VEGF modulates the effects of gonadotropins in granulosa cells

Available online at www.sciencedirect.com

Domestic Animal Endocrinology 38 (2010) 127–137

VEGF modulates the effects of gonadotropins in granulosa cells L.K. Doyle, C.A. Walker, F.X. Donadeu ∗ Division of Developmental Biology, The Roslin Institute and Royal (Dick) School of Veterinary Studies, University of Edinburgh, Roslin BioCentre, Roslin, Midlothian EH25 9PS, Scotland, UK Received 11 May 2009; received in revised form 31 August 2009; accepted 31 August 2009

Abstract Follicle selection is associated with an increase in the expression of vascular endothelial growth factor (VEGF) and its receptors in granulosa cells, however, the roles of VEGF in regulating the function of these or other non-endothelial cells in the ovary have not been explored in detail. The current study used bovine cell cultures to investigate potential roles of VEGF in the regulation of granulosa cell function during follicle development. Granulosa cells were obtained from morphologically healthy follicles 4 to 8 mm or 9 to 14 mm in diameter (corresponding to diameters before and after the establishment of dominance, respectively, during a bovine follicular wave) and exposed to a range of VEGF concentrations (1 to 100 ng/mL) encompassing concentrations found naturally in bovine dominant follicles. A concentration of VEGF of 1 ng/mL induced significant proliferation of granulosa cells from 4- to 8-mm follicles (P = 0.024) and increased the proliferative response of these cells to follicle-stimulating hormone (FSH; P = 0.045); whereas higher doses of VEGF had no effect on proliferation (P = 0.9). Treatment with VEGF induced an overall increase in mean extracellular signal-regulated kinase 1/2 (ERK1/2) phosphorylation (P = 0.02). In contrast, VEGF, alone or in combination with FSH, had no effect on expression of the steroidogenic enzyme, CYP11A1, by cells from 4- to 8-mm follicles (P = 0.9). Granulosa cells from 9- to 14-mm follicles responded to 1 ng/mL VEGF with an increase in expression of the ovulation-associated gene, PTGS2 (P = 0.003) but higher VEGF doses had no effect (P = 0.9). The PTGS2 response to 1 ng/mL VEGF was similar to that induced by treatment with luteinizing hormone (LH). Interestingly, the stimulatory effects of LH on ERK1/2 phosphorylation (P = 0.003) and PTGS2 expression (P < 0.01) in granulosa cells from 9- to 14-mm follicles were abolished (P = 0.2) by specific chemical inhibition of VEGF receptor 2 (VEGFR2). These results suggest novel and important roles of VEGF and its receptor, VEGFR2, in mediating and/or enhancing the effects of gonadotropins in granulosa cells. © 2009 Published by Elsevier Inc. Keywords: VEGF; Granulosa; LH; Cell proliferation; ERK1/2; PTGS2

1. Introduction Follicle development in monovular species involves the synchronous growth of antral follicles as follicular waves [1]. Within each wave, usually one follicle (dominant follicle) is selected to undergo pre-ovulatory maturation whereas the remaining follicles become sub-

∗ Corresponding author. Tel.: +44 131 527 4331; fax: +44 131 440 0434. E-mail address: [email protected] (F.X. Donadeu).

0739-7240/$ – see front matter © 2009 Published by Elsevier Inc. doi:10.1016/j.domaniend.2009.08.008

ordinate and eventually undergo atresia. In cattle, follicle dominance is established when the largest follicle of a wave reaches about 8.5 mm and this follicle continues to grow to about 14 to 20 mm when it normally ovulates or begins regressing [1]. The control of dominant follicle development lies within complex interactions between systemic gonadotropins and local growth factors. For example, the synergistic actions of insulin-like growth factor-1 (IGF-1) and follicle-stimulating hormone (FSH) are critical in the selection of a single follicle from the cohort of a wave [2]; whereas epidermal growth factor (EGF) plays a very important role in mediating some

128

L.K. Doyle et al. / Domestic Animal Endocrinology 38 (2010) 127–137

of the responses of the preovulatory follicle to the ovulatory luteinizing hormone (LH) surge [3]. The role of many other follicular factors remains poorly understood. One such factor is vascular endothelial growth factor (VEGF). Numerous studies have shown that VEGF and its receptors, VEGFR1 and 2, are expressed in granulosa cells, and dynamic changes in their expression have been reported throughout follicle development in primates [4], pigs [5] and cattle [6], with a consistent increase in both VEGF production and expression of VEGFR2 in dominant follicles. Studies in non-human primates have shown that intrafollicular or systemic injection of a VEGF antagonist results in significant impairment of ovulation and subsequent luteal function [4]. Further, it has been shown that intra-ovarian VEGF injection can stimulate development of pre-antral and antral follicles up to the preovulatory stage [7,8]. The mechanisms by which VEGF elicits these effects are not yet fully understood. While the actions of VEGF are undoubtedly associated with the induction of an extensive vascular system in follicular walls, studies using ruminant or porcine cells suggest that VEGF may also modulate granulosa cell function directly by promoting cell survival [6,9,10], proliferation [5] and/or migration [10]. VEGF has been shown to directly promote similar responses in a variety of other non-endothelial cell types including cells of the central nervous system, haematopoietic cells and tumour cells [reviewed in [11]]. It has been shown that, in general, cell responses to VEGF are mediated primarily through binding of VEGF to VEGFR2; whereas VEGFR1 is believed to act as a decoy receptor that regulates VEGF binding to VEGFR2 [12]. The observations that granulosa cells are a major site of VEGF synthesis within the follicle [4] and that expression of both VEGF and VEGFR2 increase in the dominant follicle reaching their maximal levels in preovulatory follicles [6] suggest potentially important non-angiogenic roles of this growth factor during the development of the dominant follicle. The aim of this study was to investigate such role(s) of VEGF by examining its effects on proliferative, steroidogenic and ovulation-associated responses of bovine granulosa cells in culture. Potential interactions between VEGF and gonadotropins in these cells were also studied. 2. Materials and Methods 2.1. Collection and culture of granulosa cells Bovine ovaries were obtained from a local abattoir and transported to the laboratory at 38 ◦ C in Phos-

