Oocyte-Specific Deletion of Pten in Mice Reveals a Stage-Specific Function of PTEN/PI3K Signaling in Oocytes in Controlling Follicular Activation

Immature ovarian primordial follicles are essential for maintenance of the reproductive lifespan of female mammals. Recently, it was found that overactivation of the phosphatidylinositol 3-kinase (PI3K) signaling in oocytes of primordial follicles by an oocyte-specific deletion of Pten (phosphatase and tensin homolog deleted on chromosome ten), the gene encoding PI3K negative regulator PTEN, results in premature activation of the entire pool of primordial follicles, indicating that activation of the PI3K pathway in oocytes is important for control of follicular activation. To investigate whether PI3K signaling in oocytes of primary and further developed follicles also plays a role at later stages in follicular development and ovulation, we conditionally deleted the Pten gene from oocytes of primary and further developed follicles by using transgenic mice expressing zona pellucida 3 (Zp3) promoter-mediated Cre recombinase. Our results show that Pten was efficiently deleted from oocytes of primary and further developed follicles, as indicated by the elevated phosphorylation of the major PI3K downstream component Akt. However, follicular development was not altered and oocyte maturation was also normal, which led to normal fertility with unaltered litter size in the mutant mice. Our data indicate that properly controlled PTEN/PI3K-Akt signaling in oocytes is essential for control of the development of primordial follicles whereas overactivation of PI3K signaling in oocytes does not appear to affect the development of growing follicles. This suggests that there is a stage-specific function of PTEN/PI3K signaling in mouse oocytes that controls follicular activation.


Introduction
In women, the 300,000-400,000 ovarian primordial follicles at menarche serve as the source of fertilizable ova for the entire duration of reproductive life [1]. In order to ensure the proper length of reproductive life, the majority of primordial follicles remain in a dormant state, and only limited numbers of them are recruited into the growing follicle pool through follicular activation. Menopause occurs when the pool of primordial follicles has become exhausted [1][2][3].
Recently, studies from our research group have suggested that the phosphatidylinositol 3-kinase (PI3K) signaling pathway in oocytes plays an important role in oocyte growth during early follicular development [4]. The PI3K pathway is a fundamental signaling pathway for regulation of cell proliferation, survival, migration, and metabolism [5]. PI3Ks are lipid kinases that phosphorylate the 39-OH group on the inositol ring of inositol phospholipids. PTEN (phosphatase and tensin homolog deleted on chromosome ten), a lipid phosphatase, dephosphorylates the inositol phospholipids and thus functions as a major negative regulator of PI3K [5].
We have shown that deletion of the Pten gene from mouse oocytes of primordial follicles using transgenic mice expressing Cre recombinase mediated by the growth differentiation factor 9 (Gdf-9) promoter, results in excessive activation of the entire pool of primordial follicles, which in turn leads to premature depletion of all primordial follicles [6]. However, such mutant mice show normal ovulation and normal litter size in young adulthood before their ovarian follicles are depleted [6].
The current study was designed to investigate the role of PTEN/PI3K-Akt signaling in oocytes of primary and further developed follicles in greater detail. We used a Cre-loxP system to delete the Pten gene from oocytes of primary follicles using transgenic mice carrying zona pellucida 3 (Zp3) promoter-mediated Cre recombinase [7]. We found that elevated PI3K-Akt signaling in oocytes of primary and further developed follicles does not affect the pool of primordial follicles. Moreover, female mice lacking Pten in oocytes of primary follicles showed unaltered follicular development, ovulation, oocyte maturation, and fertility. Our data indicate that PTEN is probably essential for regulation of follicular activation in a stage-specific fashion.

