|
|
||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Ovary |
a AgResearch, Wallaceville Animal Research Centre, Upper Hutt 6007, New Zealand
| ABSTRACT |
|---|
|
|
|---|
developmental biology, follicle, follicular development, granulosa cells
| INTRODUCTION |
|---|
|
|
|---|
The role of WT1 in gonadal development begins in the indifferent gonad, where it is strongly expressed in the genital ridge in both male and female mammals (e.g., rats [11], humans [12], sheep [13], and mice [14, 15]). The importance of WT1 during embryonic development is demonstrated by the failure of formation of any kidney or gonadal tissue in mice with targeted disruption to WT1 [16]. In these knockout mice, cells of the metanephric blastema underwent apoptosis at Day E11. The normal thickening of the coelomic epithelium (which forms the gonad) was markedly reduced compared with that of wild-type mice, and by Day E14 no gonadal remnants were visible in the urogenital ridge.
In contrast to the transient expression of WT1 in normal mouse kidneys during Days 1316 of development [17], expression in the ovary remained elevated in follicular granulosa cells throughout adult life [14, 1820]. This finding suggests that the essential functions of WT1 in the fetal kidney and gonad may also be relevant to ovarian follicular development. Little is known about the precise patterns of expression of WT1 throughout the growth of follicles in humans, sheep, or other domestic species. In rats and mice, expression of WT1 occurs in granulosa cells of primordial, primary, and secondary follicles, with expression diminishing throughout follicular development [14, 18, 19]. One possible role for WT1 is to inhibit differentiation of immature follicles [18]. This hypothesis is supported by the fact that WT1 represses transcription of several genes implicated in cellular differentiation, e.g., inhibin-
[18], insulin-like growth factor (IGF) I receptor [21], platelet-derived growth factor, IGF-II, transforming growth factor ß, and colony-stimulating factor 1 [22].
WT1 protein has at least 16 isoforms resulting from alternative splicing [23], different translational start sites, and RNA editing. Although the presence or absence of a KTS insertion and/or a 17-amino acid alternative splice site are thought to confer different functions upon WT1 [24], there is evidence that these isoforms are coordinately regulated [23, 25].
The aim of this study was to identify the cellular sites of WT1 mRNA and protein expression in the fetal lamb gonad and to determine the relationships between these cell types and those that contribute to follicular formation in sheep. In addition, we studied the expression patterns of WT1 mRNA and protein in 4-wk-old and adult sheep ovaries to determine at what specific stages of follicular development and in what ovarian cell types WT1 is likely to be important.
| MATERIALS AND METHODS |
|---|
|
|
|---|
All animal manipulations were performed with approval from the Wallaceville Animal Ethics Committee in accordance with New Zealand regulations. Romney ewes and lambs were killed with 200 mg/kg pentobarbitone (Southern Veterinary Supplies, Christchurch, New Zealand) via the jugular vein. Fetuses of known age were obtained using multiple ovulation and embryo transfer methods as previously described [5] to reduce the number of animals that needed to be slaughtered. Ovarian-mesonephric complexes or ovaries were dissected from fetuses at 24, 26, 2830, 3235, 40, 55, 75, and 135 days from the day of mating (n = 36 in each group) or from 4-wk-old lambs (n = 5). Luteal phase ovaries were collected from adult ewes at Day 10 of the estrous cycle (n = 3) and from the follicular phase 24 h after injection of prostaglandin (n = 4).
Tissues were immediately immersed in 4% (w/v) paraformaldehyde in 0.1 M PBS. After histological processing and embeddment in paraffin, 5-µm sections were produced for immunohistochemistry or in situ hybridization. Only female gonads were used in this study. The sex of fetuses recovered before Day 55 was determined by polymerase chain reaction (PCR) using primers for the SRY gene as previously described [5]. The sex of fetuses from Day 55 was determined by examination of external genitalia.
