Molecular and Cellular Endocrinologv, 72 (1990) Elsevier Scientific Publishers Ireland, Ltd.
MOLCEL
111- 120
111
02321
Cholesterol
side-chain
V. Rouiller
cleavage activity in rat fetal gonads: a limiting step for ovarian steroidogenesis
‘, M.N. Gangnerau
‘, J.L. Vayssiere
2 and R. Picon ’
’ Laboratoire de Physiologic du DCveloppement, CNRS, URA 307, Universitk Paris 7, 75251 Paris Cedex 0.5, France, and 2 Laboratoire de Biochimie CeNulaire, CoilGgede France, 75231 Paris Cedex 05, France (Received 9 April 1990; accepted 23 May 1990)
Key words: Fetus; Ovary; Testis; Pregnenolone; Cytochrome P-450,;
(Rat)
Summary The aim of this study was to examine the first step in steroidogenesis in male and female gonads of fetal rats. Pregnenolone production was measured by radioimmunoassay in organ culture, conversion of [ 3H]cholesterol to [ 3H]pregnenolone was evaluated in isolated mitochondria and cytochrome P-450,, was revealed by immunoblotting and immunocytochemical techniques. Our results clearly showed that in fetal testes (1) pregnenolone was produced in media where testes were cultured in the presence of trilostane and spironolactone, indicating an important metabolism of pregnenolone, (2) [ 3H]cholesterol was converted into [ 3H]pregnenolone in mitochondria, (3) cytochrome P-450,,, was revealed in immunoblots with a molecular weight of 50,000, (4) cytochrome P-450,, was localized in Leydig cells from 15.5-day-old fetal testes onwards. With respect to fetal ovaries, we were unable to detect any see activity, except after treatment with dibutyryl cyclic AMP. A lag period of 18 h was necessary to induce pregnenolone synthesis. However, the immunoperoxidase staining did not localize ovarian positive cells. Cytochrome P-450,,, could be revealed in postnatal ovaries by immunoblotting and some interstitial positive cells were observed with immunostaining; the reaction was enhanced in luteinizing hormone-pretreated ovaries. These data indicate that (a) the cholesterol see activity is present in fetal testes, (b) the conversion of cholesterol to pregnenolone is a limiting step for steroidogenesis in fetal ovaries. The inductive effect of the nucleotide on the enzyme suggests that the absence of gonadotrophic receptors in fetal female gonads could explain the lack of steroidogenesis before birth.
Introduction In rats, production of steroids by male gonads has been demonstrated during fetal life from 15 days onwards (Warren et al., 1973; Picon, 1976)
Address for correspondence: R. Picon, Laboratoire de Physiologic du DCveloppement, CNRS, URA 307, UniversitC Paris 7, 75251 Paris Cedex 05, France. 0303-7207/~/$03.50
whereas ovarian steroidogenesis develops only after birth (Smeaton et al., 1975; Lamprecht et al., 1976; Funkenstein et al., 1980). However, some enzymes involved in steroidogenesis have been revealed to be active in the female gonads during fetal life. Thus aromatization of testosterone to estrogen could be observed in fetal ovaries as early as 14.5 days and this biochemical property allows to differentiate ovaries from testes at this stage (Picon et al., 1985). 3P-Hydroxysteroid dehydro-
Q 1990 Elsevier Scientific Publishers Ireland, Ltd.
