| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Ovary |
Developmental Endocrinology Branch,2 NICHD, NIH, Bethesda, Maryland 20892
INRA,3 PRMD, URA CNRS 1291, 37380 Nouzilly, France
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
follicular development, granulosa cells, human chorionic gonadotropin, insulin-like growth factor receptor, ovulation
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
The IGFBPs bind IGFs in the bloodstream and extracellular fluids and thereby increase IGF half-life; IGFBPs may stabilize or neutralize IGFs in extracellular spaces and may also have IGF-independent actions (reviewed in [16]). Because of their ability to neutralize or inhibit the effects of IGF, it has been suggested that IGFBPs may induce ovarian follicular atresia. The IGFBPs are abundant in ovarian follicles [17–19, additional references in 20]), where their function remains a matter of active investigation. In vitro studies show that addition of intact IGFBP4 to the culture medium inhibits gonadotropin-induced steroid synthesis by theca and granulosa cells [21, 22]. In human and large domestic animal follicular fluid, the presence of IGFBP4 proteolytic fragments has been correlated with follicular selection (studies reviewed in [20]). Recently, the expression of an IGFBP4 protease, pregnancy-associated plasma protein A (PAPP-A), has been reported in healthy human follicles and corpora lutea [23–25]. These observations have led to the suggestion that selective proteolysis of IGFBP4 may promote the emergence of a dominant follicle. Detailed studies of cellular patterns and hormonal and/or cyclic regulation of IGF, IGFBP4, and PAPP-A expression in the human ovary have been limited for obvious reasons. To further elucidate the potential roles/interactions of these factors in human follicular development, we turned to the rhesus monkey, which has an ovarian cycle very similar to that of the human. Preliminary studies showed similar patterns of IGF system expression in the human and rhesus ovary [26], and so in the present study, we examined cellular patterns of IGF, IGFBP4, and PAPP-A gene expression during follicular development and hCG treatment in the rhesus macaque monkey.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Female rhesus monkeys (Macacca mulatta; age, 6–13 yr) from the National Institutes of Health (NIH) Poolesville colony were used in accordance with a protocol approved by the NICHD Animal Care and Use Committee. Ovariectomies were performed on regularly cycling (with at least three sequential cycles of 26–30 days in duration), late follicular phase monkeys under ketamine anesthesia via a midventral laparotomy. The animals were in Cycle Day 14 at ovariectomy. Two animals were treated with hCG (1000 IU i.m.) daily beginning on Cycle Day 7 before ovariectomy. Ovaries were snap-frozen on dry ice and stored at -70°C. Serial sections (thickness, 10 µm) were cut at -15°C and thaw-mounted onto poly-L-lysine-coated slides for in situ hybridization and immunohistochemistry.
Clones Used for Riboprobe Synthesis
The rat IGF2 and human IGFBP4 clones used for cRNA probe synthesis have been previously described [13]. The human LH receptor (LHR) cDNA was a 1250-base pair (bp) segment subcloned into pCRscript. The human FSHR cDNA was a gift from Ares Advanced Technology (Randolph, MA). This 2118-bp fragment was subcloned into pSV.Sport. The sense and antisense constructs were linearized with EcoRI and KpnI, respectively. The human aromatase 273-bp fragment was generated by polymerase chain reaction (PCR). Oligonucleotides were synthesized and used as primers. The sequence of the 20-mer 3'-oligonucleotide was 5' TTGTTGTTAAATATGATGCC 3', and the sequence of the 20-mer 5'-oligonucleotide was 5' ATACCAGGTCCTGGCTACTG 3'. The 572-bp human PAPP-A cDNA was generated from human placenta cDNA by PCR, cloned into pGEM-T Easy vector, and generously provided by Eli Adashi (University of Utah, Salt Lake City, UT) [25].
In Situ Hybridization
The 35S-labeled RNA probes were synthesized to a specific activity of approximately 2 x 108 dpm/µg in a protocol that has been previously described [13]. The specificity of the in situ hybridization results was evaluated by hybridization of parallel sections with sense probes, which produced a nearly undetectable radioactive signal. The quantification of hybridization signal was carried out in a blinded fashion. Hybridization signal was captured using NIH image v1.57 software (Bethesda, MD). Ten measurements were obtained and meaned for each cell type in each animal after subtraction of background signal. The values were compared using ANOVA. Significant differences among means were determined using ANOVA followed by the Fischer least significant difference test. Correlation between IGFBP4 and LHR expression was assessed by chi-square analysis (Statview 5.0, SAS Institute, Cary, NC).