Cloning of Ovine WT1 cDNA
Cloning of the WT1 gene and production of a probe for in situ hybridization were performed using standard methodologies. Total cellular RNA was extracted from ovine fetal (Day 55) kidney tissue. First-strand cDNA was synthesized by incubating RNA at 37°C for 1 h with mMLV-RT reverse transcriptase (Gibco BRL, Invitrogen, Auckland, New Zealand) according to the manufacturer's standard protocol and a 3' primer (GAC TAA TTC GTC TGA CCG CGC) from bases 16551675 of the human WT1 gene (accession X51630), followed by boiling for 15 min. This cDNA was subjected to PCR using the standard protocol from Roche Diagnostics (Auckland, New Zealand) with the same 3' primer and a 5' primer (GCG GCG CAG TTC CCC AAC CA) from bases 882901 of the human WT1 sequence. The PCR conditions were as follows: an initial cycle of 94°C for 3 min then 40 cycles of 94°C for 1 min (denaturing), 55°C for 1 min (annealing), and 72°C for 1 min (extension). The PCR product was ligated into the pGEMT vector (Promega, Dade Behring Diagnostics Ltd., Auckland, New Zealand) and sequenced (Waikato DNA Sequencing Facility, Hamilton, New Zealand) to confirm that the product was ovine WT1. In addition, specificity of the generated cDNA was tested by Northern blot analysis using a standard protocol [26]. This clone was used for in situ hybridization.
In Situ Hybridization
The expression of WT1 mRNA was determined using an in situ hybridization protocol previously described [5]. Slides were prepared and hybridized at 50°C with 4.5 x 105 cpm/µl of 33P-labeled sense or antisense riboprobes made from the 795-base pair WT1 insert described above. Expression of WT1 mRNA was studied in at least three different animals for each fetal age group from Day 24 to Day 75 and in neonates and adults, with at least two sense-hybridized negative controls for each age group. Hybridization on antisense slides was compared visually with hybridization on sense slides to determine which tissues were expressing WT1 mRNA.
Immunohistochemistry
To demonstrate protein localization, immunohistochemistry was performed as previously described, including antigen retrieval [5] and tyramide signal amplification (TSA) (Perkin Elmer Life Sciences, Christchurch, New Zealand) [26]. Sections were incubated with an affinity-purified rabbit polyclonal antibody raised against the the carboxy terminus of human WT1 protein (C-19; Santa Cruz Biotechnology, Santa Cruz, CA), which was applied at a concentration of 5 µg/ml. At each age group, control sections were incubated with nonimmune rabbit immunoglobulins (RbIgG; Zymed, Innovative Sciences Ltd., Christchurch, New Zealand) at the same concentration. A few control sections (from fetal Days 75 and 135 and 4-wk-old ovaries) were incubated with either WT1 or RbIgG solutions preincubated for 135 min at 22°C with 50 µg/ml of a WT1-blocking antigen (Santa Cruz Biotechnology) to demonstrate specificity of the staining. These primary antibody solutions were incubated on sections at 4°C for 16 h in a humidified chamber. The secondary antibody was biotinylated pig anti-rabbit IgG (DAKO, Med-Bio Ltd., Christchurch, New Zealand) diluted 1:500 in TSA blocking buffer containing 1% (v/v) ovine serum and 1% (v/v) porcine serum. Sections of adult ovaries were incubated with TSA for 10 min, and sections of 4-wk-old ovaries and fetal gonads were incubated for 5 min. Protein localization of WT1 was determined where specific nuclear staining levels were above those of the negative controls, i.e., in all lamb and adult ovaries and in at least three different animals for each fetal age group from Day 24 to Day 135 postmating.
Microscopy and Statistical Analysis
Using a BX50 microscope (Olympus New Zealand Ltd., Lower Hutt, New Zealand), hybridization of the WT1 antisense cRNA probe was compared visually to sense-hybridized sections under dark-field illumination to determine specific expression patterns among tissues. Immunohistochemically treated sections were examined under bright-field illumination to determine the cell-specific patterns of protein localization. Photomicrography was performed as previously described [27]. Chi-square analysis was used to determine whether the proportion of follicles expressing WT1 mRNA or protein was different between follicles <2200 µm and those >2200 µm.
