|Home | About | Journals | Submit | Contact Us | Français|
We hypothesized that vascular endothelial growth factor A (VEGFA) angiogenic isoforms and their receptors, FLT1 and KDR, regulate follicular progression in the perinatal rat ovary. Each VEGFA angiogenic isoform has unique functions (based on its exons) that affect diffusibility, cell migration, branching, and development of large vessels. The Vegfa angiogenic isoforms (Vegfa_120, Vegfa_164, and Vegfa_188) were detected in developing rat ovaries, and quantitative RT-PCR determined that Vegfa_120 and Vegfa_164 mRNA was more abundant after birth, while Vegfa_188 mRNA was highest at Embryonic Day 16. VEGFA and its receptors were localized to pregranulosa and granulosa cells of all follicle stages and to theca cells of advanced-stage follicles. To determine the role of VEGFA in developing ovaries, Postnatal Day 3/4 rat ovaries were cultured with 8 μM VEGFR-TKI, a tyrosine kinase inhibitor that blocks FLT1 and KDR. Ovaries treated with VEGFR-TKI had vascular development reduced by 94% (P < 0.0001), with more primordial follicles (stage 0), fewer early primary, transitional, and secondary follicles (stages 1, 3, and 4, respectively), and greater total follicle numbers compared with control ovaries (P < 0.005). V1, an inhibitor specific for KDR, was utilized to determine the effects of only KDR inhibition. Treatment with 30 μM V1 had no effect on vascular density; however, treated ovaries had fewer early primary, transitional, and secondary follicles and more primary follicles (stage 2) compared with control ovaries (P < 0.05). We conclude that VEGFA may be involved in primordial follicle activation and in follicle maturation and survival, which are regulated through vascular-dependent and vascular-independent mechanisms.
The primordial follicle pool has a vital role in female reproduction. Abnormal development or regulation of the primordial follicle pool can lead to ovarian dysfunction, including an impairment of reproductive capacity or premature ovarian failure [1–3]. Furthermore, manipulation of the primordial follicle pool may provide a means to increase reproductive efficiency and/or increase our understanding of follicular development.
During the late gestational period in rodents, oocytes are located within cord-like structures. These ovigerous cords consist of clusters of oocytes (oocyte cysts) surrounded by mesenchymal-like somatic cells . Primordial follicles are formed when a single layer of squamous pregranulosa cells envelop individual oocytes, disrupting the oocyte cysts [5–7]. Folliculogenesis is initiated during the first few days of life in the rodent; by Postnatal Day 3/4 (P3/4), the rat ovary consists predominantly of primordial follicles . As soon as the primordial follicle pool is formed, subsets of follicles are recruited into the growing follicle pool, with small antral follicles first appearing during the second and third weeks of life . Nonrecruited primordial follicles can remain quiescent for months or years .
Although the process of primordial follicle recruitment is incompletely understood, specific growth factors must either stimulate primordial follicles to leave the dormant state or inhibit primordial follicles from entering the growing pool. Microarray analysis has identified that Vegf mRNA expression is significantly upregulated during the primordial to primary follicle transition in postnatal rat ovaries , and in vivo injections of a vascular endothelial growth factor (VEGF) antibody have demonstrated that primordial follicles may be affected . These findings are notable because there is no vasculature surrounding primordial or primary follicles. Neither of the previous studies addressed whether the actions of VEGF were regulated indirectly through vascular development or directly at the level of the somatic cells or oocytes.
The principal angiogenic gene, VEGF, consists of the following five family members: Vegfa, Vegfb, Vegfc, Vegfd (officially called Figf, c-fos-induced growth factor), and placenta growth factor (Pgf). The best characterized of these is Vegfa. The Vegfa gene consists of eight exons, which undergo alternative splicing to form different mRNA splice variants and are translated into VEGFA protein isoforms with different numbers of amino acids. The predominant isoforms expressed in most tissues throughout the body are VEGFA_188, VEGFA_164, and VEGFA_120 . VEGFA isoforms are structurally different based upon the exons they are developed from, and these differences make them unique in their function. The larger isoforms containing exons 6 and 7 have heparin-binding domains, making them less diffusible. The smaller isoforms lack these exons and are freely diffusible. This difference in diffusion allows VEGFA isoforms to form a chemoattractant gradient to induce endothelial cell migration and the formation of vascular networks within developing organs or tumors [13, 14].
