| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
Departments of Obstetrics and Gynecology (T.E.V., M.M., A.L., J.S.T.) and Pathology (T.E.V., M.M., A.L., R.H., J.S.T.) and Research Center for Molecular Endocrinology, World Health Organization Collaborating Centre for Research on Reproductive Health, and Biocenter Oulu (S.T., O.O., V.I.), University of Oulu, FI-90014 Oulu, Finland
Address all correspondence and requests for reprints to: Juha S. Tapanainen, M.D., Ph.D., Department of Obstetrics and Gynecology, FI-90014 University of Oulu, Finland. E-mail: juha.tapanainen{at}oulu.fi.
| Abstract |
|---|
|
|
|---|
and -ß and key enzymes regulating estradiol production, 17ß-hydroxysteroid dehydrogenases (17HSDs) types 1, 2, and 7, were analyzed in human fetal ovaries. The expression of ERs was related to the development of ovarian follicles. Before the 26th week of fetal life ER
was only occasionally detected, but from then onward, its expression was detected in ovarian follicles. Consistent expression of ERß was seen from the 20th week until term. Both ER
and ERß were localized to the granulosa cells and oocytes. Expression of 17HSD1 and 17HSD7 enzymes, catalyzing the conversion of estrone to more active estradiol, was detected as early as at the 17th week of fetal life. The expression of 17HSD1 displayed a pattern similar to that of ERs and increased toward term, whereas that of 17HSD7 decreased and was negative by the 36th week. 17HSD1 was localized to the granulosa cells, whereas 17HSD7 expression was more diffuse and was found in both granulosa and stromal cells. 17HSD2, converting estradiol to less potent estrone, was negative in all samples studied. The simultaneous appearance of estrogen-converting enzymes and ERs at the time of follicle formation indicates that the machinery for estrogen action exists in fetal ovaries and suggests a possible role for estrogens in the developing ovary. | Introduction |
|---|
|
|
|---|
and ERß (2, 3). They have closely related DNA binding regions but different ligand-binding affinities (3). Their expression patterns also differ. In the ovaries ER
is expressed in the theca and interstitial cells of growing follicles, whereas ERß is mainly located in the granulosa cells (4). In the human fetus, the expression of ERs is not well known. Expression of ER
has been observed in early ovarian development (5), and ERß mRNA has been detected in midgestational human fetal ovaries (6, 7). However, their spatial and temporal expression patterns and roles in human ovarian development are unclear. Local estrogen concentrations at the tissue and cellular level are controlled by estrogen-metabolizing enzymes. We have previously shown that aromatase, which is the key enzyme converting androgens to estrogens, is expressed in human fetal ovary (8). In addition to aromatase, the 17ß-hydroxysteroid dehydrogenases (17HSDs) modulate the biological activity of sex steroid hormones by catalyzing key steps in the formation and primary metabolism of active estrogens, androgens, and progesterone (9). Hence, 17HSDs regulate the concentrations of active sex steroids able to bind to their respective steroid receptors. At present, nine 17HSD isoforms have been characterized in humans (9, 10). 17HSD1 and 17HSD7 catalyze the conversion of estrone (E1) to more active estradiol (E2) (11, 12), whereas 17HSD2 catalyzes the opposite reaction and inactivates sex hormones (13, 14). In the adult human ovary 17HSD1 is expressed in granulosa cells of growing follicles and in the corpus luteum (8, 15). Localization of 17HSD7 in the human ovary is unknown, but in rodents it is expressed mainly in the corpus luteum (16).
The role of estrogen action in ovarian development and reproductive function has been extensively studied using ER and aromatase knockout animal models (17, 18, 19, 20). However, the role of estrogen in human ovarian development is much less well known, and the developmental impact of an inactivating aromatase mutation on ovarian development is difficult to assess, owing to a possible compensatory role of maternal estrogens (21, 22). Moreover, recently discovered putative ERs that can also exist as membrane-bound forms further complicate the picture (23).
In the present study, the role of estrogen action was investigated in human ovarian development by studying the expression of ERs and the estrogen-metabolizing enzymes 17HSD1, 17HSD2, and 17HSD7 in fetal ovaries.
| Materials and Methods |
|---|
|
|
|---|
Ovaries from 10 fetuses (fetal age 1324 and 33 wk) with normal karyotype were obtained after spontaneous abortions and therapeutic abortions carried out as a result of maternal disease. In addition, eight neonates (fetal age 2640 wk) who died because of perinatal asphyxia or infection within 48 h of birth were studied. Samples with visible autolysis or abnormal karyotype were excluded from the study. All samples were fixed in 4% buffered formaldehyde for 24 h, rehydrated, and embedded in paraffin. Histological sections (4 µm) were cut and processed for immunohistochemistry and in situ hybridization. The study was approved by the Ethics Committee of Oulu University Hospital.