Follicular Classification and Measurement
To aid in follicular identification and classification, ovarian maps were drawn of sections, and nonatretic follicles were classified as previously described [27, 28] with type 1/1a follicles representing primordial follicles and types 2, 3, 4, and 5 representing primary, early secondary, preantral, and antral follicles, respectively.
Diameters of nonatretic antral follicles were measured using MicroScale image analysis software (Digithurst Ltd., Royston, Hertfordshire, U.K.) by averaging two diameters taken at right angles to each other, measured between the basement membranes. The maximal diameters for each follicle were recorded following measurements of diameters in several sections through each follicle. Healthy antral follicles (n = 51 from 12 sheep) ranging in size from 0.2 mm to 5.3 mm, were examined in this study. The largest follicles (n = 7 from six sheep), with diameters from 3 mm to 5.3 mm inclusive, all expressed the gene for aromatase [27]. Many of the follicles used in this study were the same as those used in a previous study [27], and their steroidogenic enzyme expression profiles were known.
| RESULTS |
|---|
|
|
|---|
The nucleotide sequence of ovine cDNA encoding WT1 (GenBank accession number AY115591) was 94% identical to human and pig sequences and 93% and 92% identical to rat and mouse sequences, respectively. The putative translated amino acid sequence was 99% homologous to the human WT1 isoform D (pir A38080) protein sequence and contained both alternative splice sites, i.e., the KTS site between zinc fingers 3 and 4 and the amino-terminal 17-amino acid site variation. The sheep WT1 amino acid sequence was 98% homologous to both pig and rat protein sequences and 97% to mouse protein sequences. Northern analysis using the cDNA probe (Fig. 1) found a major transcript at 2.7 kilobases (kb) with a minor transcript at 3.6 kb, showing expression in kidney, fetal kidney, corpus luteum, and ovary but not in liver or the adrenal gland.
|
Preincubation of the WT1 antibody with its antigen blocked specific nuclear staining on sections, demonstrating the specificity of the antibody to WT1 protein (Fig. 2H). In addition, no specific nuclear staining was observed when the primary WT1 antibody was replaced with nonimmune RbIgG, with the exception of the inconsistent staining in some oocyte nuclei (Fig. 2B). Sporadic nuclear staining was also observed in oocytes in antigen-blocked WT1-treated sections (Fig. 2H); thus sporadic staining in oocytes was considered nonspecific. Cytoplasmic staining by antibodies also was considered nonspecific because it was not reduced by the WT1-blocking antigen (data not shown).
|
The results of the mRNA expression and protein localization in tissues and cell types, respectively, are reported together in chronological order because the patterns of expression of WT1 mRNA and protein localization were similar.
Gonadal Development
At Day 24, WT1 mRNA was expressed strongly in the coelomic epithelial region of the mesonephros, which is characterized by a thickening of epithelial cells marking the initiation of gonadal formation (data not shown). The cells in this region had a uniform immunostaining pattern for WT1 (Fig. 2, A and B). During Days 2426, expression of WT1 mRNA was observed in the mesenchymal cells and the surface epithelium of the developing gonad (Day 26: Fig. 3, A and B), with a homogeneous pattern of protein staining in both these cell types (Day 24: Fig. 2, A and B). During this time, WT1 mRNA was also evident in many cells of the mesonephros, including the surface epithelium, epithelial cells of the giant glomeruli, mesangial cells, and mesenchymal cells (Fig. 3, A and B). At Day 30, localization of WT1 protein became heterogeneous in mesenchymal cells of the gonad and mesonephros but remained homogeneous in the surface epithelium (Fig. 2, C and D). However, in germ cells, no hybridization signal (data not shown) or protein staining (Fig. 2D) was apparent. A cluster of nucleated blood cells observed in a Day-30 sample showed no WT1 mRNA hybridization (data not shown), and such cells were also always devoid of WT1 protein whether they were located in the gonad or in the mesonephros (see Fig. 2B).
|
During the early period of ovigerous cord formation (Days 3240), the surface epithelium continued to express WT1 mRNA (Fig. 3, C and D) and protein (data not shown). Expression of both mRNA and protein was also evident in mesenchymal cells, including those associating with germ cells (i.e., pregranulosa cells). However, as previously observed, the germ cells did not express WT1 mRNA or protein (data not shown). The expression pattern in the mesonephros was similar to that observed earlier in development (Fig. 3, C and D).
Fully formed ovigerous cords at Days 55 and 75 contained germ cells, which had no WT1 mRNA or protein, and their associated pregranulosa cells, which were positive for WT1 mRNA and protein (Figs. 2, E and F, and 3, E and F). The surface epithelium also expressed WT1 mRNA and protein (Figs. 2E and 3, E and F). In addition, WT1 mRNA was evident in the mesonephros-derived cell streams and other stromal cells in the medulla and cortex of the developing gonad (data not shown). Similarly, a heterogeneous pattern of protein staining was observed in both the cell streams and other stromal cells of the gonad (Fig. 2, E and F). At Days 5575 when the mesonephros is undergoing regression, the general pattern of WT1 mRNA and protein in this tissue was the same as that observed earlier, with expression in the surface epithelium, some mesenchymal cells, and cells of the glomeruli but not tubules (data not shown).
From Day 75 when the first type 1/1a follicles were evident, almost all granulosa cells but not any of the oocytes expressed WT1 mRNA (data not shown) and protein (Fig. 2F). At Day 135 of fetal life, WT1 protein was observed in the granulosa cells of small growing follicles (Fig. 2, G and H). Other structures of the Day 135 fetal ovary with strong immunostaining for WT1 included rete ovarii, cortical stromal cells, surface epithelium, and many endothelial cells (Fig. 2, G and H). Stromal cells in the medullary region (Fig. 2G, left) had a heterogeneous pattern of immunolocalization, with a lower proportion of WT1-positive cells than in the cortical stroma. Smooth muscle cells associated with large blood vessels did not contain WT1 protein (data not shown).
Postnatal Ovarian Expression
In ovaries of 4-wk-old and adult sheep, WT1 mRNA was observed in the surface epithelium and stroma of all animals studied (Fig. 4, AF). Similarly, protein was observed in specific cell types, including surface epithelial cells, endothelial cells, most cells of the cortical stroma, some interstitial cells in the medulla, and cells of the rete (Fig. 5, A, C, E, and F). In three corpora lutea from Day 10 of the ovarian cycle (n = 3 ewes), a faint pattern of silver grains was noted over luteal tissue, denoting a low level of hybridization of the antisense WT1 probe (data not shown). In this tissue, immunohistochemistry revealed the localization of protein in many cells associated with the capillary network and therefore identified as endothelial cells (Fig. 5F). Although large luteal cells could be identified and lacked WT1 protein, the other cells in the corpus luteum that did stain were not definitively identified but were possibly small luteal cells or fibroblasts (Fig. 5F). The intensity of protein staining in the corpus luteum was less than that in stromal and granulosa cells.
|
|
Follicular Development
No WT1 mRNA was evident in oocytes at any stage of follicular growth (n
5 follicles of each type; n
5 sheep; Fig. 4, AD). Similarly, there was no specific protein staining of oocytes (Fig. 5, AD).
