Impaired ovarian development and reduced fertility in female mice deficient in Skp2


  • Abbas Fotovati,

    1. Department of Molecular and Cellular Biology, Medical Institute of Bioregulation, Kyushu University, Fukuoka, Fukuoka, Japan
    2. Laboratory of Biomedical Research, JOYUP Bio-Medicals, Nishi-Ogi-Minami Suginami-ku, Tokyo, Japan
    3. CREST, Japan Science and Technology Corporation, Kawaguchi, Saitama, Japan
    4. Department of Experimental Medicine and Medical Genetics, University of British Columbia, Child & Family Research Institute, Vancouver, BC, Canada
    Search for more papers by this author
  • Samah Abu-Ali,

    1. Laboratory of Biomedical Research, JOYUP Bio-Medicals, Nishi-Ogi-Minami Suginami-ku, Tokyo, Japan
    Search for more papers by this author
  • Keiko Nakayama,

    1. CREST, Japan Science and Technology Corporation, Kawaguchi, Saitama, Japan
    2. Division of Developmental Genetics, Center for Translational and Advanced Animal Research on Human Diseases, Tohoku University School of Medicine, Sendai, Miyagi, Japan
    Search for more papers by this author
  • Keiichi I. Nakayama

    1. Department of Molecular and Cellular Biology, Medical Institute of Bioregulation, Kyushu University, Fukuoka, Fukuoka, Japan
    2. CREST, Japan Science and Technology Corporation, Kawaguchi, Saitama, Japan
    Search for more papers by this author

Abbas Fotovati, Department of Experimental Medicine and Medical Genetics, University of British Columbia, Children and Women Hospital, Child & Family Research Institute, 950 West 28th Ave., Vancouver, BC V5Z 4H4, Canada. T:+1 778 2330705; F: +1 604 8753120; E:


p27 is a major negative regulator of somatic cellular proliferation, and its down-regulation has been shown to be associated with cancer development. Targeted disruption ofp27 results in complete loss of fertility in female mice, suggesting that it plays a significant role in the development of female gametes and the surrounding environment. We have now investigated the effect of loss of Skp2, an F-box protein that mediates ubiquitin-dependent degradation of p27, on female gamete production. The female Skp2-deficient mice showed accumulation of p27 in the ovary and severely compromised gamete development from the embryonic stage to follicular growth in the adult ovary, eventually leading to a decreased functional gamete reserve. Additional deletion of p27 resulted in relatively normal ovarian folliculogenesis, suggesting that accumulating p27 is primarily responsible for the compromised ovarian development. Embryonic ovaries of Skp2−/− mice manifested massive apoptosis as evidenced by cleavage of pro-caspase 3 and poly(ADP-ribose) polymerase-1. This in turn resulted in a significant decrease in the remaining pool of functional gametes in Skp2−/− mice shortly after sexual maturity and premature ovarian failure. The increased apoptosis seemed to be attributable to the polyploidy of granulosa cells. These results suggest that proper progression of the cell cycle, regulated by the p27-Skp2 axis, is pivotal for the maintenance of fertility, and that defects in this system may underlie the pathogenesis of abnormal gamete production and premature ovarian failure during the reproductive life of women.


Although the notion of endocrino-physiologic events that control development of fertile gametes has largely been established, the molecular mechanisms that underlie such endocrinologic events remain to be completely clarified. Given that gamete development requires proper regulation of the cell cycle in either oocytes or their supporting cells or both, understanding the role of kinases and kinase inhibitors that control the progression of the cell cycle is essential. These regulatory mechanisms appear to be more complicated in gamete cells than those in somatic cells, as they are also required to regulate the additional divisional machinery, i.e. meiosis. The proper chromosomal division during gamete production requires a highly sophisticated regulatory mechanism for producing a fertile gamete with the proper number of chromosomes. Any abnormality of these mechanisms results in depletion of gametes by apoptosis, leading to exhaustion of functional gametes and eventually impaired fertility (for review see Matzuk & Lamb, 2002; McGee & Hsueh, 2000; Wassarman & Albertini, 1994).

