|
|
||||||||
Articles |
a California Regional Primate Research Center,
b Division of Reproductive Biology,
c Department of Ob/Gyn and Institute of Toxicology and Environmental Health, University of California, Davis, California 95616
| ABSTRACT |
|---|
|
|
|---|
| INTRODUCTION |
|---|
|
|
|---|
It has been proposed that steroid hormones produced by the corpus luteum may have autocrine effects on luteal function or life span. Knobil [3] hypothesized that luteal estrogens are involved in a "self-destruct" mechanism in the nonpregnant menstrual cycle. Alternatively, it has been suggested that P4 promotes its own secretion and acts as a universal luteotropin [4]. Recently, the presence of progesterone receptors in the human and monkey corpus luteum has been confirmed by immunocytochemistry [5,6] and by the presence of mRNA in monkey luteal tissue [7]. The absence of progesterone receptors in preovulatory follicles, and their appearance in periovulatory follicles after exposure to a gonadotropin surge, support the hypothesis that LH induces progesterone receptor expression [6,8].
The high levels of P4 secreted by the corpus luteum and by granulosa-lutein cells (GLC) in culture have complicated investigation of the autocrine effects of P4. However, the development of antiprogestins has led to new information on the action of P4, especially in the endometrium [9,10], and the characterization of the progesterone receptor [11]. An additional complication of determining the effect of P4 on corpus luteum function is that during the early luteal phase in primates, the levels of progesterone receptor and its mRNA are at their lowest point and subsequently increase dramatically during the mid to late luteal phase [57]. Therefore, the autocrine effects of P4 may be different during the early and the mid to late luteal phases.
We have developed a cell culture system for human GLC that supports the timely and dynamic secretion of estrogen (estradiol-17ß; E2), P4, and relaxin (Rlx) in patterns that mimic serum patterns of secretion of these hormones during the luteal phase of the menstrual cycle [12]. To test the hypothesis that P4 is a luteotropin and is required for its own production, GLC were treated in vitro with two P4 receptor blockers, RU-486 and HRP2000, and the endocrine products P4, E2, and Rlx were measured in culture medium.
| MATERIALS AND METHODS |
|---|
|
|
|---|
Human granulosa cells were obtained from women receiving assisted reproduction treatment at Pacific Fertility Center (Sacramento, CA). The cells were a by-product of the in vitro fertilization/embryo transfer procedure and normally would have been discarded. They were provided for this study as coded samples with the identities of the women unavailable, so this research was granted an exemption from review by the university human subjects review committee. The patients received varying regimens of Metrodin (Serono, Randolph, MA) and Pergonal (Serono) and received 10 000 IU of hCG 36 h prior to follicular aspiration. Approximately 1.0 ml modified human tubal fluid medium (Irvine Scientific, Santa Ana, CA) containing Hepes buffer, antibiotics, and heparin was added to each follicular fluid aspirate during the oocyte retrieval procedure. After oocytes and cumulus masses were removed, the follicular fluid containing granulosa cells was refrigerated and transported on ice to California Regional Primate Research Center. Individual follicles were not distinguished as all granulosa cells from an individual were pooled, but cells from different subjects were not pooled.
Minimal Essential Medium (MEM; Gibco, Grand Island, NY) was modified with the following additions: sodium bicarbonate, 4.4 mg/100 ml MEM (Sigma Chemical Co., St. Louis, MO); fungizone, 1 ml/100 ml (Gibco); penicillin G, 6 mg/100 ml (Sigma); streptomycin sulfate, 6 mg/100 ml (Sigma); and 10% fetal calf serum (Hyclone, Logan, UT). Medium was filtered through a 0.22-µm sterile syringe filter (Fisher, Santa Clara, CA) and equilibrated at 37°C and 5% CO2 in air prior to use. Human CG (Pregnyl; Organon, W. Orange, NJ) was added to the culture medium in amounts as described below. Matrigel (50 µg/well; Becton-Dickinson Labware, Franklin Lakes, NJ) was applied to 4-well Nunc (Naperville, IL) dishes on the same day on which cells were collected, according to the manufacturer's directions. All plates, Matrigel, and pipettes were kept on ice during the coating procedure. Coated plates were incubated at 37°C for 30 min in 5% CO2 to set the Matrigel.
