| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Department of Animal Sciences, University of Missouri, Columbia, Missouri 65211
ABSTRACT
Ovarian follicular growth and steroidogenesis are controlled by the interaction of insulin-like growth factors (IGFs) and gonadotropins. The objective was to determine the temporal and spatial relationships for gonadotropin receptor, steroidogenic enzyme, and IGF system gene expression during the development of preovulatory porcine follicles. Sows (n = 18) were weaned and follicles were monitored by transrectal ultrasonography. Ovaries were collected from sows when the mean diameter of the preovulatory follicular cohort was approximately 2, 4, 6, or 8 mm. mRNA were measured by in situ hybridization for individual follicles within the preovulatory cohort (3 to 5 follicles per sow). Patterns of gene expression detected by in situ hybridization were confirmed by RNase protection analyses of pooled RNA samples. The amount of LH receptor mRNA and steroidogenic enzyme mRNA (17
-hydroxylase and aromatase) increased as the mean diameter of the follicular cohort increased from 2 to 6 mm, but then decreased abruptly for 8-mm follicles. Estradiol concentrations in follicular fluid closely followed the expression patterns of steroidogenic enzymes and LH receptor mRNA. FSH receptor mRNA was present in cohorts of 2-mm follicles but declined in 4-mm follicles and was undetectable in 6- and 8-mm follicles. The expression of IGF-I and type I IGF receptor mRNA were similar for follicles of 2, 4, 6, and 8 mm. In contrast, IGF-II mRNA progressively increased in follicles collected from 2-, 4-, and 6-mm cohorts, and then decreased slightly at 8 mm. Type II IGF receptor mRNA was greatest in 8-mm follicles. IGF binding protein-2 (BP-2) mRNA decreased as follicles achieved progressively larger sizes during the preovulatory period (2 to 8 mm), whereas the IGFBP-4 mRNA remained relatively low for follicles in 2- to 6-mm cohorts but then increased markedly in 8-mm follicles. In summary, temporal and spatial patterns of gene expression for gonadotropin receptor, steroidogenic enzyme, and IGF system genes were characterized in preovulatory porcine follicles by using in situ hybridization and RNase protection analyses. The unique patterns of gene expression suggest interdependence among specific genes that may be essential for preovulatory follicular development.
follicle, follicular development, growth factors
INTRODUCTION
The insulin-like growth factors (IGF-I and IGF-II), their receptors (type I and type II IGF receptors), and IGF binding proteins (IGFBP-1 to -6) constitute an endocrine, autocrine, and paracrine IGF system that regulates growth, differentiation, and apoptosis of animal cells [1–5]. Ovarian expression of IGF system genes has been demonstrated [5–7]. In general, IGF and IGFBP are localized to specific cell layers (granulosa, theca, or both) within developing follicles. The location of specific components of the IGF system, therefore, corresponds to the location of LH and FSH receptors within developing follicles [6–8]. The colocalization of IGF and gonadotropin receptor genes suggests a coordination of gonadotropin and IGF action within ovary that may control growth, differentiation, and steroidogenesis of theca and granulosa cells. Indeed, there is strong evidence for a synergistic relationship between IGF and gonadotropins for the control of ovarian follicular growth and development [4, 5, 9–11].
For the synergism of gonadotropins and IGF to occur in vivo, a temporal and spatial relationship must exist between gonadotropin receptor expression and the expression of IGF system genes. IGF-I and IGF-II are the ligands for the IGF system [1, 3, 4]. Their actions are similar because they are similar in structure and amino acid sequence. IGF-II, however, has lower potency than IGF-I for tyrosine kinase signaling through the type I IGF receptor. The relatively lower potency of IGF-II compared with IGF-I may reflect the lower affinity of IGF-II for the type I IGF receptor. Although a specific IGF-II receptor exists (type II IGF receptor), it does not activate tyrosine kinase second-messenger pathways, and its primary function may be to bind and internalize IGF-II for intracellular degradation [12]. The availability of IGF in biological systems is modulated by IGFBP. Six different IGFBPs as well as four different IGFBP-related proteins have been characterized [13]. Each IGFBP exhibits a unique pattern of tissue distribution and regulation [1, 3–5, 9, 11]. IGFBPs modulate the interaction of IGF-I and IGF-II with their receptors and may regulate cell growth and differentiation by titrating the trophic effects of IGF [1]. Taken together, the ligands, receptors, and binding proteins for the IGF system create a complex intraovarian growth factor system.
The relationship between IGF system gene expression and size of the follicle has been established for various species, including humans [3, 4, 9], laboratory animals [11], and domestic animals [5, 10]. Most of the previous studies, however, examined follicles at different stages of development (primordial through Graafian) that were found within a single ovary at various stages of the estrous or menstrual cycle. Patterns of gene expression for the IGF system have not been characterized for a uniform cohort of preovulatory follicles during a rapid period of development. Likewise, steroidogenic enzyme and gonadotropin receptor mRNA have not been studied in this context. Follicular development before weaning is suppressed in sows (Sus scrofa domestica) so that most follicles are 2 to 3 mm in diameter. After weaning, there is a period of rapid ovarian follicular development in which a cohort of 20 to 30 follicles that are 2 to 3 mm in diameter at weaning grow to 8 mm (preovulatory diameter) over a period of 3 to 5 days [14, 15]. We studied preovulatory porcine follicles as a model for gene expression during follicular growth and used in situ hybridization to specifically localize the mRNA within the follicle. The objective of the present study was to measure mRNA for gonadotropin receptors (LH receptor and FSH receptor), steroidogenic enzymes (cytochrome P450 17-hydroxylase [P45017] and aromatase [P450arom]), IGF-I, type I IGF receptor, IGF-II, type II IGF receptor, IGF binding proteins (IGFBP-2 and -4), and growth hormone (GH) receptor in dominant follicles during preovulatory follicular development. Changes in gene expression were then correlated with the maturation and differentiation of preovulatory follicles.