Two tyrosine kinase receptors, FMS-like tyrosine kinase 1 (FLT1, also known as VEGFR1) and kinase insert domain protein receptor (KDR, also known as VEGFR2 and FLK1), bind to VEGFA. The primary receptor involved in VEGFA-induced angiogenesis is KDR. Binding of VEGFA to KDR promotes endothelial cell survival, differentiation, and migration , and mice missing KDR lack endothelial cells and do not survive past midgestation . Although Flt1 knockout mice have abundant numbers of endothelial cells, they also die during embryonic development because endothelial cells are unable to assemble a functional vascular network . It has been proposed that FLT1 regulates vascular development by trapping VEGFA and suppressing VEGFA levels within an appropriate range .
Previous work in our laboratory has demonstrated that VEGFA is necessary for development of seminiferous cords and sex-specific vasculature during testis morphogenesis in the rat . In this study, we hypothesized that VEGFA is involved in early follicle development, which may be independent of vasculature development. The objectives of the present study were to determine if inhibition of both VEGFA receptors (FLT1 and KDR) or KDR alone affected vascular development and follicle progression in perinatal rat ovarian cultures.
Embryonic and postnatal ovaries were obtained from our Sprague-Dawley rat colony at the University of Nebraska-Lincoln Department of Animal Science, with founders purchased from Charles River (Wilmington, MA). Ovaries were dissected from Embryonic Day 13 (E13) to P10 rats to evaluate ovaries across the following important developmental stages: the formation of oocyte cysts, the formation of primordial follicles, and the initiation of follicular activation and development. Embryonic age was calculated from days after coitus. Postnatal age was determined using day of birth as P0. All animal procedures were approved by the University of Nebraska Animal Care and Use Committee.
Total RNA from ovaries at different ages was extracted using Tri Reagent (Sigma-Aldrich, St. Louis, MO) per the manufacturer's protocol. After isolation, total RNA was resuspended in 20 μl of diethyl pyrocarbonate water, and RT was performed on 5 μg of template using SuperScript II (Invitrogen, Carlsbad, CA) according to the manufacturer's recommended protocol . The resulting cDNA was then stored at −20°C for subsequent RT-PCR.
Primers for rat Vegfa (Table 1) were used with an annealing temperature of 58°C for 35 cycles to generate different products depending on the Vegfa mRNA isoform expressed . These primers generate products of 99 base pair (bp) for Vegfa_120, 171 bp for Vegfa_144, 231 bp for Vegfa_164, 303 bp for Vegfa_188, and 354 bp for Vegfa_205. For Flt1, a nested PCR approach with an annealing temperature of 54°C for 35 cycles was utilized to produce a 202-bp product. Outer and inner primer sets (Table 1) were designed using the PrimerQuest primer design program (Integrated DNA Technologies, Coralville, IA). A nested approach was also used with human-specific Kdr primers (Table 1) at an annealing temperature of 52°C for 30 cycles to amplify a 213-bp product . Primers for glyceraldehyde-3-phosphate dehydrogenase (Gapdh) (Table 1) were used at an annealing temperature of 60°C for 40 cycles to produce a 460-bp product. Gapdh was used as an endogenous control for RNA isolation and amplification . All PCR products were subcloned and confirmed using restriction digest analysis. The PCR products were subcloned into pCRII (Invitrogen) using the TOPO TA Cloning kit (Invitrogen) and were sequenced with primers for the T7 promoter region (data not shown). The RT-PCR was conducted on three to five different samples for each developmental time point.
Primers were designed using Primer Express 1.5 (software that accompanied the 7700 Prism sequence detector; Applied Biosystems, Foster City, CA) for rat Vegfa_120, Vegfa_164, and Vegfa_188 (Table 1). Fluorescent probes were obtained from Applied Biosystems (Table 1). Quantitative RT-PCR (QRT-PCR) for the Vegfa isoforms was performed using TaqMan Universal Master Mix (Applied Biosystems), 900 nM of both forward and reverse primers, and 200 nM of probe. Gapdh was amplified for all samples to serve as a basis for calculating relative expression. Gapdh primers and probes were obtained from Applied Biosystems. Experimental and Gapdh PCRs were carried out in separate wells in triplicate. An arbitrary value of template was assigned to the highest standard and corresponding values to the subsequent dilutions. These relative values are plotted against the threshold value for each dilution to generate a standard curve. The relative amount for each experimental and Gapdh triplicate was assigned an arbitrary value based on the slope and y-intercept of the standard curve. The average of the experimental triplicate is divided by the average of the Gapdh triplicate, and the resulting normalized values are used for statistical analysis . At least three different pools of tissue from each age group were utilized to obtain these data.
