| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Regular Article |
a Department of Medicine, Flinders University of South Australia, Bedford Park, South Australia 5042, Australia
b Howard Florey Institute, University of Melbourne, Parkville, Victoria 3010, Australia
c Institute for Hormone and Fertility Research, University of Hamburg, Hamburg 22529, Germany
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
follicle, follicular development, ovary, relaxin, steroid hormones, thecal cells
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In the majority of species, the apparent levels of RLF mRNA and protein found in nontesticular tissues is generally very low, often only detectable by reverse transcription polymerase chain reaction (RT-PCR) or very sensitive immunohistochemistry [3]. An exception is the ovary of cows and sheep, where both thecal cells and CLs are the sites of a very high level of expression, comparable to that in the testicular Leydig cells. The pattern of RLF mRNA expression in the bovine CL is very similar to that of relaxin in some other species [10, 11]. RLF may be expressed as a compensatory reaction to the apparent deletion of the relaxin gene in ruminants [12, 13]. The highest expression of RLF mRNA in the cow is in follicular theca interna [11]. RLF mRNA appears to be expressed in the theca interna at a higher level in small antral follicles than in larger, presumably preovulatory, follicles. As in the Leydig cells of the testis [5], the pattern of RLF gene expression appears to reflect the differentiation state of the ovarian thecal and luteal cells.
At the protein level, specific antibodies raised against either recombinant protein or peptide fragments have detected RLF protein in Leydig cells, CLs, and theca interna of different species [3, 6]. The expression patterns corresponded to the mRNA localization determined by in situ and Northern hybridization and by RT-PCR. Studies of RLF in theca interna have been hampered by the lack of antisera that react with RLF from species that express RLF at high levels in the theca interna, such as ruminants. The theca interna is a complex stromal layer that differentiates in bovine follicles at around the time of antrum formation. It has many cell types, including endothelial and smooth muscle cells of the vasculature, steroidogenic differentiated thecal cells, connective tissue fibroblasts, and leukocytes. It is highly plastic and expands during antrum expansion as follicles grow. In addition, the theca interna appears to play a role in one of two types of atresia that occur in bovine follicles [14].
The role of RLF in testis and ovary is not known, but studies using knockout mice, where the INSL3 gene was partially ablated, suggest that it is important for reproduction [15, 16]. The principal defect in the knockout mice was the failure of the testes to descend into the scrotum. Female homozygous mice were fertile, although litter size appeared to be slightly reduced, and most noticeably the length of the estrous cycles was extended from the usual 4–5 days to >15 days [15]. Given the consistent pattern of RLF expression in follicles and CLs, these results suggest that RLF is probably playing an important role in follicular maturation or luteal function.
We produced antisera against recombinant bovine pro-RLF and synthetic mature ovine RLF. Using both types of antisera, we examined the expression of RLF in bovine follicles during follicular growth and atresia. Atretic follicles were also further classified as undergoing antral or basal atresia. These terms refer to the location in the follicle where death of granulosa cells occurs first; the latter type appears to involve thecal destruction and macrophage infiltration as an early event [14]. The enzymes involved in the synthesis of pregnenolone, cytochrome P450 side chain cleavage (SCC), and in the synthesis of progesterone, 3ß-hydroxysteroid dehydrogenase (HSD), were also immunolocalized, thus allowing us to identify precisely when RLF is first expressed, relative to these markers of cell type and maturation, and in which cell types of the theca interna.