Aromatase activity of human granulosa cells in vitro: effects of gonadotrophins and follicular fluid

P. Guet1,3, D. Royère1,4, A. Paris2, J. Lansac1 and M.A. Driancourt3

1 Reproductive Biology Unit, Dept of Obs–Gyn & Human Reproduction, CHU Tours, Faculté de Médecine, 37 044 Tours cedex, 2 INRA Xenobiotics, Toulouse and 3 INRA–URA CNRS1291, PRMD, Nouzilly, France


    Abstract
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The aim of this study was to assess whether human dominant follicular fluid has the ability to modulate aromatase activity and/or granulosa cell proliferation. Dominant follicular fluid was obtained by laparoscopy before the luteinizing hormone surge in naturally cycling women while granulosa cells used in the tests were obtained from in-vitro fertilization patients. Aromatase was measured by the tritiated water release assay, following a 48 h incubation with follicular fluid and serum, and expressed for 5x104 granulosa cells. The effects of a range of follicular fluid or serum concentrations (2.5, 5, 10 and 20%) were compared. A decrease in aromatase activity was observed when high follicular fluid concentrations (20%) (P < 0.01) were added. Low concentrations (2.5%) of follicular fluid significantly increased cell proliferation (P < 0.01) as compared to basal values (0%). No further stimulation was however observed when concentrations increased up to 20%. Further characterization of these compounds is required to understand how they may modulate maturation of the dominant follicle.

Key words: aromatase/follicular fluid/gonadotrophins/granulosa cells/human


    Introduction
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
During the follicular phase, terminal follicular growth and maturation occur (Gougeon, 1996Go). Pre-ovulatory human follicles contain 50x106 granulosa cells and secrete large amounts of androgens and oestrogens. Steroidogenesis of the mature follicle operates according to the `two-cell, two gonadotrophins' model (Armstrong et al., 1979Go). Gonadotrophin effects on steroidogenesis are well documented: theca interna cells are stimulated by luteinizing hormone (LH) to produce aromatizable androgens (Erickson, 1993Go; Karnitis et al., 1994Go; Hillier et al., 1994Go) which are transferred to granulosa cells where they are converted to oestrogens by aromatase. This enzyme is induced by follicle-stimulating hormone (FSH) (Hillier et al., 1994Go; Smyth et al., 1995Go). However, there is also evidence suggesting that steroidogenesis can be modulated by follicular fluid compounds. For example, stimulatory and inhibitory effects of porcine follicular fluid have been identified on progesterone or oestradiol production by porcine granulosa cells (Ledwitz Rigby, 1983Go, 1987Go; Ledwitz Rigby et al., 1985Go). Interestingly, the type (positive or negative) and the magnitude of effects appeared to vary with the origin of follicular fluid and the size of the follicles providing the cells. Along a similar line, di Zerega et al. (di Zerega et al., 1982Go, 1983aGo,di Zerega et al., bGo) identified an inhibitory action of human follicular fluid on ovarian oestradiol output by aromatase. Attempts to identify the compound or compounds responsible for this effect initially used a heterospecific bioassay measuring ovarian weight and serum 17ß-oestradiol following human menopausal gonadotrophin (HMG) administration to hypophysectomized rats treated with diethylstilboestrol (di Zerega et al., 1982Go). Later, another heterospecific model was developed which monitored changes in aromatase of porcine granulosa cells (di Zerega et al., 1983aGo). However, none of these models resulted in the sequencing and identification of the active compounds.

In addition to its effect on steroidogenesis, follicular fluid also has the ability to alter cell proliferation. On one hand, human follicular fluid has been shown to be mitogenic for endothelial cells (Bryant et al., 1988Go). On the other hand, ovine and bovine follicular fluid appear to contain factors which inhibit proliferation of fibroblasts (Carson et al., 1988Go) or bovine granulosa cells (Hynes et al., 1996Go). At present, however, the net effect of human follicular fluid on human granulosa cell proliferation is unknown.

The aims of this work were to validate a homospecific test system to analyse the effects of human follicular fluid on human granulosa cells. Hence the following steps were attempted: (i) optimization of an aromatase assay using the conversion of 3H-testosterone to 3H2O by human granulosa cells in vitro, (ii) description of the kinetics (Km, Vmax) of aromatase using this in-vitro model and (iii) use of this test to check whether human gonadotrophins and human follicular fluid modulate aromatase activity or cell proliferation.