Ovaries were fixed in Bouins solution and paraffin embedded according to standard procedures . Tissues were sectioned (5 μm), deparaffinized, rehydrated, and microwaved in 0.01 M sodium citrate to boil for 5 min. After boiling, tissues were cooled for 1–2 h, and sections were blocked with 10% normal goat serum in PBS for 30 min at room temperature. Immunohistochemistry was performed as previously described . The VEGFA antibody was a rabbit polyclonal IgG raised against a peptide corresponding to amino acids 1–140 of human VEGFA (Santa Cruz Biotechnology, Santa Cruz, CA). The FLT1 antibody was a rabbit polyclonal IgG raised against a peptide at the C-terminus of human FLT1 (Santa Cruz Biotechnology). The KDR (FLK1) antibody was a mouse monoclonal IgG1 antibody raised against a peptide corresponding to amino acids 1158–1345 of mouse KDR (Santa Cruz Biotechnology). All antibodies were diluted 1:50 to 1:100 in 10% normal goat serum. As a negative control, serial sections were processed without primary antibody. Biotinylated goat anti-rabbit and goat anti-mouse secondary antibodies were diluted 1:300. Secondary antibody was detected with aminoethyl carbazole chromagen substrate solution (ZYMED Laboratories, San Francisco, CA). The same immunohistochemical procedures localized VEGFA and KDR staining to Sertoli cells, germ cell cytoplasm, and certain interstitial cells within E14 to P5 rat testes in previous investigations from our laboratory . Specificity of antibodies was determined through blocking proteins for each protein (data not shown). Immunohistochemistry for each protein was performed on at least three different sections of tissue from each age group.
Ovaries were dissected from P3/4 rats. One ovary from each animal was designated as a control, while its pair was subjected to VEGFA receptor signal transduction inhibitor treatment (VEGFR-TKI; Calbiochem, La Jolla, CA) or KDR signal transduction inhibitor treatment (V1; Calbiochem). Tyrosine kinase activity through FLT1 and KDR is inhibited by VEGFR-TKI, blocking signal transduction of VEGFA , and V1 binds KDR, blocking VEGFA-KDR interaction . Organs were cultured for 14 days as previously described . Ovaries were cultured in drops of medium on 0.4-μm Millicell-CM filters (Millipore, Bedford, MA) floating on the surface of 0.4 ml of medium in four-well plates. The medium consisted of Dulbecco modified Eagle medium/Hams F-12 medium (1:1 [vol:vol]; Gibco, Grand Island, NY), supplemented with 0.1% bovine serum albumin (BSA), 0.1% Albumax (Gibco), 5× ITS-X (insulin, sodium transferrin, sodium selenite, and ethanolamine; Gibco), 0.05 mg/ml of l-ascorbic acid (Sigma-Aldrich), penicillin, and streptomycin. Doses of VEGFR-TKI and V1 (both diluted in dimethyl sulfoxide [DMSO]) were added directly to the culture medium of the treated wells at the start of culture, and treatment was repeated daily. Similar doses of DMSO were added to the paired control wells. The medium was changed after every 2 days of culture.
We cultured P3/4 ovaries because ovaries from rats of this age consist predominantly of primordial follicles  and because similar ovarian organ culture procedures have demonstrated that primordial follicles can spontaneously initiate development to early primary follicles . We utilized a dose of 8 μM VEGFR-TKI in our organ culture system because we have demonstrated that this dose affects testis morphogenesis and vascular development in rat testis organ cultures, while not affecting germ cell mitosis . The concentration required for 50% inhibition (IC50) for single-layer cells treated with VEGFR-TKI is 2 μM for FLT1 and 0.1 μM for KDR . We are well within the IC50 for both VEGFA receptors. We utilized 15 and 30 μM doses of V1 for our cultures because it has been previously demonstrated that V1 works most efficiently between 15 and 40 μM . The IC50 for V1 is 80 μM for cell cultures . The concentrations we utilized in our experiments were well under this dose.
After culture treatment, ovaries were imaged with a Spot camera imaging capture system (Spot Advance; Diagnostic Instruments, Sterling Heights, MI). The National Institutes of Health Scion Image program (Scion Image, Frederick, MD) was utilized to obtain individual ovary areas (total number of pixels) at 40× magnification. Each ovary was outlined twice, and these two areas were averaged to obtain an accurate area measurement for each ovary . The mean area of each control ovary was set to 100%, and the mean area of each treated ovary was calculated as a percentage of its paired control. In total, 45 (VEGFR-TKI), 27 (15 μM V1), and 16 (30 μM V1) ovary pairs were imaged for area measurements.
