Thrombospondin and Vascular Endothelial Growth Factor Are

Download PDF
0 downloads 0 Views 663KB Size Report
Comment
Dec 22, 2004 - 2002; 21:7945–7956. 40. Fontanini G, Boldrini L, ... ic treatment with thrombospondin-1 enhances primary tumour radia- tion response and ...
BIOLOGY OF REPRODUCTION 72, 1071–1078 (2005) Published online before print 22 December 2004. DOI 10.1095/biolreprod.104.031120

Thrombospondin and Vascular Endothelial Growth Factor Are Cyclically Expressed in an Inverse Pattern During Bovine Ovarian Follicle Development1 James Greenaway,3 Patricia A. Gentry,3 Jean-Jacques Feige,4 Jonathan LaMarre,3 and Jim J. Petrik2,3 Department of Biomedical Sciences,3 University of Guelph, Guelph, Ontario, Canada N1G 2W1 INSERM EMI 01–05, 4 DRDC/Angio, CEA-Grenoble, 38054 Grenoble Cedex 9, France ABSTRACT Angiogenesis does not normally occur in most adult tissues. However, in the ovary, there are cyclical vascular changes including angiogenesis that involve the interaction of numerous cytokines and growth factors. Angiogenic processes are regulated by a balance between pro- and antiangiogenic factors. The purpose of this study was to determine the expression of the antiangiogenic thrombospondin family and proangiogenic vascular endothelial growth factor (VEGF) in various sizes of healthy bovine follicles. Ovaries were collected from slaughterhouse animals and healthy follicles were sorted based on size (,0.5 cm, small; 0.5–1.0 cm, medium; .1.0 cm, large). Thrombospondin (TSP) protein levels were significantly higher in small follicles. Immunohistochemistry confirmed the granulosa layer as the primary area within the follicle involved in TSP generation and that small follicles had the highest proportion of immunopositive cells. TSP-1 and -2 mRNA levels were significantly higher in small follicles than either medium or large follicles. TSP colocalized with CD36 on granulosa cells (GC) in the follicle and in cultured cells. In contrast with TSP, VEGF expression increased during growth and development of the follicle. FSH stimulated GC expression of TSP, while LH had no effect. In summary, TSP-1 and -2 were coordinately expressed in the extravascular compartment of the ovary during early follicle development. VEGF was inversely expressed, with expression increasing as follicles developed. Regulated expression and localization of these proteins suggests that they may be involved in regulating growth and development of the follicle in a novel fashion.

follicle, follicular development, granulosa cells, growth factors, ovary

INTRODUCTION

Although organ vasculature is continuously developed and organized during the neonatal and early postnatal period, in the adult, these processes are normally rare, except in pathological situations such as wound healing and inflammation. The mammalian ovary, however, has special properties in that it undergoes rapid cyclical remodeling throughout the reproductive cycle, which involves numerous tightly controlled cellular and hormonal events [1]. Angiogenesis, the formation of new capillaries from existing Supported by the Lalor Foundation, the Natural Sciences and Engineering Research Council of Canada, and the Ontario Ministry of Food and Rural Affairs. The research collaboration in this study was supported by a fellowship from the OECD (J.L.). 2 Correspondence FAX: 519 767 1450; e-mail: [email protected] 1

Received: 21 April 2004. First decision: 13 May 2004. Accepted: 16 December 2004. Q 2005 by the Society for the Study of Reproduction, Inc. ISSN: 0006-3363. http://www.biolreprod.org

blood vessels, is a complex process that occurs throughout follicular development and formation of the corpus luteum [2–4]. In the bovine ovarian cycle, a number of primordial follicles are selected to enter the cycle and develop into small antral follicles. The majority of these follicles will undergo atresia, with usually only one progressing to dominance and ovulation [5, 6]. After ovulation, theca and granulosa cells differentiate into steroidogenically active lutein cells to form the corpus luteum, which secretes progesterone to facilitate pregnancy (for review, see [7]). Research in several mammalian species indicates that, throughout this cycle, there are numerous vascular changes within the follicle that facilitate growth of the follicle, nutrient delivery, and hormone transport to support development of the oocyte [8, 9]. Follicular events are mediated by the cocoordinated expression of both extrafollicular hormones [10] and intrafollicular factors produced in an autocrine and paracrine fashion by theca and granulosa cells. Peptide growth factors such as insulin-like growth factor (IGF)-I and -II [11], transforming growth factor- b (TGF b ) [12], epidermal growth factor (EGF) [13], and vascular endothelial growth (VEGF) [14–16], have been shown to participate in growth and differentiation of cells during folliculogenesis. Factors within the extracellular matrix (ECM) have also been reported to be important in the control of follicular development and atresia. Included in this group of ECM proteins are thrombospondin (TSP)-1 and -2 [17, 18]. The TSPs are a family of five multidomain glycoproteins, of which TSP1 and -2 are grouped in the same class and have high structural homology. TSP-3 to -5 belong to a separate isoform class, with significant structural differences and variation in the procollagen homology region compared with TSP-1 and -2. Considerable work has been done on the role of TSP-1 in vascular biology. TSP-1 and -2 function in cell-cell and cell-matrix interactions and are potent inhibitors of angiogenesis (for review, see [19]). The cellular effects of the TSPs are mediated through interaction with cell-surface receptors CD36 and Integrin Associated Protein (IAP, also known as CD47) [20–22]. In the ovary specifically, isolated granulosa cells have been shown to produce TSP-1, which in turn binds to the cell surface [23]. Using immunocytochemistry and reverse transcription-polymerase chain reaction, IAP and TSP-1 have been localized to human granulosa and large luteal cells, suggesting that the ligand and receptor may contribute to autocrine control of follicle cell function [24]. We have shown previously that TSP-1 and -2 are expressed in the rodent ovary and that their expression patterns are distinct [25]. For example, TSP-1 and CD36 are coordinately expressed in the early part of the rodent ovarian cycle, whereas TSP-2 is present in the late preovulatory, early luteal phase of the rodent cycle. TSP is thought to contribute to

1071

1072

GREENAWAY ET AL.