Generation of mice with deletion of Pten in oocytes of primary and further developed follicles
To inactivate the Pten gene from oocytes of primary follicles, we crossed Pten loxP/loxP mice [8] with transgenic mice carrying Zp3 promoter-mediated Cre recombinase (Zp3-Cre mice) [7], where the Zp3 promoter drives the expression of Cre recombinase in oocytes from the primary follicle stage (Fig. 1). To determine the efficiency of deletion of Pten in oocytes of developing follicles, oocytes were collected by puncturing pregnant mare serum gonadotropin (PMSG)-treated ovaries, followed by western blot analysis. The expression of PTEN protein was found to be completely abolished in Pten loxP/loxP ; Zp3-Cre+ oocytes ( Fig. 2A).
We also confirmed the deletion of Pten gene from pups of Pten loxP/loxP ; Zp3-Cre+ females by PCR. Pten loxP/loxP ; Zp3-Cre+ female mice were mated with wild-type male mice and their offspring were genotyped for the deletion of Pten exon 5 in their tail-tip DNA. As shown in Fig. 2B, the pups showed deletion of Pten exon 5 (Pten D5) in one allele of their genomic DNA. These results indicate that the Pten gene was successfully deleted from growing oocytes of Pten loxP/loxP ; Zp3-Cre+ mice.
Elevated PI3K-Akt signaling in Pten loxP/loxP ; Zp3-Cre+ oocytes We tested whether the PI3K-Akt signaling pathway is enhanced in Pten loxP/loxP ; Zp3-Cre+ oocytes. As shown in Fig. 3, phosphorylation of Akt (at Ser473) was found to be elevated in Pten loxP/loxP ; Zp3-Cre+ oocytes as compared to control oocytes. The phosphorylation of another important site on Akt, Thr308, which is required for complete activation of Akt, was also elevated upon deletion of Pten in growing Pten loxP/loxP ; Zp3-Cre+ oocytes. As an indicator of Akt activity, the phosphorylation of Tsc2 (at Thr1462), which is an Akt substrate, was found to be elevated in Pten loxP/loxP ; Zp3-Cre+ oocytes compared to control Pten loxP/loxP oocytes. These results suggest that there is enhanced activation of PI3K-Akt signaling in oocytes of growing follicles upon deletion of Pten in Pten loxP/loxP ; Zp3-Cre+ mice.
Normal follicular development of Pten loxP/loxP ; Zp3-Cre+ mice Our previous work has shown that deletion of Pten from oocytes of primordial follicles leads to activation of the entire pool of primordial follicles, which eventually results in follicle depletion and premature ovarian failure (POF) in young adulthood [6]. To determine the function of Pten in oocytes of primary follicles, we first examined ovarian morphology of Pten loxP/loxP ; Zp3-Cre+ at postnatal day 13 (PD13), PD23, and at 16 weeks of age in comparison to control Pten loxP/loxP ovaries. We found that Pten loxP/loxP ; Zp3-Cre+ ovaries at PD13 appeared normal in morphology as compared to control Pten loxP/loxP ovaries, with primordial, primary, secondary, and some early antral follicles (Fig. 4, A and B). In comparison, in ovaries of mice with deletion of Pten in oocytes of primordial follicles (referred to as Pten loxP/loxP ; GCre+ mice), prematurely activated primordial follicles (transient follicles with enlarged oocytes surrounded by flattened pre-granulosa cells) were observed ( Fig. 4C, red arrows), as reported earlier [6]. These results show that deletion of Pten from primary follicles does not affect the primordial follicles in Pten loxP/loxP ; Zp3-Cre+ mice.
Normal resumption of meiosis in Pten loxP/loxP ; Zp3-Cre+ oocytes It has been suggested that resumption of meiosis in oocytes is regulated by the PI3K-Akt signaling pathway [9][10][11]. Resumption of meiosis in oocytes involves germinal vesicle breakdown (GVBD), chromatin condensation and segregation, and extrusion of the first polar body [12]. To determine whether the enhanced PI3K-Akt signaling has any affect on resumption of meiosis in oocytes, we examined the rates of GVBD in Pten loxP/loxP ; Zp3-Cre+ oocytes. Surprisingly, we found that the rate of GVBD in Ptendeficient oocytes was unaltered as compared to the control oocytes (Fig. 5A). These data indicate that the resumption of meiosis in oocytes was not affected by elevated PI3K-Akt signaling in Pten loxP/loxP ; Zp3-Cre+ oocytes.

Zp3-Cre+ mice
To test whether overactivation of PI3K-Akt signaling in oocytes has any effect on ovulation, Pten loxP/loxP ; Zp3-Cre+ female mice and control Pten loxP/loxP female mice were mated with wild-type males. Two-cell stage zygotes were retrieved from oviducts at E1.5 and were counted. We found that the numbers of 2-cell embryos in Pten loxP/loxP ; Zp3-Cre+ mice were similar to those in control Pten loxP/loxP mice (Fig. 5B), indicating that the rates of ovulation and fertilization are not affected by the loss of Pten in ovulated oocytes. Importantly, during a testing period from 6 to 30 weeks of age, the Pten loxP/loxP ; Zp3-Cre+ mice were completely fertile, with normal litter sizes (Fig. 5C).
The above results show that the elevated PI3K-Akt signaling in oocytes caused by loss of Pten does not affect follicular development, ovulation, and fertility in Pten loxP/loxP ; Zp3-Cre+ mice.