Although both male and female gametes originate from migrating primordial germ cells (PGCs) with active mitosis, the fate of each type of gamete, including their further proliferation, is exclusively different (Wassarman, 1994). Therefore, disruption of some genes may result in complete sterility of one sex, whereas the other sex appears completely fertile (Roy & Matzuk, 2006). p27Kip1, a CDK inhibitor acting to restrain the cell cycle, is one such gene: female p27-deficient mice are sterile, but the male mutants are fertile (Fero et al. 1996; Kiyokawa et al. 1996; Nakayama et al. 1996; Rajareddy et al. 2007). Deficiency of p27 is also associated with multiple organ hyperplasia and tumor development in the pituitary gland.

To understand the importance of regulation for p27 abundance, we have studied Skp2, a major regulator of p27 degradation in the cell (Carrano et al. 1999; Sutterluty et al. 1999; Tsvetkov et al. 1999). Skp2 is a member of the F-box proteins that determine the substrate specificity of the SCF-type ubiquitin-protein ligase complex, playing an important role in the progression of the S phase of the cell cycle by contributing to the ubiquitin-dependent degradation of p27. We have previously generated mice lacking Skp2 and have shown that the level of p27 was increased in the somatic cells of these animals (Nakayama et al. 2000). In contrast to mice lacking p27, Skp2 −/− mice showed a severely compromised cellular proliferation characterized by abnormal entry to the endoreduplication process (Nakayama et al. 2000; Minamishima & Nakayama, 2002). Skp2 −/−p27 −/− mice do not exhibit such phenotypes, suggesting that p27 accumulation is responsible for the defects (Nakayama et al. 2004) (Kossatz et al. 2004). We have now shown that Skp2−/− female mice were not completely sterile in contrast to p27 −/− females, but they showed lowered fertility rates ending with a premature cessation of reproductive activity. The results of this study suggest that accumulation of p27 due to Skp2 loss is the major factor contributing to the severe defect in gamete development from the embryonic stage to the mature ovary. This was associated with polyploidy of granulosa cells leading to their severe apoptosis. Increased apoptosis resulted in premature loss of functional gametes and, consequently, infertility of the Skp2-deficient mice.

Materials and methods


p27- and Skp2-deficient mice were generated by homologous recombination in embryonic stem cells (Nakayama et al. (1996, 2000). Mice were back-crossed into the C57BL/6 mouse line. Sexually mature mice from 2 to 12 months of age were used for this study.

Analysis of fertility

For evaluation of fertility, both wild-type and mutant (p27 −/− and Skp2 −/−) female mice were cohabited with known fertile male animals for 8 weeks and then separated. Mice were inspected daily for presence of seminal plugs, and the number and size of litters were recorded.


Immature (20–35 days of age) or adult cycling females were injected subcutaneously (s.c.) with 5 units of pregnant mare serum gonadotropin (PMSG; Calbiochem). After 48 h, mice received 5 units intraperitoneally (i.p.) of human chorionic gonadotropin (hCG; Goldline Labs, Fort Lauderdale, FL, USA) to induce ovulation. After 12–16 h, the ovulated ova were retrieved from the ampulla of the oviduct and counted.

Anatomical and histopathological analyses

For light microscopic analysis, tissue samples were fixed in either Bouin’s fixative or 4% paraformaldehyde, and after serial dehydration and xylene treatments, the samples were embedded in paraffin and sectioned (5 μm). The sections were de-paraffinized, dehydrated, rehydrated, and stained with hematoxylin and eosin according to the manufacturer’s protocol. Slides were analyzed on a Nikon Eclipse E800 microscope using either Nomarski or phase-contrast optics and were photographed using a Hamamatsu 3CCD digital camera-7780.

Non-radioactive in situ hybridization

Ovaries were collected at different stages of cyclic activity following an induced superovulation. After immersion in OCT compound, the samples were frozen in liquid nitrogen, and 5-μm sections were obtained from frozen OCT-embedded samples using cryostat at −20 °C, mounted on amino-propyl silane (APS)-coated slides, and kept at −80 °C until use. The frozen sections were fixed in 4% glutaraldehyde. Digoxigenin-dUTP-labeled riboprobes for mouse p27 were generated according to the manufacturer’s instructions (Roche Diagnostics, Mannheim, Germany). All the procedures of in situ hybridization were done using automated Ventura. Signals were detected by using an anti-digoxigenin-alkaline phosphatase-conjugated antibody (Roche Diagnostics). Staining was carried out overnight using nitroblue tetrazolium chloride (NBT) and 5-bromo-4-chloro-3-indolylphosphate p-toluidine salt (BCIP) (GIBCO/BRL). Control sections incubated with sense RNA showed no specific hybridization.