phate Buffered Saline (PBS) containing kanamycin (50 ␮g/mL; Sigma, Dorset, UK). Upon arrival at the laboratory, ovaries were briefly trimmed free of extraneous tissues and rinsed in 70% alcohol before being added to pre-warmed McCoys 5a media containing penicillin/streptomycin (Sigma). Follicles were dissected from ovarian stroma and their surface diameter was determined. Only follicles that were deemed healthy, i.e. with a well vascularized wall and transparent, ambercoloured follicular fluid without debris, were used. Granulosa cells were harvested by dissecting follicles into hemispheres and gently scraping follicle walls with blunt-ended forceps. Cells from 4- to 8-mm and 9- to 14-mm follicles (corresponding to the diameters of follicles before and after they become dominant, respectively, during a bovine follicular wave [1]) were separately pooled and cultured. These two granulosa cell populations were chosen to provide optimal responses in terms of proliferation and expression of steroidogenic genes (4- to 8-mm follicles) and expression of ovulation-associated genes (9- to 14-mm follicles) [13,14]. Cells were washed in culture media and cultured in a humidified atmosphere with 5% CO2 at 37O C in culture plates (Nuclon, Nunc, Denmark) that had been pre-coated with Fibronectin (Sigma). Cells were cultured under serum free conditions (McCoy’s 5a with 0.02 M HEPES, 3 mM L-Glutamine, 0.1% BSA, Penicillin (100 IU/mL) and Streptomycin (0.1 mg/mL) supplemented with 10 ng/mL bovine insulin, 2.5 ␮g/mL transferrin and 4 ng/mL sodium selenite) which have previously shown to provide a physiologically relevant bovine granulosa cell system in vitro based on proliferative and steroid responses [15]. 2.2. Proliferation Assay Granulosa cells from 4- to 8-mm follicles were plated in 96-well plates at a density of 50,000 cells per well in a total volume of 100 ␮l of media for 24 h prior to addition of treatments. During each experiment, the following treatments were applied in triplicate: VEGF (recombinant human; R&D Systems, Minneapolis, USA; 1, 10, or 100 ng/mL), FSH (ovine; National Hormone and Peptide Program; Torrance, California, USA; 10 ng/mL) alone or in combination with VEGF (1 ng/mL), LH (ovine; National Hormone and Peptide Program; 10 ng/mL) or media alone. FSH and LH were used as positive and negative control treatments, respectively, as granulosa cells from bovine follicles 4 to 8 mm in diameter are known to be responsive to FSH but do not express LH receptors [16,17]. Following 72 h incubation with treatments and a single wash with pre-warmed culture media, cell pro-

L.K. Doyle et al. / Domestic Animal Endocrinology 38 (2010) 127–137

liferation was measured using Celltiter 96 Aqueous One Solution Cell Proliferation Assay (Promega, Southampton, UK) according to manufacture‘s instructions. This assay uses a MTS Tetrazolium compound that is bioreduced by cells into a coloured formazan product that is soluble in culture medium. This bio-reduction takes place in metabolically active cells, thereby allowing comparative analysis of viable cell numbers. Colorimetric analyses were done using a Wallac 1420 Multilabel Counter (Wallac, Turku, Finland). To correct for background absorbance a triplicate set of control wells containing culture media only were included and their absorbance was subtracted from the absorbance values of all other samples. 2.3. Western blotting analyses To determine the effects of VEGF on extracellular signal-regulated kinase 1/2 (ERK1/2) phosphorylation, granulosa cells from 4- to 8-mm follicles were plated at a density of 500,000 cells per well in 12-well plates and allowed to settle for 24 h. Cells were then starved in culture media without supplements for another 24 h, washed and incubated in duplicate wells for 20 min at 37 ◦ C with VEGF (1, 10, 100 ng/mL), EGF (recombinant human; Sigma, Dorset UK, 100 ng/mL) as a positive control, or media alone. The incubation time was chosen to provide optimum VEGF-induced ERK1/2 phosphorylation based on data from our laboratory (not shown) and others [18,19]. After incubation with treatments all cells were washed twice with ice cold PBS and lysed with 70 ␮l of RIPA buffer (150 mm NaCl, 50 mm Tris, 1 mm EDTA, 1% Nonidet P-40, 0.5% sodium deoxycholate, 0.1% sodium dodecyl sulfate, 1 mM Na3VO4, and 1 mM NaF, pH 7.4) by gentle rocking at 4 ◦ C for 45 min. Lysates were centrifuged for 5 min at 13,000 rpm to remove cell debris. Protein lysates were stored at −20 ◦ C until required for immunoblotting. To determine the effects of VEGFR2 inhibition on ERK1/2 activation, cells from 9- to 14-mm follicles were incubated for 1 h with a selective VEGFR2 inhibitor, ZM323881 (final concentration 50 nM, Merk Chemicals Ltd., Nottingham, UK) or media alone prior to addition of VEGF (100 ng/mL), LH (100 ng/mL), EGF (100 ng/mL) or media alone for 20 min after which cells were lysed and processed as described above. Based on results of previous studies [20,21] we preliminarily tested a range of ZM323881 concentrations (10 nm to 100 nM) from which a 50 nM dose was chosen that provided optimum inhibition of VEGF responses without apparent effects on cell viability (data not show).

129

Protein lysates were assayed for protein content using the DC protein assay kit (Bio-Rad Laboratories Inc., Hemel Hempstead, UK). Twenty micrograms of protein from each lysate were heat-denatured in reducing buffer (12% SDS, 40% glycerol, 30% ␤-mercaptoethanol, 300 mM DTT, 120 mM EDTA, 1 mg/mL bromophenol blue, 375 mM TrisCl, pH 6.8) for 5 min at 100 ◦ C before being electrophoresed on a 12% SDS-Polyacrylamide gel and then transferred onto PDVF membranes. Following transfer, membranes were blocked for 2 h in a solution (10 mM Tris, 100 mM NaCl, 0.2% Tween-20) containing 5% BSA and incubated overnight at room temperature with primary antibody diluted in the same solution but with 1% BSA. All primary antibodies were from Cell Signaling Technology Inc. (Beverly, MA, USA) and they were specific for phospho-p44/42 MAP Kinase (ERK1/2; working dilution, 1:3000), p44/42 MAP Kinase (total ERK1/2; 1;1000) or ␤-Tubulin (1:1000). After primary antibody incubation, membranes were washed with Tris Buffered Saline (TBS) and then incubated with Horseradish Peroxidase (HRP)conjugated anti-rabbit anti-immunoglobulin (1:10,000; Amersham Biosciences, Buckinghamshire, UK) for 1 h at room temperature with gentle rotation. Membranes were then washed in Nonidet P-40-Tris Buffered Saline (N-TBS)/TBS and immune complexes were visualized using Super Signal West Femto maximum sensitivity detection system (Pierce Chemical Inc., Rockford, IL, USA) and a Fluoro-S Scanning System (Bio-Rad). Densitometry analyses were done using Quantity One Analysis Software (Bio-Rad). Membranes were stripped and re-blotted to allow sequential detection of phosphorylated ERK1/2, total ERK1/2 and ␤-Tubulin. Stripping was done by rinsing membranes twice for 5 min in TBS and then washing for 2 h with constant agitation in stripping solution (0.2 M Glycine pH 2.0, 0.1% SDS, 0.1% Tween-20). Membranes were then washed twice in TBS for 10 min before they were incubated in blocking buffer for 2 h and subsequently re-probed with primary antibody. Detection of total ERK1/2 allowed normalisation of ERK/1 phosphorylation values across treatment groups whereas ␤-Tubulin was used as a loading control. 2.4. Quantitative PCR (qPCR) analyses Granulosa cells were plated in 12-well plates at a density of 500,000 cells per well and allowed to settle for 24 h. Cells were then incubated with treatments in duplicate wells for 48 h before RNA was harvested. For induction of CYP11A1, cells from 4- to 8-mm follicles were treated with media alone, VEGF (1, 10 or 100 ng/mL), FSH (10 ng/mL, a dose known to induce

130

L.K. Doyle et al. / Domestic Animal Endocrinology 38 (2010) 127–137

CYP11A1 expression in these cells [16]) alone or in combination with VEGF (1 ng/mL), or LH (10 ng/mL). For induction of PTGS2, granulosa cells from 9- to 14-mm follicles were incubated with media alone, VEGF (1, 10 or 100 ng/mL) or LH (100 ng/mL, a dose shown to effectively induce PTGS2 expression in these cells [14]) alone or in combination with VEGF (1 ng/mL). In some experiments, cells were pre-incubated with VEGFR2 inhibitor (ZM323881) for 1 h prior to addition of treatments. Total RNA was extracted using RNeasy Mini Kit (Quiagen, Sussex, UK). RNA concentration and quality were determined using a NanoDrop (ND-1000) spectrophotometer (NanoDrop Technologies Inc, Wilmington, DE, USA). For first strand cDNA synthesis, random primers (Promega) and SuperScript III Reverse Transcriptase (Invitrogen Corp., Carlsbad, CA, USA) were used and the resulting cDNA was stored at −20 ◦ C until required for qPCR. Oligonucleotide primers for qPCR of CYP11A1 (forward, 5 -AGA-GAA-TCC-ACT-TTC-GCC-ACA-TC-3 and reverse, 5 -GGT-CTT-TCT-TCC-AGG-TTC-CTGAC-3 ), PTGS2 (5 -TCC-TGA-AAC-CCA-CTC-CCAA-3 and 5 -TGG-GCA-GTC-ATC-AGG-CAC-AG-3 ) and 18S (5 -GGG-GAA-TCA-GGG-TTC-G-3 and 5 GCT-GGC-ACC-AGA-CTT-G-3 ) were designed from bovine-specific sequences using LightCycler Probe Design software (Roche Diagnostics Ltd., Burgess Hill, UK) and their specificity was confirmed by semiquantitative PCR using freshly collected granulosa cells. QPCR was performed using Quantace SensiMix dT kit (Quantace, London, UK) and a Stratagene Mx3000p real-time PCR system (Stratagene, La Jolla, California, USA). A set of cDNA standards was used from a pool of freshly collected granulosa cells. Each unknown sample was run in duplicate at a 1:20 dilution (for 18S) or a 1:4 dilution (for CYP11A1 and PTGS2). The same PCR reaction settings (10 min at 95 ◦ C followed by 40 cycles of 15 s at 95 ◦ C, 30 s at 60 ◦ C and 30 s at 72 ◦ C) were used for all genes. QPCR results were analyzed using Mx3000p software. A measure of transcript abundance in each sample was obtained from the corresponding Ct value by extrapolation from a Ct-log plot obtained simultaneously from the standard curve. To correct for differences in total RNA abundance between samples, the values for CYP11A1 and PTGS2 were divided by the corresponding value for 18S within each sample. 2.5. Statistical Analyses For cell proliferation data, percentage increases over untreated controls were calculated in each experiment

and used for statistical analyses. For any other endpoints, responses to treatments were expressed as fold-increase over values in untreated controls, unless specifically indicated otherwise. Dixon’s tests were used to identify suspected outlier values (P < 0.01) within data sets. These values were then excluded from subsequent analyses. The Kolmogorov-Smirnov normality test was applied to each data set (P < 0.01). Data that were not normally distributed were log transformed. Data were then analyzed by the General Linear Model procedure using Minitab 15 statistical software package (Minitab Ltd. Coventry, UK) to determine main treatment effects using each individual experiment as a block. If a main effect was significant (P < 0.05) pair-wise multiple comparisons were performed using Tukey’s test to establish differences between treatment means and untreated controls. Comparisons involving only two means were done by unpaired t-tests. Each experiment was replicated at least 3 times, as indicated in Figure legends. 3. Results 3.1. Effects of VEGF on granulosa cell proliferation and ERK1/2 phosphorylation 3.1.1. Cell proliferation A dose-dependent effect of VEGF treatment on proliferation of granulosa cells from 4- to 8-mm follicles was observed (Fig. 1a). On average, granulosa cell proliferation was significantly induced by 1 ng/mL VEGF (24% increase over untreated controls; P = 0.024) but there were no detectable responses to 10 or 100 ng/mL VEGF (P = 0.9). As expected, FSH induced a significant proliferative response (40%; P < 0.0001) whereas LH did not (P = 0.9). Co-administration of FSH and VEGF (1 ng/mL) resulted in a cell proliferation response that was 2.4-fold greater than that elicited by FSH alone (P = 0.045). 3.1.2. ERK1/2 phosphorylation Exposure of cultured bovine granulosa cells to 1, 10 or 100 ng/mL VEGF resulted in a progressive mean increase in ERK1/2 phosphorylation (Fig. 1b). Although individual mean responses to VEGF were not significantly greater than the control mean (P = 0.2), the overall response to the three VEGF doses combined was significantly greater (P = 0.02). The highest mean response to VEGF was induced by the 100 ng/mL dose (2.7-fold over untreated controls) which was, on average, 47% the response induced by 100 ng/mL EGF.

L.K. Doyle et al. / Domestic Animal Endocrinology 38 (2010) 127–137

131

Fig. 1. a) Proliferation response (mean % ± SE; n = 12 experiments) of bovine granulosa cells from 4- to 8-mm follicles to different VEGF doses (1, 10 and 100 ng/mL), FSH (10 ng/mL), a combination of FSH (10 ng/mL) and VEGF (1 ng/mL), and LH (10 ng/mL). There was a main effect of treatment (P < 0.0001). b) Representative immunoblots and mean (± SE) levels of ERK1/2 phosphorylation after treatment of bovine granulosa cells from follicles 4 to 8 mm with different doses of VEGF (1, 10 and 100 ng/mL) or EGF (100 ng/mL; n = 6 experiments). There was a main effect of treatment (P = 0.002). Within each graph, means with different superscripts (a, b, c) are different (P < 0.05).

3.2. Effects of VEGF on expression of genes associated with steroidogenesis and ovulation

3.3. Effects of inhibition of VEGFR2 on LH-induced ERK1/2 phosphorylation and PTGS2 expression

3.2.1. CYP11A1 Expression of CYP11A1 in cultured bovine granulosa cells collected from follicles 4 to 8 mm did not change in response to VEGF concentrations of 1, 10 or 100 ng/mL (P = 0.9; Fig. 2). In contrast, FSH (P < 0.001), but not LH (P = 0.2) significantly increased the expression of CYP11A1. The response to FSH (about 2.8-fold relative to controls) was not significantly affected by co-stimulation with VEGF (1 ng/mL; P = 0.9).

3.3.1. LH-induced ERK1/2 phosphorylation The effects of treatment with LH, VEGF or EGF in the presence or absence of the VEGFR2 inhibitor,

3.2.2. PTGS2 There was a dose-dependent effect of VEGF treatment on PTGS2 expression by granulosa cells from follicles 9 to 14 mm (Fig. 3). Specifically, PTGS2 expression levels increased, relative to untreated controls, with the 1 ng/mL VEGF dose (approximately 1.8 fold; P = 0.003) but not with the 10 or 100 ng/mL VEGF doses (P = 0.9). The response to 1 ng/mL VEGF was similar to that induced by LH. However, costimulation of cells with LH and VEGF (1 ng/mL) did not increase PTGS2 expression relative to untreated controls (P = 0.9).

Fig. 2. CYP11A1 expression (mean ± SE; n = 13 experiments) in response to different doses of VEGF (1, 10 and 100 ng/mL), FSH (10 ng/mL), LH (10 ng/mL) or a combination of FSH (10 ng/mL) and VEGF (1 ng/mL) in cultured granulosa cells from 4 -to 8-mm follicles. There was an overall effect of treatment (P < 0.0001). Means with different superscripts (a, b) are different (P < 0.05).

132

L.K. Doyle et al. / Domestic Animal Endocrinology 38 (2010) 127–137

ZM32388, on ERK1/2 phosphorylation in granulosa cells from 9- to 14-mm follicles are shown (Fig. 4). LH induced an increase in ERK1/2 phosphorylation in the absence (P = 0.003) but not in the presence (P = 0.2) of VEGFR2 inhibitor (Fig. 4a). As expected, treatment with VEGF induced an increase in ERK1/2 phosphorylation (P < 0.01) but this increase was abolished (P = 0.2) when cells were preincubated with VEGFR2 inhibitor (Fig. 4b). In contrast, treatments with EGF induced an increase in ERK1/2 phosphorylation both in the absence (P < 0. 01) and the presence (P < 0. 01) of VEGFR2 inhibitor (Fig. 4c). Fig. 3. PTGS2 expression (mean ± SE; n = 8 experiments) in response to different doses of VEGF (1, 10 and 100 ng/mL), LH (100 ng/mL) or a combination of LH (100 ng/mL) and VEGF (1 ng/mL) by cultured bovine granulosa cells from 9- to 14-mm follicles. A main effect of treatment was significant (P < 0.002). Means with different superscripts (a, b) are different (P < 0.05).