Discussion
In recent years, the PI3K pathway in primary oocytes has been suggested to have important roles in controlling the activation and early development of ovarian follicles [4,[13][14][15]. Activation of the PI3K pathway in oocytes by somatic cell-produced growth factors, such as Kit ligand (KL, or stem cell factor, SCF), results in activation of Akt, which is thought to mediate oocyte growth during follicular activation and early development [15]. Our previous study provided direct functional evidence that the initial recruitment of primordial follicles is tightly regulated by PI3K signaling in oocytes, as oocyte-specific deletion of Pten at the primordial follicle stage leads to global activation of all primordial follicles [6].
In the present study, to investigate how the PTEN/PI3K-Akt signaling in oocytes affects follicular development and oocyte maturation we deleted Pten in oocytes of primary and further developed follicles by using Zp3 promoter-mediated Cre transgenic mice. As expected, we found that the pool of primordial follicles was unaffected throughout the reproductive life of Pten loxP/loxP ; Zp3-Cre+ mice. Moreover, we observed normal follicular development, ovulation, and fertility in Pten loxP/loxP ; Zp3-Cre+ mice, indicating that the PI3K signaling in oocytes is essential for follicular activation in a stage-specific fashion.
Foxo3a, which is a transcription factor and also an Akt substrate, has been shown to suppress follicular activation [16]. Oocytespecific deletion of Foxo3a and Foxo3a-deficient (Foxo3a 2/2 ) mice show similar phenotypes of premature activation of the entire primordial follicle pool [16,17], as we demonstrated in Pten loxP/loxP ; GCre+ mice [6]. Interestingly, Foxo3a was found to be specifically required for the prevention of primordial follicles from activation [18]. Thus, these results support the notion that the PI3K signaling in oocytes is probably more important for follicular activation in a stage-specific way.
It has been suggested that PI3K-Akt signaling plays an important role in the meiotic maturation of oocytes [9][10][11]19]. Inhibition of Akt activation in oocytes by a PI3K-specific inhibitor LY294002 results in suppressed GVBD and polar body extrusion in mouse oocytes [19], and inhibition of Akt phosphorylation leads to inactivation of cyclin-dependent kinase 1 (Cdk1), which in turn causes inhibition of meiosis resumption [11]. Furthermore, activation of phosphodiesterase 3A (PDE3A) by Akt-mediated phosphorylation has been suggested to play a role in the control of PDE3A activity in the maturation of mammalian oocytes [9]. The above studies suggest the importance of PI3K-Akt signaling in the oocyte for resumption of meiosis and maturation of mouse oocytes. At the same time, however, results from the current study have shown that oocytes from Pten loxP/loxP ; Zp3-Cre+ mice undergo normal maturation, ovulation, and fertilization, and the mutant mice were completely fertile with normal litter sizes, indicating that the loss of Pten in oocytes does not apparently affect the fertility of the mice.
Results from our current work with the two mutant mouse models, the Pten loxP/loxP ; GCre+ mice and Pten loxP/loxP ; Zp3-Cre+ mice that carry Pten deficiency in oocytes of primordial and primary follicles, respectively, indicate that PTEN/PI3K-Akt signaling in oocytes is critically important for maintenance of the primordial follicle pool. On the other hand, further detailed and preferably in vivo studies are needed to elucidate the roles of PTEN and PI3K signaling in oocytes during oocyte maturation.

Mice
The Pten loxP/loxP mice [8] and Zp3-Cre transgenic mice where the Cre recombinase is specifically expressed in oocytes of primary and further developed follicles (Fig. 1) [7], were obtained from the Jackson Laboratory (Bar Harbor, MN). Both strains of mice have a C57BL/6J genomic background. After multiple rounds of crossing, mice with Pten loxP/loxP ; Zp3-Cre+ genotype and control female mice with a Pten loxP/loxP genotype were obtained for experiments.
The mice with deletion of Pten in oocytes of primordial follicles (referred to as Pten loxP/loxP ; GCre+ mice) were obtained by crossing Pten loxP/loxP mice with transgenic mice carrying Gdf-9 promotermediated Cre recombinase, which is specifically expressed in oocytes of primordial and further developed follicles (Fig. 1) [20], as previously described [6].
The mice were housed under controlled environmental conditions with free access to water and food. Illumination was on between 0600 and 1800 h. All experiments were conducted with the approval of the regional ethical committee of Umeå University, Sweden.