All sections were treated for antigen retrieval by autoclaving in 0.01 m citric acid solution followed by incubation in a 50% methanol solution with 2% H2O2 to block endogenous peroxidases in the tissue. The sections were subjected to immunostaining with rabbit polyclonal antibodies to (anti-) p27 (C-19; Santa Cruz Biotechnology, Santa Cruz, CA, USA). Proliferating cells were detected with rabbit polyclonal anti-PCNA (Santa Cruz Biotechnology). Goat anti-rabbit (diluted 1 : 500; Sigma) secondary antibodies were applied to the sections in a species-specific manner for 12 h; the sections were then incubated for 3 h in ExtrAvidin (diluted 1 : 1000; Sigma), and 3,3 diaminobenzidine chromagen was used to produce latent staining. Sections were washed in phosphate-buffered saline (PBS) and 0.2% Triton X-100 (3 × 10 min) between each blocking and antibody step. After a brief contra-staining, the sections were mounted onto chrome alum-coated glass slides, dehydrated in ascending grades of alcohol to xylene, and coverslipped using Hystomount (Hughes and Hughes, Somerset, UK) mounting media.


Ovaries of superovulated animals were collected at various times after PMSG administration and later, after hCG administration and were then lysed with P-40 lysis buffer (0.2% Nonidet P-40 NaCl Tris, pH 7.4). The lysate was sonicated briefly before centrifugation at 15 000 g for 15 min at 4 °C. The supernatant was collected, and aliquots of 50 μg of proteins were loaded into each well and separated by SDS/PAGE followed by blotting and visualization of the proteins with anti-p27 (C-19; Santa Cruz Biotechnology) as described previously (Nakayama et al. 1996). Rabbit anti-cleaved-caspase-3 and poly(ADP-ribose) polymerase-1 (PARP) and cyclin D1 antibodies were from Cell Signaling Technology, and anti-cyclin E (M-20), glyceraldehyde 3-phosphate dehydrogenase (GAPDH) (Fl-335) antibodies were from Santa Cruz Biotechnology, Inc. Antibody against Skp2 was from Zymed. image-j software was used for analyzing the signal intensity of cleaved caspase3, PARP and p27.

RNA extraction and real-time reverse transcriptase (RT)-PCR

RNA extraction and real-time RT-PCR was performed as previously described (Fotovati et al. 2006). Total RNA was extracted from ovaries using Isogen-LS reagent (Nippon Gene, Toyama, Japan) and digested with DNase I (Sigma-Aldrich). Total RNA (1 μg) was then reverse transcribed using random hexamer priming and SuperScript II reverse transcriptase (Toyobo, Osaka, Japan). cDNA (100 ng) was amplified in a real-time PCR using SYBR Green Mix (PE Applied Biosystems, Warrington, UK) and 200 nmol L−1 primer for p27. The real-time PCR reactions were done in an ABI PRISM Model 7700 Sequence Detector (Applied Biosystems, Foster City, CA, USA) under the following conditions: 50 °C for 2 min, 95 °C for 1 min followed by 35 cycles at 95 °C for 15 s and 60 °C for 1 min. The sequences of primers were as follows: p27 upstream 5′-TCG CAG AAC TTC GAA GAGG-3′ and downstream 5′-TGA CTC GCT TCT TCC ATA TCC-3′ (Kwon et al. 2003); 28S rRNA upstream 5′-TTG AAA ATC CGG GGG AGA G-3′ and downstream 5′-ACA TTG TTC CAA CAT GCC AG-3′ (Hammoud et al. 2009). The levels of p27 were compared with that of 28S rRNA gene, and the relative expression of p27 in each animal was compared to the maximal level detected in a single experimental setting.

Isolation, culture and staining of granulosa cells

Adult female mice (8 weeks old) were injected intraperitoneally with PMSG and sacrificed 48 h later. All animal studies complied with the Kyushu University Animal Experimentation Act. The ovaries were excised and cleaned of surrounding connective tissue, and granulosa cells were isolated from the ovaries as previously described (Campbell, 1979). In brief, the ovaries were suspended in Dulbecco’s modified Eagle’s medium (DMEM)-F12 supplemented with 0.2% bovine serum albumin and 10 mm HEPES buffer solution (Invitrogen Life Technologies, Gaithersburg, MD, USA) and were incubated under 5% CO2 at 37 °C first with 6.8 mm ethylene glycol tetraacetic acid (EGTA) for 15 min and then with 0.5 m sucrose for 5 min. The granulosa cells were then released from the tissue using a 25-gauge needle and cultured under 5% CO2 at 37 °C in DMEM-F12 supplemented with 10% fetal bovine serum, penicillin (100 IU mL−1), and streptomycin (100 μg mL−1). Nuclei were stained by incubating the cultured cells for 1 min with a 1 : 1000 dilution of Hoechst33342, and the cells were then examined with a fluorescence microscope (Nikon Eclipse E800 microscope) and photographed using a Hamamatsu 3CCD digital camera-7780.

Flow cytometric analysis of DNA content

Isolated granulosa cells (1 × 106) were washed in PBS, re-suspended in 200 μL of PBS, and fixed by the gradual addition of 800 μL of 100% ice-cold ethanol. The fixed cells were washed twice with PBS, re-suspended in 1 mL of PBS, treated with RNase (40 μg mL−1) for 10 min at room temperature, and then stained with propidium iodide (25 μg mL−1). The DNA content of the cells was determined by flow cytometry with a FACSCaliber instrument and cellquest software (Becton Dickinson, San Jose, CA, USA).


Expression of p27 fluctuates during follicular development

To investigate the abnormalities associated with an increased or decreased level of p27, we first examined the abundance of p27 protein in normal ovaries. Both in situ hybridization and immunohistochemistry revealed that p27 is mainly expressed in granulosa cells of the pre-ovulatory ovarian follicles at the levels of mRNA (Fig. 1A) and protein (Fig. 1B–D), both of which seem to fluctuate during various steps of ovarian cycle. Data from in situ hybridization were closely related to those of previous studies (Burns et al. 2001). During the starting steps of folliculogenesis, the granulosa cells of primary follicles lack the p27 expression. However, with progression of the follicular growth, the expression of p27 gradually increased and accumulated. Furthermore, p27 was highly expressed in corpora lutea. Centrally located ova also express a considerable amount of p27.

Figure 1.

 Expression of p27 in normal ovary. The expression of both mRNA (A) and protein (B–D) of p27 was evaluated using in situ hybridization and immunohistochemistry, respectively. The p27 mRNA expressing follicles are shown. Corpora lutea with high level of p27 are circled by dotted lines. Scale bars: 1 mm (A,B,D); 100 μm (C).

Anatomical characteristics of ovary

Macroscopic observation revealed that the size of the ovary from Skp2−/− mice was markedly smaller than that of wild-type mice (Fig. 2A). This hypoplastic change was restored by additional deletion of p27 from Skp2−/− mice (Skp2−/−p27 −/− mice). Furthermore, on the basis of the gross examination, most ovaries collected from Skp2−/− mice contained slightly fewer mature follicles than those from wild-type animals.

Figure 2.

 Gross and microscopic structure of wild-type and mutant ovaries. (A) Macroscopic comparison of the ovaries of wild-type (WT), Skp2−/−, and Skp2−/−p27−/− mice at 2 months of age. Scale bar: 1 mm. (B–D) Representative light micrographs of ovarian sections of wild-type (B), Skp2−/− (C), and Skp2−/−p27−/− (D) female mice at 2 months of age. CL, corpus luteum. Scale bars: 1 mm.

For understanding the nature of gross ovarian abnormalities in Skp2 −/− mice, we inspected the microscopic structure of ovaries (Fig. 2B–D). Wild-type ovaries contained various cellular compartments including ovarian follicles with different degrees of maturation and corpora lutea (Fig. 2B), whereas the production of such growing follicles appeared to be seriously compromised in Skp2−/− ovaries (Fig. 2C). In addition, the follicular production showed a significant reduction with age (see below). As observed in ovarian size, the reduction of follicular number in Skp2 −/− ovaries was recovered in Skp2−/−p27 −/− mice (Fig. 2D). However, there was no mature corpora lutea in Skp2−/−p27 −/− mice as seen in p27−/− mice, consistent with the notion that p27 is a downstream effector of Skp2.

Expression of p27 during estrous cycle in wild-type and Skp2-deficient mice

To understand the relationship between p27 and Skp2 in various genotypes used in this study, we first evaluated the level of both proteins in ovaries from embryonic day E18.5 of all genotypes (Fig. 3A). To address the main question of this study, the status of p27 and related proteins was further evaluated in E18.5 Skp2−/− mice (Fig. 3B). Immunoblot analysis revealed a considerable accumulation of p27, cyclin D1 and cyclin E in Skp2-deficient ovaries (Fig. 3B). In contrast, there was no significant difference in the level of mRNA of p27 (Fig. 3C). We also examined the expression of p27 in response to PMSG injection that induces folliculogenesis in adult ovaries. For this purpose, ovaries of 8-week-old mice treated with PMSG were collected at 6, 12, 24 and 48 h time points. mRNA and protein level of p27 were then evaluated by real-time RT-PCR and immunoblotting, respectively (Fig. 3D,E). In both wild-type and Skp2−/− mice, mRNA expression of p27 followed a similar pattern, with the maximal level occurring at 48 h (Fig. 3D). However, changes in p27 protein levels showed a different pattern between Skp2−/− and Skp2+/+. In wild-type ovaries, p27 protein expression seemed to be low in the early follicular stage and accumulated toward the follicular-luteal interphase and later (Fig. 3E). In Skp2−/− mice, however, the initial abundance of p27 at the follicular phase was greater than that of wild-type controls, although p27 expression at the steady-state seems to be comparable between wild-type and Skp2−/− mice.

Figure 3.

 The level of p27 at various stages of ovarian cycle in wild-type and mutant mice. (A) Skp2 and p27 protein levels in various genotypes. GAPDH was used as internal control. (B) The levels of p27, Skp2 and cyclin D1 and cyclin E were monitored by immunoblotting with the indicated antibodies. The extracts from wild-type (Skp2+/+) and Skp2-deficient (Skp2−/−) ovaries in embryos at E18.5 were subjected to immunoblot analysis. (C) Real-time RT-PCR did not show any significant difference in p27 mRNA expression in Skp2+/+ vs. Skp2−/−. (D,E) The levels of p27 mRNA and protein during various stages of ovarian cycle were also examined at the indicated time points after PMSG injection by real-time RT-PCR and Western blotting. Ovarian extracts were prepared from 8-week-old wild-type (WT), Skp2−/−, and p27−/− mice. (D) p27 mRNA levels increased significantly after 48 h of PMSG injection in both Skp2+/+ and Skp2−/− (*P < 0.05). (E) Immunoblotting showed an increased level of p27 after 24–48 h in both Skp2+/+ and Skp2−/−. However, higher levels of p27 protein after 6 and 12 h of PMSG was only seen in Skp2−/−. Ovaries from P27/ were used as negative control. GAPDH was used as internal control.

Premature ovarian failure in Skp2-deficient mice

We thus compared the reproductive abilities of the wild-type and genetically manipulated animals. As has been described previously (Fero et al. 1996; Kiyokawa et al. 1996; Nakayama et al. 1996), the lack of p27 completely compromised female fertility. To determine the fertility status of mice, 8-week-old female littermates from Skp2−/− and Skp2+/+ (n = 8) were cohabited with a known fertile wild-type male for 8 weeks. The number of pregnant mice and the litter size were recorded. Although Skp2−/− female mice were not completely sterile, both pregnancy rate and litter size were significantly reduced compared with wild-type controls (Table 1). Additional disruption of p27 in female Skp2−/− mice did not improve fertility, suggesting that Skp2 is an upstream regulator of p27.

Table 1.   Fertility rate of female mice of each genotype.
GenotypePregnancy rate (n = 8)
Total (pregnant)
Litter size (pups/litter**)
  1. Both pregnancy rate and litter size were determined for 8 weeks.

  2. *Mice with litter size as low as one pup per litter are also included.

  3. **Litter size was determined using the same eight mice used for estimation of pregnancy rate.