MATERIALS AND METHODS
Animals and Tissue Preparation
The University of Missouri Animal Care and Use Committee approved the study procedures. Eighteen multiparous pregnant sows were individually housed in farrowing crates within the Animal Science Research Center (ASRC) beginning 7 days before expected parturition. Sows farrowed at the ASRC and piglets were weaned at 21 days of age. Beginning 2 days before weaning, sow ovaries were examined daily by ultrasound through the wall of the rectum and the growth of preovulatory follicles was monitored. Ovaries were collected by mid-ventral laparotomy after weaning when the mean diameter of the preovulatory follicular cohort was approximately 2 mm (mean of 2.3 ± 0.1 mm with a range of 2 to 3 mm; n = 5 sows), 4 mm (mean of 4.6 ± 0.1 mm with a range of 4 to 5 mm; n = 4 sows), 6 mm (mean of 6.3 ± 0.2 mm with a range of 6 to 7 mm; n = 4 sows), or 8 mm (mean of 8.2 ± 0.1 mm with a range of 8 to 9 mm; n = 5 sows). Sows at the 8-mm stage were displaying behavioral signs of estrus. Ovaries were cut into blocks with each block containing at least two follicles of the desired sizes. One block was frozen in liquid nitrogen and stored at -80°C until sectioning for in situ hybridization. Follicular fluid was aspirated from the remaining blocks for radioimmunoassays and ligand blotting. Two to four follicles were collected and frozen at -80°C for extraction of total cellular RNA. All follicles were frozen within 20 min of ovariectomy. The diameter of follicles within each block was confirmed with a ruler before freezing. A single blood sample was collected from the jugular vein at the time of ovariectomy. Plasma was harvested by centrifugation and stored at -20°C.
Radioimmunoassay of Plasma and Follicular Fluid
Progesterone radioimmunoassay Concentrations of progesterone in follicular fluid were analyzed by using progesterone kits (Coat-A-Count; Diagnostic Products Corp., Los Angeles, CA). The assay was performed according to the manufacturer's instructions and validated for porcine follicular fluid. One microliter of porcine follicular fluid was diluted in 100 µl of zero standard provided with the assay kit. Dilutions (0.5, 1, or 2 µl) of follicular fluid from either healthy or atretic follicles were parallel to the standard curve. The addition of 0.05, 0.2, or 1 ng of progesterone to the diluted samples yielded an average recovery of 84%. Assay sensitivity of a 1-µl sample was 10 ng/ml. The intra-assay coefficient of variation was 6%.
Estradiol radioimmunoassay Concentrations of estradiol in follicular fluid were measured by radioimmunoassay as described by Kirby et al. [16]. The estradiol assay was modified because follicular fluid samples (1 µl) were diluted in 100 µl assay buffer and were measured directly without extraction. The procedure was validated for porcine follicular fluid. Dilutions (0.5, 1, or 2 µl) of follicular fluid from either healthy or atretic follicles were parallel to the standard curve. The addition of 1, 5, or 10 pg of estradiol to the diluted samples yielded an average recovery of 113%. Assay sensitivity of a 1-µl sample was 0.25 ng/ml. The intra-assay coefficient of variation was 10.2%.
LH radioimmunoassay Plasma LH was measured by validated radioimmunoassay. Antibody and ligand were kindly donated by Dr. A.F. Parlow (National Hormone and Pituitary Program, Torrance, CA). Plasma (200 µl) was incubated with 200 µl anti-porcine LH (pLH) (AFP-15103194; 1:400 000 dilution in protein assay buffer [PAB] [0.1% gelatin, 0.01% thimersol, 0.01 M PO4, 0.9% NaCl, pH 7.2] with normal rabbit serum [1:300]) at 4°C for 24 h. On Day 2, 100 µl of PAB containing approximately 20 000 cpm [125I]-pLH (AFP-10714B) were added and the incubation continued for an additional 24 h at 4°C. Precipitation of the antibody complexes was performed on Day 3 with the addition of 100 µl goat anti-rabbit antiserum (1:50 dilution in PAB; Antibodies, Inc., Davis, CA) and 200 µl of a solution of 12.5% polyethylene glycol (average 8000 molecular weight). The tubes were incubated for 1 h at room temperature. Final centrifugation was at 3000 x g for 30 min at 20°C. Supernatant was decanted and the pellet was counted for 1 min. Concentrations of LH in unknown samples were estimated from a standard curve (0.02, 0.04, 0.08, 0.16, 0.31, 0.63, 1.25, 2.5, and 5 ng/tube) using pLH (USDA pLH B2). Increasing volumes of porcine plasma (100, 200, and 300 µl) resulted in a displacement curve that was parallel to the standard curve. Addition of different masses of pLH to the assay (0.06, 0.125, and 0.25 ng/tube) resulted in an average recovery of 114%. The intra-assay coefficient of variation was 9.1%.
Measurement of IGF-I and IGFBP in Follicular Fluid
IGF-I radioimmunoassay Concentrations of IGF-I in follicular fluid were measured by validated radioimmunoassay [17]. The sensitivity of the assay was 9.8 pg/ml follicular fluid. All samples were analyzed in a single assay and the intra-assay coefficient of variation was 4.0%.