Follicular fluid from each subject was divided equally into 15-ml disposable, sterile centrifuge tubes and centrifuged at 300 x g for 5 min and then at 500 x g for an additional 5 min. This created a firm layer of granulosa cells on top of a red blood cell pellet. The layers of granulosa cells were combined in a sterile 15-ml centrifuge tube. About 4 ml MEM was added and the cells were gently aspirated through a 1.0-ml disposable pipette tip to break up clumps. Aliquots (1 ml) of this cell suspension were layered onto 1.0 ml 40% Percoll (Sigma) in PBS columns in 15-ml centrifuge tubes and centrifuged at 500 x g for 30 min. The cell layer was removed from each Percoll column and combined in a sterile 15-ml centrifuge tube. Cells were washed twice with 510 ml fresh MEM and centrifuged for 10 min at 300 x g. The pellet was resuspended in 24 ml of MEM, and cells were filtered through an 89-µm polyester filter (Spectra/Mesh, Laguna Hills, CA). Cells were counted on a hemacytometer, brought to a final concentration of 1 x 105 cells/ml in MEM, and plated at 5 x 104 cells per well (0.5 ml). Cells had attached after 24 h, and medium was changed to remove remaining debris. Medium was changed daily in all experiments and stored frozen until assay for hormone concentrations.
Verification of Viability and Cell Number During Culture
Cells were prepared, plated, and cultured as described above with multiple wells for each patient so that cell number and viability could be measured on wells that were replicates of those used for hormone production measurements. Estimates of viability were obtained using trypan blue (0.4%; Gibco) exclusion on an Olympus (Tokyo, Japan) CK2 microscope at x200. On Day 16, medium was removed from the well, and cells were rinsed 3 times with cold PBS (Sigma). One milliliter of Matrisperse (Fisher) was added to each well to free cells from the Matrigel, and cells were scraped into a centrifuge tube. The well was rinsed with an additional 1 ml of Matrisperse that was placed in the tube and kept on ice for 1 h. Cells were centrifuged for 5 min at 500 x g, and the pellet was resuspended in 100 µl PBS. Cells were counted on a hemacytometer.
Human CG Stimulation Protocols
Two protocols of hCG administration to the culture medium were utilized to simulate two different luteal-phase events using 2 replicate wells from 7 subjects. The first protocol (Baseline) was a constant baseline dose of hCG to simulate a normal nonconceptive luteal phase. A baseline concentration of 0.02 IU/ml was selected from the dose-response study on the basis of its ability to maintain physiological profiles of steroid and Rlx secretion [12]. Human CG concentrations were held at baseline for the 20 days of culture. A second protocol (Rescue) was used to simulate early pregnancy beginning on Day 9 of culture. Human CG was maintained at the baseline concentration (0.02 IU/ml) for Days 1 through 8 of culture and was then doubled each day until Day 16 of culture. On Days 1720 of culture the hCG concentration was maintained at the highest hCG level (5.12 IU/ml). The Rescue protocol was used to determine whether additional hCG could overcome the effects of progesterone receptor antagonists.
Progesterone Receptor Antagonist Protocols
In the first experiment, cells from 7 different women were used. The progesterone receptor antagonists, RU-486 (Roussel-UCLAF, Paris, France) and HRP2000 (National Institutes of Health), were obtained from the World Health Organization. RU-486 was dissolved in ethanol to a concentration of 25 mM and stored at 4°C as a stock solution. HRP2000 was used in the same manner but was dissolved in dimethyl sulfoxide. RU-486 and HRP2000 were added directly to medium in the cell culture wells (1 µl of 25 mM stock to 500 µl medium for a final concentration of 50 µM). Antagonists were added on Days 8 and 9 of culture in the first experiment (late groups), which included either Baseline hCG or Rescue hCG protocol levels (a total of 8 treatment groups). In the second experiment using cells from a second group of 7 women, RU-486 and HRP2000 were added on Days 2 and 3 of culture (Early groups) to cells with only Baseline hCG treatment.
Assays
Although cell culture medium was collected daily, hormone measurements were performed on alternate days. E2 and P4 were measured on even days using commercial kits (Diagnostic Products Corp., Los Angeles, CA) as previously reported [12]. Rlx was measured on odd days by an enzyme immunoassay as previously reported [13] using reagents generously provided by Connetics Corp. (Palo Alto, CA).