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The cDNA for the bovine RLF-encoding transcript [12] was recloned into the expression vector pH6Ex5, a derivative of pGEX-5T [17], such that the pro-form, lacking the signal peptide, was directly colinear and in-frame with the N-terminal hexahistidine Tag incorporated into the multiple cloning site of the vector. The N-terminus of the resulting fusion protein was thus MSPIHHHHHHLVPRGSAQEA—, whereby the AQEA residues represent the postulated N-terminus of the RLF propeptide comprising the complete B, C, and A domains in order [11]. Correct plasmid construction was confirmed by sequencing. Competent Escherichia coli BL21(DE3) pLysS bacteria (Novagen, Madison, WI) were transformed with this plasmid construct, grown to an A575 of 0.5, induced with 0.4 mM isopropyl thiogalactopyranoside, and further incubated for 1–1.5 h. Longer induction times led to a considerable loss of the recombinant product. Bacteria were harvested by centrifugation, and the pellets stored at -70°C. Bacterial pellets were thawed, sonicated (30 sec) in 15 ml of sonication buffer (50 mM sodium phosphate, pH 8.0, 300 mM NaCl, 1% Tween-20), and centrifuged for 30 min at 4°C to collect the cell pellet containing inclusion bodies. This pellet was resuspended in 10 ml of buffer A (6 M guanidinium hydrochloride, 0.1 M sodium phosphate, 10 mM Tris-HCl, pH 8.0), sonicated (2 x 45 sec on ice), and then incubated for 1 h at room temperature with gentle mixing. This homogenate was centrifuged again, and the supernatant was incubated with 5 ml of 50% (w/v) Ni-NTA agarose for 1 h at room temperature. Following chelation of the recombinant protein, the agarose was transferred to a column, progressively washed with 35 ml of buffer A, 20 ml of buffer B (8 M urea, 0.1 M sodium phosphate, 10 mM Tris-HCl, pH 8.0), and 20 ml of buffer C (8 M urea, 0.1 M sodium phosphate, 10 mM Tris-HCl, pH 6.3), and eluted with sequential 3-ml aliquots of buffer C with the addition of 0.25 M imidazol. These eluate fractions were monitored by SDS-PAGE and stained with Coomassie blue R250. The recombinant protein was clearly evident as a prominent 14.5-kDa band, and a Western blot using anti-5His antibodies (no. 34660, 1:1000; Qiagen, Hilden, Germany) verified its identity. Because of the presence of other larger nonspecific proteins in these eluate fractions, the fractions were combined and subjected to a second round of Ni-NTA chelation chromatography. The resulting recombinant protein was estimated by SDS-PAGE to be >80% pure. The resulting eluates were then dialyzed overnight at 4°C against PBS (Dulbeco, without Ca2+ and Mg2+, no. 14200; Life Technologies, Karlsruhe, Germany), whereby the recombinant protein precipitated out. The precipitate was centrifuged, the pellet was resuspended in 1.5 ml of PBS, and the protein concentration was measured (total yield of 0.5–1.0 mg per 500 ml bacterial culture). Polyclonal antibodies were then generated in rats in our Hamburg laboratory using the suspended recombinant protein as immunogen, as previously described [5]. Of three sera showing specific Leydig cell immunoreactivity in preliminary tests, antiserum R33 indicated the highest specific titer and was used for all further studies. An additional antiserum (W3) was raised in a rabbit against a synthetic peptide (CGGPRWSSEEDG) from the predicted B domain of bovine RLF. This antiserum was used here only to immunopurify recombinantly expressed bovine RLF.
Antisera to Ovine RLF B Chain
Sheep RLF A and B chains were assembled by Fmoc solid-phase peptide synthesis and purified by conventional reverse-phase HPLC. The RLF A/B heterodimer was then produced by combining the chains in solution at high pH as previously described [18]. The purity and identity of the peptides was confirmed by chemical characterization including mass spectrometry [18]. The A/B heterodimer was used for testing the antibody by dot blotting.
Antibodies were raised in three rabbits in our Melbourne laboratory using limpet hemocyanin conjugated to the B chain only [19]. The specificity of the antiserum was tested by dot blotting against the RLF B chain and the complete RLF molecule. For the antibody (designated HFA) with the highest affinity for the RLF peptide, the IgG fraction was isolated using a Protein A column (HiTrap; Amersham Pharmacia, Uppsala, Sweden).