Immunohistochemistry
Paraffin sections were deparaffinized in xylene and hydrated gradually through a series of ethanol dilutions and then washed in water. Endogenous peroxidase activity was blocked with 3% hydrogen peroxide in methanol. Nonspecific binding was prevented by using fetal calf serum at a 1:5 dilution in PBS. Primary antibody was applied to the sample and it was incubated overnight. ER
was detected using mouse monoclonal antibody (NCL-ER-6F11) raised against the N-terminal (A/B) region of human ER
(working concentration 1:50 in PBS; Novocastra, Newcastle-upon-Tyne, UK). An antigen-absorbed antibody control was not used because the antibody for ER
is well characterized. ERß was detected using two different antibodies: rabbit polyclonal antibody (PAI-310) raised against a synthetic peptide corresponding to C-terminal amino acid residues 467485 of rat ERß (working concentration 1:50 in PBS; Affinity BioReagents Inc., Golden, CO) and a rabbit monoclonal antibody raised against human ERß (working concentration 1:300 in PBS; Santa Cruz Biotechnology, Santa Cruz, CA). For negative controls, PBS was used instead of primary antibody. The sections were washed in PBS before addition of the secondary antibody and incubation for 30 min. As the secondary antibodies, biotinylated rabbit antimouse and goat antirabbit immunoglobulins (Vectastain Elite ABC kit, Vector Laboratories, Burlingame, CA) were used against the respective primary antibodies. After washing in PBS, the sections were incubated for 30 min with avidin-conjugated alkaline phosphatase (Vectastain Elite ABC Kit). The sections were washed in PBS followed by treatment with liquid diaminobenzidine and a wash in water. Counterstaining was performed with hematoxylin. The sections were dehydrated through a series of ethanol dilutions and cleared in xylene.
Staining with monoclonal ERß antibodies was controlled by using antigen-absorbed antibodies, as previously described (8). Control peptides were purchased from Santa Cruz Biotechnology and corresponding antibodies were incubated with a 5-fold excess (by weight) of the blocking peptide overnight at 4 C. The neutralized antibodies were used according to normal immunostaining procedures.
In situ hybridization
Probes for in situ hybridization were prepared from a 376-bp fragment (nucleotides 1376) of human 17HSD1 cDNA (24), a 380-bp fragment (nucleotides 191570) of human 17HSD2 cDNA (13), and an 832-bp fragment (nucleotides 39870) of the human 17HSD7 cDNA (12) cloned in pGEM-4Z plasmids (Promega, Madison, WI) and used as templates for in vitro transcription. Sense and antisense [35S]CTP-labeled RNA probes were transcribed with T7 or SP6 RNA polymerases using linearized plasmids as templates. Specific activities of the synthesized RNA probes were approximately 6 x 106 cpm/µl.
The in situ hybridization reactions were performed as previously described (25). Hybridization signals were detected after 15-d (17HSD7 after 21 d) exposure to autoradiographic NTB2-emulsion (Eastman Kodak, Rochester, NY) at 20 C. The slides were developed using D-19 developer and Unifix solution (Eastman Kodak), following the instructions of the manufacturer. Nuclei were further stained with Hoechst 33258 (Sigma-Aldrich Finland, Helsinki, Finland), after which the slides were mounted with glycergel (Dako A/S, Glostrup, Denmark), except for the 17HSD7 slides, which were stained with hematoxylin and eosin and mounted with Pertex (Histolab, Gothenburg, Sweden).
Evaluation of samples
All samples were analyzed by two independent observers (immunohistochemistry: T.E.V. and M.M.; 17HSD1 and 17HSD2 in situ hybridizations: T.E.V. and O.O.; 17HSD7 in situ hybridizations: T.E.V. and S.T.). Immunohistochemical sections were analyzed using light microscopy, and in situ hybridizations were analyzed using dark-field and light microscopy. The results were evaluated by using a standard histograde system: , negative; ±, moderately positive; +, positive; ++, strongly positive; and +++, very strongly positive.
| Results |
|---|
|
|
|---|
immunohistochemistry
In the two youngest samples (13 and 14 wk) studied, no ER
immunoreactivity was detected. Between wk 14 and 24, only occasional weak staining was observed in areas in which follicles had started to develop (Table 1
). In other areas in which follicular formation had not yet been completed, ER
staining was negligible and seemed to be confined to granulosa/pregranulosa cells and oocytes (Fig. 1A
). From the 26th week onward, nearly all oocytes were individually enveloped by a granulosa cell layer. At this stage immunostaining of ER
was moderate and localized to both the follicular granulosa cells and oocytes. However, near term there was more individual variation, and in two samples aged 37 and 39 wk, the staining was negligible. Immunoreactivity of ER
was evident only in the follicular granulosa cells and oocytes; no staining was observed in the stroma (Fig. 1A
). Furthermore, the immunostaining was clearly localized to the cytoplasm and only occasional nuclear staining was detected.
|
|
Immunoreactivity of ERß was negative until wk 20, except in one sample from a 14-wk-old fetus showing weak staining. From the 20th week onward, immunostaining was seen in most of the follicles (Fig. 1B
). Similarly to ER
, ERß staining was primarily localized to follicular granulosa cells, but oocytes also showed positive staining. Immunostaining increased gradually and reached its peak between wk 27 and 40. During this period strong staining was seen in almost all follicles. Thereafter the staining weakened, and just before birth there was only weak immunoreactivity in some follicles. Similarly to ER
, ERß immunostaining was observed only in follicular cells, whereas stromal cells were unstained.
17HSDs
The expression pattern of 17HSD1 mRNA in human fetal ovaries was somewhat similar to that of estrogen receptors (Table 1
). No expression was detected before follicular development had started and follicular structures were observed. Weak expression was seen as early as at 17 wk, but after the 22nd week, strong expression was observed in all samples. The expression was clearly localized to the granulosa cells, whereas oocytes were negative. In the few growing follicles that were seen during the last trimester, 17HSD1 mRNA expression was clearly up-regulated (Fig. 2
). No 17HSD2 mRNA expression was detected in any of the fetal ovarian samples.
|
| Discussion |
|---|
|
|
|---|
and ERß was seen from the 20th week onward. Both ERs were localized in the granulosa cells and also in the oocytes. The highest expression of both ER
and ERß was observed at the beginning of the last trimester. The expression patterns of the ERs were similar, but the immunostaining of ERß was more pronounced than that of ER
. Similarly, ERß is the dominant ER in the adult ovary, in which it is expressed mainly in the granulosa cells of growing follicles, whereas ER
is localized to theca and interstitial cells (4). Weak expression of 17HSD1 was detected at the age of 17 wk, whereafter it increased toward term. However, the highest expression of 17HSD7 was observed as early as at wk 17, this being the earliest time point at which it was studied. After 17 wk 17HSD7 expression decreased, and it was negative at the 36th week. Expression of 17HSD2 was negative in all samples studied.
Neither of the ERs was clearly present in the ovary before the 20th fetal week. Therefore, it seems likely that estrogen action is not essential for early ovarian development or follicle formation. This is supported by knowledge gained from studies on ER knockout animals (17, 18, 19, 20). The ovaries of
ERKO, ßERKO, and
ßERKO mice develop normally (18, 19, 20). Although
ERKO female mice are infertile, they show normal follicular development until preovulatory stages, in which follicular growth is arrested (20). ßERKO female mice seem to suffer only from mild functional ovarian disorders because they are capable of ovulating and bearing offspring (18). On the other hand, as expected,
ßERKO mice have the most severely disturbed phenotype: ovarian development seems to follow a normal path at least to the stage at which primordial follicles are developed. The finding that estrogen receptors are predominantly expressed in the human ovary only after the follicles have developed indicates that early ovarian development in humans is similarly independent of estrogen action. However, after the development of ovarian follicles the ovaries of
ßERKO mice start to express various abnormal features. In postnatal life their ovaries undergo redifferentiation from follicles into structures that resemble seminiferous tubules of the testis (19).
Interestingly, the expression of ERs in fetal mouse ovary seems to differ from that seen in humans. ER
seems to be the dominant ER in mouse ovary during fetal life, whereas ERß expression is negligible or negative (26). The present data suggest that this might not be the case in the human fetal ovary. These species-specific differences in ER expression may reflect differences in the role of ERs during ovarian development.
Women with aromatase deficiency display disturbed ovarian function, with morphological findings similar to those in polycystic ovary syndrome and in
ERKO mice (27, 28). It seems that early ovarian development and formation of follicles occurs properly, and the observed defects are the result of a lack of estrogen action in postnatal life. Although the present findings and our previous observation that aromatase is expressed in the fetal ovary (8) suggest that the fetal ovary is capable of producing estrogens, the actual role of estrogens in ovarian development is difficult to assess, owing to the compensatory role of maternal estrogens during fetal life (28).