the regulation of angiogenesis as well as follicular and luteal development in the rodent ovary [26]. VEGF is a potent proangiogenic factor and interacts with its receptors, Flk-1/KDR and Flt-1, to stimulate angiogenesis in the follicle and corpus luteum [27, 28]. VEGF is differentially expressed during follicular development, with reduced expression in early phases of the cycle and increased expression as the follicles grow and develop [29]. VEGF-induced angiogenesis is required for follicle growth, development, and viability [30, 31]. Although the biological roles of TSP have been investigated in a number of physiological systems and cell lines, the pattern of expression during bovine ovarian follicular development has not been characterized. TSP and VEGF are known to be inversely expressed in other normal and pathological situations [32–35]. However, this relationship has not been evaluated in the ovary until now. This study was therefore designed to test the hypothesis that TSP-1 and -2 are produced and secreted in the bovine follicle at various stages throughout the ovarian cycle and that they are expressed inversely to the proangiogenic factor VEGF. The results from this study will provide insight into some of the cellular mechanisms involved in regulating folliculogenesis during the bovine ovarian cycle. MATERIALS AND METHODS

Animals and Tissue Collection Ovaries were collected from Holstein cows from a local abattoir. The ovaries were collected at 378C and transported from the slaughterhouse in an insulated container at approximately 48C in 10 mM phosphate buffered saline (PBS) at a pH of 7.2 within 1 h of collection. The ovaries were rinsed in 70% (vol/vol) ethanol and were selected and separated based on follicle size in a similar fashion as has been reported previously [36]. Follicles of a diameter greater than 1.0 cm were classified as large follicles, those with a diameter of between 0.5 and 1.0 cm were considered to be medium size, and those follicles having a diameter of less than 0.5 cm were small. Follicular fluid was collected by aspiration with a 25-gauge needle and 1.0-ml tuberculin syringe. Following centrifugation, the supernatant was stored at 2708C for future protein determination. For analysis of TSP protein and mRNA, cells were isolated from the small, medium, and large follicles as follows. Follicles were slashed with a scalpel and immersed in Dulbecco modified essential medium (DMEM) containing Ham F12 in a 1:1 ratio supplemented with 1% (vol/vol) fungizone, 2% (vol/vol) gentamycin, 1% (vol/vol) heparin, and 2% (vol/vol) penicillin/ streptomycin. Granulosa cells (GCs) were pelleted with centrifugation and, for RNA experiments, were snap frozen in liquid nitrogen and stored at 2708C. For protein experiments, isolated cells were lysed with the addition of 500 ml lysis buffer (250 mg sodium deoxycholate, 5 mg phenylmethyl sulfonyl fluoride, 5 mg aprotinin, 500 ml Nonidet P-40, 500 ml 10% [wt/vol] SDS) and stored at 2708C. For the preparation of bovine platelet lysates, blood was collected from normal, mature Holstein cows maintained at the Ponsonby Research Station, University of Guelph, under the guidelines of the Canadian Council on Animal Care. Blood was drawn by puncture of the external jugular vein with an 18-gauge needle and 60ml syringe, to which 6 ml of 3.8% (wt/vol) sodium citrate had been added as an anticoagulant. Platelets were isolated from whole blood with centrifugation and lysed with 500 ml lysis buffer (see above). Tissues to be used for immunohistochemistry were fixed in 10% neutral buffered formalin (Fisher Scientific, Nepean, ON, Canada) overnight at 48C and were processed according to normal histological procedures. All procedures were approved by the Animal Care Committee of the University of Guelph, in accordance with the Canadian Council of Animal Care guidelines on the care and use of experimental animals.

In Vitro Culture To determine the effect of gonadotropin stimulation on TSP expression, in vitro experiments were performed on freshly isolated bovine granulosa cells. Briefly, slaughterhouse ovaries were collected and rinsed in 70% ethyl hydroxide. Granulosa cells were isolated from small, medium, and large follicles and plated in DMEM/F12 supplemented with 10% fetal

bovine serum and 2% penicillin/streptomycin (Gibco BRL, Burlington, ON, Canada). At 70% confluence (approximately 24 h), cells were switched to serum-free DMEM/F12 (supplemented with 2% penicillin/ streptomycin) in the presence or absence of various concentrations of gonadotropins (FSH [Bioniche, Belleville, ON]: 10 ng/ml, 50 ng/ml, 100 ng/ ml LH [Sigma, Oakville, ON]: 10 ng/ml, 50 ng/ml, 100 ng/ml) for 24 h. After treatment, GCs were lysed and total protein was isolated for Western blot analysis. TSP protein concentration was quantified using densitometry against bovine platelet lysate (bPL) controls. For immunofluorescence experiments, GCs were isolated from small bovine follicles and cultured on 22- 3 22-mm coverslips in the bottom of six-well culture dishes (Gibco BRL) in DMEM/F12 with 10% until cells were attached (approximately 6 h). Cells were then rinsed in PBS, fixed in 10% (vol/vol) neutral buffered formalin, and processed for immunofluorescence experiments.