Immunohistochemistry
Immunohistochemistry for cell proliferation Ki67 was performed by the avidin-biotin-immunoperoxidase technique. The mouse anti-human Ki67 sera (Boehringer Mannheim, Indianapolis, IN) detects a nonhistone protein associated with the nuclear matrix expressed in all proliferating cells in the G1, S, G2, and M phases of the cell cycle. Fresh-frozen tissue sections were fixed and incubated overnight at 4°C with a 1:50 dilution of mouse anti-human Ki67 antibody or with 1% PBS as control. Thereafter, tissue sections were treated with biotinylated sheep anti-mouse immunoglobulin G (1:40) for 30 min at room temperature followed by a 45-min incubation with the avidin-biotin-peroxidase complex (Vectastain ABC Elite Peroxidase Kit; Vector Laboratories, Burlingame, CA). The antigen-antibody complex was visualized by incubation with freshly prepared 3,3'-diaminobenzidine (DAB substrate kit; Vector Laboratories) and the tissue counterstained with methyl green.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
A single dominant follicle was apparent in one ovary from each of the three normal, untreated, midcycle monkeys. The dominant follicle was identified on the basis of size (maximum diameter, >2 mm) and aromatase expression by the membrana granulosa (Fig. 1). The dominant follicle seen in Figure 1A is relatively small compared with the other two samples, one of which was in the process of ovulation at the time of collection (Fig. 1I). The animal from which the ovary shown in Figure 1A was obtained tended to have long cycles (28–30 days), so this follicle probably represents a slightly earlier stage of maturation. One or more degenerating corpora lutea from previous cycles were found in each ovary (Fig. 1, A and E). A relatively large and healthy-appearing corpus luteum was found in the nondominant ovary from each animal, indicative of a recent ovulation. Approximately 20 small to medium-sized (diameter, 500–1500 µm) antral follicles were found in each ovary. Sequential sections, taken approximately every 300 µm throughout the whole of each ovary, were analyzed for expression of the proliferation-specific Ki67 antigen, LH, FSHR, and aromatase mRNAs to characterize each follicle in terms of growth and selection potential for comparison with IGF system component expression.
|
The IGF1 mRNA is not detected in any follicles of the midcycle monkey ovary (not shown). The IGF1 receptor mRNA is abundant in granulosa cells and less abundant but present in theca cells of all follicles in the monkey ovary, similar to patterns seen in other species [4, 13, 14, 26]. The IGF2 mRNA is abundant in the thecal compartments of all antral follicles (Fig. 1, B, F, and J). It is localized in the theca interna and in stromal elements of the theca externa but not in the granulosa of nondominant antral follicles (Fig. 2). The IGFBP4 mRNA is also expressed in the theca of developing follicles (Figs. 1 and 2). The pattern, however, is not the same as that of IGF2. First, IGFBP4 expression is more selective, localized in the theca only in healthy, growing follicles, whereas IGF2 is detected in both thecal layers of virtually all antral follicles.
|
The health status of antral follicles was determined by the presence or absence of aromatase and FSHR expression. Of approximately 130 antral follicles examined in the six ovaries, 45 were aromatase- and FSHR-positive; 42 of these also demonstrated thecal IGFBP4 expression. The IGFBP4 mRNA is detected only in the steroidogenic theca interna of small to medium-sized antral follicles, whereas IGF2 is expressed in both thecal compartments (Fig. 2). Comparing IGFBP4 and LHR localization shows that these two mRNAs are colocalized in the theca interna of the same subset of antral follicles, with less than 2% of follicles being either LHR-positive and IGFBP4-negative or the converse (the correlation between IGFBP4 and LHR expression in antral follicles is significant by chi-square analysis, P < 0.0001).
The IGF2 and IGFBP4 mRNAs are not detected in granulosa cells in most follicles (Figs. 2 and 3). Interestingly, it appears that IGFBP4 expression first appears in granulosa cells in parallel with the appearance of LHR expression, seen first in mural cells (Fig. 3, C and D). Abundant IGF2 and IGFBP4 mRNA levels are found in periovulatory granulosa (Figs. 1 and 4). Figure 4 shows IGF2 and IGFBP4 gene expression in granulosa cells of a healthy preovulatory follicle demonstrating proliferation and aromatase, FSHR, and LHR expression, showing that these are healthy, mature granulosa cells. It is worth noting that aromatase, IGFBP4, and the gonadotropin receptors demonstrate a very homogeneous pattern of expression in the membrana granulosa. The IGF2 mRNA, in contrast, is rather patchy in distribution, suggesting that it could actually be expressed by invading stromal elements rather than by granulosa cells (Figs. 1J and 4E).
|
|
The PAPP-A mRNA is detected in granulosa cells of follicles of all sizes, whereas IGFBP4 expression is confined to the theca of nondominant follicles (Fig. 5). No anatomical correlation was apparent between PAPP-A and LHR expression.
|
Ovaries from hCG-Treated Monkeys
As we have shown above, IGFBP4 expression and LHR expression appear to be highly correlated in the monkey ovary. The IGFBP4 mRNA is selectively expressed in LHR mRNA-positive theca interna cells of smaller follicles and in both theca and granulosa cells of periovulatory follicles. Within the individual follicle, IGFBP4 expression also closely tracks LHR expression, such as by appearing in the mural before the antral granulosa cells (Fig. 3). The observed coincidence between IGFBP4 LHR gene expressions suggested that LH might regulate IGFBP4 expression. To test this hypothesis, two regularly cycling monkeys were treated with hCG for 7 days before ovariectomy at Cycle Day 14. The goal of this treatment was to evaluate the effects of pharmacological doses of the LH analog on IGFBP4 expression in vivo at a cycle stage matched to that of our normal control animals. Morphological evaluation showed that one of the hCG-treated monkeys had two large, dominant follicles, one in each ovary, whereas the other had a single dominant follicle. The number of small to medium-sized antral follicles was similar to that of the controls, whereas each ovary demonstrated several relatively hypertrophied corpora lutea. The IGFBP4 levels were markedly increased after hCG treatment in developing follicles (both theca and granulosa) and corpora lutea (Fig. 6 and Table 1). The hybridization signal showed a greater than 2-fold increase in IGFBP4 mRNA (P = 0.005). The PAPP-A mRNA also appeared to be increased by hCG treatment (Fig. 6), but the effect was not statistically significant (P = 0.26). The IGF2 gene expression, in contrast to IGFBP4, was detected in the theca of many follicles where LHR was not present, primarily in the theca externa (where LHR expression is never found), and thus did not correlate with LHR expression. The IGF2 mRNA was detected in microvessels but not in corpora lutea granulosa or theca lutein cells, and it was not altered by hCG treatment (data not shown).
|
|
Interestingly, little LHR expression was found in the residual corpora lutea of the normal midcycle ovary, but after several days of hCG treatment, corpora lutea LHR expression was markedly increased (Fig. 6 and Table 1). The LHR expression in antral follicles, however, showed a small but significant decrease after hCG treatment.
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
In terms of cycle duration and all known aspects of follicular development and function, the rhesus monkey is thought to provide an excellent model for human ovarian function. The ovaries examined in the present study were obtained from young, healthy, regularly cycling monkeys at closely timed midcycle stages. Their reproductive health is reflected by the presence of robust dominant follicles in all monkeys in addition to large cohorts of healthy antral follicles and the presence of recent corpora lutea. A unique feature of this study is the application of a battery of assays (Ki67, aromatase, and LHR and FSHR expression) to characterize the developmental state and gonadotropin responsiveness of each follicle. The IGFBP4 also appears to be expressed by healthy theca interna and corpora lutea in human ovaries [13, 14]. More detailed, cycle-stage specific information regarding the porcine ovary shows that IGFBP4 mRNA is concentrated in the theca of growing follicles, in the granulosa cells of mature dominant follicles, and in the corpora lutea of luteal phase ovaries, also tracking LHR expression [27]. The IGFBP4 is also concentrated in the theca of healthy, growing follicles in the bovine ovary and is increased by LH/hCG treatment [28]. The LH/hCG treatment also increases IGFBP4 expression in the murine ovary [29, 30]. Thus, LH/hCG treatment stimulates ovarian IGFBP4 expression in rodents, large animals, and primates, and IGFBP4 expression closely parallels LHR expression in healthy developing follicles and corpora lutea in large animals, including nonhuman primates and, likely, the human. The expression of LHR and IGFBP4 by the same cells suggests that this may be a direct effect by LH on IGFBP4 mRNA levels, or it could be an indirect effect of increased differentiation of LHR-expressing cells.
A refinement of the view that IGFBP4 promotes follicular atresia involves the idea that selective proteolysis of IGFBP4 occurs in ovulatory follicles. This hypothesis is based on the observation of abundant IGFBP4 fragments in the follicular fluid from human dominant follicles and IGFBP4 proteolytic activity in fluid from large animal dominant follicles (reviewed in [20]). The enzyme implicated in IGFBP4 proteolysis in follicular fluid has recently been identified as PAPP-A [23, 24]. The PAPP-A is present in the circulation of pregnant women bound to another protein known as the proform of the major eosinophil basic protein (proMBP), which inhibits its proteolytic activity. The source and function of this protein complex in the circulation is unclear. The PAPP-A in ovarian follicular fluid is produced by granulosa cells and may be associated with proMBP [23, 24], and its proteolytic activity is stimulated by IGFs 1 and 2 [30]. The PAPP-A-mediated IGFBP4 cleavage yields two major products of approximately 10 and 17 kDa, which have not been evaluated in terms of potential IGF-independent effects on follicular function.
The view that IGFBP4 proteolysis is required for follicular selection implies that the binding protein is constitutively expressed and inhibits IGF2 and, hence, the development of all follicles except the select few that express PAPP-A, which allows follicular development to proceed by degrading IGFBP4. The present study shows, however, that IGFBP4 is selectively expressed in healthy follicles, whereas PAPP-A seems to be more widely expressed. Moreover, as shown in the present study and previously [23, 24, 30], IGFBP4 expression is enhanced by gonadotropins, whereas this is not apparent for PAPP-A. A recent study on human surgical specimens [23] suggests, however, that PAPP-A is selectively expressed in healthy antral follicles and corpora lutea. This divergence could be explained by a species difference in ovarian PAPP-A expression in human and nonhuman primates, or it could reflect the fact that all the monkeys were in midcycle but that the gynecological surgery patients may have been in different cycle stages or even not cycling at all. An alternative explanation for the increased abundance of IGFBP4 fragments in dominant follicle fluid may be that the substrate, IGFBP4, is more abundant in LH-selected follicles. Moreover, if PAPP-A proteolytic activity is enhanced by IGF2 [30], then the greater proteolytic activity in dominant follicles may reflect the abundance of IGF2 in these follicles.
In the intense focus on inhibitors, and on inhibitors of inhibitors (with a potential tertiary inhibitor, proMBP, on the horizon), one could lose sight of the supposed primary effectors of follicular development, IGFs. If IGFs are, in fact, important effectors of follicular selection, then selective regulation of IGF expression seems to be a more direct, economical mechanism than invoking multiple layers of inhibitory factors. And in fact, granulosa cell IGF1 or IGF2 expression appears to be selective for healthy follicles in most species, although perhaps not in ruminant species.
It seems unlikely that the highly selective expression of IGFBP4 in healthy follicles and corpora lutea and its regulation by LH are consistent with a role in follicular demise. Why would IGFBP4 gene expression be concentrated in healthy developing follicles and corpora lutea if its role is to inhibit follicle selection? Moreover, it seems unlikely that LH would stimulate the production of a factor that inhibits LH-induced steroidogenesis or follicular selection. An alternative interpretation for these observations and for the finding that PAPP-A and IGFBP4 proteolytic products are abundant in dominant follicles is that PAPP-A is involved in posttranslational processing of IGFBP4 and that an IGFBP4-derived peptide promotes LH-induced steroidogenesis and/or luteinization. Exogenous intact IGFBP4, which binds and inhibits IGF action, inhibits steroidogenesis in vitro, but a proteolytic product generated in vivo may, as a matter of speculation, have IGF-independent effects that enhance LH activity. Further studies are required to investigate potential effects of IGFBP4 peptides on follicular development and ovarian steroidogenesis.
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
Received: 11 July 2002.
First decision: 5 August 2002.
Accepted: 3 January 2003.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  M. Matsui, B. Sonntag, S. S. Hwang, T. Byerly, A. Hourvitz, E. Y. Adashi, S. Shimasaki, and G. F. Erickson Pregnancy-Associated Plasma Protein-A Production in Rat Granulosa Cells: Stimulation by Follicle-Stimulating Hormone and Inhibition by the Oocyte-Derived Bone Morphogenetic Protein-15 Endocrinology, August 1, 2004; 145(8): 3686 - 3695. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. J. Spicer Proteolytic Degradation of Insulin-Like Growth Factor Binding Proteins by Ovarian Follicles: A Control Mechanism for Selection of Dominant Follicles Biol Reprod, May 1, 2004; 70(5): 1223 - 1230. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