WT1 mRNA was expressed throughout the ovarian cortex, where many type 1 and type 1a follicles were present (Fig. 4, AF), and the signal appeared to be associated with both granulosa and stromal cells (Fig. 4, A and B). Expression of WT1 in both granulosa and stromal cells was confirmed by immunohistochemistry, with WT1 protein observed in granulosa cells of almost all type 1 and type 1a follicles (n = 94 follicles; Fig. 5A). Granulosa cells in all type 2 to type 4 follicles expressed WT1 mRNA and protein (n
5 follicles of each type; n
5 sheep; Figs. 4, AD, and 5, AC).
Nonatretic antral follicles with a range of known diameters were studied to determine the nature of WT1 expression in type 5 follicles. All follicles that were up to 1537 µm in diameter showed both WT1 mRNA and protein in granulosa cells (n = 29; Figs. 4, E and F, and 5, CE). In contrast, only 4 of 11 follicles with diameters between 1538 and 2280 µm (inclusive) expressed WT1 mRNA and strongly expressed WT1 protein, whereas the other 7 follicles showed little or no gene expression (Fig. 4, E and F). Two of these seven follicles contained a few granulosa cells with WT1 protein (data not shown). None of the follicles >2285 µm had any detectable WT1 gene expression or protein localization (n = 11 follicles from 7 animals; Fig. 5E). Thus, the proportion of <2200-µm follicles expressing WT1 mRNA or protein was greater than the proportion of >2200-µm follicles expressing WT1 mRNA or protein (P < 0.0001).
With in situ hybridization, mRNA encoding WT1 was not evident in the theca interna of any antral follicle (Fig. 4, E and F). However, a weak signal was observed over the theca externa layer and medullary stromal cells. With immunohistochemistry, the cells within the theca interna layer were devoid of WT1 protein staining except for a few specific cells near the basement membrane. These cells always appeared to be associated with blood vessels and often had the characteristic shape of endothelial cells (Fig. 5E).
| DISCUSSION |
|---|
|
|
|---|
The sequential patterns of WT1 mRNA and protein for the fetal sheep gonad during early formation show a very similar pattern to that described for the developing human gonad up to 15 wk of gestation [12]. In sheep, the undifferentiated gonad first appears between Day 23 and Day 24 after mating as a thickening of the coelomic epithelium on the mesonephros. In our study, the coelomic epithelium of the gonadal ridge and glomerular corpuscles of the mesonephros in sheep expressed WT1, as has been found in humans [12]. As the undifferentiated ovary develops further, mesenchymal and epithelial cells of the giant glomeruli are thought to migrate into the gonad [31]. In support of this hypothesis, the expression of WT1 was observed in mesonephric mesenchymal cells and giant glomerular epithelial cells and mesenchymal cells of the gonad. Similar to other species, germ cells and oocytes did not express WT1. During the formation of ovigerous cords, germ cells become associated with pregranulosa (mesenchymal) cells that express WT1. After ovigerous cord formation and consistent with the hypothesis that pregranulosa cells are epithelial in origin [32], WT1 was observed in both the surface epithelium and the pregranulosa cells within cords. Expression continued to be observed in granulosa cells after follicular formation, whereas oocytes were never observed to express WT1.
Organized migrations of cells from the mesonephros into the gonad continue after ovigerous cord formation. Although these cells are unlikely to contribute to the pregranulosa cell pool after ovigerous cord formation in sheep [32], they probably are important for the development of other parts of the gonad, such as stromal cells, rete, and the vascular system. The heterogeneous pattern of protein staining observed in the mesenchymal cells of both the mesonephros and the gonad is consistent with the hypothesis that the stromal, rete, and many of the vasculature cell types in the gonad are of mesonephric origin [10, 33].
As follicles begin to grow, granulosa cells continue to express WT1 through to the type 5 (antral) stage. This process is in contrast to expression in granulosa cells of rat follicles up to only the secondary stage of development [18] but is similar to that in monkeys [19]. In all species, the decrease of expression of WT1 is correlated with the maturation of ovarian follicles [18, 19]. The differences among species in the stage of follicular growth when the reduction of WT1 expression occurs probably depends on the differentiation state of the granulosa cells in follicles of a particular type. For example, granulosa cells from late preantral follicles in the rat are likely to be more mature than those from a similar follicle in the ewe.