IGFBP ligand blotting Follicular fluid IGFBP was measured by ligand blotting with 125I-IGF-II as described by Lucy et al. [18]. The amount of IGFBP (pixel density) was quantified from autoradiographs by using GP Tools (Biophotonics Corp., Ann Arbor, MI).
In Situ Hybridization
The cDNA fragments used for production of ribonucleotide probes were bovine, ovine, or porcine (Table 1). In each case, the cDNA integrity and orientation were verified by dideoxy DNA sequencing. Plasmids were linearized by digestion with appropriate restriction enzymes prior to their use for ribonucleotide probe synthesis. Both antisense and sense RNA probes were transcribed from linearized cDNA templates using a transcription kit (Stratagene, La Jolla, CA). The probes were labeled with 35S-UTP (DuPont, Wilmington, DE) and purified by centrifugation over a Sephadex G-50 column and used for in situ hybridization within 1 to 2 days.
|
Twelve-micron sections of follicular tissue were cut at -20°C and stored at -80°C in desiccated, airtight boxes. The labeled probes were diluted with hybridization buffer (1x Denhardts solution, 50% formamide, 0.3 M NaCl, 10 mM Tris-HCl [pH 8.0], 1 mM EDTA, 50 mM dithiothreitol, 500 µg/ml yeast tRNA, and 10% dextran sulfate) to about 2 x 107 cpm/ml.
For hybridization, sections were removed from -80°C and remained at room temperature for 10 min. The sections were placed into 4% formaldehyde in PBS for 5 min and into 2x SSC (1x SSC = 0.15 M NaCl and 0.017 M sodium citrate, pH 7.0) for 2 min. Acetylation was performed in 0.25% acetic anhydride in 0.1 M triethanolamine (pH 8.0) for 10 min, and then slides were treated for 2 min with 2x SSC. Dehydration was performed by treating the slides with 60%, 80%, 95%, and 100% ethanol for 2 min each. The sections were then immersed in chloroform for 5 min, rinsed again in 100% and 95% ethanol, and dried at room temperature.
Hybridization was performed by covering sections with 100 µl diluted probes in a humidified oven at 55°C for at least 16 h. After hybridization, the sections were washed twice (15 min each) with 2x SSC in a shaking water bath at 55°C. The sections were then treated with ribonuclease A (Sigma Chemical Company, St. Louis, MO; 50 µg/ml in 2x SSC) for 1 h with slow agitation in a 37°C water bath. Washing was done at 55°C. Wash conditions were 2x SSC containing 0.1% ß-mercaptoethanol (ß-ME) for 15 min, 1x SSC/0.1% ß-ME for 15 min, 50% 2x SSC/50% formamide/0.1% ß-ME for 30 min, and 0.1x SSC/0.1% ß-ME for 30 min. After washing, the sections were dehydrated with 60%, 80%, 95%, and 100% ethanol and air-dried.
Air-dried slides were dipped in Kodak NTB-2 emulsion (Eastman Kodak, Rochester, NY) in total darkness and were exposed at 4°C for 3 days (P45017 and P450arom mRNA) or for 2 wk (LH receptor, FSH receptor, GH receptor, IGF-I, type I IGF receptor, IGF-II, type II IGF receptor, IGFBP-2, and IGFBP-4). The slides were then developed, counterstained with hematoxylin and eosin, and mounted for microscopic examination. Sections from the same follicles were used for in situ hybridization with different RNA probes. For each probe, two sections were hybridized with the antisense probe and one section was hybridized with the sense probe.
RNA Preparation and Ribonuclease Protection Assay
Whole follicles (theca and granulosa cell layers) that were representative of follicles within each follicular cohort (2 to 3 mm [2-mm category], 4 to 5 mm [4-mm category], 6 to 7 mm [6-mm category], or 8 to 9 mm [8-mm category]) were dissected from the ovaries of each sow. Three to five follicles were collected from each sow. Follicular tissue was pooled within sows before RNA isolation. Total cellular RNA was isolated according to the TRIZOL procedure (Gibco BRL, Gaithersburg, MD) and dissolved in water. RNA concentration and purity were determined by calculating the ratio of absorbencies at 260 and 280 nm. A sample of RNA (2.5 µg) was electrophoresed in a 1% agarose gel in Tris-borate/EDTA buffer (0.089 M Tris-borate and 0.002 M EDTA) and stained with ethidium bromide (0.5 µg/ml) to verify the integrity and quantity. RNA samples were stored at -80°C.
Plasmids containing cDNA for porcine LH receptor, porcine FSH receptor (gift from Dr. R.L. Matteri, USDA-ARS, University of Missouri, Columbia, MO), porcine GH receptor 1A, and porcine glyceraldehyde-3-phosphate dehydrogenase (GAPDH) [21] were linearized by digestion with appropriate restriction enzymes. Other cDNA plasmids used for the production of ribonucleotide probes for ribonuclease protection assay (RPA) were the same as those used for in situ hybridization. The digested plasmids were extracted with phenol/chloroform, and ethanol precipitated. Antisense ribonucleotide probes were generated by using an RNA transcription kit (Stratagene). Linearized plasmids (approximately 200 ng) were incubated with SP6, T7, or T3 RNA polymerase (Promega, Madison, WI), 32P-rCTP (New England Nuclear, Boston, MA), appropriate buffers, and nucleotides to yield ribonucleotide probes that were antisense to the specific mRNA. The size of the riboprobe was verified by comparison to a radiolabeled 100-bp ladder.