Reagents, antibodies, and immunological detection methods
The rabbit polyclonal antibodies to Akt, phospho-Akt (Ser473), tuberin/Tsc2, rabbit monoclonal antibody to phospho-tuberin/ Tsc2 (Thr1462), and mouse monoclonal antibody to PTEN were obtained from Cell Signaling Technologies (Beverly, MA). Mouse monoclonal antibody to phospho-Akt (Thr308) was purchased from BD Bioscience (Franklin Lakes, NJ). PMSG and mouse monoclonal antibody to b-actin were purchased from Sigma-Aldrich Sweden AB (Stockholm, Sweden). Western blots were carried out according to the instructions of the suppliers of the different antibodies and were visualized using the ECL Plus Western Blotting Detection System (Amersham Biosciences, Uppsala, Sweden).

Histological analysis
Ovaries were fixed in 4% paraformaldehyde, dehydrated, and embedded in paraffin. The paraffin-embedded ovaries were sectioned at 8-mm thickness and stained with hematoxylin for Signaling studies in Pten loxP/loxP ; Zp3-Cre+ oocytes showed elevated levels of p-Akt (Ser473), p-Akt (Thr308), and p-Tsc2 (Thr1462) as compared to Pten loxP/loxP oocytes. Levels of total Akt, Tsc2, and b-actin were used as internal controls. 100-150 oocytes were used for each lane. All experiments were repeated at least three times and representative results are shown. doi:10.1371/journal.pone.0006186.g003 morphological observation. Ovarian follicles at different developmental stages were categorized based on the well-accepted standards established by Pedersen and Peters [21]. Growing transient follicles were defined as follicles that have obviously enlarged oocytes but which are still enclosed in flattened pregranulosa cells. Morphology images were taken with a light microscope (Leica).

Oocyte collection and culture
To stimulate follicular development, immature 23-day-old female mice were injected intraperitoneally with 5 IU of PMSG.
The ovaries were collected 48 h later and were dissected free of fat and connective tissue. Oocytes were collected by puncturing the ovaries with a sharp needle in M2 medium (Sigma), selected using a mouth pipette, and then cultured further in M16 medium (Sigma) at 37uC in an atmosphere of 5% CO 2 for 6 h to determine the rate of GVBD.

Natural ovulation and embryo culture
Adult female mice were housed with wild-type males, and vaginal plugs were checked every morning. Embryonic day 0.5 (E0.5) refers to the day when a vaginal plug was found. The mated females were sacrificed at E1.5, and 2-cell embryos were recovered from their oviducts and cultured further in KSOM medium (Specialty Media). The numbers and stages of development of the embryos were recorded.

Statistical analysis
All experiments were repeated at least 3 times. For comparisons of oocyte numbers, differences between the two groups were calculated with Student's t-test, and a difference was considered significant if P,0.05.  week-old PMSGprimed Pten loxP/loxP ; Zp3-Cre+ mice or control Pten loxP/loxP mice were collected in M2 medium. They were cultured further in M16 medium for GVBD analysis. GVBD rate was scored after a culture period of 6 h. Normal GVBD rates were observed in Pten mutant oocytes as compared to control oocytes. The numbers of oocytes used (n) are shown. (B) Two-cell embryos were collected at E1.5 and were cultured in KSOM medium supplemented with amino acids. Numbers of 2-cell embryos per female (mean6SEM) mouse were counted. The numbers of mice used (n) and the results of statistical analysis are shown. No significant difference (NS) was seen between Pten loxP/loxP ; Zp3-Cre+ mice and control mice. (C) Normal litter size of Pten loxP/loxP ; Zp3-Cre+ females as compared to Pten loxP/loxP control mice (mean6SEM). The numbers of litters (n) from 10 mice of each genotype and results of statistical analysis are shown. NS, not statistically significant. doi:10.1371/journal.pone.0006186.g005