Dot Blot and Western Blot Analyses
For dot blot analyses, peptides or recombinant proteins were blotted onto nitrocellulose membranes (Millipore, Bedford, MA) dampened with distilled water. Membranes were allowed to dry for 30 min, were dampened again with distilled water, and were washed with 20 mM Tris buffer (pH 7.5) containing 0.5 M NaCl (TBS) for 5 min. The membranes were incubated for 1 h in 10% skim milk powder in TBS, washed briefly with TBS, and incubated with the relevant RLF antibodies in 3% skim milk powder in TTBS (TBS containing 0.05% Tween 20) for 1 h at room temperature on an orbital shaker. The membranes were then washed three times with TTBS for 10 min each time and incubated with anti-rabbit horseradish peroxidase-linked secondary antibody (1:2500; Biorad, Sydney, Australia) or anti-rat horseradish peroxidase-linked secondary antibody (1:2500, Chemicon, Temecula, CA) in 3% skim milk powder in TTBS for 1 h. Membranes were washed three times in TTBS for 10 min each time, and antibody-peptide complexes were visualized by chemiluminescence (ECL Western blotting detection kit; Amersham Pharmacia).
Western blot analyses were also carried out on recombinantly expressed pro-RLF that had been purified by immunoprecipitation. Two microliters (
20 ng) and 10 µl (100 ng) of gel-purified recombinant pro-bovine RLF were incubated with 5 µl of either rabbit preimmune serum or serum W3 in 490 µl of 10 mM Tris-HCl buffer, pH 7.5, containing 150 mM NaCl and 0.05% Tween 20, for 1.5 h at 4°C on a rotating wheel. Then 50 µl of Protein-A-Sepharose Fast Flow Suspension (50% slurry; Amersham Pharmacia) was added, and the incubation was continued for 1.5 h. Bound immune complexes were pelletted by low-speed centrifugation for 20 sec. Pellets were then washed three times in 10 mM Tris-HCl buffer, pH 7.5, containing 150 mM NaCl and 0.05% Tween 20 and then once in 50 mM Tris-HCl buffer, pH 7.5, and finally resuspended in 30 µl of 50 mM Tris-HCl buffer, pH 7.5, containing 1% SDS and 100 mM dithiothreitol. After heating to 95°C for 5 min and briefly centrifuging, 20 µl of the supernatant was subjected to SDS-PAGE and Western blotting using antibody R33, following standard procedures as per dot blotting.
Additional Antibodies
Serum from a rabbit immunized against bovine SCC purified from adrenal mitochondria [20] was obtained from OXYgene (Dallas, TX) as a gift from Dr. E. Simpson. Mouse monoclonal antibody FDO66Q was initially raised against JEG choriocarcinoma cells [21] and IgG purified from mouse ascites fluid. This anitbody recognizes an epitope very similar to FDO161G [22, 23], which recognizes human 3ßHSD type I. Type II is expressed in ovaries and is 94% homologous with type I in humans. FDO66Q binds to human ovarian follicles and Leydig cells (B. Kalionis, personal communication), indicating that it cross-reacts with 3ßHSD type II.
Tissues and Histology
Antral follicles were harvested from the ovaries of young nonpregnant cows (Bos taurus) slaughtered at the local abattoir in South Australia and were transported to the laboratory in transport medium on ice. For small follicles (2–5 mm), 2 were collected per ovary, and for large follicles (6–17 mm), 1, 2, or 3 follicles were collected per ovary from 32, 2, and 1 ovary, respectively. Follicles and adhering stroma were dissected from each ovary and snap-frozen in Tissue Tek OCT compound (Miles Inc., Elkhart, IN). Small (2–5 mm, n = 87) or large (6–17 mm, n = 21) frozen follicles were bisected, and one half was immersed in 2.5% glutaraldehyde, postfixed in osmium tetroxide, and embedded in epoxy resin. Sections were cut at a thickness of 1 µm, stained with methylene blue, and examined by light microscopy. Thirteen large follicles (6–17 mm) were frozen in OCT only. Antral follicles were classified by light microscopy of methylene blue-stained sections where available or of frozen sections if resin-embedded sections were not available.