    Materials and methods
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Cell culture
Granulosa cells were collected from pre-ovulatory follicles during oocyte retrieval for in-vitro fertilization (IVF). Briefly, stimulation protocols combined down-regulation with a gonadotrophin-releasing hormone (GnRH) agonist [triptorelin, 0.1 mg/day s.c. (Decapeptyl®; Ipsen, Paris, France) or buserelin, 0.6 mg/day, s.c. (Suprefact®; Hoechst, Paris, France)] followed by gonadotrophins, 1–4 ampoules/day (HMG = Neopergonal®; Serono, Paris, France or Humegon®; Organon, Paris, France; FSH = Metrodine HP®; Serono). Gonadotrophin treatment was initiated when pituitary function was down-regulated (17ß-oestradiol concentrations <50 pg/ml, no follicles visualized by ultrasonography) about 10–20 days later. Gonadotrophins were injected until one follicle, 15–18 mm in diameter, could be visualized and plasma 17ß-oestradiol concentrations were 500–1000 pmol/ml per follicle >13 mm in diameter, with at least one follicle reaching 18 mm in diameter. Finally, human chorionic gonadotrophin (HCG, 10 000 IU i.m., gonadotrophin chorionique®; Endo, Organon) was administered 34–36 h before oocyte retrieval was realized. After isolation of cumulus oocyte complexes (used for IVF), follicular fluids were pooled, centrifuged (400 g, 10 min) and most of the red blood cells were discarded from the pellet using two layers of discontinuous percoll gradient (40%, 60% in Ham's medium, Gibco-BRL; Life Technologies, Cergy Pontoise, France) and centrifugation (400 g, 20 min). In order to remove as many red blood cells as possible, the 40% fraction was then treated with haemolytic medium (NH4Cl 10 mmol/l in Tris HCl pH 7.5; Sigma, Isles d'Abeau, France). Following centrifugation, the pellet was washed in minimum essential medium (MEM, Sigma-Aldrich Company, St Quentin Fallavier, France). Cell viability was determined using Trypan Blue dye exclusion. Granulosa cells were seeded at 5x104 viable cells per well in 96-multiwell plates (Falcon, Becton-Dickinson, Oxnard, CA, USA) in 200 µl MEM per well and incubated (37°C, 5% CO2 in air). Following a 24 h plating period, the medium was aspirated, and fresh medium added. Cells were further cultured for 60 h. This protocol, used to assess gonadotrophin and follicular fluid effects, was based on the pretrials detailed below.

Measurement of aromatase activity
Aromatase activity was assessed by measuring the conversion of [1ß, 2ß] 3H-testosterone to 3H2O as previously reported (Gore Langton and Dorrington, 1981Go).

After a 48 h culture period with the test substances (in the absence of the labelled substrate), cells were cultured for 12 h with 0.5 µmol of 1ß 2ß 3H-testosterone (Dupont de Nemours, Les Ulis, France). At the end of culture, medium was removed and stored at –20°C. Cell numbers in each well were measured (see below). To estimate aromatase activity, 100 µl of culture medium were loaded on C18-Sep Pak cartridges (Waters, Milford, MA, USA) and sequentially eluted with 3 ml of H2O and 3 ml of methanol to separate 3H2O from [1ß, 2ß] 3H-testosterone. Radioactivity in each fraction was counted in a scintillation counter (Packard Research, Rungis, France). Blanks were included in all cultures and radioactivity in blanks was deduced from the 3H2O counts. Aromatase activity was estimated as dpm 3H2O/(dpm 3H2O + [1ß, 2ß] 3H-testosterone) per 5x104 cells at the end of culture.

In order to determine cell numbers at the end of culture, cells were fixed by addition of 100 µl of 10% glutaraldehyde diluted in the culture medium and numbers then assessed according to Kueng et al. (1989). Briefly, following a 20 min fixation, the plates were washed with deionized water, dried and stained by addition of 100 µl of a 0.1% solution of crystal violet dissolved in 200 mmol/l 2[N-Morpholino]ethanesulphonic acid (Mes; Sigma-Aldrich Company), pH 6.0. The plates were again stirred for 20 min at room temperature, excess dye was removed by extensive washing with deionized water and plates were again dried. After crystal violet solubilization with 100 µl of 10% acetic acid in water, optical density (OD) was measured with a plate reader (Titertek Multiscan MKII®; Labsystems, Les Ulis, France) at a wavelength of 600 nm.

Chromatographic analysis of the samples
Radioactive steroids contained in the incubation medium were purified and concentrated on 1 g C18-Supelclean® LC-18 SPE cartridges (Supelco, St Germain-en-Laye, France) in order to discard 3H2O and specifically to elute retained steroids with methanol acidified with 1% acetic acid.

The purified radioactive steroids were analysed by high performance liquid chromatography (HPLC) on a Philips 4100 apparatus equipped with a rheodyne Model 7125 injector and connected to a Philips 4021 diode array detector (both from Philips, Eindhoven, Belgium). Testosterone metabolites were separated by HPLC on a reverse-phase ODS 2 column (180x4.6 mm, 3 µm) (Interchim, Montlucion, France) by means of a methodology derived from that of Wood et al. (Wood et al., 1983Go). Radioactive samples were collected on a Gilson-202® fraction collector (Gilson, Villiers de Bel, France) and were measured on a Packard 2200 Spectrophotometer® (Packard Research). Identification of the different steroidal products resulting from the metabolization (hydroxylation, aromatization or 17-oxidoreduction) was achieved by comparison with authentic standards.

In contrast, however, to what had been done previously with sheep follicular walls (Driancourt et al., 1996Go), attempts to purify sufficient amounts of hydroxylated and/or aromatized metabolites were unsuccessful.

Validation of the aromatase assay
Several specific points were tested to set up adequate culture conditions and validate the test: (i) whether 3H2O accumulation was linear at specific times following addition of [1ß, 2ß] 3H testosterone in the medium was assessed by obtaining samples at 3–12 h, following a 24 h plating period; (ii) whether aromatase activity was steady during the duration of culture was assessed by measurement of aromatase activity ([1ß, 2ß] 3H-testosterone added during 12 h), after 0–72 h following a 24 h plating period; (iii) to choose optimum cell numbers per well and optimum [1ß, 2ß] 3H-testosterone concentrations per well, the effects of cell numbers (6500–100 000 cells per well) on aromatase activity were also checked, as were the effects of altering the amounts of substrate provided (10–9 mol/l to 10–6 mol/l). For this purpose, the duration of culture was 48 h followed by a 12 h incubation with 3H-testosterone.

Kinetic characteristics
Following a 24 h plating period, aromatase kinetics were measured by adding a range of [1ß, 2ß] 3H-testosterone concentrations (10–6 mol/l, 5x10–7 mol/l, 10–7 mol/l, 10–8 mol/l, or 10–9 mol/l) to wells containing 50 000 cells and obtaining samples of culture medium at 3 h intervals for 12 h. The amount of 3H2O and cell numbers were determined at each stage. Km and Vmax were calculated as described by Michaelis (1922). This experiment was replicated twice (pool 1 and pool 2).

Modulation of aromatase by gonadotrophins and follicular fluids
Study of the effects of gonadotrophins on aromatase activity
Granulosa cells were obtained from four women, pooled and seeded in triplicate at 50 000 cells per well. After a 24 h plating period, medium was aspirated and replaced with new medium containing 0, 12.5, 25, 50 ng/ml of human FSH (Metrodine HP®; Serono, France) or similar amounts of human LH (provided by Dr Y.Combarnous) or a combination of both hormones. After 48 h of treatment, 3H-testosterone was added for 12 h. Aromatase activity for each treatment was then determined and expressed according to the cell number found per well at the end of culture. This experiment was replicated twice.

Characterization of follicular fluids
Human follicular fluid was collected by laparoscopy, around the time a dominant follicle may have been present (according to the time lag from the previous menses), from the largest follicle in 10 regularly menstruating women who had tubal or unexplained infertility problems. Oestradiol and {Delta}4 androstenedione concentrations were measured in these follicular fluids by radioimmunoassays (125I-Estradiol Coatria®; 3H {Delta}4 androstenedione radioimmunoassay kit®; Biomerieux SA, Marcilly l'Etoile, France, Table IGo). At the same time, blood samples were collected. Oestradiol, progesterone, and LH concentrations were measured in these samples by radioimmunoassays (125I Estradiol Coatria®; 125I Progesterone Coatria®; 125I human LH Coatria®; Biomerieux SA, Table IGo). The stage of the cycle when the follicular fluids were obtained was established from 17ß-oestradiol, progesterone and LH concentrations in blood samples and from 17ß-oestradiol, {Delta}4 androstenedione concentrations in follicular fluids. Only dominant follicular fluid was used for the cultures. This was identified as follows: serum oestradiol concentrations were in excess of 500–1000 pmol/l, those of progesterone were lower than 0.5 nmol/l, and those of LH lower than 180% of basal concentration, while the ratio between {Delta}4 androstenedione and oestradiol concentrations in follicular fluid was <1 (Gougeon, 1996Go). Aliquots of 200 µl of follicular fluid of the largest follicle were prepared for each sample and stored at –20°C, after blood cells had been removed by centrifugation (400 g, 20 min).


View this table:
[in this window]
[in a new window]
 
Table I. Intra- and inter-assay coefficients of variation of radioimmunoassays used, 125I Estradiol Coatria®, 3H{Delta}4 androstenedione radioimmunoassay kit®, 125I Progesterone Coatria®, 125I human luteinizing hormone (LH) Coatria® (Biomerieux SA, Marcilly l'Etoile, France)
 
Effects of follicular fluids on aromatase activity
To compare the effects of dominant follicular fluid with those of its control serum, granulosa cells were obtained from three or four women. Different concentrations of both fluids (0, 2.5, 5, 10, 20%) were added to the culture media. The effect of each treatment was assessed on triplicate wells with 50 000 cells seeded per well. After plating, culture medium containing the above concentrations of serum or follicular fluid was added for 48 h. At this stage, [1ß, 2ß] 3H-testosterone (5x10–7 mmol/l) was added for 12 h. At the end of culture, the amounts of 3H2O and cell number for each well were assessed. This experiment was replicated six times since six samples of dominant follicular fluid were available (see Results).

Specificity of follicular fluid effects
To ascertain whether follicular fluid effects detected in the previous studies were specific of aromatase, follicular fluid effects on two other steroidogenic enzymes were analysed. All reagents used were obtained from BioMerieux SA. Granulosa cells cultured for 48 h in the 10% serum or 10% dominant follicular fluid were immediately frozen in lysis medium (10 mmol/l KCl/10 mmol/l Tris/0.5 mmol/l EDTA), to which protease inhibitors had been added (1 µmol/l PMSF (phenyl-methyl-sulphonyl-fluoride), 100 µmol/l TPCK (N tosyl-L-phenylalanine-chloromethyl-ketone) and 10 µmol/l TLCK (N{alpha}-P-tosyl-L-lysine-chloromethyl-ketone). Equal amounts of granulosa cell proteins (assessed according to Bradford, 1976Go) were used for Western blot analysis. The samples were heated for 1 min with electrophoresis sample buffer [1.25 mol/l Tris containing 15% (v/v) glycerol, 10% (w/v) sodium dodecyl sulphate (SDS) pH6.8] at 90°C and subjected to electrophoresis using 15x13 cm polyacrylamide SDS gels (4% stacking gels and 10% separating gels). After electrophoresis, the proteins were transferred to a nitrocellulose membrane (Biorad, Ivry sur Seine, France) in a 1.2% (w/v) glycine/0.25% (w/v) Tris/20% (v/v) methanol/H2O buffer by incubation overnight at 4°C. After completion of transfer, the quality of each blot was checked by staining the nitrocellulose with 0.2% (w/v) Ponceau S stain (Sigma, France).