After imaging, ovaries were fixed in Bouins solution and paraffin embedded for histology (described herein) or fixed in 4% paraformaldehyde for whole-mount immunohistochemistry. VEGFR-TKI organ culture ovaries were fixed overnight at 4°C. Ovaries were then washed in PBS for 2 h at room temperature and incubated overnight at 4°C in blocking buffer (PBS, 5% BSA, and 0.1% Triton X-100; Sigma-Aldrich). After 1 h at room temperature, ovaries were incubated overnight at 4°C with primary antibody and washed with washing buffer (PBS, 1% BSA, and 0.1% Tween 20; Sigma-Aldrich) for 4–5 h at 4°C (changing the buffer once), followed by two 1-h rinses of washing buffer at room temperature. The first primary antibody utilized was a mouse monoclonal IgG1 raised against rat platelet endothelial cell adhesion molecule (PECAM1) or CD31 (1:50 dilution in blocking buffer; BD Pharmingen, San Jose, CA). Ovaries were then incubated overnight at 4°C with Cy5-conjugated secondary antibody (1:500 dilution in blocking buffer; Jackson Laboratories, West Grove, PA) and washed with washing buffer as with the primary antibody. These blocking, incubation, and washing steps were repeated using an antilaminin primary antibody and a Cy2-conjugated secondary antibody. The second primary antibody was a rabbit polyclonal immunoglobulin raised against rat laminin (1:100; DAKO, Carpinteria, CA).
A different whole-mount immunohistochemistry procedure was utilized for the V1 organ culture ovaries to accommodate Alexa Fluor secondary antibodies in accord with previously reported methods [7, 30]. Ovaries were fixed for 1 h in 4% paraformaldehyde and rinsed twice in washing buffer (PBS with 0.1% Triton X-100 [WB-PT]) for 5 min, followed by an additional 1-h rinse at room temperature. After being blocked in blocking buffer 1 (BB1) and WB-PT with 5% BSA for 1 h, ovaries were incubated with PECAM1 antibody (1:50 dilution in BB1) overnight at 48°C. The following morning, ovaries were rinsed three times in blocking buffer 2 (BB2) and WB-PT with 1% BSA for 30 min before being incubated with secondary antibodies (Alexa Fluor 488 and 647, diluted 1:200 in BB1; Molecular Probes, Eugene, OR) for 3–4 h at room temperature. Ovaries then received four 30-min washes in BB2, followed by a final rinse with PBS. VEGFR-TKI and V1 organ culture ovaries were mounted in Gel/Mount (Biomeda, Foster City, CA) for subsequent confocal microscopy.
VEGFR-TKI organ culture images were collected using a Bio-Rad MRC1024ES confocal laser scanning microscope (Bio-Rad Laboratories, Hercules, CA). Whole ovaries were scanned through a series of Z-sections at 100×, 200×, and 600× magnifications (thickness of sections, 10, 5, and 3 μm, respectively) with green, red, and merged channels to determine positive staining for PECAM1 and laminin. Organ depths were estimated at 100× magnification by the number of 10-μm Z-steps required to scan through the entire organ. V1 organ culture images were collected using an Olympus Fluo-View 500 confocal laser scanning microscope (Olympus America, Center Valley, PA). Ovaries were scanned at 100× magnification (thickness of sections, 3 μm) with the red channel to determine positive staining for PECAM1. Organ depths were estimated at 100× magnification by the number of 3-μm Z-steps required to scan the entire organ.
Red channel confocal images at 600× magnification (VEGFR-TKI) or 100× magnification (V1) were used to analyze vascular density of control and treated organ culture ovaries. Vascular density or staining index was quantified using Scion Image. Densitometry was performed on three fields for each organ. For V1 ovaries, the three fields were analyzed using a digital zoom at three times the original image that was acquired at 100× magnification. Within each field, the staining index was defined as the number of pixels exceeding an arbitrary gray scale value. The mean staining index for each organ was defined as the average of the staining indexes from all three fields. The mean vascular density for each control ovary was set to 100%, and the mean of each treated ovary was calculated as a percentage of its paired control. In total, 17 (VEGFR-TKI and 15 μM V1) and 19 (30 μM V1) ovary pairs were analyzed for vascular density.
Organ culture ovaries not used for whole-mount immunohistochemistry were fixed and embedded as already described. Embedded ovaries were sectioned (5 μm), deparaffinized, stained with hematoxylin-eosin, and rehydrated. With bright-field microscopy, follicles from each ovary were classified as stages 0–4 . A stage 0 primordial follicle consists of an oocyte surrounded by a single layer of squamous pregranulosa cells. A stage 1 early primary follicle consists of an oocyte surrounded by a single layer of cells composed of a combination of pregranulosa and granulosa cells. A stage 2 primary follicle consists of an oocyte surrounded by a single layer of cuboidal granulosa cells. A stage 3 transitional follicle is developing a second layer of granulosa cells, while a stage 4 follicle has theca cells beginning to organize around the granulosa cell layers . For the VEGFR-TKI organ cultures, three ovary pairs were imaged at 400× magnification for follicle quantification. The middle 12 histology sections were evaluated for each ovary, and follicles were staged and counted from three nonoverlapping cortical areas. Middle sections were utilized to count follicles over a full cross-section of ovarian tissue and to evaluate differences in follicle numbers over comparable regions between treated and control ovaries. From the VEGFR-TKI experiments, we determined that three ovarian sections were as accurate as 12 in determining accurate counts of follicle stages. Thus, for the V1 organ cultures, six ovary pairs were imaged at 200× magnification, and the middle three histology sections were evaluated for each ovary. Follicle staging and counting were performed independently by two individuals, and discrepancies in counts, if any, were averaged. The mean number of follicles in each stage of development per area counted was statistically analyzed between control and treated ovaries. Additionally, the mean number of follicles from each treated ovary was expressed as a percentage of its paired control, and the percentages from all ovary pairs were analyzed. The mean number of each follicle stage was also expressed as a percentage of the total number of follicles from each area. These percentages were statistically analyzed between control and treated ovaries.
All data were analyzed by one-way ANOVA using JMP software (SAS Institute, Cary, NC). Student t-test was used to compare mean normalized QRT-PCR values between different developmental ages. Student t-test and Dunnett test were used to compare ovarian area, vascular density, and follicle counts between control and treated organs. Differences in data were considered statistically significant at P < 0.05 unless otherwise stated.
Conventional RT-PCR was used to evaluate mRNA expression during early ovarian development. Nine developmental time points were evaluated (E13, E14, E16, E18, P0, P3, P4, P5, and P10). Both Vegfa_120 and Vegfa_164 mRNA was detected at all time points evaluated (Fig. 1). Messenger RNA from Vegfa_188 was detected at all developmental time points except E14 (Fig. 1). The Vegfa_205 mRNA isoform was distinctly present at P10, with a faint band detected at E18 (Fig. 1). The upper bands from E16 and P0 were cut out of the gel and were determined to be Vegfa_188 and not Vegfa_205 through subcloning and sequencing. Messenger RNA for Flt1 and Kdr was also detected throughout all time points evaluated (Fig. 1). Quantitative RT-PCR demonstrated differential abundance of Vegfa_120, Vegfa_164, and Vegfa_188 during prenatal and perinatal development of the ovary (Fig. 2). Specifically, Vegfa_120 levels increased from embryonic ages to P0; after birth, levels declined from P0 to P3 (Fig. 2A) (P < 0.05). Messenger RNA levels for Vegfa_164 increased from E13 to E18, declined after birth, and then increased again at P5 (Fig. 2B) (P < 0.05). Levels for Vegfa_188 mRNA peaked at E16 (Fig. 2C) (P < 0.05).
After confirmation of Vegfa, Flt1, and Kdr mRNA expression during early ovarian development, immunohistochemistry was utilized to localize expression to specific cell types. Immunohistochemistry was performed on ovaries from P0, P3, P4, P6, and P10 rats. Staining for VEGFA protein was localized to oocyte cysts and to the pregranulosa or granulosa cells of follicles from primordial through antral stages (Fig. 3, A, D, and G). In preantral and antral follicles, VEGFA staining was also detected in the cytoplasm of oocytes and in theca cells (Fig. 3G). No staining was identified in ovaries processed without primary antibody (Fig. 3J). Expression of FLT1 (Fig. 3, B, E, and H) and KDR (Fig. 3, C, F, and I) protein was localized to oocyte cysts and to the oocytes and pregranulosa or granulosa cells of primordial through antral-stage follicles. Staining for FLT1 and KDR was also identified in theca cells of advanced-stage follicles (Fig. 3, H and I) but was not detected in negative control ovaries (Fig. 3, K and L).
To determine the function of VEGFA in follicle progression, P3/4 rat ovarian organ cultures were treated with VEGFR-TKI, a VEGFA tyrosine kinase signal transduction antagonist that inhibits both KDR and FLT1. Ovaries treated with VEGR-VEGFR-TKI (Fig. 4D) had a 25% reduction in ovarian area compared with their paired controls (Fig. 4, A and G) (P < 0.0001). Depth of cultured ovaries, determined by the total number of 10-μm Z-series confocal microscopy images taken of each organ, was not statistically different between treated and control ovaries.
Whole-mount immunohistochemistry and confocal microscopy were utilized to further evaluate the role of VEGFR-TKI treatment on ovarian cultures. Laminin staining was used to localize basement membranes  and thus outline individual follicles (Fig. 4, B and E). VEGFA signal transduction inhibition did not appear to alter follicle formation or organization (Fig. 4E). PECAM1 staining was used to localize endothelial cells  and thus identify the vasculature within ovaries (Fig. 4, C and F). Ovarian vascular density was diminished by 94% in treated ovaries (Fig. 4, F and H) compared with controls (Fig. 4C) (P < 0.0001).
The number of follicles in control ovaries was compared with VEGFR-TKI-treated ovaries to determine differences due to treatment. Ovaries treated with VEGFR-TKI (Fig. 5B) had 119% more primordial (stage 0) follicles and 43% more total follicles per area evaluated than control ovaries (Fig. 5, A and C) (P < 0.0001). VEGFR-TKI treatment also resulted in 40% less transitional (stage 3) and secondary (stage 4) follicles (Fig. 5C) (P < 0.005). Furthermore, treated ovaries consisted of 51% primordial follicles and 49% developing follicles (stages 1–4), while control ovaries had 32% primordial follicles and 68% developing follicles (Fig. 5F) (P < 0.0001). Specifically, VEGFR-TKI-treated ovaries consisted of 19% more primordial follicles, 8% less early primary follicles (stage 1), and 7% less transitional and secondary follicles (Fig. 5E) (P < 0.003). Evaluation of total follicle counts from treated ovaries as a percentage of control ovaries did not reveal any significant differences (Fig. 5D).
To determine if VEGFA was primarily working through KDR, we utilized V1, a signal transduction inhibitor specific for KDR, at two different doses. Ovaries treated with V1 had 18% (15 μM) and 15% (30 μM) reductions in ovarian area (Fig. 6, D and G) (P < 0.0001) compared with their paired controls (Fig. 6A). The depth of cultured ovaries was not statistically different between 15 μM V1 and paired controls; however, there tended to be a reduction in organ depth with the 30 μM V1 dose (P < 0.06) (data not shown).
Whole-mount immunohistochemistry with PECAM1 staining was conducted to determine effects of V1 treatment on vascular density. Vascular density for the ovaries treated with 15 μM V1 was not statistically different from that of their paired controls (data not shown). However, there was a 37% increase in vascular density from merged Z-sections in ovaries treated with 30 μM V1 (Fig. 6, E, F, and H) (P < 0.04) compared with their controls (Fig. 6, B and C). Because the V1 treatment at 30 μM reduced the depth of ovaries, we also analyzed the middle 3-μm slice from each ovary to normalize for the loss in overall ovarian depth. Vascular density for the middle slice of the 30 μM V1-treated ovaries (Fig. 6, E and F) was not different compared with their paired controls (Fig. 6, B, C, and H).
The number of follicles in organ culture ovaries treated with V1 was compared with their paired controls to determine effects of treatment. Ovaries treated with 15 μM V1 had no alteration in follicle numbers (data not shown). However, ovaries treated with a 30 μM V1 dose had 35% less early primary (stage 1) follicles and 31% less transitional (stage 3) and secondary (stage 4) follicles per section than control ovaries (Fig. 7A) (P < 0.05). Evaluation of total follicle counts from treated ovaries as a percentage of control ovaries revealed a 30% reduction in early primary follicles in V1-treated ovaries (Fig. 7B) (P < 0.03). The percentage of primordial and developing follicles (stages 1–4) in treated ovaries was not different from control ovaries (Fig. 7D); however, V1-treated ovaries consisted of 7% less early primary follicles and 6% more primary (stage 2) follicles (Fig. 7C) (P < 0.05).
The present study demonstrates a role for VEGFA and its receptors in follicle progression in the perinatal rat ovary. Inhibition of VEGFA signal transduction through two tyrosine kinase inhibitors affecting both KDR and FLT1 or KDR alone resulted in different effects on vascular density and initial follicle activation. Both inhibitors affected secondary follicle progression, supporting previous studies [33–35] on the importance of VEGFA in follicle development. However, each inhibitor had novel effects on early-stage follicle numbers, implicating different signal transduction pathways or a greater dependence on inputs from vasculature. The VEGFR-TKI inhibitor (which blocks KDR and FLT1) dramatically arrested vascular development and increased primordial follicle numbers, suggesting that VEGFA may be a regulator of initial follicle activation. In contrast, the KDR-specific inhibitor, V1, did not alter primordial follicle numbers or vascular development but reduced early primary, transitional, and secondary follicle numbers. This suggests that VEGFA, through KDR, directly regulates granulosa or oocyte function. Taken together, the results in this study demonstrate a role for VEGFA in follicle activation and progression through vascular-dependent and vascular-independent mechanisms.