3.3.2. LH-induced PTGS2 expression An LH-induced increase in PTGS2 expression in granulosa cells from 9- to 14-mm follicles (P < 0.01) was abolished when cells were pre-incubated with VEGFR2 inhibitor (P = 0.9; Fig. 5a). In contrast, VEGFR2 inhibition did not affect (P = 0.9) the induction of CYP11A1

Fig. 4. Representative immunoblots and levels of ERK1/2 phosphorylation (mean ± SE, n = 3 experiments) in granulosa cells from 9- to 14-mm follicles exposed to a) LH (100 ng/mL), b) VEGF (100 ng/mL) or c) EGF (100 ng/mL) in the presence or absence of the VEGFR2 inhibitor, ZM323881 (50 nM). For each experiment, ERK1/2 phosphorylation values for each group were calculated relative to the value for the group treated with VEGF, LH or EGF alone, as shown in each graph. A main effect of treatment was significant (P < 0.0001) in all cases. Within each graph, means with different superscripts (a, b) are different (P < 0.05).

L.K. Doyle et al. / Domestic Animal Endocrinology 38 (2010) 127–137

Fig. 5. a) PTGS2 expression (mean ± SE; n = 3 experiments) in granulosa cells from 9- to 14-mm follicles exposed to LH (100 ng/mL) in the absence or presence of the VEGFR2 inhibitor, ZM323881 (50 nM). There was an overall effect of treatment (P = 0.001). b) CYP11A1 expression (mean ± SE; n = 3 experiments) in granulosa cells from 4to 8-mm follicles exposed to FSH (10 ng/mL) in the absence or presence of ZM323881 (50 nM). There was an overall effect of treatment (P = 0.009). Within each graph, means with different superscripts (a, b) are different (P < 0.05).

by FSH in granulosa cells from 4- to 8-mm follicles (Fig. 5b). 4. Discussion Treatment of primates with VEGF antagonists has demonstrated that VEGF is an absolute requirement for the completion of follicle development [4]. The observations that 1) the avascular granulosa cell layer is a major site of VEGF production within the follicle [4], 2) these cells express VEGF receptors [6], and 3) the expression of both VEGF and VEGFR2 in granulosa cells markedly increase in selected (dominant) follicles [6,22] led us to more precisely investigate in vitro the direct effects that physiological levels of VEGF, alone or in combination with gonadotropins, may have on specific responses of granulosa cells collected from late-antral stage follicles. Proliferative and steroidogenic responses to VEGF were investigated using granulosa cells from follicles with diameters (4 to 8 mm) corresponding to shortly before the beginning of dominance during a bovine follicular wave [1], whereas potential effects of VEGF on

133

the peri-ovulatory increase in PTGS2 expression were assessed in cells collected from dominant-size follicles (9 to 14 mm). Follicle status in this study was determined solely on the basis of morphological appearance [23] and therefore granulosa cell populations, particularly for the 4- to 8-mm category, likely included cells not only from growing follicles but also from follicles in early stages of atresia. However, in agreement with previous studies using the same bovine cell culture system [15], the consistent responses to gonadotropin treatment in terms of proliferation, CYP11A1 expression and PTGS2 expression indicated that these populations were functionally representative of those found during the development of a follicular wave and therefore adequate for the purposes of this study. VEGF concentrations in bovine follicular fluid were reported to be about 1 ng/mL in follicles 5 to 7 mm in diameter and concentrations increased to 5 ng/mL in 12to 14-mm follicles [24]. In another study [25] mean intrafollicular VEGF concentrations in estrogen-active follicles with a mean diameter of 8.6 mm (corresponding to the beginning of dominance) were 1.2 ng/mL, whereas estrogen-inactive follicles (mean diameter, 11.3 mm) had a mean VEGF concentration of 0.5 ng/mL. In comparison, VEGF levels in follicular fluid from porcine preovulatory follicles have been reported to range between 3 and 10 ng/mL [26,27]. Thus, the doses of VEGF used in this study (1, 10 and 100 ng/mL) were chosen to encompass the expected range of concentrations found within the developing bovine dominant follicle. The finding in our study that VEGF dose-dependently induced proliferation of bovine granulosa cells is consistent with results with porcine granulosa cells [5] and with a variety of other non-endothelial cell types (reviewed in [11]). In the present study, a concentration of VEGF of 1 ng/mL, which is close to levels naturally found in early bovine dominant follicles, resulted in a significant proliferative response of granulosa cells; whereas VEGF concentrations of 10 or 100 ng/mL abolished all proliferative effects. Previous studies investigating bovine granulosa cell responses to VEGF only considered VEGF concentrations of 50 ng/mL or higher [6,9] and found that VEGF increased the survival of granulosa cells but had no effects on cell proliferation. We speculate that the failure to observe any proliferative effects in those studies was due to the exposure of cells to VEGF levels above the range of concentrations that would be expected to stimulate cell proliferation. Coadministration of FSH and VEGF (1 ng/mL) resulted in a 2.4-fold increase in cell proliferation relative to FSH treatment alone, thus indicating that VEGF potentiates the effects of FSH. A synergistic effect of FSH and VEGF