112
genase was also detected in fetal female gonads (Picon et al., 1988). In adult tissues, the side-chain cleavage (see) of cholesterol to pregnenolone is considered to be the rate-limiting step in the synthesis of steroid hormones (Hall and Koritz, 1965; Hall, 1966; Anderson and Mendelson, 1985). Mediated by a single mitochondrial cytochrome P-450, this step is known to be an important site of regulation by trophic hormones (Waterman and Simpson, 1985). Luteinizing hormone (LH) receptors are present in testes from 15-day-old fetuses onwards (Gangnerau et al., 1982) whereas gonadotrophic receptors cannot be revealed in fetal ovaries (Siebers et al., 1977; Smith-White and Ojeda, 1981). Stimulation of estradiol-17P release from rat ovaries with LH was observed only from the beginning of week 2 after birth (Lamprecht et al., 1976). The aim of this study was to investigate the first step of steroidogenesis to give some insight into the mechanism controlling pregnenolone production in male and female gonads during fetal life. We therefore assessed the production of pregnenolone during culture of the fetal gonads; measured the conversion of radiolabelled cholesterol to pregnenolone in isolated mitochondria and examined the effect of dibutyryl cyclic AMP ((Bu),cAMP) on see activity in fetal ovaries. The presence of cytochrome P-450,,, was investigated with techniques of immunoblotting and immunohistochemistry. Materials and methods Animals Fetuses and newborn of the Wistar strain were used. Gestational age was based on the timing of ovulation as previously described (Gangnerau et al., 1982). Parturition occurred during the daytime of day 22 of gestation. Materials [4,7-3H]pregnenolone (spec. act. 19 Ci/mmol) and [1,2( n)- 3H]cholesterol (spec. act. 40 Ci/mmol) were obtained from the Radiochemical Centre (Amersham, U.K.). [4-‘4C]Pregnenolone (spec. act. 57 mCi/mmol) was purchased form NEN Research Products (Boston, MA, U.S.A.). Unlabelled steroids, (Bu),cAMP, spironolactone (3-[3-0x0-
7~-acetylthio-l7~-hydroxy-4-androstene-l7~-yl]propionic acid-y-lactone), bovine serum albumin (BSA), and 3,3’-diaminobenzidine tetrahydrochloride (DAB) were obtained from Sigma (St. Louis, MO, U.S.A.). Aminoglutethimide (Caminophenyl3-ethyl-2,6_piperidinedione) was obtained from Ciba-Geigy (Rueil-Malmaison, France). Acrylamide and bisacrylamide were purchased from Serva (Heidelberg, F.R.G.). Dithiothreitol (DTT) was obtained from Boeringher Mannheim (Meylan, France) and Tween 80 from Merck (Darmstadt, F.R.G.). Medium 199 with Earle’s salts (M199) was obtained from Eurobio (Paris, France). Tissue culture dishes were purchased from Falcon Plastics (Oxnard, CA, U.S.A.). Trilostane (4cY,S-epoxy-17/%hydroxy-3-keto-5cY-androstan-2acarbonitrile) was a gift from the Sterling-Winthrop Research Institute (Rensselaer, NY, U.S.A.) and ovine LH (NIH-LH-S19, 1.01 units/mg) was kindly supplied by Dr. P.G. Condliffe (NIAMDD. Bethesda, MD, U.S.A.). The antiserum obtained from a single bleed of rabbits immunized with pregnenolone-succinyl-bovine serum albumin was purchased from Steranti Research (St. Albano, U.K.). Rabbit antibody to bovine cytochrome P450,,, prepared by immunization with the protein purified from adrenocortical mitochondria of cattle (Defaye et al., 1982) was a generous gift from Dr. E. Chambaz (INSERM U244, Grenoble, France). Biotinylated IgGs raised against rabbit immunoglobulins and avidin-biotin-peroxidase complex (Vectastain reagents) were purchased from Vector Laboratories (Burlingame, CA, U.S.A.). Estimution of pregnenolone production by organ culture The pregnant rats were anesthetized by an intraperitoneal injection (90 pl/lOO g) of sodium pentobarbital (Lathevet, Paris, France) and fetuses were delivered aseptically by cesarean section. The fetal gonads were immersed in Medium 199 (100 pi/ovary and 500 pi/testis) supplemented with glutamine (0.35%), gentamicin (40 pg/ml), and penicillin (200 IU/ml), in the presence or absence of (Bu),cAMP (1 mM), spironolactone (10 PM) and trilostane (30 PM). The media were removed every 24 h and assayed for pregnenolone without extraction. Ovaries were combined (2-6) and ho-
113
mogenized in ethyl acetate in a Kontes homogenizer at 4’ C. The homogenates were extracted twice with 500 1.11ethyl acetate with centrifugation and the combined extracts were evaporated under a stream of dry N,. The resulting residue was taken up in 350 ~1 of phosphate-buffered saline (pH 7) with 0.1% gelatin before the assay. The radioimmunoassay (RIA) of pregnenolone was performed by a procedure similar to that of the testosterone assay (Picon, 1976). The antiserum cross-reacted with progesterone (5.9%), less than 0.3% with 17a-hydroxypregnenolone, 17~hydroxyprogesterone and 20a-dihydroprogesterone and less than 0.01% with cholesterol. The limit of sensitivity was estimated to be 0.25 pmol in the radioimmunoassay. Conversion of [‘H]cholesterol