The membrane was then washed with Tris buffered saline (TBS: 10 mmol/l Tris containing 0.1% Tween 20; pH 7.4). The blot was then immersed in blocking buffer (1 mol/l Tris with 5% milk powder/0.2% Nonidet P-40; pH7.4) for 1 h. The membrane was then incubated with primary antibodies in blocking buffer for 1.5 h at 37°C and subsequently washed with TBS. The primary antibodies were polyclonal rabbit sera anti-P450scc (a gift from Dr B.Hales, Chicago University, USA) and anti-3 ßHSD (donated by Professor I.J.Mason, University of Edinburgh, UK) at a 1:1000 dilution. This was followed by incubation with a 1:1000 dilution of peroxidase-labelled goat anti-rabbit antibody (Pasteur, Sanofi, France). Bands corresponding to the steroidogenic enzymes were visualized using chemiluminescence (Amersham, Courtaboeuf, France) and were quantified by densitometric analysis (Kepler Software, Large Scale Biology, Rockville, TN, USA) as previously described (Driancourt et al., 1998).

Analysis of data
The data are expressed as mean ± SEM. Regression analysis was used to validate the aromatase test (effects of duration of culture, of 3H-testosterone concentrations, of cell number). One-way analysis of variance (ANOVA) and two-way ANOVA were used to compare effects of follicular fluids and of gonadotrophins on aromatase activity and on cell proliferation. Quantitative differences in the darkness of the bands found on the Western blots were identified by t-test.


    Results
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Chromatographic analysis of samples
As shown in Figure 1Go, the HPLC profile of the incubation samples revealed a major peak eluting with a retention time close to that of oestrone or 17ß-oestradiol (51.6 min and 52.7 min respectively). Furthermore, some minor metabolites could be identified according to their retention time, such as 6ß-hydroxytestosterone (16.5 min), 19-hydroxytestosterone (17.4 min) and 2{alpha}-hydroxytestosterone (34.8 min). This demonstrated that 3H2O production was associated with 17ß-oestradiol production.



View larger version (17K):
[in this window]
[in a new window]
 
Figure 1. Separation by high performance liquid chromatography of the labelled metabolites of [1ß, 2ß] 3H-testosterone present in culture medium. 6ß-OH-T = 6ß-hydroxytestosterone; 19-OH-T = 19-hydroxytestosterone; 2{alpha}-OH-T = 2{alpha}-hydroxytestosterone; T = testosterone; E1 = oestrone; E2ß = oestradiol-17ß.

 
Validation of the aromatase assay
Effects on tritiated water accumulation of the time lag between [1ß, 2ß] 3H-testosterone addition and collection of the samples were evaluated to establish the optimum incubation time with labelled testosterone (Figure 2Go). Regression analysis showed a linear increase in the accumulation of 3H2O over the period 0–12 h (n = 12, r2 = 0.612, P < 0.01). Hence an incubation with [1ß, 2ß] 3H-testosterone of 12 h was selected for further use.



View larger version (11K):
[in this window]
[in a new window]
 
Figure 2. Effect of incubation time (3 h, 6 h, 9 h, 12 h) on the metabolism of 3H-testosterone to 3H2O (expressed by tritiated water and all radioactivity ratio per 50 000 cells). Each mean was estimated from triplicate samples.

 
Aromatase activity (expressed as a percentage of 0 h culture activity, following a 24 h plating period) for specific durations of culture was steady from 0–48 h after a 24 h plating period. In contrast, at 72 h, aromatase activity was reduced to 55% of the activity measured at time 0. Hence, in all further experiments, incubation time was limited to 48 h.

The changes in aromatase activity according to the number of cells seeded showed a linear increase in aromatase activity (expressed per well), when cell numbers increased from 6500 to 100 000 cells (n = 21, r2 = 0.81, P < 0.01) (Figure 3Go). Hence, to allow space for the cells to proliferate during culture, 50 000 cells were seeded per well in all further experiments.



View larger version (12K):
[in this window]
[in a new window]
 
Figure 3. Effect of cell number (6500, 12 500, 25 000, 50 000, 75 000, 100 000) on aromatase activity (expressed per 50 000 cells). Each mean was estimated from triplicate samples.

 
Amounts of tritiated water produced by granulosa cells were related by logarithmic regression (n = 12, r2 = 0.506, P < 0.01) to the amounts of precursor provided (Figure 4Go). For concentrations over 5x10–7 mol/l, no further increase in tritiated water was obvious. Hence an intermediary concentration of [1ß, 2ß] 3H-testosterone (5x10–7 mol/l) was used in all further experiments.



View larger version (12K):
[in this window]
[in a new window]
 
Figure 4. Effect of tritiated testosterone concentration (10–6 mol/l, 5 x 10–7 mol/l, 10–7 mol/l, 10–8 mol/l) on tritiated water release (expressed per 50 000 cells). Each mean was estimated from triplicate samples.