The enzyme 17HSD1 is the key regulator of estrogen metabolism in the adult human ovary, converting E1 into more potent E2 (11, 29). Whereas 17HSD7 performs a similar function to that of 17HSD1 in catalyzing E1 to E2, it is less efficient (12). In addition to the conversion of E1 to E2, human 17HSD7 has other substrates in steroid hormone metabolism (12) and it participates in cholesterol biosynthesis (30). The expression of 17HSD1 had a pattern clearly similar to those of the ERs. The enzyme was first observed at the 17th fetal week, and clear expression of both ER
and ERß was detected shortly thereafter. The expression of 17HSD1 increased toward term, but that of 17HSD7 had a different pattern: it was seen at the 17th fetal week, and it decreased toward term. Its expression was negligible or negative after the 28th fetal week when the expression patterns of 17HSD1 and ERß reached their maxima. The localization of 17HSD1 was clearly restricted to the granulosa cells, whereas no expression was found in the oocytes. The localization of 17HSD7 was not as clear as that of 17HSD1, but it was mainly found in the stromal and pregranulosa cells. The fact that 17HSD7 has an important role in the rodent ovary but a less clear function in humans underlines species-specific differences in the reproductive system. Its expression in early fetal ovaries, however, emphasizes its possible function in estrogen metabolism during human ovarian development (12, 16).
The concurrence of follicle formation with the expression of estrogen-converting enzymes and ERs could indicate that estrogens play a physiological role in late fetal life (31). We have previously discovered that aromatase is expressed in the human fetal ovary, indicating that the capacity to produce estrogens from precursor steroids exists already before birth (8). Together with the present finding, that 17HSD1 is expressed from midgestation onward, the presence of aromatase further augments the hypothesis that during the last trimester fetal ovary is capable of producing estrogens and E2. To further underline the importance of fetal ovary as a functioning estrogen-producing organ, 17HSD2, which inactivates E2 to less potent E1, was completely absent in all analyzed stages of ovarian development.
The present work demonstrates that ERs and estrogen-converting enzymes are expressed in the human fetal ovary. The high expression of ER
, ERß, and 17HSD1 during the last trimester of fetal life and the highest expression of 17HSD7 during midgestation suggest different roles of 17HSD1 and 17HSD7 in development. Furthermore, the lack of 17HSD2 and increasing activity of 17HSD1 during gestation reflect increased estradiol production toward term. The simultaneous appearance of ERs and 17HSD1 and 17HSD7 further indicates that the machinery for estrogen action already exists during fetal life and suggests a possible role of estrogens in the developing ovary after follicle formation.
| Footnotes |
|---|
First Published Online March 22, 2005
Abbreviations: E1, Estrone; E2, estradiol; ER, estrogen receptor; HSD, hydroxysteroid dehydrogenase.
Received September 14, 2004.
Accepted March 3, 2005.
| References |
|---|
|
|
|---|
and ß mRNA expression: impact of development and estrogen. Mol Cell Endocrinol 149:153161[CrossRef][Medline]
(ER-
) and ß (ER-ß) mRNAs in normal ovary, ovarian serous cystadenocarcinoma and ovarian cancer cell lines: down-regulation of ER-ß in neoplastic tissues. J Clin Endocrinol Metab 83:10251028
and ER-ß isoforms in normal endometrial and endometriosis-derived stromal cells. Mol Hum Reprod 5:651655
(ER-
) and ß (ER-ß) mRNA in the midgestational human fetus. J Clin Endocrinol Metab 82:35093512
-hydroxysteroid dehydrogenase activity. J Biol Chem 268:1296412969
and ß. Science 286:23282331
gene in female mice: characterization of ovarian responses and phenotype in the adult. Endocrinology 140:27332744
and not of estrogen receptor ß in the female rat reproductive tract. J Mol Endocrinol 28:8797[Abstract]
This article has been cited by other articles:
![]() |
C. Wang, E. R. Prossnitz, and S. K. Roy G Protein-Coupled Receptor 30 Expression Is Required for Estrogen Stimulation of Primordial Follicle Formation in the Hamster Ovary Endocrinology, September 1, 2008; 149(9): 4452 - 4461. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. Montani, M. Penza, M. Jeremic, G. Biasiotto, G. La Sala, M. De Felici, P. Ciana, A. Maggi, and D. Di Lorenzo Genistein is an Efficient Estrogen in the Whole-Body throughout Mouse Development Toxicol. Sci., May 1, 2008; 103(1): 57 - 67. [Abstract] [Full Text] [PDF] |
||||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Endocrinology | Endocrine Reviews | J. Clin. End. & Metab. |
| Molecular Endocrinology | Recent Prog. Horm. Res. | All Endocrine Journals |