RPA was performed to confirm the mRNA pattern observed by in situ hybridization. Total cellular RNA were combined across sows to create a pooled RNA sample for the 2-, 4-, 6-, or 8-mm categories. Individual RNA samples from each sow were not analyzed by RPA. Ten or 20 µg from each pool were analyzed by using the RPA II kit (Ambion Inc., Austin, TX). Protected mRNA were identified by their electrophoretic mobility through a 6% acrylamide, 8 M urea gel (Acryl-A-Mix-6; Promega). Gels were dried and autoradiography was performed by using BIOMAX film (Eastman Kodak) at -80°C with intensifying screens. The probe length and protected fragment length were verified by comparison to 32P-labeled DNA standards. Each RPA contained a negative control (yeast tRNA) to confirm the specificity of the hybridization. The GAPDH probe was used to confirm equal loading of mRNA.
Statistical Analyses
In situ hybridization intensity (mRNA amount) was measured using the Bioquant Image Analysis System (R&M Biometrics, Inc., Nashville, TN). The system measured the number of pixels within a marked area that were occupied by silver grains. Four fields at 90° angles were measured for each cell layer (granulosa, theca, or both) within the follicle for the two sections hybridized to the antisense probe and the one section hybridized to the sense probe. Measurements were not made for cell layers that were not expressing the mRNA (i.e., visually not above background). Specific hybridization intensity was defined as the average hybridization intensity for the two sections hybridized to the antisense probe minus the average hybridization intensity for the section hybridized to the sense probe (background). The procedure was repeated for cell layers of all follicles within the cohort that were represented by the slide (two to five follicles per slide). Healthy follicles (intact granulosa cell layer with intact basement membrane and no invagination of the theca into the granulosa) that were part of the cohort (2 to 3 mm [2-mm category], 4 to 5 mm [4-mm category], 6 to 7 mm [6-mm category], or 8 to 9 mm [8-mm category]) were included in the analyses. Follicles that were not part of the cohort (i.e., preantral follicles) or follicles classified as atretic (i.e., thin, fragmented granulosa cell layers with free-floating granulosa cells) were not included in the analyses. The number of follicles analyzed for each probe ranged from 15 to 39 for the 2-mm category, 9 to 16 for the 4-mm category, 8 to 11 for the 6-mm category, and 8 to 13 for the 8-mm category. The data were then averaged for each follicular cell layer from each sow (a sow was the experimental unit) and the mean value for each sow was subjected to statistical analysis. Data were analyzed by the GLM procedure of SAS [24]. The dependent variable was mRNA expression of specific genes (percentage occupied pixels determined by the quantification of in situ hybridization). The model included the effect of follicular stage (2-, 4-, 6-, or 8-mm diameter), cell type (granulosa or theca), and the stage by cell type interaction. Some genes were expressed only in one cell layer. Therefore, a statistical model that included only the main effect of follicular stage was used. Plasma or follicular fluid concentrations of hormones and IGFBP were also tested for the main effect of follicular stage. Main effect means were separated by using the Duncan's multiple range test. If an mRNA was found in two cell layers, then the data were sorted by cell layer so that the effect of follicular stage could be tested and a multiple range test could be used to examine changes in gene expression within each cell layer at different stages of development. Results are reported as least square means and SEM. Statistical significance was defined as P < 0.05. Tests of significance with probabilities greater than 0.1 were not significant. Tendencies for statistical significance were 0.05 < P < 0.1.
Pooled follicular mRNA samples (whole follicles that included theca and granulosa cell layers) from follicles representing each follicular stage were analyzed by RPA so that the general pattern of mRNA expression (increasing, decreasing, or no change) could be confirmed by a second, independent mRNA analysis. A statistical analysis could not be performed because only one pooled RNA sample for each follicular stage was analyzed in each RPA. Each RPA contained a GAPDH control to confirm equal loading.
RESULTS
Hormone and IGFBP Concentrations in Plasmaand Follicular Fluid
Plasma LH concentrations Concentrations of LH in plasma from the single blood sample collected at ovariectomy did not differ (P > 0.05) for sows with follicles at the 2-, 4-, or 6-mm stages (Fig. 1). In sows with follicles at the 8-mm stage, however, plasma LH concentrations were increased (P < 0.05) when compared with sows with follicles at either 2- or 4-mm stages of development.
|
Follicular fluid estradiol and progesterone concentrations There was an effect of follicular stage on follicular fluid estradiol (P < 0.03) and progesterone (P < 0.01) concentrations (Fig. 2). Concentrations of estradiol in follicular fluid were greatest for follicles at the 6-mm stage and were similar for follicles at the 2-, 4-, and 8-mm stages of development. Concentrations of progesterone in follicular fluid were similar for follicles at the 2-, 4-, and 6-mm stages but were increased for 8-mm follicles.
|
Follicular fluid IGF-I and IGFBP concentrations Concentrations of IGF-I in follicular fluid were similar for antral follicles at all stages of development (Fig. 3; P > 0.1). Ligand blotting of follicular fluid samples demonstrated a 43–40-kDa IGFBP (presumably IGFBP-3) and a 34-kDa IGFBP (presumably IGFBP-2; Fig. 4A). The ligand blots did not show IGFBP below 34 kDa. There was a tendency for an effect of follicular stage on the amount of IGFBP-3 (P < 0.07) because IGFBP-3 was greater in follicles at the 4- and 6-mm stages compared with follicles at the 2-mm stage of development (Fig. 4B). The follicles at the 8-mm stage were intermediate for IGFBP-3. The amount of IGFBP-2 was greatest for follicles at the 4-mm stage of development and least for follicles at the 6-mm stage of development (Fig. 4B; P < 0.07). The follicles at the 2- and 8-mm stages were intermediate for IGFBP-2.