 
Kinetic characteristics
The effect of altering the amount of substrate on the amount of 3H2O generated per 50 000 granulosa cells is shown in Figure 5Go. Km ranged from 168 nmol to 57.8 nmol (pool 1 and pool 2 respectively) and Vmax ranged from 3.08 nmol/h to 0.67 nmol/h (pool 1 and pool 2 respectively).



View larger version (16K):
[in this window]
[in a new window]
 
Figure 5. Effect of [1ß, 2ß] 3H-testosterone concentration (10–6 mol/l: {circ}, 5 x 10–7 mol/l: {triangleup}, 10–7 mol/l: {square}, 10–8 mol/l: +) on the amount of 3H2O generated by granulosa cells present in pool 1.

 
Modulation of aromatase by gonadotrophins and follicular fluids
Effects of gonadotrophins on aromatase activity
The effect of different concentrations of FSH and LH on aromatase activity is shown in Figure 6Go. A significant effect of treatment (P < 0.03), a significant effect of the hormone concentration used (P < 0.01), and a significant interaction between treatment and concentrations (P < 0.01) were detected. At 10 ng/ml, aromatase activity was maximally stimulated by the combined presence of FSH + LH (P < 0.01). In contrast, LH and FSH alone, although exhibiting stimulatory effects (P < 0.01 and P = 0.03) were less effective than FSH + LH. Interestingly, raising the hormone concentrations to 25 ng/ml did not increase aromatase to further activity. It even decreased aromatase in the LH stimulated wells (P < 0.05). In the LH treated group, increasing concentrations over 40 ng/ml never managed to increase aromatase again. The same conclusion also stood for the combined administration of FSH + LH.



View larger version (32K):
[in this window]
[in a new window]
 
Figure 6. Effects of graded doses of follicle stimulating hormone (FSH: {blacksquare}), luteinizing hormone (LH: {blacksquare}) and FSH + LH ({blacksquare}) on aromatase activity (measured by the percentage of label converted to 3H2O per 50 000 cells) (a versus b, 2 versus 3, 1 versus 3: P < 0.01; x versus y, 1 versus 2: P < 0.05). Four replicates were used to generate this graph.

 
Effects of follicular fluids on aromatase activity
Six out of the 10 follicular fluids obtained were judged to be aspirated at the dominant stage as previously described. Effects of different concentrations of each of these follicular fluids on aromatase activity are presented in Figure 7Go. Significant effects of treatment (serum versus follicular fluids, P < 0.01) and concentration of these substances (P < 0.01) were detected, together with a significant interaction between concentrations and treatments (P < 0.01). A doubling of aromatase activity (P < 0.01) was induced by the addition of 2.5% serum. These high levels were maintained when higher concentrations were used. In contrast, 2.5, 5 and 10% follicular fluid did not affect aromatase while higher amounts (20%) significantly depressed aromatase activity (P < 0.01). As a consequence, aromatase activity in the presence of 20% of follicular fluid was 45% of the control values while it was three-fold lower than in the presence of serum.



View larger version (16K):
[in this window]
[in a new window]
 
Figure 7. Effects of follicular fluid and its control serum on aromatase activity (measured by the percentage of label converted to 3H2O per 50 000 cells). Follicular fluid {square}; serum {blacksquare}. Significance of different concentrations is indicated with different letters, within each treatment (a versus c: P < 0.05; x versus y: P < 0.01). Six replicates, each using follicular fluid and serum of a specific patient contributed to this graph.

 
Effects on cell number at the end of culture in the presence of serum versus follicular fluid at concentrations ranging from 0–20% are shown in Figure 8Go. No significant difference between serum and follicular fluid effects was found, but a significant effect of the concentration of these substances (P < 0.01) together with a significant interaction between concentration and treatment (P = 0.04) were found. Both serum and dominant follicular fluid increased cell number at all concentrations tested compared to control values.



View larger version (14K):
[in this window]
[in a new window]
 
Figure 8. Effects of follicular fluid and its control serum on cell number at the end of culture. Follicular fluid {square}; serum {blacksquare}. Significance of different concentrations is indicated on the figure with different letters within each treatment (a versus b, x versus y: P < 0.05). Differences between treatments (follicular fluid versus serum) did not reach statistical significance. Six replicates, each using follicular fluid and serum of a specific patient contributed to this graph.

 
Steroidogenic enzymes
Following 1D polyacrylamide gel electrophoresis (PAGE) and Western blotting, P450scc and 3ß-HSD could be visualised in the samples of cellular proteins at 50 and 48 kDa, respectively. A representative blot, probed with an antibody raised against P450scc is shown in Figure 9Go. Densitometric analysis of the bands obtained after Western blotting failed to detect any difference between granulosa cells incubated with serum versus follicular fluid on the amounts of P450scc (59 900 ± 10 500 versus 85 000 ± 24 000) or 3ß-HSD (54 100 ± 13 600 versus 45 700 ± 17 000) (arbitrary units, for serum versus dominant follicular fluid respectively).



View larger version (45K):
[in this window]
[in a new window]
 
Figure 9. Amounts of P450scc present in granulosa cells cultured in the presence of 10% serum (S) (five left-hand lanes) or 10% follicular fluid (F) (five right-hand lanes). Molecular weights (MW, kDa) are indicated on the left.

 

    Discussion
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
The main finding of this study is the demonstration that human follicular fluid obtained from a dominant follicle (as opposed to a non-dominant follicle – see below) contained a factor or factors which inhibited aromatase activity of these cells and/or stimulate granulosa cell proliferation.

