Chengyu Liat, Zhaojun Liuat, Weijian Lia, Liangliang Zhanga, Jilong Zhoub, Minghong Suna, Jiaqi Zhoua, Wang Yaoa, Xuan Zhanga, Honghui Wanga, Jingli Taoa, Ming Shena,*, Honglin Liua,*

Abstract
Recent evidence has indicated that follicular vascularization is critical to ovarian follicle development and survival. Follicle-stimulating hormone (FSH),a gonadotropin that induces follicular growth and development, also acts as the major survival factor for antral follicles. FSH has been reported to stimulate angiogenesis in the theca layers mediated in part by the vascular endothelial growth factor A (VEGFA) and the transcription factor hypoxia inducible factor 1α (HIF-1α). However,it remains largely undetermined whether FSH-dependent growth and survival of antral follicles relies on FSH-induced vascularization. Here, we first demonstrated that induction ofangiogenesis through the FSH-HIF-1α-VEGFA axis is not required Institute of Medicine for FSH-stimulated follicular growth in mouse ovary. FSH increased the total number of bloodvessels in mouse ovarian follicles, which was correlated with elevated expression of VEGFA and HIF-1α in granulosa cells (GCs). In contrast, blocking of follicular angiogenesis using inhibitors against the HIF-1α-VEGFA pathway repressed vasculature formation in follicles despite FSH administration. Interestingly, by measuring follicular size and ovarian weight, we found that the suppression ofangiogenesis via HIF-1α-VEGFA pathway did not influence FSH-mediated follicular growth. However, inhibition of FSH- induced follicular vascularization by PX-478,a small-molecule inhibitor that suppresses HIF- 1α activity, blocked ovulation and triggered atresiain large follicles. On the other hand, PX- 478 injection reducedoocyte quality via impairing the meiotic apparatus, showing a prominently defective spindle assembly and actin dynamics. Collectively, our findings unveiled avascularization-independent effect of FSH on follicular growth; whereas follicular survival, ovulation, and oocyte development relies on FSH-mediated angiogenesis in the follicles.

Keywords: FSH / vascularization / follicular growth / atresia / ovulation / oocyte quality

1.Introduction
Angiogenesis, the process of new bloodvessels formation from pre-existing vascular beds, is crucial for ovarian follicle maintenance and growth in female reproduction (1,2). Aberrant vascularization within the ovary was reported to cause infertility, anovulation, or polycystic ovary syndrome (PCOS) (3). In mammals, it has been demonstrated that angiogenic factors secreted by granulosa cells (GCs) play essential roles in follicular development and survival via inducing angiogenesis in the theca layers (4,5). Particularly, vascular endothelial growth factors (VEGF) represent potent and specific stimulators that promote vascular endothelial cell proliferation plus angiogenesis (6). VEGF also contributes to mammalian follicular growth and corpus luteum formation (2,7). The VEGF family consists of VEGFA, VEGFB, VEGFC, VEGFD, VEGFE, and VEGFF, of which VEGFA (hereafter referred to as VEGF) is the most prominent proangiogenic factor in the ovary (8). Following puberty, cyclical increases in serum gonadotropins stimulates follicular development, rupture, and release of maturedoocytes. During each menstrual cycle, FSH provides the key stimulus that promotes ovarian follicle growth to the preovulatory stage (9). Recent studies revealed that FSH also inducesangiogenesis in the follicles (10). As reported, elevated mRNA expression of VEGF was detected in GCs within secondary and tertiary follicles, whose development is becoming dependent on FSH stimulation (11).

Further investigations elucidated that FSH indeed facilitatesangiogenesis in the peripheral follicle thecal cells mediated in part by the VEGF (10,12).Growing evidence supports hypoxia as a proangiogenic stimulus and its relationship with VEGF production has been documented (13). Previous studies indicate a progressive deficient supply of oxygen during follicular growth, which was accompanied by an increase of VEGF production by GCs in developing follicles (14). In fact, hypoxia inducible factor-1α (HIF-1α), the key mediator in hypoxic response, has been identified as a major transcriptional regulator of VEGF in ovarian follicles (15). Under normoxia conditions, HIF-1α undergoing posttranslational modifications is rapidly degraded through the ubiquitin– proteasome pathway (16,17). During hypoxic stimulation, the stabilization of HIF-1α initiates the transcriptional activation of VEGF production (14). In cultured GCs, it has been shown that FSH enhances HIF-1α activity through the PI3-kinase/AKT-dependent activation of mTOR and that HIF-1α activity is required for the upregulation of FSH target genes inculding VEGF (18). Moreover, the mutation of hypoxia response element in the Vegfa promoter blocked FSH-regulated Vegfa expression in mouse follicles (7). Together, these studies indicate that FSH acts through HIF-1 to stimulate VEGF expression, and this is presumed to promote ovarian follicle angiogenesis. Nevertheless, it remains unclear whether the growth and development of ovarian follicles is dependent on the proangiogenic role of FSH in this process. Herein, we aimed to daily new confirmed cases further define the role and mechanism of HIF-1α-VEGF axis in the regulation ofangiogenesis by FSH in vivo. In addition, we determined whether FSH- mediated follicular development, survival, and ovulation are relevant to its proangiogenic actions through this signaling cascade. Furthermore, we sought to study the physiological effects of FSH-induced vascularization on oocyte development.

2. Material and Methods
2.1 Reagents and antibodies
PX-478 (an inhibitor of HIF-α activity; S7612) was purchased from Selleck Chemicals (Houston, TX, USA). TUBA1A (2125) (19) antibody was obtained from Cell Signaling Technology (Beverly, MA, USA). Antibodies against CD34 (ab81289) (20), CD31 (ab28364) (21), and HIF-1α (ab179483) (22) were purchased from Abcam (Cambridge, MA, USA). caspase-3 antibody (19677-1-AP) (23) was bought from proteintech (Chicago, IL, USA). VEGF antibody (orb11554) (24) was purchased from biorbyt (San Francisco, CA, USA). Anti-α- tubulin-FITC (F2168) (25) and anti-phalloidin-TRITC (P1951) (26) antibodies were obtained from Sigma-Aldrich (St. Louis, MO, USA). The detailed information of the antibodies could also be found in Table 1. FSH and hCG was purchased from Ningbo Second Hormone Factory (Ningbo, Zhejiang, China).

2.2 Study design and sample collection
Three-to-four-week-old female ICR mice (Qing Long Shan Co., Animal Breeding Center, Nanjing, China) were group-housed in a temperature-controlled (22±2°C) room with a 12:12 h light:dark cycle (lights on from 7:00 a.m. to 7:00 p.m.) and hadad libitum access to water and food. Mice were randomly assigned to Control, PX-478-injected, FSH-injected, and FSH+PX-478-injected groups. 18 mice were included in each group. To induce follicular growth, mice in FSH-injected group were injected intraperitoneally (i.p.) with FSH at 7:00 a.m. and 7:00 p.m. for 2 days at a dose of 10 IU on day 1 and 5 IU on day 2. Concomitantly, the control mice were injected with 0.9% saline. In PX-478 and FSH+PX-478-injected groups,mice were i.p. injected with PX-478 (0.2 mg/mouse) (27) at 7:00 a.m., 3:00 p.m., and 11:00 p.m. for 2 days. 12 mice from each of these groups were killed to obtain the ovaries. The left ovaries were weighed, fixed with 4% paraformaldehyde and embedded in paraffin for subsequent histological examination, including haematoxylin and eosin (H&E) staining, immunofluorescent staining, and immunohistological staining. GCs were isolated from the right ovaries for qRT-PCR and western blotting analysis. Next, the remaining 6 mice in each group were injected i.p. with 10 IU human chorionic gonadotropin (hCG) to induce the ovulatory process. At 10 h after hCG injection, 3 mice in each group were sacrificed to obtain oocytes for immunofluorescent staining. At 15 h after hCG injection, the remaining 3 mice in each group were killed. The left ovaries were used for H&E staining. Oocytes were collected from the oviducts and right ovaries for immunofluorescent staining or counting the number of ovulated eggs. All the animal experiments were conducted in accordance with the guidelines of the Animal Ethical Committee at Nanjing Agricultural University.

2.3 H&E staining
Haematoxylin and eosin (H&E) staining was performed as previously described (28). Briefly, at 48 h after FSH injection and 15 h after hCG injection, ovaries collected from mice were fixed with 4% paraformaldehyde, paraffin-embedded, serially sectioned to 5 μmthickness, and then mounted on glass slides. After deparaffinization and rehydration, sections were stained with H&E (Nanjing Jianchen Institute of Biological Engineering, Nanjing, China) and representative images were taken under a dot slide-digital virtual microscopy (Olympus, Tokyo, Japan). To calculate the number of different sized follicles, the images of ovarian sections were imported into the CaseViewer software for measuring follicular sizes and counting follicular numbers.

2.4 Immunohistochemistry and immunofluorescent staining
Mice ovaries (n=4 in each treatment group) collected as mentioned above were embedded in paraffin, serially sectioned to approximately 5 μm, and mounted on glass slides. The sections next to the middle of selleck compound the ovaries (with maximum diameter) were selected for histological analysis, while one section of each ovary was used as the representative images for the immunofluorescence and TUNEL assay. Briefly, the ovarian sections were deparafinized in xylene, rehydrated, and retrieved by microwave heating with buffer of citrate (10 mM Sodium citrate, 0.05% Tween-20, pH6.0) for 0.5 h. Endogenous peroxidase activity was eliminated by exposure with 3% H2O2 (Sigma-Aldrich, 216763-100 ml) for 10 min. After 1 h of blocking with 1% bovine serum albumin (BSA), sections were incubated with antibodies against VEGF (1:500), HIF-1α (1:50), CD34 (1:3000), CD31 (1:50) orcaspase-3 (1:100), and corresponding secondary antibodies. For immunohistochemistry, the immunoreactive signals were visualized using the 3, 3′-diaminobenzidinechromogen solution (Sigma-Aldrich, D8001). The nuclei were counterstained in hematoxylin (Sigma- Aldrich, H9627) prior to dehydration and coverslip placement. For immunofluorescent staining, the ovarian sections were mounted with VECTASHIELD Mounting Medium plus DAPI and examined under a laser-scanning confocal microscope (Carl Zeiss, Oberkochen, Germany).

2.5 qRT-PCR assay
Ovaries collected from mice after 2 days of FSH injection as mentioned above were individually transferred into 35-mm Petri dishes containing PBS and then punctured with a syringe to release GCs from ovarian follicles under a surgical dissecting microscope (Olympus, Tokyo, Japan). After centrifugation at 2000 g for 5 min, GCs were rinsed with PBS, lysed using TRIzol reagent (Invitrogen), and the total RNA was reverse transcribed into cDNA using the PrimeScriptTM RT Master Mix (Takara) following the manufacturer’sinstructions. qRT-PCR was performed using AceQ® qPCR SYBR® Green Master Mix (Vazyme, Nanjing, China) in the ABI QuantStudio5 system (Applied Biosystems,Foster City, CA, USA). Each reaction was performed with a total volume of 20 μl, consisting of 10 µL of 2 × AceQqPCR SYBR Green Master Mix, 0.4 µLof each 5′- and 3′-primer (10 µM), 0.4 µL of 50 × ROX Reference Dye 1, and 8.8 µL of sample cDNA. The primers were designed using the online Primer–BLAST software from NCBI, and the primer sequences were shown in Table 2. The qRT-PCR conditions were 1 cycle at 95℃ for 5 min, followed by 40 cycles of 10 sec at 95℃, and 30 sec at 60℃. The melting program was performed at 95℃ for 15 sec, 60℃ for 60 sec, and 95℃ for 15 sec.

2.6 TUNEL assay
The TUNEL staining of ovarian section was performed using the In Situ Cell Death Detection Kit (Roche Applied Science, Mannheim, Germany) following the procedures as described previously (29). After TUNEL reactions, the ovarian sections were counterstained with DAPI and mounted with VECTASHIELD on glass slides. Fluorescent images were obtained using a laser-scanning confocal microscope (Carl Zeiss, Oberkochen, Germany). The optical density was evaluated in each GC with ImageJ 1.42q software (National Institutes of Health, Bethesda, USA).

2.7 Confocal imaging of meiotic apparatus
Denudedoocytes were fixed with 4% paraformaldehyde in PBS for 1 h, rinsed three times in PBS, rehydrated, and then transferred to the permeabilization solution (1%Triton X-100, 20 mM HEPES, pH7.4, 3 mM MgCl2, 50 mM NaCl, 300 mM sucrose, 0.02% NaN3 in PBS) for 8-12 h. After 1 h of blocking with 3% BSA at room temperature, oocytes were incubated with antibodies against α-tubulin-FITC (1:200) orphalloidin-TRITC antibody (1:200) at 4°C overnight, followed by incubation with an corresponding secondary antibody for 1 h. Oocytes were then mounted with VECTASHIELD Mounting Medium plus DAPI and observed under alaser-scanning confocal microscope (Carl Zeiss, Oberkochen, Germany).

2.8 Western blotting
Total proteins extracted from mouse GCs using radioimmune precipitation assay buffer (RIPA; Beyotime Institute of Biotechnology, Shanghai, China) were separated by electrophoresis on a 4-20 % Sure PAGE gel (Genscript, Nanjing, China) and transferred to PVDF membranes (Millipore, Bedford, MA) by electroblotting. After blocking with TBST (Solarbio) containing 5% bovine serum albumin for 1 h, the membranes were treated with primary antibodies against VEGF (1:200), HIF-1α (1:1000), and caspase-3 (1:800) in blocking solution at 4℃ overnight. The membranes were then washed in TBST three times, and incubated with an appropriate secondary antibody (1:2000) at room temperature for 2 h. The immunoreactive signals were detected using WesternBright ECL HRP substrate kit(Advansta) according to the manufacturer’s directions. The relative expression of target proteins was normalized to TUBA1A, which was served as the control for loading.

2.9 Statistical analysis
The experiments were repeated at least three times, as detailed in the figure legends.Experimental data were presented as mean ± S.E. Statistical differences between two groups were evaluated by non-parametric Mann-Whitney test (GraphPad Prism 5), and between multiple groups using non-parametric Kruskal-Wallis test (GraphPad Prism 5). P values < 0.05 were considered to be statistically significant.

3.Results
3.1 FSH inducesangiogenesis in mouse ovarian follicles, and escalates the expression of
VEGF and HIF-1α in follicular granulosa cells (GCs) To examine whether FSH stimulatesangiogenesis in follicles, we collected ovaries from mice subjected to FSH injection. Using hematoxylin and eosin (H&E) staining, we observed that mice injected with FSH displayed more bloodvessels in ovarian follicles compared with those injected with saline (Fig. 1aand b). In accordance with this, immunofluorescence staining of the endothelial cell marker CD31 and CD34 also implicated a significant induction of vascular development in follicles of FSH-primed mice (Fig. 1c-f). Moreover, histological analysis revealed increased level of follicular VEGF after FSH administration (Fig. 1gand h). Specifically, VEGF-positive staining was concentrated in granulosa cells (GCs) (Fig. 1g). Consistently, data from qRT-PCR and western blotting assay further confirmed that FSH activates VEGF expression within ovarian GCs (Fig. 1i-k). We also measured follicular levels of HIF-1α following FSH injection. As shown in Fig. 1l-o, compared with the control group, mice primed with FSH exhibited significant higher expression of HIF-1α in ovarian GCs. Taken together, these results suggested that the angiogenic effect of FSH might be relevant to the induction of HIF-1α and VEGF in follicular GCs.

3.2 Inhibition of HIF-1α activity blocks FSH-induced VEGF expression and angiogenesis in ovarian follicles
To investigate the involvement of HIF-1α in the regulation of bloodvessel formation by FSH in the ovary, FSH-primed immature mice were intraperitoneally (i.p.) injected with PX- 478, a HIF-1α antagonist that acts by inhibiting HIF-1α expression (Fig. 2a). As show in Fig. 2b and c, the protein level of HIF-1α was significantly reduced in GCs collected from mice with PX-478 injection. Concomitantly, both western blot analysis and qRT-PCR results showed that, PX-478 effectively abolished the induction of VEGF expression by FSH in ovarian GCs (Fig. 2dande). Consistent with this, immunohistochemical staining of HIF-1α and VEGF in mouse ovarian sections further confirmed that inhibition of HIF-1α activity repressed FSH-activated VEGF expression within ovarian follicles (Fig. 2f and g). We next detected the effects of PX-478 on follicular angiogenesis under the same conditions. By assessing the number of bloodvessels and the expression of CD31 and CD34, it was observed that mice received both FSH and PX-478 injection displayed lower angiogenic activity compared with those injected with FSH alone (Fig. 2h-k). Based on these data,we proposed that the activation of the HIF-1α-VEGF axis is required for FSH-mediated follicular angiogenesis.

3.3 FSH-mediated follicular growth is independent of follicular angiogenesis
FSH is believed as the key stimulus that promotes follicles growth and development (9). To determine whether FSH-mediated follicular growth relies on its proangiogenic effects,histopathological analysis of the ovaries were done in immature mice injected with FSH plus PX-478. Ovaries from mice received both FSH and PX-478 injections showed no clear differences in the ovarian weight compared with those injected with FSH alone (Fig. 3a). In addition, both the ovarian size and the number of antral follicles were morphologically indistinguishable between these two groups (Fig. 3b and c). Moreover, a quantitative analysis of the follicular size also revealed no definite influence of PX-478 on the average diameter of ovarian follicles in FSH-injected mouse (Fig. 3d). Considering the role of FSH-HIF- 1α-VEGF signaling in regulating follicular angiogenesis as mentioned above, our data thus indicated that bloodvessel formation might not be required for FSH-induced follicular growth.

3.4 Inhibition ofangiogenesis prevents ovulation in FSH-injected mice
As reported, the suppression of HIF-1α-VEGF pathway is supposed to reduce litters production in female mice due to compromised ovulation (7,30). Although our findings indicated a role of FSH in stimulating follicular growth without vasculatureformation, it remains unclear whether these follicles are normally developed to become capable of ovulation. Therefore, we examined the histology of the ovaries of PX-478 injected mice subjected to a superovulation protocol. As shown in Fig. 4a, at 15 h after hCG treatment, ovaries from FSH-injected mice were comprised primarily of corpus luteum, indicating that ovulatory follicles had ruptured and undergone luteinization. In striking contrast, the ovaries from the PX-478-injected mice harbored numerous unruptured follicles with trapped oocytes. By counting the number of oocytes in oviduct and corpus luteum in the ovary, we further confirmed that PX-478 blocks gonadotropin-induced ovulation (Fig. 4b and c). These results indicated that the induction ofangiogenesis through FSH-HIF-1α-VEGF pathway is essential for ovulation in mouse ovaries.

3.5 Blockingangiogenesis negates the protective effects of FSH on follicular survival
It is well established that FSH is the primary pro-survival factor for antral follicles (28). To determine whether the blockade ofangiogenesis might exert any influence on FSH-induced follicular survival, we examined apoptotic signals in ovaries of mice primed with FSH and/or PX-478. Using the terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) assay, we found that mice received both FSH and PX-478 injection exhibited significantly increased apoptosis in ovarian GCs compared with those injected with FSH alone (Fig. 5a-c). In addition, PX-478 injection markedly enhanced caspase-3 expression in GCs within large follicles of FSH primed mice (Fig. 5d). Moreover, western blotting analysis of activated caspase-3 in GCs further confirmed a vastly greater incidence of apoptosis induced by PX-478 injection (Fig. 5e and f). These results indicated that the inhibition of vascularization by antagonizing the HIF-1α-VEGF axis in GCs might compromise the protective effects of FSH on follicle viability.

3.6 The ovulation disturbance caused by PX-478 injection leads to defective oocyte development in FSH-treated mice
Given the role of FSH and hypoxia in regulating oocyte development and maturation (31,32), we next investigated whether PX-478-mediated inhibition ofangiogenesis and the resultant anovulation might affect FSH-mediatedoocyte development. Immature mice were intraperitoneally (i.p.) injected with FSH and PX-478 for two consecutive days, and hCG was then injected to stimulate ovulation (Fig. 2a). In response to this sequential hormonal treatment, the preovulatory follicles typically rupture at approximately 12-14 h after hCG administration. To assess the influence of ovulation status on oocyte development, we chose to conduct our experiments using ovaries collected at 10 h (representing preovulatory stage) and 15 h (representing postovulatory stage) after hCG administration respectively. As shown in Fig. 6a, at the preovulatory stage, mice received both FSH and PX-478 injections showed no clear differences in the ratio of dead oocytes compared with those injected with FSH alone. Dead oocytes were only found in the PX-478 group and absolutely none were found in mice receiving other treatments (Fig. 6a). However, increased rates of dead oocytes were observed within both ovaries and oviducts in FSH- and PX-478-injected mice at the postovulatory stage (Fig. 6b).

To illustrate the morphological difference between normal oocytes and dead oocytes, we provided a representative image of oocytes collected in oviducts from both FSH- and PX-478-treated mice at 15 h after hCG injection (Fig. 6c). During oocyte maturation, actin plays a critical part in spindle positioning and cortical reorganization. To evaluate whether the detrimental effect of PX-478 on the oocyte development involves the actin dynamics, phalloidin was used to label the F-actin. As shown in Fig. 6dande, at the preovulatory stage, oocytes collected from both FSH and FSH+PX- 478-injected mice displayed strong signals of actin staining that was concentrated evenly on the plasma membrane. However, at the postovulatory stage, ovarian oocytes from FSH- injected mice exhibited a discontinuous distribution of actin filaments after PX-478 administration (Fig. 6f and g).Accurate control of spindle assembly is obligatory for orderly meiosis during oocyte maturation. To ask whether PX-478 treatment might cause abnormal spindle assembly in the ovulatedoocytes, spindle morphologies were observed by immunostaining with anti-α- tubulin-FITC antibody. In FSH-treated mice, oocytes displayed atypical barrel-shape spindle apparatus with strong signals of tubulin staining (Fig. 6h and i). By contrast, a higher incidence of defective spindles was observed in oviducts from mice primed with FSH plus PX-478 (Fig. 6hand i). Together, these results showed that the blockade of FSH-mediated superovulation by PX-478 treatment might cause developmental defects in the oocytes.

4. Discussion
During follicular growth，the rapid proliferation of GCs in the growing follicles in conjunction with the devoid of vasculature in the interior mural granulosa layers establish a local hypoxic environment within the follicles (33). The hypoxic conditions prevent HIF-1α degradation, resulting in accumulation of the HIF-1α in the nucleus and subsequent induction of HIF-target genes like VEGF. Recent studies suggested that HIF-1α activity in GCs is required for FSH-induced VEGF expression in mice (7); however the authors failed to alter ovarian vasculature in their experiments, since their data were obtained from VEGF transgenic mice, where the angiogenesis machinery has already been manipulated. Our current data showed that FSH injection significantly enhanced expression of HIF- 1α, VEGF, and CD34 in mouse ovaries, accompanied by greatly increased number of bloodvessels in ovarian follicles. In striking contrast, the intraperitoneally administration of PX-478, a HIF-1α specific antagonist, blocked FSH-induced VEGF expression, and impaired the angiogenic activity of FSH in the ovary. Indicatively, our findings provide the direct evidence demonstrating that the activation of VEGF expression via HIF-1α contributes to FSH- mediated angiogenesis in ovarian follicles.Angiogenesis is associated with follicular development (11). FSH,agonadotropin secreted by the anterior pituitary, is believed to provide the key stimulus that promotes ovarian follicle growth to the preovulatory stage (9). On the other hand, FSH could stimulate angiogenesis in the follicle (10,12).

However, it remains unclear if the proangiogenic effects of FSH are correlated with its actions on promoting follicular growth. In the present study, we succeeded in blocking FSH-induced follicular angiogenesis by inhibiting the HIF-1α-VEGF axis using PX-478. Unexpectedly, in FSH-injected mice, neither the ovarian weight, nor the number of preovulatory follicles, or the follicular size was significantly decreased following PX-478 administration. Therefore, our findings might bring forward a new understanding of FSH -mediated follicular growth through anangiogenesis independent manner.In addition, our data showed that that in the PX-478 group, vessel formation is diminished, and VEGF protein level was almost undetectable. Nevertheless, the FSH and PX478 treated group, although showing statistically significant lower levels for VEGF protein and vessel quantification than FSH group, did not fully reproduce the values in PX-478 group. In fact, the VEGF protein and average number of vessels per follicle were increased ( ≈ 2-fold), when compared to PX-478, reaching levels similar to those observed in controls. These findings might suggested a HIF-1α-independent VEGF secretion and vessel formation in the ovary. Previous studies indicated that the HIF-1α-VEGF pathway might be involved in ovulation (7,30). We thus asked whether the blockade ofangiogenesis using PX-478 could affect ovulation capability in mice subjected to gonadotropin treatment.

The results showed that, at 15 hpost hCG injection, the number of ovulatedoocytes in the oviducts and the formation of corpus luteum were signifcantly reduced in mice injected with FSH plus PX-478 compared with those injected with FSH alone, suggesting that the activation of vascularization through FSH-HIF-1α-VEGF axis is necessary for ovulation. On the other hand, FSH functions as a critical survival signal for ovarian follicles (34). The process of follicle atresiais characterized by follicular bloodvessel network reduction (35), indicating that FSH- induced follicular angiogenesis might be essential to its prosurvival effects. GCs apoptosis is considered as the major cause of follicular atresia (29). Our current data showed that the PX-478 injection markedly increased apoptotic signals in ovarian GCs from FSH-primed mice, implicating that the formation of bloodvessels is required for FSH-mediated follicular survival. Based on these clues, we proposed follicular atresia as one possible cause of anovulation when FSH-stimulated follicles develop in the absence ofangiogenesis, although more in vivo experiments should be performed to investigate the connection between atresia and ovulation in future studies. Both FSH and hypoxia have been reported to regulate oocyte development and maturation (31,32). Since HIF-1α acts as a keystimulatorofangiogenesis during hypoxic response (15), we tested whether the inhibition of HIF-1α-induced angiogenesis by PX-478 might exert any influence on FSH-mediatedoocyte development. We observed that oocytes collected in preovulatory follicles from mice received both FSH and PX-478 injections showed no clear differences in oocyte viability and oocyte meiotic activity compared with those injected with FSH alone, suggesting that angiogenesis is not required for FSH-induced oocyte development before the preovulatory stage. However, at the postovulatory stage, PX-478 negated FSH-mediated protection for oocytes retrieved from FSH primed mice.These results indicated that the defects in oocyte development occur after ovulation. In other words,the irregular ovulation itself might bring about developmental abnormality in the oocytes. Since ovulation depends on FSH-induced vascularizationas mentioned above, we speculated that although the proangiogenic effects of FSH could not directly affect oocyte maturation, it might act through ovulation to determine whether the oocyte will continue to develop normally at the postovulatory stage.


In summary, this study showed for the first time that FSH could stimulate follicular growth without vasculatureformation. However, these follicles does not normally developed, as evidenced by increased GCs apoptosis and compromised ovulation, which might consequently lead to defects in oocyte development. These findings further define the role and mechanism of FSH in regulating follicular development and provide novel insights into the treatment procedures of anovulatory infertility.