134

L.K. Doyle et al. / Domestic Animal Endocrinology 38 (2010) 127–137

was also reported in relation to bovine cumulus-oocyte complexes [24]. Any conclusions on the physiological relevance of the present results need to be taken with caution as potential interactions of VEGF with known follicular mitogens, such as estrogens, were not considered, however, we can speculate that the differential increases in VEGF and VEGF receptor expression in granulosa cells during follicle selection [22,24] may act in combination with FSH to enhance cell proliferation thus promoting the continuous growth of the dominant follicle in a similar manner to what has been shown for other growth factors such as IGF-1 [2]. Presumably, once VEGF concentrations (and VEGFR2 expression in granulosa cells) reach their highest levels in preovulatory follicles, a stimulatory effect on cell proliferation may no longer occur, a suggestion that is in keeping with the cease in cell proliferation associated with the terminal differentiation of follicular cells during the periovulatory period [28]. Since growth factors typically induce their mitogenic effects through activation of ERK1/2 [29] we investigated whether this pathway was activated by VEGF in granulosa cells. The association between ERK1/2 phosphorylation and cell proliferation in response to 1 ng/mL VEGF was consistent with the role of ERK1/2 in cell proliferation responses. However, higher concentrations of VEGF continued to induce ERK1/2 activation despite their inhibitory effects on cell proliferation. The reason for this apparent dissociation between ERK1/2 activation and proliferative responses to VEGF is not known but may be explained by the reported roles of ERK1/2 signaling in the regulation of both proliferation and differentiation of granulosa cells [30]. Results of previous studies suggest that the dose-dependent effects of LH on proliferation and differentiation responses by these cells may involve specific time-related differences in ERK1/2 phosphorylation patterns induced by different concentrations of LH [31]. Whether a similar mechanism may account for the observed effects of different doses of VEGF in the present study should be determined in the future by more precisely characterising ERK1/2 responses following VEGF stimulation. We then examined the potential involvement of VEGF in another granulosa cell response associated with follicle selection, CYP11A1 expression. A predominant feature of the selection process is a pronounced increase in steroid production by the dominant follicle [2]. In cattle, this increase is largely accounted for by an increase in the expression of CYP11A1 in granulosa and theca cells [13]. In the present study, in contrast to FSH, VEGF did not induce CYP11A1 expression at any concentration tested. The physiological significance of this finding

cannot be fully ascertained because estradiol production in response to VEGF was not measured. In a previous study, VEGF induced changes in estradiol production by porcine granulosa cells in a follicle size-dependent manner [5], however, whether those changes were associated with changes in CYP11A1 expression was not analyzed. The precise role of VEGF in follicular cell steroidogenesis and whether such a role varies in different species warrants further investigation. To identify wider roles of VEGF in the development of dominant follicles, we also examined the effects on a well-characterized granulosa cell response associated with ovulation, PTGS2 expression. PTGS2 is a component of the temporal inflammatory response associated with ovulation and as such is only transiently induced in response to the ovulatory gonadotropin surge, and this occurs both in vivo and in vitro [14]. VEGF has been shown to be involved in upregulation of PTGS2 in endothelial cells [32] and non-endothelial cells including bovine oviductal cells [33]. Interestingly, PTGS2 increased in response to treatment with 1 ng/mL of VEGF in the present study suggesting that VEGF, in addition to LH, may mediate the follicular inflammatory response associated with ovulation. The reason for the absence of a PTGS2 response in cells simultaneously stimulated with VEGF and LH is not known. A previous study failed to show any synergistic effects of LH on VEGF-induced PTGS2 by bovine oviductal cells [33]. Although the present finding on the combined effects of LH and VEGF on PTGS2 expression by granulosa cells from dominant-size follicles was not further explored, it is conceivable that overstimulation of PTGS2-inducing cell pathways in these cells may have a negative effect on PTGS2 expression. This possibility, which should be tested in the future by performing appropriate dose-response studies, would be consistent with a physiological need for fine-tuned intracellular mechanisms that ensure that the peri-ovulatory increase in PTGS2 is only transient. The result of similar induction of PTGS2 expression in cells exposed to VEGF or LH alone in the present study raised the possibility of a potential interplay between VEGF and LH in preovulatory follicles similar to that described between LH and the EGF network [3]. In particular, EGF receptor activation has been shown to mediate the LH induction of specific ovulatoryassociated responses, among them PTGS2 expression. To address that possibility, we evaluated the effects of specific VEGFR2 inhibition with ZM323881 on LH induction of PTGS2 in granulosa cells. The effects on LH-induced activation of the ERK1/2 pathway, which has been shown to mediate PTGS2 expression in gran-

L.K. Doyle et al. / Domestic Animal Endocrinology 38 (2010) 127–137

ulosa cells [34], were also investigated. We found that addition of ZM323881 abolished the stimulatory effects of LH on ERK1/2 phosphorylation in granulosa cells. In addition, inhibition of VEGFR2 prevented LH from inducing PTGS2 expression in these cells. The effects of ZM323881 on LH responses were attributable to specific inhibition of VEGFR2 because although ZM323881 simultaneously and efficiently prevented the activation of ERK1/2 by VEGF it had no effect on ERK1/2 activation by EGF. This was expected as the concentration of VEGFR2 inhibitor used was over 1000-fold lower than the IC50 for inhibition of other receptor tyrosine kinases [>50 uM, [35]]. In addition, VEGFR2 inhibition had no effect on the CYP11A1 response of granulosa cells to FSH, as indicated by similar gene expression in the presence and absence of inhibitor, a result that would be consistent with the specificity, at least under our experimental conditions, of the apparent interaction between the LH and VEGF signalling systems. Although the present results provide preliminary evidence that the VEGF system may be involved in mediating the LHinduced periovulatory increase in PTGS2 expression, future studies will be needed to confirm these findings and to address the mechanistic basis of any potential LH-VEGF interactions. A more definitive conclusion on the extent of VEGF mediation of LH actions during ovulation will require analysing additional LH-induced peri-ovulatory responses such as progesterone synthesis. Of interest, in a recent study VEGF was shown to mediate the induction of cell-matrix interaction proteins in human luteinized granulosa cells by human chorionic gonadotropin [10], consistent with the suggested involvement of the VEGF system in LH-induced periovulatory events. A potential mechanism by which VEGFR2 may mediate LH-induced PTGS2 responses in granulosa cells may involve, at least partly, an increase in VEGF expression induced by LH [36]. This mechanism, however, may not explain the observed effects of VEGFR2 on more immediate (within 20 min) cell responses to LH (ERK1/2 phosphorylation). Studies with rat granulosa cells showed that the activation of another receptor tyrosine kinase, the EGF receptor, was necessary for an immediate gonadotropin-induced increase in ERK1/2 phosphorylation and that this effect involved the action of metalloproteinases (MMPs) that induce the release of EGF receptor ligand from extracellular matrix (ECM) stores [31]. Similar to EGF receptor ligands, VEGF bioactivity in tissues is regulated through MMP cleavage of VEGF from ECM stores that results in an increase in smaller (bioactive) VEGF isoforms that can freely access their cognate receptors in the cell surface [12]. It

135

would therefore be of interest to investigate the effects of MMP inhibitors on LH transactivation of VEGFR2 signaling. Additionally or alternatively, LH-mediated VEGFR2 activation may involve intracellular signals such as Src tyrosine kinases which have been shown to be activated by gonadotropins during granulosa cell differentiation [37]. In summary, VEGF at levels (1 ng/mL) similar to those found in developing dominant follicles acted to induce proliferation of bovine granulosa cells from 4 -to 8-mm follicles and to enhance the effect of FSH on cell proliferation. The stimulatory effect of VEGF on proliferation was associated with an increase in intracellular ERK1/2 activation. However, VEFG had no effects on the expression of the steroidogenic enzyme, CYP11A1, by these cells. Treatment of granulosa cells from 9- to 14-mm follicles with 1 ng/mL of VEGF up-regulated the expression of PTGS2 to levels similar to those induced by treatment with LH. In addition, pre-treatment with a selective VEGFR2 inhibitor blocked the ability of LH to induce ERK1/2 phosphorylation and to stimulate PTGS2 expression in those cells. The results of this study suggest potentially important roles of VEGF, independent of its angiogenic effects, in the regulation of granulosa cell function at different stages during the development of the dominant follicle. Studies should follow to confirm the proposed novel role of VEGFR2 in mediating peri-ovulatory responses induced by the gonadotropin surge similar to what has been described in relation to EGF receptors. Acknowledgements We thank Charis Hogg for excellent technical assistance. This study was supported by a Royal (Dick) School of Veterinary Studies scholarship to L.K. Doyle (E06008). References [1] Ginther OJ, Beg MA, Donadeu FX, Bergfelt DR. Mechanism of follicle deviation in monovular farm species. Anim Reprod Sci 2003;78(3–4):239–57. [2] Beg MA, Ginther OJ. Follicle selection in cattle and horses: role of intrafollicular factors. Reproduction 2006;132(3):365–77. [3] Conti M, Hsieh M, Park J-Y, Su Y-Q. Role of the epidermal growth factor network in ovarian follicles. Mol Endocrinol 2006;20(4):715–23. [4] Fraser HM, Wulff C. Angiogenesis in the primate ovary. Reprod Fertil Dev 2001;13(8):557–66. [5] Grasselli F, Basini G, Bussolati S, Tamanini C. Effects of VEGF and bFGF on proliferation and production of steroids and nitric oxide in porcine granulosa cells. Reprod Domest Anim 2002;37(6):362–8.

136

L.K. Doyle et al. / Domestic Animal Endocrinology 38 (2010) 127–137

[6] Greenaway J, Connor K, Pedersen HG, Coomber BL, LaMarre J, Petrik J. Vascular endothelial growth factor and its receptor, Flk-1/KDR, are cytoprotective in the extravascular compartment of the ovarian follicle. Endocrinology 2004;145(6):2896–905. [7] Danforth DR, Arbogast LK, Ghosh S, Dickerman A, Rofagha R, Friedman CI. Vascular endothelial growth factor stimulates preantral follicle growth in the rat ovary. Biol Reprod 2003;68(5):1736–41. [8] Shimizu T, Iijima K, Miyabayashi K, Ogawa Y, Miyazaki H, Sasada H, Sato E. Effect of direct ovarian injection of vascular endothelial growth factor gene fragments on follicular development in immature female rats. Reproduction 2007;134(5):677–82. [9] Kosaka N, Sudo N, Miyamoto A, Shimizu T. Vascular endothelial growth factor (VEGF) suppresses ovarian granulosa cell apoptosis in vitro. Biochem Biophys Res Commun 2007;363(3): 733–7. [10] Rolaki A, Coukos G, Loutradis D, DeLisser HM, Coutifaris C, Makrigiannakis A. Luteogenic hormones act through a vascular endothelial growth factor-dependent mechanism to upregulate ␣5␤1 and ␣v␤3 integrins, promoting the migration and survival of human luteinized granulosa cells. Am J Pathol 2007;170(5):1561–72. [11] Duffy AM, Bouchier-Hayes DJ, Harmey JH. Vascular endothelial growth factor (VEGF) and its role in non-endothelial cells: autocine signaling by VEGF. In: Harmey JH, editor. VEGF and Cancer. Eurekah.com and Kluwer Academic/Plenum Publishing; 2004. p. 133–44. [12] Park JE, Keller GA, Ferrara N. The vascular endothelial growth factor (VEGF) isoforms: differential deposition into the subepithelial extracellular matrix and bioactivity of extracellular matrix-bound VEGF. Mol Biol Cell 1993;4(12):1317–26. [13] Tian X, Berndtson A, Fortune J. Differentiation of bovine preovulatory follicles during the follicular phase is associated with increases in messenger ribonucleic acid for cytochrome P450 side-chain cleavage, 3 beta-hydroxysteroid dehydrogenase, and P450 17 alpha-hydroxylase, but not P450 aromatase. Endocrinology 1995;136(11):5102–10. [14] Bridges PJ, Komar CM, Fortune JE. Gonadotropin-induced expression of messenger ribonucleic acid for cyclooxygenase-2 and production of prostaglandins E and F2␣ in Bovine preovulatory follicles are regulated by the progesterone receptor. Endocrinology 2006;147(10):4713–22. [15] Gutierrez C, Campbell B, Webb R. Development of a longterm bovine granulosa cell culture system: induction and maintenance of estradiol production, response to follicle- stimulating hormone, and morphological characteristics. Biol Reprod 1997;56(3):608–16. [16] Silva J, Price C. Insulin and IGF-I are necessary for FSH-induced cytochrome P450 aromatase but not cytochrome P450 side-chain cleavage gene expression in oestrogenic bovine granulosa cells in vitro. J Endocrinol 2002;174(3):499–507. [17] Bao B, Garverick HA. Expression of steroidogenic enzyme and gonadotropin receptor genes in bovine follicles during ovarian follicular waves: a review. J Anim Sci 1998;76(7):1903–21. [18] Shu X, Wu W, Mosteller RD, Broek D. Sphingosine kinase mediates vascular endothelial growth factor-induced activation of Ras and mitogen-activated protein kinases. Mol Cell Biol 2002;22(22):7758–68. [19] Rafiee P, Heidemann J, Ogawa H, Johnson N, Fisher P, Li M, Otterson M, Johnson C, Binion D, Cyclosporin. A differentially inhibits multiple steps in VEGF induced angiogenesis in human

[20]

[21]

[22]

[23]

[24]

[25]

[26]

[27]

[28]

[29]

[30]

[31]

[32]

[33]

microvascular endothelial cells through altered intracellular signaling. Cell Communication and Signaling 2004;2(1):3. Kim DS, Franklyn JA, Boelaert K, Eggo MC, Watkinson JC, McCabe CJ. Pituitary Tumor transforming gene (PTTG) stimulates thyroid cell proliferation via a vascular endothelial growth factor/kinase insert domain receptor/inhibitor of DNA binding-3 autocrine pathway. J Clin Endocrinol Metab 2006;91(11):4603–11. Xiao Z, Kong Y, Yang S, Li M, Wen J, Li L. Upregulation of Flk-1 by bFGF via the ERK pathway is essential for VEGFmediated promotion of neural stem cell proliferation. Cell Res 2007;17(1):73–9. Ginther OJ, Gastal EL, Gastal MO, Checura CM, Beg MA. Doseresponse study of intrafollicular injection of insulin-like growth factor-I on follicular fluid factors and follicle dominance in mares. Biol Reprod 2004;70(4):1063–9. Yang MY, Rajamahendran R, Morphological. Biochemical identification of apoptosis in small, medium, and large bovine follicles and the effects of follicle-stimulating hormone and insulin-like growth factor-I on spontaneous apoptosis in cultured bovine granulosa cells. Biol Reprod 2000;62(5):1209– 17. Einspanier R, Schönfelder M, Müller K, Stojkovic M, Kosmann M, Wolf E, Schams D. Expression of the vascular endothelial growth factor and its receptors and effects of VEGF during in vitro maturation of bovine cumulus-oocyte complexes (COC). Mol Reprod Dev 2002;62(1):29–36. Grazul-Bilska AT, Navanukraw C, Johnson ML, Vonnahme KA, Ford SP, Reynolds LP, Redmer DA. Vascularity and expression of angiogenic factors in bovine dominant follicles of the first follicular wave. J Anim Sci 2007;85(8):1914–22. Barboni B, Turriani M, Galeati G, Spinaci M, Bacci ML, Forni M, Mattioli M. Vascular endothelial growth factor production in growing pig antral follicles. Biol Reprod 2000;63(3):858–64. Galeati G, Spinaci M, Govoni N, Zannoni A, Fantinati P, Seren E, Tamanini C. Stimulatory effects of fasting on vascular endothelial growth factor (VEGF) production by growing pig ovarian follicles. Reproduction 2003;126(5):647–52. Robker RL, Richards JS. Hormonal control of the cell cycle in ovarian cells: proliferation versus differentiation. Biol Reprod 1998;59(3):476–82. Byrne AM, Bouchier-Hayes DJ, Harmey JH. Angiogenic and cell survival functions of vascular endothelial growth factor (VEGF). J Cell Mol Med 2005;9(4):777–94. Su Y-Q, Nyegaard M, Overgaard MT, Qiao J, Giudice LC. Participation of mitogen-activated protein kinase in luteinizing hormone-induced differential regulation of steroidogenesis and steroidogenic gene expression in mural and cumulus granulosa cells of mouse preovulatory follicles. Biol Reprod 2006;75(6):859–67. Andric N, Ascoli M. A delayed gonadotropin-dependent and growth factor-mediated activation of the extracellular signal-regulated kinase 1/2 cascade negatively regulates aromatase expression in granulosa cells. Mol Endocrinol 2006;20(12):3308–20. Tamura M, Sebastian S, Gurates B, Yang S, Fang Z, Bulun SE. Vascular endothelial growth factor up-regulates cyclooxygenase2 expression in human endothelial cells. J Clin Endocrinol Metab 2002;87(7):3504–7. Wijayagunawardane MPB, Kodithuwakku SP, Yamamoto D, Miyamoto A. Vascular endothelial growth factor system in the cow oviduct: a possible involvement in the regulation

L.K. Doyle et al. / Domestic Animal Endocrinology 38 (2010) 127–137 of oviductal motility and embryo transport. Mol Reprod Dev 2005;72(4):511–20. [34] Shimada M, Hernandez-Gonzalez I, Gonzalez-Robayna I, Richards JS. Paracrine and autocrine regulation of epidermal growth factor-like factors in cumulus oocyte complexes and granulosa cells: key roles for prostaglandin synthase 2 and progesterone receptor. Mol Endocrinol 2006;20(6):1352–65. [35] Whittles CE, Pocock TM, Wedge SR, Kendrew J, Hennequin LF, Harper SJ, Bates DO. ZM323881, a novel inhibitor of vascular endothelial growth factor-receptor-2 tyrosine kinase activity. Microcirculation 2002;9(6):513–22.

137

[36] Christenson LK, Stouffer RL. Follicle-stimulating hormone and luteinizing hormone/chorionic gonadotropin stimulation of vascular endothelial growth factor production by macaque granulosa cells from pre- and periovulatory follicles. Journal of Clinical Endocrinology Metabolism 1997;82(7):2135–42. [37] Wayne CM, Fan H-Y, Cheng X, Richards JS. Follicle-stimulating hormone induces multiple signaling cascades: evidence that activation of rous sarcoma oncogene, ras, and the epidermal growth factor receptor are critical for granulosa cell differentiation. Mol Endocrinol 2007;21(8):1940–57.