The Vegfa gene is complex and produces multiple isoforms through alternative splicing. The different isoforms have unique functions during vascular development in many organs . In the present study, we detected expression of smaller Vegfa isoforms early, with larger isoforms appearing later during ovarian development. Further analysis with QRT-PCR demonstrated that Vegfa_188 mRNA was most abundant at E16, with elevated amounts of Vegfa_164 and Vegfa_120 after E18 and P0, respectively. Around E18 in the rat, primordial germ cells have formed germline cysts that are ceasing to proliferate and are undergoing meiosis. Germline cysts are composed of clusters of oogonia connected by intracellular bridges formed through incomplete cytokinesis. These cell clusters synchronously proliferate, and waves of oogonia enter meiosis in a nonsynchronous manner until they are arrested around birth [30, 36]. Thus, during embryonic development, VEGFA_188 may aid in mitosis, transition into meiosis, or allow for survival of oogonia directly or indirectly through reorganization of vasculature.
There is no information to suggest that ovarian vasculature undergoes reorganization during the time when oocyte cysts are present. The VEGFA_188 isoform is the least diffusible isoform, and it increases branching of vasculature and development of capillaries from larger vasculature. The VEGFA_164 isoform recruits endothelial cells, aiding in the formation of major blood vessels [13, 37]. Thus, greater expression of Vegfa_188 and Vegfa_164 may be necessary to increase branching of vasculature as large vessels form between nests to aid in maintenance and survival of oogonia before cyst breakdown.
After birth and meiotic arrest, oocyte cysts undergo a programmed breakdown in which pregranulosa cells invade, divide cytoplasm, and surround oocytes to form primordial follicles [7, 30]. During this time, there is a tremendous amount of oocyte loss . In our study, greater concentrations of Vegfa_120 were detected after P0. Because the highly diffusible VEGFA_120 isoform is primarily thought to be involved in endothelial cell recruitment , it is possible that VEGFA_120 may be involved in the oocyte cyst to primordial follicle transition if vascular reorganization is necessary. To date, there have been no studies to indicate that vascular development is involved in the breakdown of oocyte cysts. However, estrogen, estrogen-like compounds, and progesterone affect oocyte cyst breakdown, increase the number of multioocytic follicles, and interfere with assembly of primordial follicles [38–43]. Estrogen and progesterone have been demonstrated to regulate the Vegfa gene to cause differential expression of Vegfa isoforms in many other tissues [44–52]. Thus, VEGFA could be a candidate gene involved directly in the process of oocyte cyst breakdown or in vascular reorganization that results from oocyte cyst breakdown and primordial follicle formation.
Nonangiogenic survival roles have been attributed to VEGFA in several different systems [53, 54], and survival genes such as Bcl2  have been demonstrated to be upregulated by VEGFA. Therefore, it is possible that VEGFA may act to reduce oogonial loss during the oocyte cyst to primordial follicle transition. In the present study, we localized VEGFA, FLT1, and KDR to oocyte cysts at P0. Previous investigators have also reported that Bcl2 was localized to oocyte cysts in porcine ovaries at similar stages of development . We speculate that VEGFA is acting in a novel nonangiogenic manner to upregulate survival factors during the oocyte cyst to primordial follicle transition to allow for increased numbers of primordial follicles.
Previous immunohistochemical studies have determined that VEGFA, KDR, and FLT1 were present in oocytes of primordial and primary follicles [57–59] and in granulosa and theca cells [57, 60–63] from adult rats, humans, cattle, and primates. In the present study of postnatal rat ovaries, oocyte cysts, pregranulosa cells, and granulosa cells expressed VEGFA. Furthermore, VEGFA was present in theca cells and the cytoplasm of oocytes of preantral and antral follicles. FLT1 and KDR were localized to oocyte cysts, pregranulosa cells, granulosa cells, and oocytes of all follicle stages, as well as theca cells of more advanced-stage follicles. Localization of VEGFA and its receptors to nonvascular cells (granulosa and oocyte) and vascular cells (theca) within the follicle implies angiogenic and nonangiogenic roles for VEGFA in the regulation of follicle development.
Postnatal ovarian organ culture data from the present study support a role for VEGFA in initial primordial follicle activation. The VEGFR-TKI inhibitor, which blocked both KDR and FLT1, resulted in increased numbers of primordial follicles compared with the cultured controls, suggesting that primordial follicle activation was impaired without VEGFA. In contrast, in vivo studies in which VEGFA was neutralized with VEGFA or receptor antibodies resulted in a depletion of primordial follicles in the adult mouse ovary . This suggests that the adult ovary or an in vivo system may act differently in response to VEGFA inhibition compared with perinatal organ cultures. For example, organ culture ovaries are continuously exposed to nutrient-rich media, while in vivo ovaries are likely exposed to smaller quantities of nutrients because of decreased vasculature from inhibition of VEGFA. Treatment with VEGFR-TKI also resulted in an increase in the total number of follicles. Based on the follicle counts for each stage of development, this increase appears to be the result of fewer follicles being stimulated to leave the primordial follicle pool and thus fewer growing follicles being lost through atresia.
In support of the data in the present study, microarray analysis of perinatal rat ovaries revealed that the Vegfa gene is upregulated during the primordial to primary follicle transition . In addition to VEGFA, a number of growth factors have already been identified through in vitro culture systems or knockout mice models to be important in primordial follicle progression [28, 43, 64–70]. For example, knockout of PTEN, a suppressor of PIK3, within the oocytes of mice results in activation of all primordial follicles during early adulthood and premature ovarian failure . Because VEGFA has been shown to activate the PIK3 signaling pathway [72–74] and because the present study localized VEGFA and its receptors to primordial follicles, VEGFA signaling through the PIK3 pathway may be involved in regulating primordial follicle activation.
Primordial follicle numbers were not affected by the KDR-specific inhibitor (V1), but it increased the percentage of primary follicles. Because control ovarian organ cultures spontaneously activate, we speculate that the KDR-specific inhibitor allowed for primordial follicle progression but that follicles were lost after the primary stage through apoptosis or atresia. A recent study  involving VEGFA treatment of bovine cortical cultures demonstrated enhanced transition of follicles from the primary to secondary stages. Thus, inhibition of VEGFA signaling through KDR may prevent VEGFA from promoting the progression of follicles beyond the primary stage.
The different results obtained from these two inhibitors may be due to effects on different signal transduction pathways or effects on vascular density. Because KDR has been postulated to be the primary receptor involved with VEGFA in the initial vascularization of many tissues , the lack of effect on vascular density from the KDR-specific inhibitor was surprising. In contrast, the VEGFR-TKI inhibitor dramatically arrested vascular development and arrested follicles at the primordial stage. The present study supports previous work that suggests FLT1 is necessary for the formation of a functional vascular network , although there is no information to support a role for vascular development in primordial follicle activation. Vasculature may be necessary to transport growth factors and nutrients to primordial follicles to help transition them to later stages of development. Even if there is no blood flow through the vasculature in organ cultures, the development of vasculature may provide an indirect method for angiogenic factors to affect follicular development. It is also possible that strict regulation of VEGFA levels through FLT1 binding may allow for primordial follicle activation through completely nonvascular mechanisms.
Both inhibitors affected the number of follicles that progressed to the secondary stage. Later-stage follicles require vascularization of theca cells to provide increased nutrients to the growing follicle. Inhibition of VEGFA through neutralization of VEGFA or VEGFA receptors has been widely studied in many different species. In most cases, development of later-stage follicles is disrupted, and ovulation is impaired [33–35]. Therefore, these previous data, in combination with the results of the present study, indicate that VEGFA may be a potential target gene involved in regulation of early stages of follicle initiation, as well as later stages of follicle development.
Treatment with both inhibitors also resulted in an overall reduction in ovarian area. We speculate that this change in area is the result of a greater composition of primordial follicles, which are smaller than developing follicles, in treated ovaries compared with controls. It is not clear why an increase in area (horizontal growth along the filter) was seen in some ovaries without an increase in depth (vertical growth above the filter); however, it is likely the result of our organ culture system. Ovaries were cultured in a drop of media on top of floating filters. The filter may have provided more structural support for horizontal growth, and the limited amount of media within the drop may have restricted vertical growth.
In conclusion, the present study supports a role for VEGFA in early follicle development. Expression patterns of VEGFA, FLT1, and KDR suggest that VEGFA could potentially be involved in nonangiogenic or angiogenic events regulating oocyte cyst breakdown to development of later-stage follicles. Ovarian organ culture data with two different inhibitors suggest that VEGFA can act through vascular-dependent mechanisms to affect primordial follicle activation and through vascular-independent mechanisms to alter numbers of primary follicles. A reduction in the number of early-stage follicles would reduce the number of follicles developing to the preantral stage, affect the pool of follicles available for ovulation, and impair the reproductive capacity of females. Our study demonstrates a novel role for VEGFA in the recruitment of primordial follicles into the growing follicle pool, as well as a potential survival factor for primary and later-stage follicles through vascular-dependent and vascular-independent mechanisms.
We would like to thank Dr. Jennifer Wood for critical review of the manuscript. We would also like to thank the undergraduate and graduate students who helped maintain the rat colony for their assistance in these studies. An additional thank-you is extended to Dr. Joe Zhou and the microscopy core at the University of Nebraska-Lincoln for their assistance in obtaining the confocal microscopy images.
1Supported in part by NIH 5R01HD051979, by an NSF Epscore Women in Science grant, and by Tobacco Funds for Biomedical Research from the state of Nebraska. A contribution of the University of Nebraska Agricultural Research Division, Lincoln, Nebraska.