Immunohistochemistry
Tissue sections were cut from OCT-embedded bovine follicles using a CM 1800 cryostat (Leica Microsystems Pty. Ltd., Victoria, Australia), collected on glass slides treated with 0.01% poly-L-lysine hydrobromide (P-1524; Sigma Chemical Co., St. Louis, MO), and stored at -20°C until use. Unfixed sections of bovine follicles were dried under vacuum for 5 min and then incubated in 10% normal donkey serum (D-9663; Sigma) in antibody diluent containing 0.55 M sodium chloride and 10 mM sodium phosphate (pH 7.1) for 20 min. Sections were then incubated overnight with a combination of primary antibodies (rat anti-bovine pro-RLF, R33, 1:200; mouse anti-human 3ß-HSD, FDO66Q, 1:1000) or with rabbit anti-bovine SCC (1 µg/ml IgG) alone. Secondary antibodies used were fluorescein isothiocyanate (FITC)-conjugated AffiniPure donkey anti-rat IgG (cat. 712-096-153, 1:100) and biotin-SP-conjugated AffiniPure donkey anti-mouse IgG (715-066-151, 1:100) or biotin-SP-conjugated AffiniPure donkey anti-rabbit IgG (711-066-152, 1:200), followed by Cy3 conjugated-streptavidin (016-160-084, 1:100), all from Jackson ImmunoResearch Laboratories (West Grove, PA), in antibody diluent. Immunostaining for pro-RLF was colocalized with 3ß-HSD (n = 86), and staining for SCC was undertaken on another section (n = 66). Pro-RLF was coimmunolocalized with mature RLF (HFA antiserum, 1:20) as above (n = 11), secondary antibodies employed were Cy3-conjugated AffiniPure donkey anti-rabbit IgG (711-166-152, 1:100) and biotin-SP-conjugated AffiniPure donkey anti-rat IgG (712-066-153, 1:100), followed by FITC-conjugated streptavidin (016-090-084, 1:100). All incubations were carried out at room temperature in a humidified chamber. Following incubation with primary or secondary antibodies or streptavidin-conjugated reagents, sections were washed (3 x 5 min) in hypertonic PBS containing 0.274 M sodium chloride, 5.4 M potassium chloride, and 10 mM sodium phosphate, pH 7.2.
Light Microscopic Observations and Photography
Sections of bovine ovary stained with methylene blue were examined using an BX50 microscope (Olympus Optical Co. Ltd., Tokyo, Japan) and photographed with a SC35 camera attachment (Olympus) and FP-4 125 black-and-white film (Ilford Imaging UK Ltd., Cheshire, U.K.). Sections processed for immunofluorescence staining were observed and photographed with a Vanox AHBT3 fluorescence microscope (Olympus) with a C35AD-4 camera attachment (Olympus) and photographed with T-Max 400 black-and-white film (Kodak, Rochester, NY).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Polyclonal antisera were raised in rats against bovine pro-RLF recombinantly expressed in E. coli. Of these antisera, antiserum R33 had the highest specific titer in immunohistochemistry of bovine testis sections and was used in the present study. In dot blot analysis (Fig. 1) against the same batch of nickel-agarose-purified recombinant pro-RLF used as the immunogen and against the chemically synthesized mature RLF (ovine A and B chains covalently linked), antibody R33 bound to both forms, although more strongly to the pro-form (Fig. 1). Additional recent blots (not shown) observed that this antibody bound to the ovine A chain but not detectably to the ovine B chain. To additionally confirm the specificity of this antiserum, Western blots analyses of E. coli recombinantly expressed, gel-purified pro-bovine RLF, which was further purified by immunoprecipitation with another antibody directed against the B chain of bovine RLF (W3), clearly showed that this antiserum detects RLF pro-form at 14 kDa (Fig. 2).
|
|
Rabbit antiserum HFA was raised against the ovine B chain. The ovine RLF B chain is identical to the bovine form except for one less Glu residue at position 30, and the ovine A chain is identical to the bovine form except for a conservative substitution at position 4 of Val for Ile. The antibody clearly bound to the mature form of RLF (ovine A and B chains covalently linked; Fig. 1). Binding to pro-RLF was tested using pro-RLF recombinantly expressed in E. coli or Cos cells. In the Cos cells, the pro-RLF was not further proteolytically processed into C and A/B fragments, as determined by Western analysis using antisera to the N-terminal extension containing the FLAG sequence (data not shown). Binding of HFA antibody to either of these bovine pro-RLF preparations was not significantly greater than that to a nonspecific control on dot blotting (Fig. 1).