to [3H]pregnenolone by mitochondria Mitochondria were prepared from cultured or fresh gonads. The gonads were homogenized in 3 ml of Mersmann medium (Mersmann et al., 1972). Homogenization and subsequent centrifugation steps were carried out at 4” C. The homogenate was centrifuged for 10 min at 480 x g and the supernatant was then centrifuged at 7700 X g for 10 min. The mitochondrial pellet was washed twice and resuspended in the isolation medium. The incubation was performed in a shaking water bath at 37 o C in the incubation medium composed of 9 mM MgCl,, 50 mM Tris, 49 mM KCl, 22 mM KH,PO,, 0.5% BSA and 10 mM succinate (pH 7.4). The reaction was initiated by the addition of 150 ~1 to 250 ~1 of mitochondrial pellet to the incubation medium (final volume 2 ml) supplemented with spironolactone (10 PM), trilostane (30 PM) and [ 3H]cholesterol (3.6 PCi). Unlabelled cholesterol (1 PM) was added to the medium in order to reduce possible differences between the endogenous pools present in the gonads. The reaction was stopped by the addition of 2 ml of chloroform-methanol (2/l). [i4C]Pregnenolone (400 dpm) was added to each tube to determine recovery and extraction was performed for 60 min. A two-phase system was established by addition of 2 ml of chloroform and 1 ml of water. The chloroform phase was reduced to a small volume. A preliminary separation of cholesterol, pregnenolone and progesterone was achieved by Sephadex
LH-20 partition chromatography in the solvent system hexane-benzene-methanol (85/5/4). Unlabelled pregnenolone (25 pg) was added as carrier, pregnenolone was further purified by silica gel G thin-layer chromatography in the system chloroform-ether (75/15). Pregnenolone was visualized with iodine vapor. The metabolites were scraped off, extracted with 1 ml of methanol overnight and counted in 8 ml of EP scintillation fluid (Beckmann, Glenrothes, U.K.) in an Intertechnique liquid scintillation spectrometer. Blank values obtained at time 0 of incubation were estimated and subtracted from the other values in each experiment. Protein immunoblotting Sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis was performed essentially as described by Laemmli (1970). The separation gel contained 15% polyacrylamide (acrylamide-bisacrylamide, 75/l). The stacking gel was composed of 5 % polyacrylamide (acrylamide-bisacrylamide, 38/l). Samples were solubilized in 50 mM Tris,
izl Day 1 I Day2
a +
+
+
spironolactone
(l3u),cAMP
-
+
+
lrilostane
-
+
Fig. 1. Production and content of pregnenolone in rat female gonads from 185day-old fetuses in the presence of (Bu),cAMP (1 mM), spironolactone (10 PM) and trilostane (30 PM). Ovaries were cultured for 48 h. The amount of pregnenolone produced in the medium was estimated every 24 h and the pregnenolone content at the end of the culture by RIA. Each column represents the mean f SEM of the number of explantations indicated in parentheses. The dashed lines correspond to the limit of sensitivity (RIA sensitivity X dilution factor). (- - - -) for production; (__........) for content. **p < 0.01, * **p < 0.001 (Scheffe F-test).
114
5% glycerol, 50 mM dithiothreitol, 1% SDS, and boiled 5 min. Proteins were transferred from acrylamide gel onto nitrocellulose paper (Hybond C, Amersham) electrophoretically (Multiphor II/Novablot from LKB) as described by Towbin et al. (1979). Transfers were performed at 0.8 mA/cm2 of gel during 1 h. The efficiency of transfer was determined by silver staining of the remaining proteins on the gel, according to Morissey (1981). Retention of proteins in the blots was checked by brief Ponceau Red staining. The paper sheets were then incubated for 1 h at room temperature in a phosphate buffer saline (PBS) containing 5% dried skim milk (PBS-milk). After saturation, sheets were directly incubated overnight at room temperature with antibody against cytochrome P-450,,,, diluted (l/200) in PBS-milk. Sheets were then rinsed in PBS with 0.1% Tween 80 (PBS-T) (three 15 min washings) before incubation at room temperature for 60 min with 35Slabelled mouse anti-rabbit IgG (0.3 pCi/ml in PBS-T). After three 15 min washings in PBS-T, sheets were dried. The specific bands corresponding to the cytochrome P-450,,, were visualized by autoradiography. Immunoperoxidase
staining
Tissue blocks were fixed for 2 h in Carnoy’s fixative (absolute ethanol-chloroform-glacial acetic acid, 6/3/l). After dehydratation in absolute ethanol and 1-butanol, the tissue was embedded in paraffin wax and 7.5 pm sections were mounted on uncoated slides. Deparaffinized sections were
TABLE
510
20
TIME
510
(min)
TIME
Fig. 2. Time course of the conversion of [ 3H]cholesterol to [3H]pregnenolone by mitochondria. A: From 21.5-day-old fetal testes. B: From 18.5-day-old fetal ovaries precultured for 48 h in the presence of (Bu),cAMP (1 mM). In parentheses: number of experiments. Each assay was performed in duplicate.