Immunohistochemistry/Immunofluorescence Histological sections (5 mm) of small, medium, and large follicles from bovine ovaries collected at slaughter were cut from paraffin blocks with a rotary microtome and mounted on glass microscope slides (Superfrost Plus; Fisher Scientific). Immunohistochemistry was performed to localize TSP and VEGF within ovaries at various stages of the ovarian cycle using a modification of the avidin-biotin peroxidase method [37]. Briefly, slides were incubated for 24 h at 48C in a humidified chamber with mouse antihuman TSP (1:400 dilution; Medicorp Inc., Montreal, QC) or rabbit polyclonal anti-human VEGF (1:600 dilution; Santa Cruz Biotechnology, Santa Cruz, CA). Antisera was diluted in 0.01 M PBS (pH 7.5) containing 2% (wt/vol) BSA and 0.01% (wt/vol) sodium azide (100 ml/slide). All subsequent incubations were at room temperature. Biotinylated anti-mouse or anti-rabbit IgGs (both 1:100 dilution; Vector Laboratories, Burlingame, CA) were diluted in the same buffer and applied for 2 h. The slides were then washed in PBS and incubated with avidin and biotinylated horseradish peroxidase (1:30 dilution) (Extravidin, Sigma Chemical Co., St Louis, MO). Peptide immunoreactivity was localized by incubation in fresh diaminobenzidine tetrahydrochloride (DAB tablets, 10 mg; Sigma) with 0.03% (vol/vol) hydrogen peroxide for 2 min. Tissue sections were counterstained with Carazzi hematoxylin for 1 min. Tissues were dehydrated and placed under a coverslip with Permount (Fisher Scientific). The percentage of immunopositive versus immunonegative follicular cells was calculated with integrated morphometry software (Metamorph; Universal Imaging Inc., Downington, PA). For analysis of colocalization of TSP and CD36, follicular tissue was collected and freshly isolated GCs from small follicles were allowed to attach to coverslips and were fixed in 10% neutral buffered formalin. Tissues and cells were incubated with a combination of rabbit polyclonal anti-TSP (1:600 dilution; Santa Cruz Biotechnology) and mouse monoclonal IgG raised against human CD36 (1:500 dilution; Santa Cruz Biotechnology) overnight at 48C. Samples were then incubated with the appropriate secondary antibodies (all 1:100 dilution; Vector Laboratories) conjugated to fluorescein isothiocyanate (for TSP) or Texas Red (for CD36) for 2 h at room temperature. Nuclei were counterstained with DAPI (49,69-diamidino-2-phenylindole), antifade (Vectashield, Vector Labs) was applied, and the specimens were imaged with a fluorescent microscope. To establish specificity of staining in histology and fluorescence experiments, primary antisera was substituted with nonimmune serum and secondary antibody was omitted. In each case, staining was abolished.

Western Blot Analysis TSP and VEGF proteins were detected and quantified using Western blot analysis. Protein samples (20 mg of total protein extracted from bovine follicular fluid, 5 mg for granulosa cell lysates, 2 mg for platelet lysates) were subjected to SDS-PAGE using an 8% separating gel. Pure human TSP (0.54 mg—a generous gift from Dr. Steven L. Gonias, University of Virginia), purified bovine platelet lysate (0.2 mg) and purified VEGF (5 mg; Santa Cruz Biotechnology) were used as positive controls. For VEGF Westerns, 5 mg VEGF (165aa) (Santa Cruz) was loaded as a positive control for each membrane. The separated proteins were electrotransferred to polyvinylidene fluoride membrane (Millipore, Billerica, MA). Membranes were blocked overnight with 5% (wt/vol) skim milk at 48C and then were incubated for 2 h at room temperature with mouse anti-human TSP (1:500 dilution; Medicorp Inc., PQ, Canada) or anti-VEGF (1:600 dilution; Santa Cruz Biotechnology) on a rocking platform. After washing with TTBS (TBS, 1% (vol/vol) Tween 20) and reblocking, peroxidaseconjugated sheep anti-mouse (1:800 dilution) or anti-rabbit (1:1000 dilution) was added for 1 h at room temperature on a rocking platform. Reactive protein was detected with ECL chemiluminescence (Boehringer

INVERSE EXPRESSION OF VEGF AND TSP IN THE FOLLICLE

FIG. 1. Quantitation of TSP protein in bovine follicular fluid. A) Western blot of 20 mg total protein from follicular fluid from small, medium, and large bovine follicles. Purified human TSP (hTSP) was run as a positive control, and bPL was used to determine antibody reactivity with bovine proteins. Estimated molecular weight of TSP is indicated. B) Densitometric measures of TSP in bovine follicles quantified against bovine platelet lysate values (bars represent mean 6 SD). Bars with different letters are statistically different (P , 0.05); n 5 3 different blots.

1073

FIG. 2. Quantitation of TSP protein in bovine granulosa cell lysates. A) Western blot of 5 mg total protein from lysed granulosa cells from small, medium, and large bovine follicles. B) Densitometric measures of TSP in GC lysates quantified against bovine platelet lysate values (bars represent mean 6 SD). Bars with different letters are statistically different (P , 0.05); n 5 3 different blots.

for differences between means, was used to assess statistical differences between the percentage of immunopositive cells at the various stages of follicle development.

Manheim, Framingham, MA) and Konica medical x-ray film (Wayne, NJ). Computer-assisted densitometry was performed and values presented as band intensities relative to positive controls. After ECL detection, membranes were stained with Coomassie Blue to ensure equal protein load.

RESULTS

Northern Blot Analysis

Determination of TSP in Follicular Fluid and Granulosa Cell Lysates

Granulosa cells were isolated from small, medium, and large bovine follicles using needle aspiration and centrifugation. Isolated granulosa cells were immersed in cold Trizol reagent (Gibco BRL, Burlington, ON, Canada) and lysed by repeated pipetting. Total GC RNA was isolated according to the protocol provided by Gibco BRL. Twenty micrograms of GC RNA from each of the individual aspirates from small, medium, and large follicles was subjected to electrophoresis in a 0.8% formaldehyde/agarose gel. Following separation, RNA was transferred to nylon membrane (Hybond-N, Amersham Pharmacia Biotech, Buckinghamshire, UK) using passive capillary transfer. Membranes were ultraviolet cross-linked and hybridized for a minimum of 2 h in prehybridization buffer (63 SSPE, 0.5% SDS, 53 Denhardt reagent, 2 mg salmon sperm DNA, 50% formamide). The cDNA probes for bovine TSP-1 and TSP-2 (kindly provided by Dr. P. Bornstein, Department of Biochemistry and Medicine, University of Washington) were radiolabeled using [a32P] dCTP (Amersham, Oakville, ON, Canada) and Rediprime random primer labeling (Amersham). The radiolabeled probes were hybridized overnight in hybridization buffer. Membranes were washed twice for 15 min in 23 saline-sodium citrate (SSC), 1% SDS at 558C for 20 min. Membranes were blotted dry and analyzed in a Molecular Imager GS250 (Bio-Rad, Richmond, CA). To determine equivalency of load and transfer, membranes were stripped by washing in 0.013 SSC, 0.5% SDS for 30 min at 908C and rehybridized with a probe for 7S ribosomal RNA.