|
|
Messenger RNA Expression
Figures 5 through 10 show photographs of follicular walls from representative preovulatory follicles within each stage of development (2, 4, 6, or 8 mm). The sections were illuminated by darkfield microscopy. In each case, sense controls were negative for gene expression and, in the interest of brevity, the sense pictures are not shown. Likewise, brightfield photographs are not shown because their inclusion would lead to an excessive number of photographs. To the right of each of Figures 5 through 10 are bar graphs for pixel density measured from in situ hybridization. RPA results (pooled RNA samples for each follicular diameter category) are shown as an inset within the graph and were not statistically analyzed because pooled RNA samples were tested. Probability values, therefore, represent statistical tests of data collected from in situ hybridization experiments. The reported location of mRNA expression (granulosa or theca cells) is for antral follicles examined in the present study (2, 4, 6, or 8 mm).
|
Gonadotropin receptor mRNA FSH receptor mRNA was detected in granulosa cells of follicles at the 2- and 4-mm stages of development (Fig. 5A). Expression of FSH receptor mRNA was greatest in 2-mm follicles. The signal was less in 4-mm follicles and was undetectable in follicles at the 6 mm and 8 mm stages of development (P < 0.01). A similar pattern of mRNA expression (reduced FSH receptor mRNA in follicles at larger stages of development) was observed when pooled RNA samples were analyzed by RPA.
Expression of LH receptor mRNA was detected in granulosa and theca interna cells (Fig. 5B) and was affected by the stage of follicular development (P < 0.001). Expression of LH receptor mRNA in granulosa and theca interna cells was greatest for follicles at the 6-mm stage of development. There was a follicular stage by cell type interaction (P < .05) because, in follicles at the 6-mm stage of development, the expression of LH receptor mRNA was greater in granulosa compared with theca interna cells, whereas both cell layers had similar expression in follicles at the 2-, 4-, and 8-mm stages. Ribonuclease protection analyses of pooled RNA samples showed maximal LH receptor mRNA in follicles at the 6-mm stage of development.
Steroidogenic enzyme mRNA
P45017 was detected in theca interna cells and P450arom was detected in granulosa cells (Fig. 6, A and B). The expression of P45017 and P450arom closely followed the expression pattern of LH receptor mRNA and was affected by the stage of follicular development (P < 0.01 and P < 0.001, respectively). Both P45017 and P450arom mRNA were maximal in follicles at the 6-mm stage of development. The pattern of mRNA expression (greatest mRNA expression in 6-mm follicles) for P45017 and P450arom was confirmed when pooled follicular RNA samples were analyzed by RPA.
|
IGF-I and type I IGF receptor mRNA IGF-I mRNA was detected in granulosa cells as well as theca externa cells (Fig. 7A). The amount of IGF-I mRNA expression was similar for granulosa and theca externa cells (i.e., there was no effect of cell type; P > 0.1). There was also no effect of follicular stage (2, 4, 6, or 8 mm) on IGF-I mRNA (P > 0.1). RPA showed a similar level of IGF-I mRNA expression at each follicular stage.
|
Granulosa cells expressed the type I IGF receptor mRNA (Fig. 7B). The expression of type I IGF receptor mRNA in granulosa cells was similar for follicles at different stages of development (P > 0.1). When measured by RPA, the relative amounts of type I IGF receptor mRNA were greatest in follicles at the 8-mm stage but this increase was not observed by in situ hybridization nor could the difference be explained by relatively greater GAPDH in the pooled 8-mm sample (data not shown).
IGF-II and type II IGF receptor mRNA IGF-II mRNA was detected in theca interna cells and in ovarian stroma cells (Fig. 8A). The level of IGF-II mRNA expression was similar for theca interna and stroma (P > 0.1). There was, however, a follicular stage by cell type interaction (P < 0.05) because there was no effect of follicular stage on IGF-II mRNA in stroma cells (P > 0.1), whereas there was an effect of follicular stage on IGF-II mRNA in theca interna cells (P < 0.01). In theca internae, IGF-II mRNA was greater in follicles at the 6-mm stage when compared with follicles at the 2- or 4-mm stages. Follicles at the 8-mm stage had decreased IGF-II mRNA that was intermediate between 4- and 6-mm follicles. The pattern of expression for IGF-II mRNA in theca internae was similar to that observed by RPA.
|
Type II IGF receptor mRNA was detected in granulosa and theca interna cells (Fig. 8B). There was an effect of cell type because the type II receptor mRNA was greater in granulosa compared with theca cells (14.7 ± 2.2 versus 5.4 ± 2.2, respectively; P < 0.01). Expression of mRNA was also affected by the stage of follicular development (P < 0.001). The stage by cell type interaction was not significant. For granulosa cells, the level of type II IGF receptor mRNA expression was similar in follicles at the 2-, 4-, and 6-mm stages of development, but increased when follicles reached the 8-mm stage. For theca interna cells, there was a tendency (P < 0.1) for an effect of follicular stage because the type II IGF receptor appeared to increase in follicles at the 8-mm stage of development. The increase in the type II IGF receptor mRNA in 8-mm follicles was also detected in RPA.