The anti-pro-RLF antibody R33 bound to bovine Leydig cells and a population of cells in the theca interna in Bouin-fixed and paraffin-embedded tissue (data not shown). This finding was expected, based upon results from other species [3] and identification of RNA in these bovine cells [12]. Staining was cytoplasmic. Colocalization with anti-pro-RLF R33 and anti-RLF B chain (HFA antiserum) using bovine tissues showed that pro-RLF and RLF B chain were colocalized in the theca interna of 11 healthy follicles, 2–5 mm (Fig. 3), and 11 antral atretic follicles (data not shown). Thus, all results obtained suggest that both antibodies clearly were specific for RLF moities and both reacted with bovine RLF.
|
Follicle Classification
Follicles were classified on the basis of their morphology, using epoxy resin-embedded material where available or frozen sections, as being either healthy (n = 91 for small follicles; n = 28 for large follicles) or atretic. The atretic follicles were further classified as undergoing either antral (n = 29 for small follicles; n = 16 for large follicles) or basal (n = 18 for small follicles) atresia. This classification of atresia has recently been described and characterized at the light and electron microscope level. It simplifies all types of atresia previously reported in bovine follicles [14] into two simple and distinct forms.
Antral atretic follicles had the classic features of atretic follicles (Fig. 4b). They had numerous pyknotic nuclei present in either the layer(s) of the membrana granulosa closest to the antrum or in the antrum itself in close proximity to the membrana granulosa. As atresia progressed, the first cells to die were nearest the antrum, and cell death proceeded progressively towards the follicular basal lamina. Basal atresia, in contrast to antral atresia, exhibited destruction of the most basal layer of granulosa cells (Fig. 4c) while the most antral granulosa cells remained healthy and closely opposed to each other. The granulosa cells close to the antrum contained morphologically normal nuclei. The cells in the most basal layer of the membrana granulosa were separated from each other and often from the basal lamina by large intercellular spaces. Capillaries and macrophages had breached the basal lamina and were present in the basal area [14]. As with previous observations [14], basal atresia was observed only in smaller (<5 mm) follicles.
|
Follicles were arbitrarily classified as early, middle, or late atretic. Early antral atretic follicles had pyknotic nuclei in the antra, middle atretic follicles had these nuclei in the antral layers, and late atretic follicles had advanced pyknosis, such that only one or two layers of healthy basal cells remained. Basal atresia was classified as middle or late; early stage atretic cells were difficult to differentiate from healthy cells using light microscopy. Cell death in middle basal atretic follicles occurred in the basal layer, with partial expansion of matrix there. In late basal atresia, extensive basal areas of the former membrana granulosa were occupied by fluid, cell debris, and macrophages.
SCC and 3ß-HSD Expression
SCC and 3ß-HSD were localized to cells of the theca interna (Figs. 5–7). Staining was cytoplasmic and was not evenly distributed in cells. Only a portion of the thecal cells stained positively, as expected given the large number of nonsteroidogenic cells present in the theca, in particular fibroblasts and endothelial cells. Both small and large antral follicles had thecal cells expressing both of these enzymes (Table 1). With atresia of small follicles, the theca continued to express both enzymes. With atresia of large follicles, many follicles had ceased to express 3ß-HSD and about 50% had ceased to express SCC in the theca interna (Table 1). The membrana granulosa cells generally did not express SCC or 3ß-HSD. However, on reaching large antral size the membrana granulosa of most follicles expressed 3ß-HSD and some expressed SCC. With basal atresia, the remaining antrally located granulosa cells also expressed both SCC and 3ß-HSD.
|
|
RLF Expression
Colocalization of pro-RLF with 3ß-HSD showed that cells staining positive for pro-RLF were steroidogenic thecal cells (Figs. 5–7). In small healthy follicles, a few cells contained 3ß-HSD and not pro-RLF (Fig. 5). This finding could indicate a separate population of RLF-negative 3ß-HSD-positive cells or could indicate a difference in the expression of the two molecules or differing levels of antibody affinity resulting in detection of one molecule and not the other.
Expression of pro-RLF first occurred coincidently with expression of SCC and 3ß-HSD in the theca interna of very small antral follicles when the theca interna first developed (results not shown). Pro-RLF, SCC, and 3ß-HSD all continued to be expressed in the theca interna throughout development to the stage of healthy small (2–5 mm) antral follicles (Fig. 5). Then, with atresia the expression pattern in the theca interna diverged between pro-RLF, SCC, and 3ß-HSD. With antral atresia of small follicles, pro-RLF, SCC, and 3ß-HSD all continued to be expressed in the theca interna until late atresia, when the proportion of follicles expressing pro-RLF declined (Fig. 6 and Table 1). In contrast, with basal atresia the proportion of follicles expressing pro-RLF declined markedly by middle atresia (Fig. 7 and Table 1). These differences in when pro-RLF expression declined in the two types of atresia relate to the degree to which atresia had advanced, as visually assessed. This finding does not necessarily mean that pro-RLF expression declines more quickly in one type of atresia relative to the other, since only antral atretic follicles have been examined at different time points as atresia progressed [14]. As follicles enlarged, expression patterns of pro-RLF, SCC, and 3ß-HSD in the theca interna also diverged (Table 1). As healthy follicles enlarged (>10 mm), the proportion of follicles expressing pro-RLF declined to about 50% (Table 1), unlike SCC and 3ß-HSD. Upon atresia in follicles >6 mm, which are only of the antral type, the proportion of follicles expressing RLF also declined, unlike SCC and 3ß-HSD (Table 1). The decline in RLF expression seen in atresia or during growth of follicles was both qualitative (cells staining less intensely) and quantitative (fewer positive-staining cells) (compare Figs. 5b and 6b).
|
|
Expression of pro-RLF in the theca interna was compared with expression of SCC and 3ß-HSD in the membrana granulosa (Table 2). A significant (P < 0.001, 
2 test) inverse relationship in the expression of pro-RLF in the theca interna and 3ß-HSD in the membrana granulosa was observed. No significant (P > 0.05, 2 test) relationship was observed with expression of SCC in the membrana granulosa.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In the current study follicle status was assess by histological evaluation of health and atresia [14] and by localization of the steroidogenic enzymes SCC and 3ß-HSD. Studies of the expression of 3ß-HSD [24] and SCC [25, 26] have previously been undertaken in bovine follicles. Although the thecal cells always expressed these enzymes, the granulosa cells in most follicles did not express SCC or 3ß-HSD however those in some follicles did [26]. Similar results were obtained in the present study, but with many larger follicles in this study it is clear that the expression in granulosa cells develops as the follicles enlarge to preovulatory size (>10 mm).
Two anti-RLF antisera were studied in detail. One was produced in rat against recombinantly expressed bovine pro-RLF containing the A, B, and C domains but without the predicted signal peptide, which is likely to be removed in vivo. This antibody bound to both pro-RLF and the mature form, although it bound more strongly to the pro-form than did the second antibody. Additional recent blots suggest anti-pro-RLF binds preferentially to the A rather than the B chain. Reactivity to the C chain has not been tested, although binding is also likely to occur. The second antibody was raised against the ovine B chain, which differs in only one amino acid from the bovine B chain. This second antibody clearly bound to the mature form of ovine RLF, but binding to the pro-RLF was not significantly greater than binding to a negative control. The first antiserum reacted with Leydig cells as expected. When used to coimmunostain follicles, both antisera reacted with the same cells; the pro-RLF was colocalized to 3ß-HSD-positive cells and thus RLF is expressed in steroidogenic thecal cells, similar to the situation in testis where Leydig cells express RLF [3]. RLF mRNA but not protein has been examined in bovine follicles previously [11, 12]; thus, the current results extend these findings to show that the mRNA is translated into pro-RLF. Very little is known about the processing of pro-RLF in vivo and whether it is processed through to the A/B mature form, as is relaxin. Based upon the differential binding ability of the HFA antibody on dot blots, it is tempting to speculate that pro-RLF is processed into mature RLF in bovine thecal cells. However, additional experiments are required to verify this assumption.
It has been suggested that RLF is a marker for differentiation of the theca interna [11]. In the theca interna of some follicles, not all steroidogenic cells expressed pro-RLF, whereas the converse did not occur. In addition, we observed whole follicles (atretic or large healthy follicles) in which the theca interna failed to express any pro-RLF but continued to express SCC and 3ß-HSD. The precursor small healthy follicles expressed pro-RLF, which implies that at some stage of follicle growth pro-RLF expression was switched off. The trigger for this event is not clear. One possibility is that transcription of RLF, SCC, and 3ß-HSD all cease simultaneously but that mRNA or protein for both SCC and 3ß-HSD have longer cellular half-lives, on the order of days [27], whereas those for pro-RLF have much shorter half-lives. Alternatively, the mechanisms of regulation of the gene promoters may be governed by different factors. Although cloning of the bovine RLF gene promoter (unpublished results) revealed the presence of steroidogenic factor 1 responsive elements as in the promoters of the SCC and 3ß-HSD genes, long-term primary cell culture experiments showed that LH has opposite effects on RLF and SCC gene expression. RLF was downregulated and SCC was upregulated by the gonadotropin [11], implicating divergent regulatory mechanisms.
RLF mRNA in bovine ovaries was found previously by in situ hybridization [12] to be present in the theca interna and to be expressed in cultured thecal cells [11]. Northern analysis of follicles and CLs [6] have also been carried out. Although the early CL has minimal RLF mRNA or protein, both are readily detectable later in the luteal phase [11, 28]. Thus, a decline in RLF expression occurs late in follicular development, during ovulation, or during formation of a CL. The results of the present study, examining follicles up to large preovulatory stages, suggest that the decline in RLF expression may precede the LH surge that initiates ovulation and instead may reflect the maturing status of the ovarian follicle. This pattern of expression is not unlike the preovulatory and pre-LH surge expression of SCC, 3ß-HSD, cytochrome P450 aromatase, and LH receptor accompanied by a decline in expression of FSH receptors in the membrana granulosa. However, even though these dynamic changes have been recognized in the membrana granulosa, no changes such as those associated with RLF have been previously recognized in the theca interna.
The physiological role of RLF is poorly understood, but RLF knockout mice have been developed [15, 16]. The male phenotype is characterized by failure of the testes to descend. In females, extension of the length of estrous cycle has been observed in one strain of mice [15] but not in another [16]. An accelerated increase in the number of regressed follicles and decreased levels of remaining CLs were observed in ovaries of the knockout mice [29], suggesting that in mice the RLF may have a role related to follicular atresia and luteal regression. In the current study in bovine cells, we found that RLF expression in the theca interna declined upon atresia of the follicles, particularly when atresia was more advanced, suggesting that RLF expression could be related to atresia. The RLF receptor has not been identified or localized, nor have the target cells for RLF been identified. Could the target be or include the granulosa cells, and could RLF play a negative role inhibiting 3ß-HSD expression in the membrana granulosa cells? In our study, SCC and 3ß-HSD were not coexpressed in the membrana granulosa in many follicles. What was striking was the significant (P < 0.001) inverse level of expression of RLF in the theca and of 3ß-HSD in the membrana granulosa. RLF was expressed in small and healthy antral follicles, and no expression of 3ß-HSD was observed in the granulosa cells. As follicles enlarged or entered atresia, the incidence of RLF expression among follicles declined, and the expression of 3ß-HSD in the membrana granulosa increased. This relationship may be a consequence of follicle maturation or may be the result of cause and effect. The answer awaits further investigation.
In the pig [30] and the marmoset [31, 32], the related molecule relaxin is also expressed by the cells of the theca interna and the CL. Relaxin can influence the steroidogenic capacity of thecal cells [32], granulosa cell replication [33], and insulin-like growth factor 1 production [34]. However, the classical role of relaxin is to regulate tissue remodelling by stimulating degradation of structural collagens [10]. Both the thecal layers [35] and the membrana granulosa [36] are extensively remodelled, particularly as the antral cavity of the follicle enlarges and expands the layers laterally [36]. The membrana granulosa offers little resistance for lateral expansion, but the thecal layers in healthy follicles contain structural collagens [26, 37] that must be degraded during the process of lateral expansion. Relaxin secretion in the theca could stimulate collagen breakdown. In support of this hypothsis, injection of anti-relaxin antibodies into the ovarian bursa of the rat blocked ovulation, probably because of relaxin's stimulatory effect on collagenases [38]. In ruminant species, which appear not to have a relaxin gene, RLF may substitute for relaxin, in which case thecal expression of RLF may be important for thecal remodelling during follicular growth.
Expression of RLF occurs in the steroidogenic cells of the theca interna of bovine follicles. Its expression dynamically changes with follicle development, with RLF levels highest at the early antral stages but declining well in advance of SCC or 3ß-HSD expression in the theca interna. The results presented here suggest that, as in Leydig cells, RLF in bovine thecal cells is correlated with differentiation of thecal cells but only in the early stages of the differentiation process. The answer to what RLF is doing must await the availability of more new tools, such as peptides, with which to manipulate the follicle in vivo and in vitro.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
1 This work was supported by the Flinders Medical Centre Research Foundation, the Flinders University of South Australia, and the National Health and Medical Research Council of Australia (NHMRC). R.A.D.B. was a recipient of an NHMRC Howard Florey Centenary Fellowship. Research was also funded by grants from the Deutsche Forschungsgemeinschaft to R.I. (Iv7/1-4, Iv7/9-1) and by a block grant to the Howard Florey Institute from the NHMRC. 
2 Correspondence. FAX: 61 8 8222 7521; ray.rodgers{at}adelaide.edu.au
3 Current address: Reproductive Medicine Unit, Department of Obstetrics and Gynaecology, University of Adelaide, Adelaide, SA 5000, Australia
Accepted: October 31, 2001.
Received: March 29, 2001.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
-hydroxylase cytochrome P-450 in bovine ovarian follicles. J Reprod Fertil 1986; 78:627-638This article has been cited by other articles:
|
|
 |
|
  E. Lague and J. J. Tremblay Antagonistic Effects of Testosterone and the Endocrine Disruptor Mono-(2-Ethylhexyl) Phthalate on INSL3 Transcription in Leydig Cells Endocrinology, September 1, 2008; 149(9): 4688 - 4694. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 A. Gambineri, L. Patton, R. De Iasio, F. Palladoro, U. Pagotto, and R. Pasquali Insulin-Like Factor 3: A New Circulating Hormone Related to Luteinizing Hormone-Dependent Ovarian Hyperandrogenism in the Polycystic Ovary Syndrome J. Clin. Endocrinol. Metab., June 1, 2007; 92(6): 2066 - 2073. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H. F. Irving-Rodgers, B. E. Friden, S. E. Morris, H. D. Mason, M. Brannstrom, K. Sekiguchi, N. Sanzen, L. M. Sorokin, Y. Sado, Y. Ninomiya, et al. Extracellular matrix of the human cyclic corpus luteum Mol. Hum. Reprod., September 1, 2006; 12(9): 525 - 534. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. J.K. Anand-Ivell, V. Relan, M. Balvers, I. Coiffec-Dorval, M. Fritsch, R. A.D. Bathgate, and R. Ivell Expression of the Insulin-Like Peptide 3 (INSL3) Hormone-Receptor (LGR8) System in the Testis Biol Reprod, May 1, 2006; 74(5): 945 - 953. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 R. A. Bathgate, R. Ivell, B. M. Sanborn, O. D. Sherwood, and R. J. Summers International Union of Pharmacology LVII: Recommendations for the Nomenclature of Receptors for Relaxin Family Peptides. Pharmacol. Rev., March 1, 2006; 58(1): 7 - 31. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. J. Clark, H. F. Irving-Rodgers, A. M. Dharmarajan, and R. J. Rodgers Theca Interna: The Other Side of Bovine Follicular Atresia Biol Reprod, October 1, 2004; 71(4): 1071 - 1078. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H. F. Irving-Rodgers, M. Krupa, and R. J. Rodgers Cholesterol Side-Chain Cleavage Cytochrome P450 and 3{beta}-Hydroxysteroid Dehydrogenase Expression and the Concentrations of Steroid Hormones in the Follicular Fluids of Different Phenotypes of Healthy and Atretic Bovine Ovarian Follicles Biol Reprod, December 1, 2003; 69(6): 2022 - 2028. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
R. Ivell and R. A.D. Bathgate Reproductive Biology of the Relaxin-Like Factor (RLF/INSL3) Biol Reprod, September 1, 2002; 67(3): 699 - 705. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||