Statistical Analysis Three replicates of the Western and Northern blots using different representative samples were used for statistical analysis. For Western and Northern blot analyses and culture experiments, data were subjected to one-way analysis of variance (ANOVA) followed by Tukey test when indicated by ANOVA (a 5 0.05). For in vitro experiments, data were collected from three independent cultures for each treatment group, each in triplicate. For morphometric analysis of histological images, four separate follicles for each size category were used and the entire follicle was included in the analysis. One-way ANOVA, followed by Bonferroni test

Western blot analysis was performed using a monoclonal antibody that did not distinguish between different isoforms of TSP. A representative Western blot for total TSP in follicular fluid is shown in Figure 1, and for granulosa cell lysates in Figure 2. For each of the Western experiments, equivalent amounts of total protein was loaded from follicular fluid of small, medium, and large follicles and compared with the human TSP standard and bovine TSP derived from activated platelets. Loading equivalency was verified with Coomassie Blue staining of all membranes. In all samples, TSP migrated as a single protein band of approximately 180 kDa under reducing conditions. It was clearly evident that TSP concentration was highest in both follicular fluid and granulosa cell lysates of small follicles, and this was confirmed with densitometric analysis of a minimum of three gels for each experiment. Using bovine platelet lysates as control, the relative amount of TSP protein in follicular fluid from small follicles was significantly (P , 0.05) higher than in medium and large follicles (Fig. 1). TSP protein in follicular fluid from medium follicle was significantly (P , 0.05) higher than in large follicles (Fig. 1). Again, using bovine platelet lysates as control, the TSP protein concentration in granulosa cell lysates from small follicles was higher (P , 0.05) compared with medium and large follicles (Fig. 2). Granulosa cell lysate TSP protein levels were significantly (P , 0.05) higher in medium follicles than large follicles (Fig. 2). Compared with follicular fluid, TSP protein was more highly concentrated in granulosa cell lysates at all stages of follicular development when assessed as a ratio with bovine platelet lysate TSP.

1074

GREENAWAY ET AL.

FIG. 3. Immunohistochemical localization of TSP in formalin-fixed, paraffin-embedded sections (5 mm) of small, medium, and large bovine follicles. The majority of immunopositive (brown-staining) cells were located within the GC layer, with only an occasional theca cell (TC) staining positive. Most GC in the small follicles were immunopositive, with a reduction in positive cells in the medium follicle and almost a complete absence of staining in the large follicle. As demonstrated, antibody omission controls were void of staining. Magnification, 3400. Bars (mean 6 SD) with different letters are statistically different (P , 0.05).

Immunohistochemical Localization of TSP and CD36 in Bovine Follicles and Granulosa Cells

To characterize the cellular localization of TSP, immunohistochemistry was performed on tissue sections of isolated small, medium, and large follicles. TSP was localized primarily to the granulosa cell layer of small and medium follicles, with some immunopositive theca cells (Fig. 3). The majority of granulosa cells in small follicles stained for TSP, with fewer (P , 0.05) positive cells in medium follicles, and a further reduction (P , 0.05) in staining in large follicles (Fig. 3). Immunostaining of fluorescently labeled antibodies was performed to determine whether TSP and its receptor CD36 were colocalized in follicular tissue and in vitro in cultured granulosa cells. In follicular tissues, TSP and CD36 were colocalized to the granulosa cell layer (Fig. 4A). CD36 expression followed a similar pattern of expression to TSP in small and medium follicles, with abundant expression in small follicles and reduced immunoreactivity in medium follicles (Fig. 4A). In large follicles, TSP was absent, while a small percentage of GC remained positive for CD36 (Fig. 4A). In vitro, freshly isolated granulosa cells from small follicles also demonstrated colocalization of TSP and CD36 (Fig. 4B). Evaluation of TSP-1 and -2 mRNA in Granulosa Cells

Northern blot analysis was used to determine whether granulosa cells of various sized bovine ovarian follicles were capable of synthesizing TSP and also to quantitate the expression of specific isoforms of the TSP gene. Although the antibody used for Western blot analysis and immunohistochemistry does not delineate between isoforms of TSP, specific cDNA probes for TSP-1 and -2 (provided by Dr. P. Bornstein, University of Washington [38, 39]) were used to determine the ontology of isoform expression in isolated granulosa cells from small, medium, and large bovine follicles. Representative Northern blots for TSP-1 and -2 are

FIG. 4. A) Colocalization of TSP and its cell-surface receptor, CD36 in small, medium, and large bovine follicles. The majority of bGC were immunopositive for TSP (green stain) and CD36 (red stain). TSP and CD36 were colocalized in GC of medium follicles, although there were fewer positive cells. There was a relative absence of TSP staining in large follicles, although a small percentage of GC were positive for CD36. Nuclei were counterstained with DAPI (blue stain) to visualize tissue morphometry. B) TSP and CD36 are colocalized in cultured bGC from small follicles. TSP (green stain) and CD36 (red stain) demonstrated colocalization (yellow stain) ubiquitously in culture. Cell nuclei are counterstained with DAPI (blue stain). Magnification, A 3400; B 31000.

shown in Figure 5. As was seen with the analysis of TSP protein, mRNAs for TSP-1 and -2 were more abundant in small follicles compared with medium and large follicles. Statistical analysis showed that the abundance of TSP-1 and -2 mRNAs from isolated granulosa cells of small follicles was significantly greater (P , 0.05) than in either medium or large follicles. The abundance of mRNA for both isoforms in medium follicles was significantly less (P , 0.05) than small and significantly greater (P , 0.05) than in large follicles (Fig. 5). The pattern of expression for both isoforms was similar, and there were no significant differences between TSP-1 and -2 expression in small, medium, or large follicles. Gonadotropin Regulation of TSP Expression in Isolated Granulosa Cells

Previous findings in the rat ovary indicated that TSP-1 and -2 expression was regulated, at least in part, by the gonadotropins [25]. We therefore evaluated the effect of gonadotropin stimulation on TSP expression in bovine GC. Freshly isolated bovine granulosa cells pooled from various sized follicles were cultured in DMEM/F12 in the presence or absence of FSH or LH at varying concentrations (10,

INVERSE EXPRESSION OF VEGF AND TSP IN THE FOLLICLE

1075

FIG. 5. Expression of TSP-1 and TSP-2 mRNA in granulosa cells of small (lane 1), medium (lane 2), and large (lane 3) bovine follicles detected using Northern blot analysis. Membranes were probed for 7S to detect loading consistency and TSP band intensity was quantified against 7S. Bars (mean 6 SD) with different letters are statistically different (P , 0.05).