Data Analysis
To normalize the endocrine data, the values were converted to the common logarithm for statistical analysis and averaging. Data were converted to arithmetic scale for graphing (geometric mean). Parametric tests of significance depend upon normally distributed data. Endocrine data are frequently not normally distributed, and a logarithmic transformation was used to normalize the data for these tests. Hormone values for the treatment protocols were compared by two-way repeated-measures ANOVA, and significance was determined by Student-Newman-Keuls multiple comparisons test using P < 0.05 as a significance level.
| RESULTS |
|---|
|
|
|---|
|
In Early treatment, RU-486 and HRP2000 were given for the same duration as above (48 h) but on Days 2 and 3 of culture (Fig. 2). The antagonists induced an immediate decrease in estrogen and P4 production by GLC. However, unlike those of cells in the Late-treatment protocol, the hormone concentrations in the Early-treatment cells were not different from control levels within 5 days of cessation of antiprogestin treatment. Rlx production is not normally detectable until Day 5 of culture and therefore was not affected during the antagonist treatment in the Early protocol. However, RU-486 and HRP2000 significantly delayed the initiation of Rlx production. Rlx was first detected on Days 56 in controls and on Days 910 in RU-486-treated and later in HRP2000-treated cells. Moreover, the Rlx production in the antiprogestin-treated cells was significantly lower than Rlx production in control cultures. Early treatment did not significantly affect cell number or viability (data not shown).
|
| DISCUSSION |
|---|
|
|
|---|
The effect of antiprogestins on estrogen production by GLC has not been confirmed by other investigators. However, several studies on women receiving RU-486 during various phases of the menstrual cycle show effects on serum estrogen levels. Whether administered during the follicular phase [17], the late follicular phase [18], the midluteal phase [17], or the late luteal phase [19] of the menstrual cycle, RU-486 decreased serum estrogen concentrations within 1 day of treatment. However, when RU-486 was given for the first 3 days of the menstrual cycle, serum E2, follicular phase length, and luteal phase length were unaffected [20].
The effects of progesterone receptor blockers on Rlx production either in vitro or in vivo have not been previously reported. We used Rlx as a marker of peptide secretion to determine whether the effects on the progesterone receptor were confined to steroid secretion or extended more broadly to other classes of hormones. In our studies, the disruption of Rlx production by both RU-486 and HRP2000 when given to GLC either early or later in culture suggests that P4 may have an autocrine role in the regulation of Rlx production. Moreover, after discontinuation of the antiprogestins when GLC were treated on Days 2 and 3, P4 production began to increase immediately; however, Rlx production did not increase for another 4 days. It may be significant that this delay is similar in length to that of the initial normal delay in Rlx secretion of about 5 days, suggesting that luteal cells must be exposed to P4 for several days to develop the ability to produce Rlx.
Comparing this study with other studies of GLC function in vitro is also difficult because culture conditions are not identical. However, several studies have found that RU-486 treatment results in significant declines in P4 production by GLC [16,21,22]; but Greenberg et al. [23] did not find an effect of RU-486 on P4 production. The lack of effect in the latter study may be due to the high P4 levels in culture and the relatively low dose of RU-486 (up to 100 nM). RU-486 binds to the same receptors as does P4, and it has been shown to compete for the specific binding site with P4 [9]. The studies showing an effect of antiprogestins on P4 production either utilized protocols that produced low levels of P4 or used a higher dose of RU-486 (50 µM). Interestingly, in our experiments, which utilized a higher dose of RU-486 and low hCG support during the baseline phase of the culture period, increasing the gonadotropin support to simulate early pregnancy did not overcome the effect of RU-486 on P4 production. These differences in response to low and high RU-486 levels in vitro are similar to the in vivo effect of RU-486 in women. When RU-486 is given during the midluteal phase of the menstrual cycle, serum P4 levels are significantly decreased in women receiving 100 mg/day of RU-486, while only 20% of women receiving 25 mg/day of RU-486 showed a decline in serum P4 [24].
Comparisons of in vitro and in vivo studies are difficult because of the many differences between the culture conditions of the GLC and those under which the human corpus luteum functions in vivo: 1) GLC came from follicles that had been exposed to a pattern of FSH/LH treatment different from the FSH/LH secretion in the normal human menstrual cycle; 2) none of the other constituent cells of the corpus luteum, such as theca interna or endothelial cells, were included in the cell cultures; and 3) although hCG resembles LH and may act through the same receptor, hCG is not the same as LH, and hCG treatment in vitro does not follow the pattern or pulsatile secretion of LH during the luteal phase of the menstrual cycle. In spite of these differences, GLC in vitro apparently produce E2, P4, and Rlx in patterns similar to those seen during the luteal phase of a normal menstrual cycle [12].
Previous studies have demonstrated that decreased P4 levels after trilostane administration during the periovulatory interval or the midluteal phase can result in P4 levels that are diminished throughout the luteal phase [25,26]. However, macaques given trilostane, a 3ß-hydroxysteroid dehydrogenase inhibitor, had dramatically reduced P4 levels for several days following administration on Days 6 and 7 of the luteal phase [27]. Despite the reduction in P4, Rlx secretion remained normal during the luteal phase. It is possible that the low levels of P4 remaining were adequate to provide an autocrine effect on the ovary to maintain Rlx production.
Although the effects of the two antiprogestins were similar, RU-486 was slightly more effective in decreasing hormone production in the GLC model at the doses used. The ability of these two antiprogestins to induce abortions has been previously compared in vivo in macaques [28]. Both compounds were effective at inducing abortion when given by i.m. injection, but RU-486 appeared more effective than HRP2000 when given orally. Additionally, the decrease in serum P4 levels in macaques treated with RU-486 occurred more rapidly and to a greater magnitude than in HRP2000-treated animals. The relative effectiveness of the two antiprogesterones in vivo was similar to the response of the GLC model in the present study.
The effects of these antiprogestins are not specific to the progesterone receptor. Both RU-486 and HRP2000 have glucocorticoid receptor antagonist activity [29] that may contribute to the effects of these compounds on GLC hormone production. Michael et al. [30] demonstrated that cortisol inhibits LH-stimulated steroidogenesis in human GLC and that the sensitivity of cells from different patients to cortisol depends on the intracellular level of 11ß-hydroxysteroid dehydrogenase, the enzyme that metabolizes cortisol to the inactive derivative, cortisone. That study also showed that cortisol decreased the production of pregnenolone, the precursor to ovarian steroid production. Therefore, it is possible that glucocorticoid receptor activity could have an effect on steroid hormone production that is independent of the progesterone receptor. However, the effects of cortisol noted above do not provide an explanation for the action of antiprogestins on the production of Rlx by GLC. Further studies with antiprogestins with little glucocorticoid antagonist activity will be necessary to elucidate the relative importance of progesterone and glucocorticoid receptors in the hormone production of GLC.
In summary, we found that treatment of GLC with either of 2 progesterone receptor antagonists starting late in culture produced an acute and permanent drop in steroid and peptide hormone secretion. In contrast, treatment early in the culture period resulted in only a temporary decline in steroid and a delay in the appearance of peptide secretion. This supports a role of the progesterone receptor in ovarian hormone secretion, and this role changes during luteal aging. The data presented here support the hypothesis that P4 has autocrine effects on its own production and extend this hypothesis to other ovarian steroids and peptides.
| FOOTNOTES |
|---|
1 Supported in part by National Institutes of Health Grants ESO 6198 and P51-RR00169 and a grant from the Andrew W. Mellon Foundation. ![]()
2 Correspondence. FAX: 530 752 2880; cavandevoort{at}ucdavis.edu ![]()
Accepted: August 31, 1999.
Received: August 10, 1998.
| REFERENCES |
|---|
|
|
|---|
This article has been cited by other articles:
![]() |
L. Engmann, R. Losel, M. Wehling, and J. J. Peluso Progesterone Regulation of Human Granulosa/Luteal Cell Viability by an RU486-Independent Mechanism J. Clin. Endocrinol. Metab., December 1, 2006; 91(12): 4962 - 4968. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. P. Houk, E. J. Pearson, N. Martinelle, P. K. Donahoe, and J. Teixeira Feedback Inhibition of Steroidogenic Acute Regulatory Protein Expression in Vitro and in Vivo by Androgens Endocrinology, March 1, 2004; 145(3): 1269 - 1275. [Abstract] [Full Text] [PDF] |
||||
![]() |
C.L. Swan, M.C. Agostini, P.M. Bartlewski, V. Feyles, R.J. Urban, and P.J. Chedrese Effects of Progestins on Progesterone Synthesis in a Stable Porcine Granulosa Cell Line: Control of Transcriptional Activity of the Cytochrome P450 Side-Chain Cleavage Gene Biol Reprod, April 1, 2002; 66(4): 959 - 965. [Abstract] [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |