulosa cell-derived activin signaling (i.e. all of the β subunits form activin with no inhibin to antagonize activin signaling) and oocyte-derived GDF9, which then suppress Kitl expression in the granulosa cells. The reduction in KITL signaling to the oocyte via its receptor KIT has detrimental effects upon oocyte growth and contributes to the small oocyte phenotype observed in many of the Inha−/− follicles. Additionally, the loss of inhibin αinhibits the production of factors that play important roles in regulating the rate at which follicles enter the growth phase, including AMH. We hypothesize that in addition to being a tumor-suppressor, inhibin also plays a very important role in regulating follicle dynamics such that the ovary is not subjected to precocious development. These studies emphasize the need for further research of inhibin/activin system in clinical diseases such as premature ovarian failure.
This work was supported by a Burroughs Wellcome Career Award in Biomedical Sciences, a Duncan Scholar Award from the Dan L. Duncan Center, a Caroline Wiess Law Fund for Molecular Medicine and L.E. and Josephine S. Gordy Memorial Cancer Research Fund (to S.A.P) and a NIH grant CA60651 (to M.M.M). We thank the UVA Ligand Core NICHD/NIH U541146 HD28934 for FSH measurements.
Publisher's Disclaimer: This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to our customers we are providing this early version of the manuscript. The manuscript will undergo copyediting, typesetting, and review of the resulting proof before it is published in its final citable form. Please note that during the production process errors may be discovered which could affect the content, and all legal disclaimers that apply to the journal pertain.