This conclusion is based on a bioassay which was developed and validated in the present study. Human follicular fluids and serum as well as granulosa cells from IVF patients were used to ensure the validation of this homospecific test. These cells, despite their earlier exposure to high concentrations of exogenous gonadotrophins, retain an active aromatase for a few days at least (Erickson et al., 1989Go; Wood et al., 1994Go; Földesi et al., 1998Go). Our bioassay used aromatase activity as an end point since aromatase is a key enzyme for granulosa cells which increases with follicle size and vanishes when follicles enter atresia (Driancourt et al., 1991Go). Aromatase was measured using the tritiated water release assay (Gore Langton et al., 1981) and validated for other species (cattle, sheep) (Thatcher et al., 1991Go; Driancourt et al., 1996Go respectively). One of the pretrials of this study established that, in humans as well, 3H2O formation was indeed associated with formation of 17ß-Estradiol. All parameters of the culture system were adjusted following several pretrials to ensure optimal culture conditions. The conditions used were 50 000 cells plated for 24 h, an incubation of 48 h with test substances, and a concentration of 5x10–7 mol/l of testosterone. Culture duration was never longer than 84 h (24 h plating + 48 h of action of the test substances before addition of 3H-testosterone for 12 h) because aromatase tended to decrease for longer cultures. While this is in agreement with the data reported (Wood et al., 1994Go), it contrasts with the data reported by Bernhisel et al. and Földesi et al. who demonstrated maintenance of aromatase for longer culture periods (Bernhisel et al., 1987Go; Földesi et al., 1998Go). However, serum-free conditions were used for plating in this study, while fetal bovine serum was used at the corresponding time in the above studies. Evidence that the granulosa cells remain functional in the conditions used in this study was provided (i) by the increase in cell number with time, (ii) by an aromatase activity well over background activity and (iii) by the sustained ability of gonadotrophins to stimulate aromatase activity. The magnitude of the increases in aromatase activity observed in the presence of LH and FSH were lower than the values reported using cells from unstimulated follicles (Mason et al., 1993Go). Granulosa cells from IVF patients which have been exposed to large gonadotrophin concentrations in vivo may be partly refractory to additional gonadotrophins in vitro and appear to need a longer duration of culture (7 days) to recover a full ability to respond to FSH by an increased aromatase activity (Földesi et al., 1998Go).

The approach selected in the present study used follicular fluid, a complex protein and steroid source, to check its effect on aromatase or cell proliferation. Numerous compounds present in follicular fluid (insulin-like growth factor I, activin, epidermal growth factor, interleukin-6 and leukaemia inhibitory factor) (Gougeon, 1996Go) have the ability positively or negatively to alter aromatase or cell proliferation. However, because compounds with positive and negative effects are mixed in follicular fluid, such studies do not provide clues to the net biological effect of follicular fluid on the cells of the follicle which contain it. In the present study, the dominant follicular fluid used was carefully selected to ensure it was aspirated while the follicle was in its dominance phase. Hence, the effect of follicular fluid from large (dominant) follicles was tested on granulosa cells from large (stimulated) follicles. This design was used because the question addressed dealt with the intrafollicular control of the maturation of the dominant follicle. In addition, earlier studies in swine (Ledwitz Rigby et al., 1985Go; Ledwitz Rigby, 1987Go) have underlined the complexity of interactions existing when follicular fluid obtained in a specific size group is added to granulosa cells originating from another size group.

The net effect of addition of dominant follicular fluid to human granulosa lutein cells was an inhibition of aromatase activity. Three lines of evidence support this claim: (i) aromatase in the presence of 20% dominant follicular fluid was significantly reduced compared to that observed in the presence of lower concentrations; (ii) aromatase in the presence of 20% dominant follicular fluid was three-fold lower than in the presence of 20% serum; (iii) contrasting with the effects observed with dominant follicular fluid, no change was observed in aromatase activity using non-dominant follicular fluid (Guet et al., 1997Go). This finding is in agreement with the earlier reports using heterospecific tests (di Zerega et al., 1982Go, 1983aGo,di Zerega et al., bGo). The magnitude of the inhibition observed in the present study is close to what was reported previously (di Zerega et al., 1982Go, 1983aGo). A noticeable difference between the two approaches is that, in this study, basal conditions were used while in most studies of di Zerega's group, HMG or FSH stimulated conditions were used (di Zerega et al., 1982Go, 1983bGo). However, in one of their earlier studies, di Zerega et al. (1983b) demonstrated that the inhibition of aromatase induced by follicular fluid could also be observed when basal conditions were used.

Specificity of the effect on aromatase has been demonstrated in the present study by showing that other steroidogenic enzymes (namely P450scc and 3ß-HSD) were not affected by culture in the presence of follicular fluid compared to serum. In addition, dominant follicular fluids from sheep have no effect on progesterone production by luteal cells (M.A.Driancourt, unpublished data). Hence the effect of dominant follicular fluids on aromatase is likely to be specific.

As a decrease in aromatase activity at the higher concentration of follicular fluid used was observed, cell counts at the end of the culture were assessed to check whether this could be related to a toxic effect. Surprisingly, at all concentrations tested an increase in cell numbers both in the presence of follicular fluid and serum was observed. Such a finding strongly argues against a toxic effect.

The finding of a stimulatory action of dominant follicular fluid on granulosa cell proliferation is novel. Small amounts (2.5 and 5%) of dominant follicular fluid were very potent (six- to eight-fold increase in cell numbers) in stimulating cell proliferation and about two times more potent than similar concentrations of serum. Interestingly, concentrations in excess of 5% were not more efficient than 2.5 and 5%. However human follicular fluid is known to have the ability to stimulate proliferation of other cells (endothelial ones) (Bryant et al., 1988Go).

Because the estimates of aromatase were expressed per 50 000 cells, it may be argued that part of the changes in the index of aromatase activity may be due to increased cell proliferation. This is unlikely to be the case since (i) cell proliferation was stimulated by all concentrations of dominant follicular fluid tested while aromatase activity was only reduced at 20% dominant follicular fluid, (ii) at 20% serum or dominant follicular fluid, proliferation was similar while aromatase activity was reduced by dominant follicular fluid and (iii) non dominant follicular fluid has no ability to decrease aromatase activity (Guet et al., 1997Go) when added at a concentration of 20%. Whether the positive effect on cell proliferation is produced by the same compound or compounds as those negatively affecting aromatase activity is still unknown. Such a possibility might be unlikely since proliferation and differentiation are generally assumed to be regulated by distinct mechanisms (Monniaux et al., 1994Go). In regard to the results of this study, endothelin-1 might be a suitable candidate for sustaining both inhibitory activity on aromatase and stimulating activity on proliferation of human luteinized granulosa cells (Kamada et al., 1995Go). However, such a hypothesis looks unlikely since preliminary steps of characterization found that both activities were present after dialysis with a cut off of 5 kDa, which allowed low molecular weight compounds like endothelin to be excluded. Further purification of these compounds which might be important regulators of dominant follicle function will be the next step to understand their precise physiological role.


    Acknowledgments
 
The help of the reproductive biology unit technicians and the financial help of Serono are gratefully acknowledged. We are indebted to Dr Combarnous for providing human LH for this study and to Dr Guilloteau for hormone assays in venous serum and in follicular fluids.


    Notes
 
4 To whom correspondence should be addressed Back


    References
 Top
 Abstract
 Introduction
 Materials and methods
 Results
 Discussion
 References
 
Armstrong, D.T., Goff, A.K. and Dorrington, J.H. (1979) Regulation of follicular estrogen biosynthesis. In Midgley, A.R. Jr and Sadler, W.A. Jr (eds) Ovarian Follicular Development and Function. Raven Press, New York, pp. 169–182.

Bernhisel, M.A., Holman, J.F., Haney, A.F. and Schomberg, D.W. (1987) Estrogen and progesterone production by granulosa cell monolayers derived from in vitro fertilization procedures: lack of evidence for modulation by androgen. J. Clin. Endocrinol. Metab., 64, 1251–1256.[Abstract]

Bradford, M.M. (1976) A rapid sensitive method for the quantitative of microgram quantities of protein utilising the principle of protein-dye binding. Ann. Biochem., 72, 248–254.

Bryant, S.M., Gale, J.A., Yanagihara, D.L. et al. (1988) Angiogenic, mitogenic, and chemotactic activity of human follicular fluid. Am. J. Obstet. Gynecol., 158, 1207–1214.[ISI][Medline]

Carson, R.S., Robertson, D.M. and Findlay, J.K. (1988) Ovine follicular fluid inhibits thymidine incorporation by 3T3 fibroblasts in vitro. J. Reprod. Fert., 82, 447–455.[Abstract]

di Zerega, G.S., Marrs, R.P., Roche, C.R. et al. (1982) Identification of proteins in pooled human follicular fluid which suppress follicular response to gonadotropins. J. Clin. Endocrinol. Metab., 56, 35–41.[Abstract]

di Zerega, G.S., Campeau, J.D., Nakamura, R.N. et al. (1983a) Activity of a human follicular fluid protein(s) in spontaneous and induced ovarian cycles. J. Clin. Endocrinol. Metab., 57, 838–846.[Abstract]

di Zerega, G.S., Marrs, R.P., Campeau, J.D. and Kling, O.R. (1983b) Human granulosa cell secretion of protein(s) which suppress follicular response to gonadotropins. J. Clin. Endocrinol. Metab., 56, 147–155.[ISI][Medline]

Driancourt, M.A., Gougeon, A., Royère, D. and Thibault, C. (1991) La fonction ovarienne. In Thibault, C. and Levasseur, M.C. (eds) La reproduction chez les mammifères et chez l'homme. INRA Ellipse, Paris, pp. 273–298.

Driancourt, M.A., Paris, A. and Debrauwer, L. (1996) Aromatase activity of follicular walls: technical validation and physiological control. Reprod. Fertil. Dev., 8, 875–884.[ISI][Medline]

Driancourt, M.A., Guet, P., Reynaud, K. et al. (1999) Presence of an aromatase inhibtor, possibility heat shock protein, 90, in dominant follicles of cattle. J. Reprod. Fert., 115, 45–58.[Abstract]

Erickson, G.F. (1993) Normal regulation of ovarian androgen production. Semin. Reprod. Endocrinol., 11, 307–312.[ISI]

Erickson, G.F., Garzo, V.G. and Magoffin, D.A. (1989) Insulin-like growth factor-I regulates aromatase activity in human granulosa and granulosa luteal cells. J. Clin. Endocrinol. Metab., 69, 716–724.[Abstract]

Földesi, I., Breckwoldt, M. and Neulen, J. (1998) Oestradiol production by luteinized human granulosa cells: evidence of the stimulatory action of recombinant human follicle stimulating hormone. Hum. Reprod., 13, 1455–1460.[Abstract]

Gore Langton, R.E. and Dorrington, J.H. (1981) FSH induction of aromatase in cultured rat granulosa cells measured by a radiometric assay. Mol. Cell. Endocrinol., 22, 135–151.[ISI][Medline]

Gougeon, A. (1996) Regulation of ovarian follicular development in primates: facts and hypotheses. Endocr. Rev., 17, 121–155.[ISI][Medline]

Guet, P., Driancourt, M.A., Paris, A. et al. (1997) Inhibitory effect of dominant follicular fluid on human granulosa cells: preliminary steps of characterization. J. Reprod. Fertil., 20, 36 (abst).

Hillier, S.G., Whitelaw, P.F. and Smyth, C.D. (1994) Follicular oestrogen synthesis: the `two-cell, two-gonadotropin' model revisited. Mol. Cell. Endocrinol., 100, 51–54.[ISI][Medline]

Hynes, A.C., Kane, M.T. and Sreenan, J.M. (1996) Partial purification from bovine follicular fluid of a factor of low molecular mass with inhibitory effects on the proliferation of bovine granulosa cells in vitro and on rat follicular development in vivo. J. Reprod. Fertil., 108, 185–191.[Abstract]

Kamada, S., Blackmore, P.F., Kubota, T. et al. (1995) The role of endothelin-1 in regulating human granulosa cell proliferation and steroidogenesis in vitro. J. Clin. Endocrinol. Metab., 80, 3708–3714.[Abstract]

Karnitis, V.J., Townson, D.H., Friedman, C.I. and Danforth, D.R. (1994) Recombinant human follicle-stimulating hormone stimulates multiple follicular growth, but minimal estrogen production in gonadotropin-releasing hormone antagonist-treated monkeys: examining the role of luteinizing hormone in follicular development and steroidogenesis. J. Clin. Endocrinol. Metab., 79, 91–97.[Abstract]

Kueng, W., Silber, E. and Eppenberger, U. (1989) Quantification of cell cultures on 96-well plates. Anal. Biochem., 182, 16–19.[ISI][Medline]

Ledwitz Rigby, F. (1983) Follicular fluid stimulation of estrogen secretion by immature porcine granulosa cells. Mol. Cell. Endocrinol., 29, 213–222.[ISI][Medline]

Ledwitz-Rigby, F. (1987) Local regulation of granulosa cell maturation. J. Steroid Biochem., 27, 385–391.[ISI][Medline]

Ledwitz Rigby, F., Petito, S.H., Tyner, J.K. and Rigby, B.W. (1985) Follicular fluid effects on progesterone secretion are not due to follicle stimulating hormone or steroids. Biol. Reprod., 33, 277–285.[Abstract]

Mason, H.D., Mannaerts, B., de Leeuw, R. et al. (1993) Effect of recombinant human follicle stimulating hormone on cultured human granulosa cells: comparison with urinary gonadotrophins and actions in preovulatory follicles. Hum. Reprod., 8, 1823–1827.[Abstract]

Michaelis, L. (1922) Die Wasserstoffionenkonzentration. Springer Verlag, Berlin.

Monniaux, D., Pisselet, C. and Fontaine, J. (1994) Uncoupling between proliferation and differentiation of ovine granulosa cells in vitro. J. Endocrinol., 142, 497–510.[Abstract]

Smyth, C.D., Miro, F., Howles, C.M. and Hillier, S.G. (1995) Effect of luteinizing hormone on follicle-stimulating hormone-activated paracrine signalling in rat ovary. Hum. Reprod., 1, 33–39.[Abstract]

Thatcher, W.W., Driancourt, M.A., Terqui, M. and Badinga, L. (1991) Dynamics of ovarian follicular development in cattle following hysterectomy and during early pregnancy. Dom. Anim. Endocr., 8, 223–234.

Wood, A.W., Ryan, D.E., Thomas, P.E. and Lewin, W. (1983) Regio- and stereoselective metabolism of two C19 steroids by five highly purified and reconstituted rat hepatic cytochrome P-450 isozymes. J. Biol. Chem., 258, 8839–8847.[Abstract/Free Full Text]

Wood, A.M., Lambert, A., Hooper, M.A.K. et al. (1994) Exogenous steroids and the control of oestradiol secretion by human granulosa-lutein cells by follicle stimulating hormone and insulin-like growth factor-I. Hum. Reprod., 9, 19–23.[Abstract]

Submitted on September 21, 1998; accepted on January 29, 1999.





This Article
Abstract
FREE Full Text (PDF )
Alert me when this article is cited
Alert me if a correction is posted
Services
Email this article to a friend
Similar articles in this journal
Similar articles in ISI Web of Science
Similar articles in PubMed
Alert me to new issues of the journal
Add to My Personal Archive
Download to citation manager
Search for citing articles in:
ISI Web of Science (1)
Request Permissions
Google Scholar
Articles by Guet, P.
Articles by Driancourt, M.A.
PubMed
PubMed Citation
Articles by Guet, P.
Articles by Driancourt, M.A.