In sheep, WT1 was not observed in oocytes of any follicles nor in differentiated theca cells, a finding similar to that reported for other species [18, 19]. Expression of WT1 in the surface epithelium has been observed in many species, including pigs, monkeys, rats, and mice [14, 18, 19] and fetal but not postnatal humans [34]. The high prevalence of WT1 expression in the cortical stroma of neonatal and adult sheep ovaries is in contrast to a lack of expression in stromal cells reported in humans and mice [14, 34]. In addition, expression has not been observed in luteal tissue of rats [19]. These differences might be due to differences between species and/or the sensitivity of the various detection methods.
The gene for WT1 appears to switch off in granulosa cells of ovine follicles around 2 mm in diameter. Based on studies in several species (pig, monkey, and rat), the hypothesis has been proposed that WT1 is important in slowing the proliferation of granulosa cells in the early stages of follicular growth [18, 19] and that reduced expression of WT1 is associated with the rapid growth observed during the final stages of development. However, although WT1 expression is high during the earlier slower phase of follicular growth in sheep, strong expression is still observed at the time when granulosa cells are dividing most rapidly [35]. Moreover, expression of WT1 does not stop until after the division rate of granulosa cells has begun to slow down again. Thus, WT1 probably is not important in regulating the rate of proliferation of granulosa cells in sheep. At the stage of growth when follicles reached 2.2 mm in diameter, granulosa cells began expressing cytochrome P450 side-chain cleavage (P450scc) [27]. This enzyme cleaves cholesterol to form pregnenolone (the precursor for steroid hormone synthesis). The expression of P450scc in granulosa cells of
2-mm-diameter follicles facilitates increased steroid secretion by enabling granulosa cells to assist the theca cells in the production of androgen precursors. This stage of follicular development may mark a point when the steroidogenic function of granulosa cells begins, leading to increased estradiol synthesis by the follicle and thus preovulatory follicular maturation. The development of steroidogenic capability of granulosa cells never seems to occur concomitant with WT1 expression. Because WT1 is a gene transcription suppressor, this gene may need to be switched off before granulosa cells can transform into a steroidogenic phenotype. Similarly, during gonadal development, the switch from a homogeneous pattern of WT1 expression in mesenchymal cells to a heterogeneous pattern of expression coincides with the onset of expression of the steroidogenic enzymes 3ß-hydroxysteroid dehydrogenase and 17
-hydroxylase in some of these cells [26]. In addition, cells thought to be progenitor cells for the theca interna, i.e., cells within the cell streams of the developing gonad, express WT1, but well-differentiated theca cells do not. This finding further supports the hypothesis that ovarian cells expressing WT1 are unlikely to be capable of any steroidogenic function. As yet, there is no direct evidence that WT1 suppresses steroidogenic enzyme gene expression, and this possibility requires investigation.
Messenger RNA and protein expression for the Wilms' tumor gene were observed in the gonad and mesonephros during development. WT1 mRNA and protein also was expressed in many ovarian cell types, including follicular granulosa cells, where expression was limited to cells in follicles <2.3 mm in diameter. A major role for WT1 appears to be inhibition of the differentiation of cells into a steroidogenic phenotype, but the precise functions of WT1 in ovarian cells are yet to be established.
| ACKNOWLEDGMENTS |
|---|
| FOOTNOTES |
|---|
2 Correspondence: Jennifer L. Juengel, AgResearch, Wallaceville Animal Research Centre, P.O. Box 40-063, Upper Hutt 6007, New Zealand. FAX: 64 4 922 1380; jenny.juengel{at}agresearch.co.nz ![]()
Received: 28 June 2002.
First decision: 1 August 2002.
Accepted: 16 September 2002.
| REFERENCES |
|---|
|
|
|---|
gene promoter. Mol Endocrinol 1995 9:1356-1366
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |