Diverse effects of FSH and LH on proliferation of human ovarian surface epithelial cells

Karin Ivarsson1,3, Karin Sundfeldt1,2, Mats Brännström1, Pär Hellberg1 and Per Olof Janson1

1 Institute for the Health of Women and Children, Department of Obstetrics and Gynecology and 2 Department of Physiology, Göteborg University, Göteborg, Sweden


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The purpose of this study was to evaluate the effects of FSH and LH on growth regulation of normal ovarian surface epithelial (OSE) cells harvested from both premenopausal and postmenopausal women. Ovarian surface epithelial cells were obtained through brushing of the ovarian surface during surgery. FSH and LH were added to the OSE cultures and the proliferative effects were analysed using two different culture models, non-confluent and confluent cells, and two different detection methods, [3H]thymidine incorporation and a colorimetric cell number assay. FSH lowered the OSE proliferation under non-confluent conditions (10–27%), and the inhibitory effect was most pronounced among cells from postmenopausal women (P < 0.01). In the confluent model only cells from postmenopausal women showed significantly (P < 0.05) decreased proliferation. No effects of LH on OSE cells were detected. The unexpected results of an anti-proliferative effect of FSH on OSE, and the absence of effect by LH, do not support the theory that gonadotrophins are directly involved in ovarian carcinogenesis through an enhanced proliferation of OSE cells.

Key words: epithelial/FSH/ovarian/proliferation/surface


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The ovarian surface epithelium (OSE) is the origin of ~80–90% of ovarian adenocarcinomas (Scully, 1977Go), which account for most deaths in gynaecological tumour diseases in the developed world. The surface epithelium consists of a single cell layer of epithelial cells of mesodermal origin. Despite these cells being the origin of ovarian cancer, relatively little is known about the growth regulation of the OSE cells.

The `gonadotrophin theory' for ovarian cancer development is based on the observed rising incidence rate of ovarian cancer during the years after menopause, and suggests that the malignant transformation of the OSE cells is a consequence of the increased exposure to the postmenopausal elevated concentrations of FSH and LH (Stadel, 1975Go). This hypothesis also implies that the repeated exposure to high concentrations of gonadotrophins, and other intra-ovarian ovulation-related factors during the woman's fertile period, may be of importance for the OSE transformation. This theory is supported by the relative protective effects on ovarian epithelial cancer by oral contraceptives (Centers for Disease Control Cancer and Steroid Hormone Study, 1983Go; Cancer and Steroid Hormone Study of the Centers for Disease Control and the National Institute of Child Health and Human Development, 1987Go), multiparity (Negri et al., 1991Go; Adami et al., 1994Go), and breast-feeding (Whittemore et al., 1992Go). Furthermore, the potential influence of gonadotrophins on OSE cells is indicated by the presence of FSH receptors on the OSE (Zheng et al., 1996Go).

In order to increase the knowledge concerning the growth pattern of the human OSE cells during in-vitro culture conditions, several methods have been developed for harvesting these cells from ovarian biopsies, including scraping, dissection and enzymatic digestion (Auersperg et al., 1984Go; Berchuck et al., 1992Go; Ziltener et al., 1993Go; Nakamura et al., 1994Go; Karlan et al., 1995Go; Marth et al., 1996Go). Using such techniques, some studies designed to evaluate the growth regulation of human OSE cells have been performed, but so far, no investigations regarding the role of gonadotrophins in this regard have been presented. Stimulatory effects on OSE growth by epidermal growth factor (EGF), platelet-derived growth factor, tumour necrosis factor-{alpha}, and interleukin-1ß (Rodrigues et al., 1991Go; Marth et al., 1996Go; Dabrow et al., 1998Go), as well as growth-inhibiting actions by transforming growth factor-ß (Berchuck et al., 1992Go) have been reported. In another study, the effects of steroid hormones were investigated but no significant effects on cellular growth were demonstrated (Karlan et al., 1995Go).

Since ovarian specimens are not easily accessible for research, a simplified `brushing' technique has been developed to obtain OSE cells for culture during minimally invasive surgery, thus requiring neither biopsies nor removal of the ovaries. This study presents two different culture models attempting to reflect different stages in ovarian epithelial function. To further evaluate possible biological evidence for the gonadotrophin theory behind ovarian cancer development, the effects of the two gonadotrophins, LH and FSH, on OSE cell growth in vitro were investigated using cells harvested from both preand postmenopausal women.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Patients
In total, cells from 62 women (mean age 52 years, range 29–82) undergoing gynaecological surgery for benign disease were cultured. Of these cultures, 36 were obtained through laparotomy, 24 through laparoscopy, and two through vaginal surgery.

For proliferation experiments, cells from 25 women [16 premenopausal (mean ± SEM 42.2 ± 1.9 years, range 31–54) and nine postmenopausal (62.1 ± 2.2 years, range 55–74)] were used. Of these, 10 were obtained through laparoscopy and the remaining through laparotomy.

Harvest of cells and culture conditions
To obtain OSE cells, cytobrushes (Cytobrush® Plus; Medscand AB, Malmö, Sweden) normally used for cervical cytology were used. The sterilized brushes were used either directly when the peritoneal cavity had been opened at laparotomy, introduced into the abdominal cavity via a trocar at laparoscopy, or through the vagina at vaginal hysterectomy. The brushes were then rotated and at the same time moved over the ovarian surface exerting only a slight pressure to minimize the risk of damaging or disrupting the underlying basal membrane or tunica albuginea. For each ovary, two brushes were used, one for the medial and one for the lateral aspect. This technique was approved by the Ethical Committee of Göteborg University.

The brushes were immediately placed in culture medium, MCDB 105/M199 (1:1) supplemented with 15% fetal bovine serum (FBS; Life Technologies Ltd, Paisley, UK) and penicillin–streptomycin (100 IU/ml–100 µg/ml; Life Technologies Ltd) (Kruk et al., 1990Go), and taken to the laboratory. The cells were released from the brushes by rubbing them against each other, and the cell suspension was then centrifuged for 5 min at 300 g, diluted in fresh culture media, and then seeded into two 30 mm diameter culture dishes (Falcon; Becton Dickinson, Meylan, France). The media were changed every second day until the cells reached confluence. In order to verify the epithelial origin of the cells and to identify contaminating cells, some cells were seeded in chamber slides for immunohistochemical analyses with antibodies against low molecular weight cytokeratin (1:40; AE1/AE3), vimentin (1:100), and coagulation factor VIII (1:100) (Boehringer Mannheim, Mannheim, Germany).

Immunohistochemistry
The slides were washed with 1xphosphate-buffered saline (PBS), fixed with methanol (-20°C) for 30 s, air-dried and frozen at –20°C until analysis.

The chamber slides were incubated with 5% non-fat milk for 30 min, followed by incubation with primary antibodies overnight at room temperature. After washes in PBS, the dishes were incubated with a 1:100 diluted fluorescein isothiocyanate (FITC)-conjugated goat anti-rabbit secondary antibody for 1 h (Sigma Chemicals, St Louis, MO, USA). After subsequent washes with PBS and distilled water, the dishes were mounted in 4-diazabicyclo-2,2,2-octane (Dabco; Fluka, Buchs, Switzerland) and immediately photographed with a camera attached to a Nikon Optiphot-2 microscope.

Proliferation assays
Two different culture models were used to investigate the proliferation of the OSE cells. The non-confluent model made use of cells that were seeded at low density. The confluent model used cells that were seeded at a higher density and then were allowed to reach confluence before the proliferation experiments were initiated.

The confluent cells from the primary culture were trypsinized, counted in a Bürker chamber and seeded at a density of 4000 cells per well (non-confluent cultures) or 15 000 cells per well (confluent cultures) in 96-well microtitre plates (Falcon). The cells for non-confluent culture were allowed to attach and resume growth (overnight or for 24 h), and the cells for the confluent cultures were allowed to reach confluence (~4 days), before the media were changed to the same medium containing only 1% FBS and treatment agents. The cells were exposed to FSH (0.5 IU/ml Puregon®; Organon, Oss, The Netherlands), LH (0.1 IU/ml LHADI®; Serono Laboratories Inc., Rome, Italy), or epidermal growth factor (EGF) (10 ng/ml; Life Technologies) diluted in PBS. To the control cultures equal amounts of PBS were added instead of the agent studied. The experiments were performed in six duplicate wells (non-confluent cultures) or double or triple wells (confluent cultures) depending on the number of available cells.

The cultures were pulsed with [3H]thymidine (2 µCi/ml; NEN Life Science Products, Zaventum, Belgium) for 3 h before the cells were precipitated with 10% trichloroacetic acid (TCA) at 4, 24 and 48 h after start of stimulation. The precipitates were washed twice with 5% TCA and dissolved in warm 0.1 mol/l NaOH (40°C), and the radioactivity was determined in a liquid scintillation counter (Packard, Meriden, CT, USA).

The possible change of cell number was also detected with CellTiter 96® AQueous Non-Radioactive Cell Proliferation Assay (Promega, Madison, WI, USA), referred to in this paper as the MTS method. Since 4 h of gonadotrophin exposure was estimated to be too short a time for inducing changes in cell population detectable with the MTS method, only the possible changes after 24 and 48 h were assayed with this method.

Twenty µl of the MTS reducing agent were added to every well containing 0.1 ml medium, and was incubated for 3 h, before analysis at 492 nm in a plate reader spectrophotometer (Uvmax; Molecular Devices Corp., Sunnyvale, CA, USA).

Statistical analysis
Student's paired t-test was used to evaluate the difference between control groups and treated groups. The original proliferation data were natural logarithm-transformed to enable comparison despite large differences in control values between individual experiments. Proliferation data are displayed as a percentage of control in Figures 1 and 2.GoGo



View larger version (109K):
[in this window]
[in a new window]
 
Figure 1. (a) Confluent ovarian surface epithelium (OSE) cells in culture. Phase contrast microscopy (bar = 0.01 mm). (b) Immunofluorescent staining for cytokeratin in non-confluent OSE cells (bar = 0.025 mm).

 


View larger version (37K):
[in this window]
[in a new window]
 
Figure 2. Effects of FSH on ovarian surface epithelium growth evaluated with the MTS method and thymidine incorporation (3H-Th). Effects after 24 h (grey bars) and 48 h (black bars) were compared to untreated control (striped bars) expressed as 100%. Significantly lower than control: *P < 0.05; **P < 0.01. For description of MTS method, see text.

 

    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Epithelial cell culture
Cells from 62 women were collected. In 51 of these cultures (mean age of women 50.2 years), epithelial cell layers grew to confluence, with a mean time of culture to confluence of 13 days. The cultures were judged as epithelial based on their cobblestone-like appearance (Figure 1aGo), and on a strong cytokeratin staining in the cells during the initial 2–3 weeks of culture (Figure 1bGo). Cells examined after longer periods of culture or after several subcultures appeared progressively more elongated with a parallel loss of cytokeratin immunostaining and acquisition of vimentin immunostaining. Staining for coagulation factor VIII was absent from all cultures thus indicating that endothelial cells were not present.

The remaining 11 cultures, in which no cell layers developed (mean age of woman 58.6 years), usually had a low estimated initial number of cells, or were infected with bacteria during the first weeks of culture. When comparing the ages of the women whose cells resulted in failed and successful cultures, it was found that the women whose cells resulted in failed cultures were significantly (P < 0.05) older. The growth rates in established OSE cultures did not differ statistically significantly between cells from pre- and postmenopausal women; the mean population doubling times were 29.3 and 32.4 h respectively.

No difference in culture outcome could be established between cells obtained through laparotomy and laparoscopy (16.2 and 20.8% failed cultures respectively), although the cultures from women undergoing laparotomy usually included more red blood cells in the primary culture.

In experiments parallel to the gonadotrophin studies, the proliferative capacity of the OSE cells was investigated by the use of EGF as a positive control. In these experiments, the previously described (Rodrigues et al., 1991Go) proliferative effect of EGF on OSE cells was confirmed, with growth enhancement under both culture conditions (data not shown). The non-confluent OSE cells from premenopausal women responded to EGF with significantly (P < 0.05) increased growth (146%) only after 24 h detected with [3H]thymidine incorporation. In the confluent model, increased growth (310%) was detected with [3H]thymidine incorporation after 24 h (P < 0.01). In experiments with OSE cells from postmenopausal women, a significantly (P < 0.05) increased growth was detected with both the MTS method (116%) and [3H]thymidine incorporation (146%) after 24 h. After 48 h an increased growth (180%) was detected with [3H]thymidine incorporation (P < 0.001). In the confluent model, no significant results were detected with the MTS method, while the [3H]thymidine incorporation was significantly (P < 0.01) increased (329%) after 24 h of stimulation with EGF.

Effects of FSH
Cells from 24 women were used for the studies of FSH effects in cells from premenopausal and postmenopausal women (Figure 2Go).

Premenopausal women (n = 16)
In these cultures an overall decreased growth compared to untreated controls was detected. The most significant reduction was seen in the non-confluent state (10–27%), although only the decrease after 48 h detected with [3H]thymidine incorporation was statistically significant (P < 0.05). The confluent cultures showed no significant changes in growth (n = 4 with MTS, n = 8 with [3H]thymidine).

Postmenopausal women (n = 8)
In these cultures an overall decreased growth compared to untreated controls was detected. The non-confluent model demonstrated the most pronounced growth reduction also in these cells. The thymidine method (n = 5) detected significant reductions at 24 and 48 h (P < 0.01), while the MTS method (n = 8) showed significance only at 48 h (P < 0.05). In the confluent model, the MTS method (n = 5) recorded significant decreases after 24 and 48 h (P < 0.05), while no significant changes were detected with the thymidine incorporation method (n = 8).

Effects of LH
Cells from 17 women (eight premenopausal and nine postmenopausal) were used for the studies of LH effects (Figure 3Go). In contrast to the effects seen upon exposure to FSH, the LH-induced changes were smaller and mostly non-significant. In the non-confluent model, only the cells from postmenopausal women showed decreased growth after 48 h as detected with the MTS method (P < 0.05), while the thymidine incorporation showed no significance in the same experiment.



View larger version (35K):
[in this window]
[in a new window]
 
Figure 3. Effects of LH on ovarian surface epithelium growth evaluated with the MTS method and thymidine incorporation (3H-Th). Effects after 24 h (grey bars) and 48 h (black bars) were compared to untreated control (striped bars) expressed as 100%. Significantly lower than control: *P < 0.05.

 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
In this study, a partly new method was used for collecting human OSE cells. The method involves gentle brushing of the ovary during surgery, performed either by laparotomy, laparoscopy or during vaginal hysterectomy. This method has made it possible to investigate for the first time the possible effects of gonadotrophins on cell growth in an extensive material of normal OSE cells in vitro obtained from both premenopausal and postmenopausal women. Previously used methods, including dissection and enzymatic digestion (Auersperg et al., 1984Go; Berchuck et al., 1992Go; Ziltener et al., 1993Go; Nakamura et al., 1994Go; Karlan et al., 1995Go; Marth et al., 1996Go), may yield more cells as a starting material but do also involve a risk of obtaining impure OSE cell cultures, including fibroblasts and other cells which are present beneath the exterior basal membrane of the ovary. Such contamination was, however, found to be very rare with the brushing technique of the present study. Based on the results from this study, the risk of damaging the underlying basal membrane and tunica albuginea seems to be negligible with the use of soft brushes instead of scalpels or plastic scrapers.

The harvested cells of the present study developed into cell layers that were clearly epithelial, according to both immunohistochemical and morphological criteria. The cells retained their epithelial appearance during the first two or three subcultures, but acquired a more elongated fibroblast-like appearance during later passages. This was also demonstrated by the decreasing expression of cytokeratins during longer culture. This altered phenotype represents an epithelial/mesenchymal conversion (Birchmeier and Birchmeier, 1994Go) described as a normal feature for these cells, and reflecting their low epithelial differentiation as well as mesothelial origin (Kruk and Auersperg, 1992Go; Auersperg et al., 1994Go).

The use of fertility drugs in ovulation induction and assisted reproduction encompasses both increased exposure to gonadotrophin and multiple ovulations, which may be underlying factors in ovarian carcinogenesis according to the `gonadotrophin theory' (Stadel, 1975Go) and the `incessant ovulation theory' (Fathalla, 1971Go). Several epidemiological studies have been presented, indicating such an increased risk (Whittemore et al., 1992Go; Rossing et al., 1994Go). There are, however, contradictory results from several studies (for review, see Riman et al., 1998). In this study, attempts were made to supplement epidemiological data by investigating the effects of FSH and LH on the growth rate of normal OSE cells. The results show that FSH lowers the overall proliferative activity of non-confluent cells and that LH does not change the growth pattern of these cells.

The experiments on cells from postmenopausal women showed the most pronounced effects on proliferation with significant reductions by FSH detected with both methods. The absence of a decrease after 24 h with the MTS method in the non-confluent cells may reflect that the two methods are actually measuring two different variables, which are both related to proliferation. The effect seen at both 24 and 48 h on DNA replication, detected with thymidine incorporation, is likely to appear before any detectable effects on cell numbers would be detected with the MTS method.

In this study, two different culture models were used, with the aim of reflecting two distinct stages of OSE function. The non-confluent model corresponds to the situation when the cells are growing to cover a surface, as in tissue repair after ovulation or any other insult to the ovarian surface. The confluent model may reflect a situation where the cells cover a surface without any major disturbances or changes, like the OSE cells of the postmenopausal ovary. Such cells would normally not grow unless the confluent layer was mechanically disrupted.

Thus, the use of both the non-confluent and confluent culture model in this study would broaden the experimental findings to reflect proliferation during premenopausal (non-confluent) and postmenopausal (confluent) conditions.

The confluent premenopausal cells in this study, in contrast to the non-confluent cells, did not respond significantly either to FSH or to LH. However, a growth-inhibiting effect would not lead to any detectable changes in a confluent cell layer, in which no or very little cell growth is occurring. Thus, the statistically significant decreases in cell number observed in cultures of OSE cells from postmenopausal women may indicate an apoptotic effect of FSH.

The somewhat surprising results of lowered proliferation induced by FSH are against the general dogma, since FSH is most commonly described as a proliferative agent, based on its effects on growing follicles. However, several alternatively spliced forms of the FSH receptor have recently been reported in other species (Sairam et al., 1996Go), which may also be true in humans and may provide mechanisms for different functions of FSH, depending on specific FSH receptor expression. The observed inhibitory effect of FSH on OSE growth was most pronounced among the OSE cells from postmenopausal women, indicating a changed sensitivity to this specific gonadotrophin after the menopause. This may be explained by a changed expression of FSH receptors on this cell type at the time of menopause. The only report on the presence of FSH receptors on ovarian surface epithelium (Zheng et al., 1996Go) includes only women of childbearing age, and no comparisons between women of different ages have so far been performed. The more pronounced inhibitory effect on cells from postmenopausal women seen in this study may be a result of a shift to a less strict endocrine functional control in the postmenopausal period, permitting the OSE to respond to factors that earlier in life were not involved in growth regulation. A speculation on this phenomenon is that a diminished proliferation of OSE induced by FSH after menopause is biologically relevant since the ovaries become smaller in size after the menopause, and that it may even be part of a mechanism protecting against uncontrolled growth.

In contrast to FSH, LH only induced significant growth inhibition in one of all models, methods and time-points tested. This effect, seen after 48 h of exposure in the postmenopausal non-confluent model detected with the MTS method, is difficult to interpret since the thymidine incorporation experiments did not show any corresponding effects. There appear to be no reports on the presence of LH receptors on normal human ovarian surface epithelium, whereas LH (human chorionic gonadotrophin) receptors, demonstrated as specific binding, were found to be present in 28% of malignant ovarian epithelial tumours (Rajaniemi et al., 1981Go). FSH receptors have also been detected in ovarian epithelial tumours (Kammerman et al., 1981Go; Nakano et al., 1989Go). The presence of gonadotrophin receptors in portions of the epithelial ovarian cancer, as well as FSH receptors in normal OSE, would suggest a direct involvement of gonadotrophins in epithelial transformation in at least some of these tumours. However, the normal OSE cells in the current study were inhibited by FSH and not influenced by LH, in contrast to reports of enhanced proliferation of ovarian cancer cells by gonadotrophins (Simon et al., 1983Go). The results of the present study are in line with the results of an epidemiological study (Helzlsouer et al., 1995Go), in which high gonadotrophin concentrations were found to be protective against ovarian cancer. The findings of the current study do not support the gonadotrophin theory in which the gonadotrophins are suggested to influence the growth and directly induce transformation of OSE cells. However, there may be other explanations for the coincidence of elevated gonadotrophin concentrations and increased rate of ovarian cancers after menopause. A recent report showed the influence of gonadotrophins on neovascularization of ovarian tumours, and also the connection to expression of vascular endothelial growth factor (Schiffenbauer et al., 1997Go), a well known angiogenic factor presumably involved in ovarian tumour biology (Mesiano et al., 1998Go). This implies that the role of the gonadotrophins is to facilitate the growth of already existing microtumours through improved blood supply, rather than to induce the epithelial cell transformation itself.

In conclusion, the results from this study do not indicate an involvement of gonadotrophins in the transformation of OSE cells through effects on the proliferation rate, although they do not exclude their role for regulation of tumour growth later in the cancer process.


    Acknowledgments
 
This work was supported by grants from the Swedish Medical Research Council (11607 to M.B.) and by the research foundations of King Gustaf V Jubilee Clinic, Hjalmar Svensson, Nilsson, and Assar Gabrielsson.


    Notes
 
3 To whom correspondence should be addressed at: Department of Obstetrics and Gynecology, Göteborg University, Sahlgrenska University Hospital, 413 45 Göteborg, Sweden. E-mail: karin.ivarsson{at}medfak.gu.se Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Adami, H.O., Hsieh, C.C., Lambe, M. et al. (1994) Parity, age at first childbirth, and risk of ovarian cancer. Lancet, 344, 1250–1254.[ISI][Medline]

Auersperg, N., Siemens, C.H. and Myrdal, S.E. (1984) Human ovarian surface epithelium in primary culture. In Vitro, 20, 743–755.[ISI][Medline]

Auersperg, N., Maines-Bandiera, S.L., Dyck, H.G. et al. (1994) Characterization of cultured human ovarian surface epithelial cells: phenotypic plasticity and premalignant changes. Lab. Invest., 71, 510–518.[ISI][Medline]

Berchuck, A., Rodriguez, G., Olt, G. et al. (1992) Regulation of growth of normal ovarian epithelial cells and ovarian cancer cell lines by transforming growth factor-ß. Am. J. Obstet. Gynecol., 166, 676–684.[ISI][Medline]

Birchmeier, W. and Birchmeier, C. (1994) Mesenchymal-epithelial transitions. BioEssays, 16, 305–307.[ISI][Medline]

Cancer and Steroid Hormone Study of the Centers for Disease Control and the National Institute of Child Health and Human Development (1987) The reduction on risk of ovarian cancer associated with oral-contraceptives use. N. Engl. J. Med., 316, 650–655.[Abstract]

Centers for Disease Control Cancer and Steroid Hormone Study (1983) Oral contraceptive use and risk of ovarian cancer. J. Am. Med. Assoc., 249, 1596–1599.[Abstract]

Dabrow, M.B., Francesco, M.R., McBrearty, F.X. et al. (1998) The effects of platelet-derived growth factor and receptor on normal and neoplastic human ovarian surface epithelium. Gynecol. Oncol., 71, 29–37.[ISI][Medline]

Fathalla, M.F. (1971) Incessant ovulation | a factor in ovarian neoplasia? Lancet, ii, 163.

Helzlsouer, K.J., Alberg, A.J., Gordon, G.B. et al. (1995) Serum gonadotropins and steroid hormones and the development of ovarian cancer. JAMA, 274, 1926–1930.[Abstract]

Kammerman, S., Demopoulos, R.I., Raphael, C. et al. (1981) Gonadotropic hormone binding to human ovarian tumors. Hum. Pathol., 12, 886–890.[ISI][Medline]

Karlan, B.Y., Jones, J., Greenwald, M. et al. (1995) Steroid hormone effects on the proliferation of human ovarian surface epithelium in vitro. Am. J. Obstet. Gynecol., 173, 97–104.[ISI][Medline]

Kruk, P.A. and Auersperg, N. (1992) Human ovarian surface epithelial cells are capable of physically restructuring extracellular matrix. Am. J. Obstet. Gynecol., 167, 1437–1443.[ISI][Medline]

Kruk, P.A., Maines-Bandera, S.L. and Auersperg, N. (1990) A simplified method to culture human ovarian surface epithelium. Lab. Invest., 63, 132–136.[ISI][Medline]

Marth, C., Zeimet, A.G., Herold, M. et al. (1996) Different effects of interferons, interleukin-1ß and tumor necrosis factor-{alpha} in normal (OSE) and malignant human ovarian epithelial cells. Int. J. Cancer, 67, 826–30.[ISI][Medline]

Mesiano, S., Ferrara, N. and Jaffe, R.B. (1998) Role of vascular endothelial growth factor in ovarian cancer: inhibition of ascites formation by immunoneutralization. Am. J. Pathol., 153, 1249–1256.[Abstract/Free Full Text]

Nakamura, M., Katabuchi, H., Ohba, T. et al. (1994) Isolation, growth and characteristics of human ovarian surface epithelium. Virchows Arch, 424, 59–67.[ISI][Medline]

Nakano, R., Kitayama, S., Yamoto, M. et al. (1989) Localization of gonadotropin binding sites in human ovarian neoplasms. Am. J. Obstet. Gynecol., 161, 905–910.[ISI][Medline]

Negri, E., Franceschi, S., Tzonou, A. et al. (1991) Pooled analysis of 3 European case-control studies: I. Reproductive factors and risk of epithelial ovarian cancer. Int. J. Cancer, 49, 50–56.[ISI][Medline]

Rajaniemi, H., Kauppila, A., Ronnberg, L. et al. (1981) LH (hCG) receptor in benign and malignant tumors of human ovary. Acta Obstet. Gynecol. Scand. Suppl., 101, 83–86.[Medline]

Riman, T., Persson, I. and Nilsson, S. (1998) Hormonal aspects of epithelial ovarian cancer: review of epidemiological evidence. Clin. Invest., 49, 695–707.

Rodrigues, G.C., Berchuck, A., Whitaker, R.S. et al. (1991) Epidermal growth factor receptor expression in normal ovarian epithelium and ovarian cancer II. Relationship between receptor expression and response to epidermal growth factor. Am. J. Obstet. Gynecol., 164, 745–750.[ISI][Medline]

Rossing, M.A., Daling, J.R., Weiss, N.S. et al. (1994) Ovarian tumours in a cohort of infertile women. N. Engl. J. Med., 331, 771–776.[Abstract/Free Full Text]

Sairam, M.R., Jiang, L.G., Yamey, T.A. et al. (1996) Follitropin signal transduction: Alternative splicing of the FSH receptor gene produces a dominant negative form of receptor which inhibits hormone action. Biochem. Biophys. Res. Commun., 226, 717–722.[ISI][Medline]

Schiffenbauer, Y.S., Abramovitch, R., Meir, G. et al. (1997) Loss of ovarian function promotes angiogenesis in human ovarian carcinoma. Proc. Natl Acad. Sci. USA, 94, 13203–13208.[Abstract/Free Full Text]

Scully, R.E. (1977) Ovarian tumours—a review. Am. J. Pathol., 87, 686–720.[ISI][Medline]

Simon, W.E., Albrecht, M., Hansel, M. et al. (1983) Cell lines derived from human ovarian carcinomas: growth stimulation by gonadotropic and steroid hormones. J. Natl Cancer Inst., 70, 839–845.[ISI][Medline]

Stadel, B.V. (1975) Letter: The ethiology and prevention of ovarian cancer. Am. J. Obstet. Gynecol., 123, 772–774.[Medline]

Whittemore, A.S., Harris, R. and Intyre, J. (1992) Characteristics relating to ovarian cancer risk: collaborative analysis of 12 US case-control studies. II. Invasive epithelial ovarian cancers in white women. Collaborative Ovarian Cancer Group. Am. J. Epidemiol., 136, 1184–203.[Abstract]

Zheng, W., Magid, M.S., Kramer, E.E. et al. (1996) Follicle-stimulating hormone receptor is expressed in human ovarian surface epithelium and fallopian tube. Am. J. Pathol., 148, 47–53.[Abstract]

Ziltener, H.J., Maines-Bandiera, S., Schrader, J.W. et al. (1993) Secretion of bioactive interleukin-1, interleukin-6, and colony-stimulating factors by human ovarian surface epithelium. Biol. Reprod., 49, 635–641.[Abstract]

Submitted on June 12, 2000; accepted on October 6, 2000.