50, 100 ng/ml) for 24 h. Low (10 ng/ml) FSH concentrations induced a modest (P , 0.05) increase in the amount of TSP protein released. Treatment of bovine GC with 50 and 100 ng/ml FSH, however, resulted in a significantly (P , 0.05) greater increase in TSP protein expression compared with nontreated controls (Fig. 6A). Treatment with LH did not alter TSP protein levels at any concentration compared with nontreated controls (Fig. 6B). Expression and Localization of VEGF During Follicle Development

Western blot analysis for VEGF was performed on 10 mg GC lysate protein separately isolated from small, medium, and large bovine follicles. In contrast with TSP, there was an apparent increase in the amount of VEGF protein during growth of the follicle (Fig. 7). The antibody used detects the 189, 165, and 121 amino acid splice variants of VEGF. The visible bands correspond to VEGF165 and VEGF121. Both subunit bands were used in calculating densitometry. Using relative densitometry of band intensity against purified VEGF-positive control, small follicles exhibited a band intensity of 112.3 6 22.2 U. There was a significant (P , 0.05) increase in VEGF protein to 410.5 6 106.3 U in medium and a further significant (P , 0.05) increase to 864.2 6 306.4 U in large follicles (Fig. 7). Immunohistochemical experiments were performed to determine the localization of VEGF within the follicles during growth and development (Fig. 8). Morphometric analysis was performed on these immuno-sections to determine the percentage of VEGF-positive granulosa and theca cells in small, medium, and large follicles. In small follicles, there were only a few VEGF-positive granulosa (16% 1 9%) and

FIG. 6. Bovine granulosa cell production of TSP after treatment with FSH (A) and LH (B). Western blot analysis was performed after in vitro treatment with FSH (A) and LH (B). Densitometry was performed with bPL as control. Bars (mean 6 SD) with different letters in the FSH treatment group are statistically different (P , 0.05). Protein concentrations in the LH-treatment groups were not statistically different.

theca (23% 1 11%) cells. In medium follicles, there was a significant (P , 0.05) increase in the percentage of immuno-positive granulosa (61% 6 12%) and theca (66% 6 14%) cells (Fig. 8). In large follicles, there was a further significant (P , 0.05) increase in granulosa (99% 6 14%) and theca (89% 6 8%) cells expressing VEGF (Fig. 8). VEGF immunoreactivity was restricted to granulosa (black arrows), theca (red arrows), and endothelial cells (green arrows) of adjacent blood vessels in the bovine ovary (Fig. 8). DISCUSSION

The purpose of this study was to characterize the expression patterns of TSP-1 and -2 and their specific cell surface receptor CD36 as well as VEGF in the bovine ova-

1076

GREENAWAY ET AL.

FIG. 7. VEGF protein increases in granulosa cell lysates from small, medium, and large bovine follicles. A) Western blot analysis of GC lysates protein from freshly isolated GC from bovine follicles. Bands represent 165 and 121aa variants of VEGF. B) Graph demonstrating relative densitometry of VEGF protein band intensity compared against the VEGF standard (165aa) in small, medium, and large follicles. Bars (mean 6 SD) with different letters are statistically different (P , 0.05).

ry during follicular development. We discovered a novel expression of members of the TSP family in the bovine ovary during follicle development. Based on relationships established in other systems, such as malignant tumors [32– 35, 39–41] and our own studies with rats [25], we hypothesized that TSP and VEGF might display differential expression during the development of bovine ovarian follicles. We also hypothesized, based on our previous results in the rat [25], that TSP expression would be subject to regulation by gonadotropins. It is well recognized in the field of cancer-treatment research that TSPs can be important inhibitors of angiogenesis, primarily through their inhibition of proangiogenic genes such as VEGF [42–49]. In the present study, the observed decrease in TSP content and expression in medium and large bovine follicles compared with small follicles is associated with a concomitant increase in VEGF expression. TSP protein derived from granulosa cell lysates and immunohistochemical localization of protein in the granulosa cell layer revealed that a major source of the follicular-fluid TSP was from endogenous production and secretion within the granulosa cells lining the follicular antrum. As TSP is a large, inducible glycoprotein with an approximate size of 450 kDa and as we detected mRNA for TSP-1 and -2 in granulosa cells, it is very likely that the follicular TSP observed here was due primarily to endogenous production within granulosa cells. Although there were some immunopositive theca cells in developing follicles, the contribution of these cells to accumulation of TSP in follicular fluid remains to be determined. TSP appears to have an inverse relationship with tumor formation and vascularization, where deletion of the gene results in tumorigenesis and neovascularization [50], while its pres-

FIG. 8. Localization of VEGF during bovine follicular development. A) Immunohistochemical images of VEGF protein localization in small, medium, and large bovine follicles. There was very little VEGF immunostaining in granulosa (black arrows) or theca (red arrows) layers in small follicles. There was an increase in staining in both GC and TC in medium follicles and a further increase in large follicles. VEGF was also localized to endothelial cells of blood vessels in the theca layer and ovarian stroma (green arrows). B) Graph representative of the percent immunopositive cells in small, medium, and large follicles. Magnification 3400. Bars (mean 6 SD) with different letters are statistically different (P , 0.05).

ence lowers tumor microvessel density and induces focal areas of necrosis and apoptosis within the tumor [47, 48]. Our present results using bovine ovarian tissue is in agreement with the observations of others using human and porcine ovarian tissues [29, 51], and our observations using rat ovarian tissue [25] demonstrates that VEGF is weakly expressed during early ovarian follicular development and becomes more pronounced in granulosa and theca cells as the dominant follicle(s) develop to the preovulatory stage. Further, in rats, exogenous stimulation with VEGF results in increased formation of small, preantral follicles [16]. Whether or not this phenomenon occurs in other species remains to be determined. Studies with tissue from various species indicate that VEGF may have a number of functions in the ovary. For example, VEGF stimulates angiogenesis in follicles and corpora lutea of at least some mammals [27, 28]. Recently, we have shown that VEGF has a cytoprotective role in the extravascular granulosa cell compartment of the ovary [52]. In addition, VEGF increases vascular permeability [53] and changes in ovarian follicular vascular permeability are associated with increased expression of VEGF in response to stimulation by exogenous gonadotropins in vivo [54, 55]. While these observations have led to speculation regarding the role of VEGF in the reproductive disorder know as ovarian hyperstimulation syndrome [54, 55], our present results indicate that it is important to remember that increased expression and production of VEGF, and thus increased vascular permeability, appears to be an important, normal aspect of healthy developing follicles in cows as well as other species [25, 27, 28]. Previously, we demonstrated and characterized TSP and CD36 expression in the rat ovary during follicular and luteal development [25]. As in the present study, TSP-1, -2,

INVERSE EXPRESSION OF VEGF AND TSP IN THE FOLLICLE

and CD36 are expressed predominantly by rat granulosa cells. However, in contrast with the indistinguishable patterns of expression of TSP-1 and -2 by bovine granulosa cells in the present study, TSP-1 and -2 are differentially regulated in the rat ovary, with TSP-1 expressed early in the development of rat antral follicles and TSP-2 expressed later during the ovulatory/early luteal phase [25]. The reason for this difference in isoform-specific expression patterns for bovine and rat ovarian follicle models is not yet understood. Recent studies in cattle have demonstrated, however, that TGFb is highly expressed in early antral follicles and is significantly decreased in large follicles [56]. These observations regarding TGFb appear to mirror the expression of TSP-1 and -2 that we have observed in bovine ovarian follicular tissue in the present study. In contrast, TGFb is expressed in rodent ovarian tissue in a noncyclic manner throughout the rat reproductive cycle [57]. Whether or not there is a correlation between the expression of ovarian TGFb and the expression of TSPs in either bovine or rodent ovaries remains to be determined. Further, the observation that TGFb can stimulate TSP-1 and -2 expression in bovine adrenal tissue [58] provides evidence to support further studies regarding the potential role of TGFb in regulating TSP expression in bovine follicles. Another interspecies difference appears to be the role of gonadotropins in the regulation of ovarian follicular TSP-1 and -2 expression. In the rat [25], LH appears to play the major role in regulating the positive expression of ovarian TSP-1 and -2. In distinct contrast, FSH appears to play the major role in stimulating bovine ovarian TSP-2 and -2 expression in small bovine follicles. In further contrast with rat follicles, TSP production by bovine granulosa cells did not increase in response to LH in vitro. This latter observation may be due to the fact that the bovine granulosa cells that were subjected to stimulation by LH in vitro were collected from follicles at various stages of development, including small- and medium-sized follicles. Such follicles are not mature and may not yet have undergone the critical point in granulosa cell differentiation where FSH and estradiol induce the expression of gonadotropin-responsive granulosa cell LH receptors. In contrast, gonadotropin-responsive LH receptors are present early in the development of rat ovarian antral follicles [59, 60]. Thus, species differences in the timing of LH receptors in granulosa cells during the rodent and bovine reproductive cycles may provide another partial explanation for the differences observed in rat and bovine granulosa cell expression of TSP isoforms in response to LH. The significance of the apparent inverse relationship between the pattern of TSP isoform expression and VEGF by bovine ovarian granulosa cells during normal follicular development remains to be determined. Based on the results of this study, we propose that TSP-1 and -2 may act in autocrine and/or paracrine fashions to regulate cellular events that occur during the early phase of follicular development. It seems reasonable that the expression of TSP isoforms by bovine granulosa cells in the early stages of follicular development would serve to suppress vascularization of the granulosa cell compartment to optimize preovulatory granulosa cell proliferation and follicular growth, which, in turn, would produce a postovulatory corpus luteum capable of maximal progesterone production to support the initial stages of a potential pregnancy. However, the significant differences in patterns of expression of TSP isoforms in developing follicles of rodents and cows and in the regulation of these patterns of expression by gonado-

1077

tropins underscores the importance of further research regarding the expression and the roles of TSPs in higher order mammals of domestic importance as well as in primates, including women. ACKNOWLEDGMENTS We would like to thank members of the Animal Biotechnology Embryo Laboratory at the University of Guelph for ovary collection and Michelle Ross for technical support.

REFERENCES 1. Tsafriri A. Ovulation as a tissue remodeling process. Adv Exp Med Biol 1995; 377:121–140. 2. Redmer DA, Grazul AT, Kirsch JD, Reynolds LP. Angiogenic activity of bovine corpora lutea at several stages of luteal development. J Reprod Fertil 1988; 82:627–634. 3. Plendl J. Angiogenesis and vascular regression in the ovary. Anat Histol Embryol 2000; 29:257–266. 4. Acosta TJ, Hayashi KG, Ohtani M, Miyamoto A. Local changes in blood flow within the preovulatory follicle wall and early corpus luteum in cows. Reproduction 2003; 125:759–767. 5. Yang MY, Rajmahendran R. Involvement of apoptosis in the atresia of nonovulatory dominant follicle during the bovine estrous cycle. Biol Reprod 2000; 63:1313–1321. 6. Hendriksen PJ, Gadella BM, Vos PL, Mullaart E, Kruip TA, Dieleman SJ. Follicular dynamics around the recruitment of the first follicular wave in the cow. Biol Reprod 2003; 69:2036–2044. 7. Richards JS, Russell DL, Ochsner S, Hsieh M, Doyle KH, Falender AE, Lo YK, Sharma SC. Novel signaling pathways that control ovarian follicular development, ovulation, and luteinization. Recent Prog Horm Res 2002; 57:195–220. 8. Fraser HM, Lunn SF. Angiogenesis and its control in the female reproductive system. Br Med Bull 2000; 56:787–797. 9. Ferrara N, Frantz G, LeCouter J, Dillard-Telm L, Pham T, Draksharapu A, Giordano T, Peale F. Differential expression of the angiogenic factor genes vascular endothelial growth factor (VEGF) and endocrine gland-derived VEGF in normal and polycystic human ovaries. Am J Pathol 2003; 162:1881–1893. 10. Richards JS, Jahnsen T, Hedin L, Lifka J, Ratoosh S, Durica JM, Goldring NB. Ovarian follicle development: from physiology to molecular biology. Rec Prog Horm Res 1987; 43:231–272. 11. Geisthovel F, Moretti-Rojas FJ, Asch RH. Insulin-like growth factors and thecal-granulosa cell function. Human Reprod 1990; 5:785–799. 12. Fournet N, Weitsman SR, Zachow RJ, Magottin DA. Transforming growth factor-b inhibits ovarian 17a-hydroxylase activity by a direct non-competitive mechanism. Endocrinology 1996; 137:166–174. 13. Maruo T, Ladines-Llave CA, Samoto T, Matsuo H, Manalo AS, Ito H, Mochizuki M. Expression of epidermal growth factor and its receptor in the human ovary during follicular growth and regression. Endocrinology 1993; 132:924–931. 14. Zimmermann RC, Hartman T, Kavic S, Pauli SA, Bohlen P, Sauer MV, Kitajewski J. Vascular endothelial growth factor receptor 2-mediated angiogenesis is essential for gonadotropin-dependent follicle development. J Clin Invest 2003; 112:659–669. 15. Shimizu T, Jiang JY, Iijima K, Miyabayashi K, Ogawa Y, Sasada H, Sato E. Induction of follicular development by direct single injection of vascular endothelial growth factor gene fragments into the ovary. Biol Reprod 2003; 69:1388–1393. 16. 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; 68:1736–1741. 17. Adams JC. Thrombospondins: multifunctional regulators of cell interactions. Annu Rev Cell Dev Biol 2001; 17:25–51. 18. Hornebeck W, Emonard H, Monboisse JC, Bellon G. Matrix-directed regulation of pericellular proteolysis and tumor progression. Semin Cancer Biol 2002; 12:231–241. 19. Hogg PJ, Hotchkiss KA. Role of thrombospondins in vascular biology. In: Berndt MC (ed.), Platelets, Thrombosis and the Vessel Wall. Toronto, ON, Canada: Harwood Academic Publishers; 2000: 253–266. 20. Gao AG, Lindberg FP, Finn MB, Blystone SD, Brown EJ, Frazier WA. Integrin-associated protein is a receptor for the C-terminal domain of thrombospondin. J Biol Chem 1996; 271:21–24. 21. Pfander D, Cramer T, Deuerling D, Weseloh G, Swoboda B. Expres-

1078

22.

23. 24.

25.

26.

27.

28.

29.

30.

31.

32.

33.

34.

35.

36.

37.

38.

39.

GREENAWAY ET AL.

sion of thrombospondin-1 and its receptor CD36 in human osteoarthritis cartilage. Ann Rheum Dis 2000; 59:448–454. Carron JA, Wagstaff SC, Gallagher JA, Bowler WB. A CD36-binding peptide from the thrombospondin-1 can stimulate resorption by osteoclasts in vitro. Biochem Biophys Res Commun 2000; 270:1124– 1127. Dryefus M, Dardik R, Suh BS, Amsterdam A, Lahav J. Differentiation-controlled synthesis and binding of thrombospondin to granulosa cells. Endocrinology 1992; 130:2565–2570. Higuchi T, Fujiwara H, Yamada S, Tatsumi K, Kataoka N, Itoh K, Maeda M, Fujita J, Fujii S. Co-expression of integrin-associated protein (IAP/CD47) and its ligand thrombospondin-1 on human granulosa and large luteal cells. Mol Hum Reprod 1999; 5:920–926. Petrik JJ, Gentry PA, Feige J-J, LaMarre J. Expression and localization of thrombospondin-1 and -2 and their cell surface receptor, CD36 during rat follicular development and formation of the corpus luteum. Biol Reprod 2002; 67:1522–1531. Bagavandoss P, Sage EH, Vernon RB. Secreted protein, acidic and rich in cysteine (SPARC) and thrombospondin in the developing follicle and corpus luteum of the rat. J Histochem Cytochem 1998; 46: 1043–1049. Ferrara N, Chen H, Davis-Smyth T, Gerber HP, Thuy-Nhung N, Peers D, Chisholm V, Hillan KJ, Schwall RH. Vascular endothelial growth factor is essential for corpus luteum angiogenesis. Nat Med 1998; 4: 336–340. Shimizu T, Jiang JY, Sasada H, Sato E. Changes of messenger RNA expression of angiogenic factors and related receptors during follicular development in gilts. Biol Reprod 2002; 67:1846–1852. Barboni B, Turriani M, Giovanna G, Spiinaci M, Bacci ML, Forni M, Mattioli M. Vascular endothelial growth factor production in growing pig antral follicles. Biol Reprod 2000; 63:854–864. Watson ED, Al-Zi’abi MO. Characterization of morphology and angiogenesis in follicles of mares during spring transition and the breeding season. Reproduction 2002; 124:227–234. Van Blerkom J, Antczak M, Schrader R. The developmental potential of the human oocyte is related to the dissolved oxygen content of follicular fluid: association with vascular endothelial growth factor levels and perifollicular blood flow characteristics. Hum Reprod 1997; 12:1047–1055. Kaio E, Tanaka S, Oka S, Hiyama T, Kitadai Y, Haruma K, Chayama K. Clinical significance of thrombospondin-1 expression in relation to vascular endothelial growth factor and interleukin-10 expression at the deepest invasive tumor site of advanced colorectal carcinoma. Int J Oncol 2003; 23:901–911. Baltaci S, Orhan D, Gogus C, Filiz E, Tulunay O, Gogus O. Thrombospondin-1, vascular endothelial growth factor expression and microvessel density in renal cell carcinoma and their relationship with multifocality. Eur Urol 2003; 44:76–81. Kwak C, Jin RJ, Lee C, Park MS, Lee SE. Thrombospondin-1, vascular endothelial growth factor expression, and their relationship with p53 status in prostate cancer and benign prostate hyperplasia. BJU Int 2002; 89:303–309. Doll JA, Reiher FK, Crawford SE, Pins MR, Campbell SC, Bouck NP. Thrombospondin-1, vascular endothelial growth factor and fibroblast growth factor-2 are key functional regulators of angiogenesis in the prostate. Prostate 2001; 49:293–305. Yang MY, Rajmahendran R. Morphological and 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; 62:1209–1217. Hsu SM, Raine L, Fanger H. Use of avidin-biotin peroxidase complex (ABC) in immunoperoxidase techniques: a comparison between ABC and unlabeled antibody (PAP) procedures. J Histochem Cytochem 1981; 29:577–580. Armstrong LC, Bjorkblom B, Hankenson KD, Siadak AW, Stiles CE, Bornstein P. Thrombospondin 2 inhibits microvascular endothelial cell proliferation by a caspase-independent mechanism. Mol Biol Cell 2002; 13:1893–1905. Hawighorst T, Oura H, Streit M, Janes L, Nguyen L, Brown LF, Oliver G, Jackson DG, Detmar M. Thrombospondin-1 selectively inhibits early-stage carcinogenesis and angiogenesis but not tumor lymphan-

40.

41. 42. 43. 44. 45.

46.

47.

48. 49. 50.

51.

52.

53. 54.

55.

56. 57. 58.

59. 60.

giogenesis and lymphatic metastasis in transgenic mice. Oncogene 2002; 21:7945–7956. Fontanini G, Boldrini L, Calcinai A, Chine S, Lucchi M, Mussi A, Angeletti CA, Basolo F, Bevilacqua G. Thrombospondins I and II messenger RNA expression in lung carcinoma: relationship with p53 alterations, angiogenic growth factors, and vascular density. Clin Cancer Res 1999; 5:155–161. Shafiee A, Penn JS, Krutzsh HC, Inman JK, Roberts DD, Blake DA. Inhibition of retinal angiogenesis by peptides derived from thrombospondin-1. Invest Opthalmol Vis Sci 2000; 41:2378–2388. Lawler J. Thrombospondin-1 as an endogenous inhibitor of angiogenesis and tumor growth. J Cell Mol Med 2002; 6:1–12. Vailhe B, Feige JJ. Thrombospondins as anti-angiogenic therapeutic agents. Curr Pharm Des 2003; 9:583–588. Bocci G, Francia G, Man S, Lawler J, Kerbel RS. Thrombospondin1, a mediator of the antiangiogenic effects of low-dose metronomic chemotherapy. Proc Natl Acad Sci U S A 2003; 100:12917–12922. Zhang YW, Su Y, Volpert OV, Vande Woude GF. Hepatocyte growth factor/scatter factor mediates angiogenesis through positive VEGF and negative thrombospondin 1 regulation. Proc Natl Acad Sci U S A 2003; 100:12718–12723. Rofstad EK, Henriksen K, Galappathi K, Mathiesen B. Antiangiogenic treatment with thrombospondin-1 enhances primary tumour radiation response and prevents growth of dormant pulmonary micrometastases after curative radiation therapy in human melanoma xenografts. Cancer Res 2003; 63:4055–4061. Jin RJ, Kwak C, Lee SG, Lee CH, Soo CG, Park MS, Lee E, Lee SE. The application of an anti-angiogenic gene (thrombospondin-1) in the treatment of human prostate cancer xenografts. Cancer Gene Ther 2000; 7:1537–1542. Miyanaga K, Kato Y, Nakamura T, Matsumura M, Amaya H, Horiuchi T, Chiba Y, Tanaka K. Expression and role of thrombospondin-1 in colorectal cancer. Anticancer Res 2003; 22:3941–3948. Lawler JW, Slayter HS, Coligan JE. Isolation and characterization of a high molecular weight glycoprotein from human blood platelets. J Biol Chem 1978; 253:8609–8616. Hawighorst T, Velasco P, Striet M, Hong YK, Kyriakides TR, Browth LF, Bornstein P, Detmar M. Thrombospondin-2 plays a protective role in multistep carcinogenesis: a novel host anti-tumor defense mechanism. EMBO J 2001; 20:2631–2640. Kamat BR, Brown LF, Manseau EJ, Senger DR, Dvorak HF. Expression of vascular permeability factor/vascular endothelial growth factor by human granulosa and theca lutein cells. Role in corpus luteum development. Am J Pathol 1995; 146:157–165. Greenaway J, Connor K, Pedersen HG, Coomber BL, LaMarre J, Petrik J. VEGF and its receptor, Flk-1/KDR are cytoprotective in the extravascular compartment of the ovarian follicle. Endocrinology 2004; 145:2896–2905. Behzadian MA, Windsor LJ, Ghaly N, Liou G, Tsai NT, Caldwell RB. VEGF-induced paracellular permeability in cultured endothelial cells involves urokinase and its receptor. FASEB J 2003; 17:752–754. Gomez R, Simon C, Remohi J, Pellicer J. Administration of moderate and high doses of gonadotropins to female rats increases ovarian vascular endothelial growth factor (VEGF) and VEGF receptor-2 expression that is associated to vascular hyperpermeability. Biol Reprod 2003; 68:2164–2171. Wang TH, Horng SG, Chang CL, Wu HM, Tsai YJ, Wang HS, Soong YK. Human chorionic gonadotropin-induced ovarian hyperstimulation syndrome is associated with up-regulation of vascular endothelial growth factor. J Clin Endocrinol Metab 2002; 87:3300–3308. Nilsson EE, Doraiswamy V, Skinner MK. Transforming growth factorbeta isoform expression during bovine ovarian antral follicle development. Mol Reprod Dev 2003; 66:237–246. Teerds KJ, Dorrington JH. Immunohistochemical localization of transforming growth factor-beta 1 and -beta 2 during follicular development in the rat ovary. Mol Cell Endocrinol 1992; 84:R7–R13. Negoescu A, Lafeuillade B, Pellerin S, Chambaz EM, Feige JJ. Transforming growth factor beta stimulates both thrombospondin-1 and CISP/Thrombospondin-2 synthesis by bovine adrenocortical cells. Exp Cell Res 1995; 217:404–409. Ojeda SR, Advis JP, Andrews WW. Neuroendocrine control of the onset of puberty in the rat. Fed Proc 1980; 39:2365–2371. Richards JS, Uilenbroek JT, Jonassen JA. Follicular growth in the rat: a reevaluation of the roles of FSH and LH. Adv Exp Med Biol 1979; 112:11–26.