IGF binding protein mRNA Granulosa and theca interna cells expressed IGFBP-2 mRNA (Fig. 9A) and the expression of IGFBP-2 mRNA was greater in granulosa cells compared with theca cells (39.8 ± 2.4 versus 31.3 ± 2.4, respectively; P < 0.05). Furthermore, IGFBP-2 expression in granulosa cells appeared greatest in those cells closest to the basement membrane. The follicle stage by cell type interaction was not significant because the amount of IGFBP-2 mRNA was decreased in both granulosa (P < 0.001) and theca interna (P < 0.001) cells of large follicles (6- and 8-mm stages of development) relative to the amount of mRNA observed in small follicles (2- and 4-mm stages of development). RPA also showed a decrease in IGFBP-2 mRNA in large follicles (6 and 8 mm) relative to small follicles (2 and 4 mm).
|
IGFBP-4 mRNA was expressed in granulosa and theca interna cells (Fig. 9B). There was an effect of cell type because theca interna cells had greater IGFBP-4 than granulosa cells (14 ± 1.7 versus 2.9 ± 1.7, respectively; P < 0.001). The effect of follicular stage on IGFBP-4 mRNA amount was significant for theca interna (P < 0.001), but was not significant for granulosa (P > 0.1) cells. Expression of IGFBP-4 mRNA in theca interna cells was greater in follicles at the 8-mm stage when compared with follicles at the 2-, 4-, or 6-mm stages. The amount of IGFBP-4 mRNA appeared to increase in granulosa cells of 8-mm follicles as well, but this increase did not achieve statistical significance. The increase in IGFBP-4 mRNA in the theca interna of 8-mm follicles was reflected in the RPA.
GH receptor mRNA GH receptor mRNA was expressed in granulosa and theca interna cells at extremely low levels that were barely detectable above background (approximately 1% pixel density; Fig. 10). There was an effect of cell type because granulosa cells had greater GH receptor mRNA than theca interna cells (1.1 ± 0.3 versus 0.3 ± 0.3, respectively; P < 0.05). The effect of follicular stage and the stage by cell type interaction were not significant, although there was an apparent decline in GH receptor mRNA in theca interna that approached statistical significance (P < 0.14). The presence of low-level expression of GH receptor mRNA within ovarian follicles and the decline with advancing follicular stage was confirmed by RPA.
|
DISCUSSION
We measured the mRNA expression for gonadotropin receptors, steroidogenic enzymes, and IGF system genes during the maturation and differentiation of preovulatory follicles in the sow. Four developmental patterns of gene expression were found. The first pattern of gene expression was correlated with the development of the estrogenic capacity of the follicle. Four of 11 genes (P45017, P450arom, LH receptor, and IGF-II) were maximally expressed in 6-mm follicles when the greatest concentrations of estradiol in follicular fluid were observed. The increase in P45017 and P450arom mRNA in estrogenic follicles suggests that these mRNA are correlated with their respective proteins that have essential roles in follicular estrogen biosynthesis [25]. The increase in steroidogenic enzymes was associated with an increase in LH receptor mRNA in both granulosa and theca interna cells (Fig. 5B). The localization of LH receptor mRNA in both granulosa and theca interna cells was similar to cattle [20, 26], but pigs and cattle differed with respect to the temporal expression of LH receptor in the granulosa cell layer. In cattle, the expression of LH receptor in granulosa cells occurs during the initiation of follicular dominance, when follicles are relatively mature [20]. In pigs, the expression of LH receptor in granulosa cells occurred at 2 mm; a relatively immature follicular stage. Perhaps LH receptor expression in the granulosa cell layer at an early developmental stage enables the development of multiple dominant follicles in pigs through an LH-dependent mechanism [14, 15].
The P45017, P450arom, and LH receptor mRNA underwent a precipitous decline at the 8-mm stage after sows were observed in estrus and apparently had an LH surge (Fig. 1). IGF-II mRNA, however, remained elevated in 8-mm follicles. A decrease in steroidogenic mRNA after the LH surge was also observed in previous studies of cattle [27] and pigs [25]. Likewise, a decrease in LH receptor mRNA in ovarian follicles was observed in the rat after hCG treatment [28] and, collectively, the data suggest that the LH surge causes a major shift in gene expression. The mechanisms that led to the loss of P45017, P450arom, and LH receptor mRNA did not affect IGF-II mRNA, however, because IGF-II mRNA did not undergo a significant decline in 8-mm follicles. Assuming that IGF-II mRNA and protein are correlated, then the maintenance of IGF-II gene expression suggests that IGF-II may have additional functions in either ovulation or luteinization. Indeed, IGF-II has been shown to stimulate steroidogenesis in luteinized granulosa cells [29].
The second developmental pattern of gene expression was inversely correlated with the diameter of the follicles (i.e., a decrease in mRNA expression as follicles progressed to larger sizes [2 through 8 mm]). This pattern of gene expression was observed for FSH receptor, IGFBP-2, and GH receptor (theca cell layer). The loss of FSH receptor mRNA in large antral follicles confirms our previous observations [6] and suggests that FSH receptor mRNA is not required for terminal growth of antral follicles in the porcine ovary. The down-regulation of FSH receptor specifically occurs in developing preovulatory follicles because the few small antral follicles (1 to 2 mm) present on the ovary of sows at the 6- and 8-mm stages were expressing FSH receptor (data not shown). The pattern of FSH receptor gene expression in preovulatory porcine follicles was different from what was classically described for laboratory animals and cattle. In rats [30] and cattle [20, 26], the pattern of FSH receptor mRNA expression was similar to that of LH receptor mRNA during the estrous cycle because FSH receptor mRNA increased as follicles grew and became estrogenic. The data for FSH receptor mRNA in the present study (decreased FSH receptor mRNA in follicles larger than 2 mm) agree conceptually with those of Driancourt et al. [31] who measured gonadotropin responses in vivo and found that 1.1- to 2-mm porcine follicles were FSH-dependent but porcine follicles larger than 2 mm were LH-dependent. Likewise, our IGFBP-2 mRNA observations were in agreement with porcine studies of IGFBP-2 protein in antral follicles (i.e., decreased IGFBP-2 in larger-diameter follicles) [32] and extend those observations by showing changes in mRNA by in situ hybridization. Cattle undergo similar declines in IGFBP-2 but differ slightly from pigs because IGFBP-2 mRNA was not detected in theca cells of cattle [23, 33].
The 8-mm follicles were somewhat variable for IGFBP-2 in follicular fluid because two sows (lanes 3 and 5 of the 8-mm samples) had a strong IGFBP-2 signal on the ligand blot (Fig. 4A). The follicular fluid estrogen to progesterone ratios for these two sows were similar to other sows in the 8-mm group. Compared with other sows in the 8-mm group, however, their follicular fluid IGF-I protein was twofold to fourfold greater, and the IGF-I mRNA in granulosa was increased. Ovulation in the pig is poorly timed relative to estrus and the LH surge [34]; therefore, these two sows could have been at a more advanced stage (closer to ovulation) relative to the other sows in the 8-mm group. Additional studies with more precisely timed collections of 8-mm follicles may be necessary to clarify specific changes in IGF-I and IGFBP-2 during the late preovulatory and early postovulatory phases of ovarian development.
In situ hybridization analyses showed a tendency for GH receptor to decrease in the theca cell layer during the development of preovulatory follicles (Fig. 10). RPA also suggested a decrease in GH receptor in larger preovulatory follicles. We classified the GH receptor mRNA, therefore, as a gene that underwent a decline in expression during preovulatory follicular development and include it with the other two genes in this category (FSH receptor and IGFBP-2). The GH receptor mRNA was expressed at very low levels (approximately 1% pixel density) that were difficult to detect above background. The expression pattern was also somewhat variable between follicles (note the large standard error bars for granulosa in Fig. 10), and this variability combined with the very low expression level made it difficult to detect changes in mRNA. The physiological significance of extremely low GH receptor expression in pig ovary is not known. We did not measure GH receptor protein or binding, so we cannot directly correlate our mRNA observations with receptor protein or action. In vitro, GH amplified the actions of IGF-I on steroidogenesis in porcine granulosa cells [35–37]. A similar mechanism may exist in vivo within the preovulatory follicle. If true, the data from the present study suggest that the mechanism is most important in small follicles (smaller than 4 mm) when GH receptor mRNA (and presumably protein) is greatest.
The third developmental pattern was constitutive gene expression (i.e., no change in expression for 2- through 8-mm follicles). This was observed for two of the 11 genes that we studied (IGF-I and type I IGF receptor). We also observed more variability in gene expression for these mRNA across individual sows (note the relatively large standard errors in Fig. 7). The pattern of IGF-I and type I IGF receptor gene expression agreed with studies in several species [6, 7, 38–40] and supported our observation (as well as those made by others [32]) that IGF-I mRNA in follicles (Fig. 7A) and IGF-I protein in follicular fluid (Fig. 3) are not correlated with follicular growth. This does not mean that IGF-I is not physiologically important within the preovulatory follicle. Indeed, numerous gonadotropin-dependent and gonadotropin-independent actions of IGF-I have been described for both granulosa and theca cells [3–5, 9–11]. Under normal physiological conditions, a threshold of IGF-I protein in follicular fluid may be met by local (paracrine/autocrine) and endocrine sources of IGF-I [41]. The lack of developmental regulation for type I IGF receptor gene expression (Fig. 7B) suggests that IGF-I is required throughout the preovulatory period and highlights the importance of IGF-I and the type I IGF receptor for all phases of follicular growth that we studied (2 to 8 mm). The potential for IGF-I action in large follicles is enhanced by the coordinated decrease in IGFBP-2 gene expression (Fig. 9) and protein (Fig. 4) because IGF-I action in preovulatory follicles may be greater when IGFBP are decreased [42].
The fourth developmental pattern of gene expression was increased gene expression in 8-mm follicles. Sows with 8-mm follicles had expressed estrus and had elevated LH concentrations that were suggestive of an LH surge (Fig. 1). The follicles were not estrogenic and had the highest concentrations of progesterone when compared with follicles at the 2- through 6-mm stages (Fig. 2). Based on the follicular fluid estradiol and progesterone concentrations, we conclude that the 8-mm follicles were luteinized. Two of 11 genes (IGFBP-4 and type II IGF receptor) had increased mRNA expression in 8-mm follicles. We measured IGFBP-2 and IGFBP-4 mRNA because these represent the major IGFBP produced within granulosa and theca cells [23]. IGFBP-4 mRNA was expressed in theca interna with low but detectable expression in granulosa cells (Fig. 9B). Unlike IGFBP-2 mRNA, which decreased during follicular growth, IGFBP-4 mRNA was similar in 2-, 4-, and 6-mm follicles and increased in both granulosa and theca interna cells of 8-mm follicles. LH stimulated IGFBP-4 mRNA in vitro [23] and the increase in IGFBP-4 that we observed may have been caused by the LH surge in sows with 8-mm follicles. Given the large increase in IGFBP-4 mRNA immediately preceding ovulation (8-mm follicle) and assuming mRNA and protein are correlated, we predict that IGFBP-4 may play a role in the luteinization of ovarian cells and ovulation, or both. Future studies should address the role of IGFBP-4 in these processes.
There was also a large increase in type II IGF receptor in 8-mm follicles (Fig. 8B). The type II IGF receptor may signal through a G-protein and activate adenylate cyclase to increase intracellular cAMP [43]. Although this mechanism for type II IGF receptor second-messenger signaling is appealing, it has never been demonstrated within nontransfected ovarian cells. A more likely function for the type II IGF receptor is to bind IGF-II and target IGF-II to intracellular lysosomes [12]. This targeting mechanism may prevent tumor formation in cells that overexpress IGF-II [12] and could control cell growth and prevent tumor formation during luteinization. This potential function for the type II IGF receptor in ovarian cells should be examined in future studies.
In summary, coordination of gonadotropin receptor, steroidogenic enzyme, and IGF system gene expression was demonstrated during a rapid period of follicular growth. Four developmental patterns of gene expression were observed for the ovarian mRNA. The first developmental pattern of gene expression was correlated with the estrogenic capacity of the follicle and included genes for estrogen biosynthesis, IGF-II, and LH receptor. The second developmental pattern of gene expression was inversely correlated with follicular growth and included genes with decreased expression in large follicles (FSH receptor, IGFBP-2, and GH receptor [theca cell layer]). The third developmental pattern was constitutive (i.e., no obvious developmental pattern; IGF-I and type I IGF receptor), and the fourth developmental pattern was increased expression after estrus and the LH surge (IGFBP-4 and type II IGF receptor). These data demonstrate the dynamic nature of ovarian gene expression with genes of specific classes undergoing unique expression patterns in support of follicular growth. Future studies should test a wider array of genes (both mRNA and corresponding proteins) for their roles in follicular growth.
FOOTNOTES
First decision: 8 December 1999.
1 The USDA National Research Initiative Competitive Grants Program supported this research. Grant CSREES 96-35203-3255 awarded to M.C.L. Contribution from the Missouri Agricultural Experiment Station. Journal Series 12,997. 
2 Correspondence: Matthew C. Lucy, 164 Animal Science Research Center, University of Missouri, Columbia, MO 65211. FAX: 573 882 6827; lucym{at}missouri.edu
3 Current address: Lincoln University, Jefferson City, MO 65102.
Accepted: May 1, 2000.
Received: November 16, 1999.
REFERENCES
-hydroxylase, and cytochrome P450 aromatase in bovine follicles during the first follicular wave. Endocrinology 1995; 136:981–989.[Abstract]-hydroxylase and P450 aromatase in preovulatory bovine follicles decrease after the luteinizing hormone surge. Endocrinology 1993; 132:2239–2245.This article has been cited by other articles:
|
|
 |
|
  L K Doyle, C O Hogg, E D Watson, and F X Donadeu Seasonal effects on the response of ovarian follicles to IGF1 in mares Reproduction, November 1, 2008; 136(5): 589 - 598. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. C. Lucy The bovine dominant ovarian follicle J Anim Sci, March 1, 2007; 85(13_suppl): E89 - E99. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. J. Bracken, R. P. Radcliff, B. L. McCormack, D. H. Keisler, and M. C. Lucy Decreased follicular size during late lactation caused by treatment with charcoal-treated follicular fluid delays onset of estrus and ovulation after weaning in sows J Anim Sci, August 1, 2006; 84(8): 2110 - 2117. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C. Agca, J. E Ries, S. J Kolath, J.-H. Kim, L. J Forrester, E. Antoniou, K. M Whitworth, N. Mathialagan, G. K Springer, R. S Prather, et al. Luteinization of porcine preovulatory follicles leads to systematic changes in follicular gene expression Reproduction, July 1, 2006; 132(1): 133 - 145. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. D. Guthrie The follicular phase in pigs: Follicle populations, circulating hormones, follicle factors and oocytes J Anim Sci, June 1, 2005; 83(13_suppl): E79 - 89. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. Jiang, K. M. Whitworth, N. J. Bivens, J. E. Ries, R. J. Woods, L. J. Forrester, G. K. Springer, N. Mathialagan, C. Agca, R. S. Prather, et al. Large-Scale Generation and Analysis of Expressed Sequence Tags from Porcine Ovary Biol Reprod, December 1, 2004; 71(6): 1991 - 2002. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
C.J. Corbin, F.M. Moran, J.D. Vidal, J.J. Ford, T. Wise, S.M. Mapes, V.C. Njar, A.M. Brodie, and A.J. Conley Biochemical Assessment of Limits to Estrogen Synthesis in Porcine Follicles Biol Reprod, August 1, 2003; 69(2): 390 - 397. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. Monget, S. Mazerbourg, T. Delpuech, M.-C. Maurel, S. Maniere, J. Zapf, G. Lalmanach, C. Oxvig, and M. T. Overgaard Pregnancy-Associated Plasma Protein-A Is Involved in Insulin-Like Growth Factor Binding Protein-2 (IGFBP-2) Proteolytic Degradation in Bovine and Porcine Preovulatory Follicles: Identification of Cleavage Site and Characterization of IGFBP-2 Degradation Biol Reprod, January 1, 2003; 68(1): 77 - 86. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 J. A. Arraztoa, P. Monget, C. Bondy, and J. Zhou Expression Patterns of Insulin-Like Growth Factor-Binding Proteins 1, 2, 3, 5, and 6 in the Mid-Cycle Monkey Ovary J. Clin. Endocrinol. Metab., November 1, 2002; 87(11): 5220 - 5228. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Eimerl and J. Orly Regulation of Steroidogenic Genes by Insulin-Like Growth Factor-1 and Follicle-Stimulating Hormone: Differential Responses of Cytochrome P450 Side-Chain Cleavage, Steroidogenic Acute Regulatory Protein, and 3{beta}-Hydroxysteroid Dehydrogenase/Isomerase in Rat Granulosa Cells Biol Reprod, September 1, 2002; 67(3): 900 - 910. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