Wild-type8 (8)7.25 ± 1.03
Skp2 −/− 8 (6*)1.75 ± 1.38
p27 −/− 8 (0)0
Skp2 −/− p27 −/− 8 (0)0

The low fertility in female Skp2−/− mice declined with age, eventually resulting in early cessation of reproductive activity of female animals (Fig. 4A). We did not find any abnormality in uteri or oviducts of aged Skp2−/− mice (data not shown). Furthermore, the number of ova recovered from superovulated mice was reduced gradually with age (Fig. 4B). Histological examination of ovaries also revealed that wild-type mice at 6 months of age showed normal structures of cycling ovaries (Fig. 4C), whereas the Skp2−/− ovaries showed a progressive disappearance of the normal ovarian follicular structure (Fig. 4D) and replacement of fibrocyte-like cells (Fig. 4D, red arrow in enlarged selected area). Ultrastructural analysis of ovaries also confirmed the presence of many fibrocyte-like cells in Skp2−/− ovaries, compared with normal ova-containing wild-type ovaries (Fig. 4E,F). Given the characteristics of both ova production and ovarian pathology, we concluded that Skp2−/− ovaries suffer from premature ovarian failure.

Figure 4.

 Premature ovarian failure in Skp2-deficient mice. (A) The fertility of female Skp2+/+ (open diamonds), Skp2+/− (closed squares), and Skp2−/− (closed circles) mice is shown at the indicated ages. Data are expressed as the percentage of fertile females in seven or eight animals per genotype. (B) The ova recovered from superovulated mice were counted, and the percentage was calculated according to maximal ova production. *Skp2−/− vs. Skp2+/+ at same age (< 0.05). (C,D) Light microscopic analysis of ovaries isolated from a 6-month-old wild-type (C) and Skp2−/− (D) littermate after induction of ovulation. The Skp2−/− ovary was depleted of oocyte and filled with stromal connective tissue (asterisks). Fibrocyte-like cells are shown (red arrow). (E,F) Ultrastructural analysis of the ovaries of 6-month-old wild-type (E) and Skp2−/− (F) mice. The ovary of the mutant animal shows a predominance of fibroblasts relative to gametes (open arrowheads). Scale bars: 1 mm (C,D) 5 μm (E), and 10 μm (F).

The effect of p27 balance on ovarian follicular proliferation

As p27 plays an important role in the control of the cell cycle, we next questioned whether the deficiency or excess of p27 affects cellular proliferation of gonads. To this end, the mitotic activities of the cellular structures of gonads in both models were evaluated by immunostaining for PCNA.

Adult Skp2−/− ovaries contained significantly smaller numbers of follicles with PCNA-positive cells (Fig. 5B) compared with wild-type ovaries (Fig. 5A). In contrast, p27−/− mice showed a high proliferation rate in an increased number of primary follicles (Fig. 5C), which was significantly higher than the follicular pool normally recruited for folliculogenesis during each ovarian cycle in wild-type animals (Fig. 5A). In Skp2−/−p27−/− mice (Fig. 5D), the number of PCNA-positive follicles was comparable to that of wild-type animals. The number of PCNA-positive follicles in wild-type and mutant mice was counted and statistically analyzed (Fig. 5E). There was a significantly lower number of PCNA-positive cells in Skp2−/− mice compared with all other genotypes. In contrast, p27−/− ovaries contained significantly greater numbers of PCNA-positive follicles than those in all other genotypes.

Figure 5.

 Proliferation status of cells in wild-type and mutant ovaries. (A–D) Immunohistochemical analysis of adult ovaries from wild-type (A), Skp2−/− (B), p27 −/− (C), and Skp2−/−p27 −/− (D) mice at 2 months of age. Scale bars: 1 mm. (E) The number of proliferating follicles was counted (= 3 in each genotype). *Skp2−/− vs. WT, p27−/− and Skp2−/−p27 −/− (< 0.05). **p27 −/− vs. WT, Skp2−/− and Skp2−/−p27 −/− (< 0.05).

Increased apoptosis in Skp2-deficient ovary

As apoptosis is the major corrective process removing the follicles that are not selected for ovulation, we examined whether apoptosis was responsible for loss of ovarian follicles. TUNEL staining revealed an increased frequency of apoptosis in Skp2−/− ovaries in various stages of folliculogenesis compared with that of physiological apoptosis occurring in wild-type ovaries (Fig. 6A–C). The increased apoptosis seemed to be responsible for the significant loss of follicular resources.

Figure 6.

 Increased apoptosis associated with polyploidy in granulosa cells. (A–C) Increased level of apoptosis in the gonads of Skp2−/− mice. TUNEL staining of ovarian sections of Skp2+/+ (A) or Skp2−/− (B) mice at 4 months of age. The number of TUNEL-positive cells was increased in growing follicles (arrowheads) of Skp2−/− mice, leading to their early atresia, whereas only physiological apoptosis was detected in normal atretic follicles. (C) The number of apoptotic follicles was quantified. Synchronized ovaries from Skp2−/− and Skp2+/+ showed an activated apoptotic machinery in Skp2−/−. *Skp2−/− vs. wild-type (WT) < 0.05. (D) Increased levels of apoptotic markers, cleaved caspase-3, and PARP indicated comparatively higher levels of apoptosis in ovaries from Skp2−/− mice. (E–H) Endoreplication in ovarian follicular granulosa cells in Skp2−/− mice. facs analysis of granulosa cells isolated from ovarian follicles of Skp2+/+ (E) and Skp2−/− (G) showed considerable polyploidy in Skp2−/− mice. Nuclear staining of these cells with DAPI also demonstrated morphological evidences of nuclear polyploidy in Skp2−/− (H) vs. Skp2+/+ (F). Scale bars: 100 μm (A,B,F,H).

The absence of p27 has been associated with suppression of atresia through inhibition of apoptosis in embryonic ovaries (Rajareddy et al. 2007). For evaluation of the apoptotic process in Skp2−/− mice, the wild-type and Skp2−/− embryonic ovaries at E18.5 were subjected to immunoblot analysis for caspase-3 and PARP. Although a substantial amount of cleaved caspase-3 and PARP was detected in wild-type ovaries, reflecting extensive apoptosis physiologically occurring in many embryonic follicles at this stage, the extent of cleaved caspase-3 and PARP in Skp2−/− ovaries is modestly but reproducibly greater than that in wild-type ovaries (Fig. 6D).

Isolated granulosa cells were subjected to analysis of DNA content. Granulosa cells from wild-type mice mostly contained diploid (2n) cells (Fig. 6E). However, Skp2−/− cells manifested a significant polyploidy (Fig. 6G), i.e. the proportion of diploid cells was reduced, and the population of tetraploid (4n) and cells with more than 4n was increased. Examination of the nuclear morphology of granulosa cells by 4’,6-diamidino-2-phenylindole (DAPI) staining revealed that the cell population harvested from ovaries of Skp2−/− mice contained a large number of the cells with an abnormal nuclear size and shape (Fig. 6F,H). Collectively, the loss of Skp2 results in deregulation of p27 level, which may in turn give rise to abnormal ploidy and increased apoptosis in the ovaries. Thus, the p27-Skp2 axis substantially contributes to normal development of ovary and maintenance of fertility.


In this paper, we studied the significance of proper proteolysis of p27 mediated by SCFSkp2 ubiquitin-protein ligase in the development of the female gametes utilizing Skp2−/− female mice. These results showed that lack of Skp2 was associated with accumulation of p27, which severely affected female functional gamete development from the embryonic stage to the reproductive age. Considering the critical role of p27 in the progression of the cell cycle, it was expected that the proper balance of p27 synthesis and turnover is vital in gonads in which cell cycle progression is believed to be strictly controlled. Therefore, as a general feature of cell cycle regulators, not only the deficiency of p27 but also its excessive accumulation is expected to impair the progression of the cell cycle, production of functional gametes, and consequently, fertility.

Female gametes, before being fully matured and ready for fertilization, undergo several developmental steps, including a long dormant stage. The molecular events of these steps, which are associated with a considerable loss of gametes, are not fully understood. Additionally, in contrast to males, production of female gametes after sexual maturity in most mammalians is not continuous and is synchronized by so-called ovarian cycles. Adjusting such a process within a highly active cell cycle is dependent on the cell-specific expression of several cyclins and their cyclin-dependent kinase (Cdk) activator as well as Cdk inhibitor such as p27 (Bayrak & Oktay, 2003).

Ovarian development was severely impaired in mice lacking p27 (Kiyokawa et al. 1996; Nakayama et al. 1996). Others have also shown that decreased expression of p27 results in hyperplasia of the ovaries and increased frequency of gonadal tumors in mutant models lacking inhibin α (Inha−/−) (Cipriano et al. 2001). The absence of p27 resulted in uncontrolled cellular proliferation of granulosa cells in ovarian follicles, leading to excessive follicular proliferation. In contrast, development of functional gametes was substantially compromised from a very early stage of embryonic ovarian development in Skp2−/− mice. This is attributed to p27 accumulation in Skp2−/− mice. The inhibitory effect of accumulated p27 on female gametes is observed not only in the embryonic stage but also in the adult stage, given that follicular proliferation is significantly compromised in adult cyclic animals, which limits the size of follicular growth waves and finally results in reduced production of functional gametes. Furthermore, a large subset of these follicles do not complete their development and degenerate prematurely. Oocyte-depleted areas of ovaries are eventually replaced with connective tissues, which resembles so-called premature ovarian failure (POF) in humans. Human POF with hypergonadotrophic hypogonadism occurs in 1% of women (Goswami & Conway, 2005). The syndrome is a very heterogeneous clinical disorder probably due to the complex genetic networks controlling human folliculogenesis. The mechanisms of this syndrome are under investigation (Fassnacht et al. 2006). There are also studies regarding possible involvement of cell cycle regulators in reducing functional gamete reserves and development of POF (Vital-Reyes et al. 2006). Therefore, abnormalities of major controlling mechanisms of cell cycle progression such as Skp2 and especially p27 in highly proliferative ovarian cells might also be involved in POF.

In mammals, gametes develop in a microenvironment of supporting stromal cells of somatic origin that interact with the former through autocrine and paracrine mechanisms as well as direct cell-to-cell interactions (Matzuk et al. 2002). In Skp2−/− mice, an increased frequency of apoptosis in ovarian follicles is apparent, further limiting the number of growing follicles that reach the final stage of ovulation. In fact, an active apoptotic process was found in the embryonic stage of ovaries as evidenced by increased cleavage of caspase 3 and PARP.

Ovarian granulosa cell-specific ablation of retinoblastoma (Rb) gene in inhibin-α deficient mice was recently shown to result in an increase in Skp2 expression, possibly leading to decreased p27 protein (but not mRNA) level in these mice (Andreu-Vieyra et al. 2008). This was associated with increased apoptosis in ovarian tumors observed in these mice. In contrast, our results showed that Skp2 deficiency results in accumulation of p27 protein, which is associated with aberrant apoptosis, at least in embryonic gonads. Interestingly, depletion of Skp2 by RNA interference in a cancer model also resulted in increased apoptosis (Jiang et al. 2005). The high prevalence of apoptosis in ovarian granulosa cells was accompanied by an increased frequency of polyploidy in the granulosa cells of Skp2-deficient animals. Aneuploidy may also develop in the cellular contents of ovarian follicles of Skp2−/− mice by a mechanism similar to that operative in other somatic cells of these animals, that is, endoreplication caused by the accumulation of p27 (Minamishima & Nakayama, 2002; Nakayama et al. 2004). Indeed, impaired degradation of p27 during G2 phase in Skp2−/− cells may result in suppression of Cdc2 activity and consequent inhibition of entry into M phase (Nakayama et al. 2004). Given the important role of granulosa cells in maintaining the female gametes (Matzuk & Lamb, 2002), follicles with apoptotic granulosa cells likely fail to progress to later stages of development and eventually undergo follicular atresia. Apoptosis seemed to be even more extensive in growing follicles, affecting the primordial follicular reserve of ovaries and leading to follicle-depleted ovaries.

In conclusion, these results suggest the importance of p27 in coordinated proliferation and differentiation of ovarian cellular compartments and the molecular pathways involved in synchronization between hormonal and cellular events leading to female functional gamete production.


The authors would like to acknowledge Ms Anna H. Reeves for thoughtful review and commentary regarding this manuscript. This research was supported by CREST, Japan Science and Technology Corporation (A.F., K.I.N. and K.N.), Japan Society for the Promotion of Science (JSPS) (A.F., K.I.N.), and funds provided by JoyUp Biomedicals, Fukuoka, Japan (A.F., S.A.-A.).

Author contributions

A.F. performed major animal and pathological analysis; S.A.-A. completed Western blotting of Fig. 3 as well as real-time RT-PCR; K.N. developed Skp2 and p27 KO mice; K.I.N. wrote most of the manuscript and supervised the project.