rehydrated and rinsed with PBS and exposed for 30 min to 10% ovalbumin to block non-specific reactions. The sections were then incubated for 2 h in appropriate dilutions of anti-P-450,,,-IgG diluted in PBS, pH 7.4 containing 0.1% ovalbumin. After washing in PBS, the sections were incubated for 30 min in biotinylated goat secondary antibody, rinsed again in PBS and incubated for 30 min in avidin-biotin-peroxidase complex. Peroxidase was revealed in diaminobenzidine, 0.05% in PBS containing 0.035% hydrogen peroxide. Finally, the sections were rinsed in distilled water, dehydrated and mounted. Sections of fetal testes were included in each series to serve as positive control of the reaction. Negative controls were obtained with IgG from non-immune serum, the
1
CONVERSION
OF (‘HICHOLESTEROL
Each experiment
was performed
in duplicate
No. of testes
Protein content (pg/testis)
Expt. 1
40
134
Expt. 2
42
141
Expt. 3
38
160
TO [‘HJPREGNENOLONE
BY TESTES
FROM
21.5-DAY-OLD
FETUSES
(values in parentheses).
[ 3Hlpregnenolone
20
(min)
(dpm/mg
protein
Crude homogenate
Mitochondrial fraction
3,350 (3,250-3,470) 2,330 (2,560-2,100) 5,227 (5,060-5,395)
32,580 (33,5Ot-31,660) 15,785 (14,930-16,640) 35,832 (30,150-41,515)
per min) Microsomal fraction 448 487 (540-435)
Cytosolic fraction ND ND
115
specific staining
antibody diluted down to extinction or exhausted by an excess of bovine antigen. All reactions were performed 4%C room temperature.
of Pat
spironolactone enhanced the production of pregnenolone and the production was found to increase with culture time. Spironolactone also enA
(a)
R
C
D
F
E
G
Protein determination Protein content of homogenates and of mitochondrial pellets was determined according to the technique described by Bradford (1976). Bovine serum albumin was used as standard.
70
451
c
Statistical analysis All values are means f SEM. Results were analyzed by the Student t-test or by one-way analysis of variance (ANOVA) (Scheffe F-test).
.
5
E
Results Pregnenolone detected by radioimmunoassay in vitro A first series of experiments was performed with testes from 16.5-day-old fetuses in vitro. The production of pregnenolone was undetectable in control media (n = 4). Therefore further metabolism of pregnenolone was inhibited with spironolactone, an inhibitor of 17ol-hydroxylase (Cigorraga et al., 1978) and trilostane, an inhibitor of 3fl-hydroxysteroid dehydrogenase (Sir@-Asa et al., 1982). For each fetus, one gonad was cultured in the presence of spironolactone and the contralateral testis in the presence of spironolactone and trilostane. The production of pregnenolone reached 15 + 2 pmol/gonad per 24 h (n = 3) in the presence of spironolactone and 50 + 3 pmol/gonad per 24 h (n = 3) in the presence of both inhibitors (p < 0.01, paired test). Then, ovaries from l&S-day-old fetuses were explanted for 48 h. Pregnenolone produced in the media every 24 h and pregnenolone content of the ovaries at the end of the culture were assayed. No production of pregnenolone was noted during the 2 days of culture in basal medium (M199) (n = 2) and in the media supplemented with spironolactone (n = 10) or with spironolactone and trilostane (n = 20). Pregnenolone was not detected in the corresponding organ extracts. When the fetal ovaries were cultured in the presence of (Bu),cAMP (Fig. l), a very low production of pregnenolone was detectable in the media and in the organ extracts. The addition of
(b)
A
0
C
D
E
Fig. 3. Immunoisolation of cytochrome P-454,,. For each sample, proteins (150 pg) were separated by SDS-polyacrylamide gel electrophoresis and electroblotted onto nitrocellulose. Cytochrome P-450,, was immunoisolated as described in Materials and Methods. (a) From fetal gonads: autoradiograms of purified cytochrome P-450,,, (A), homogenate of fetal heart (B), homogenates of ovaries from 18.5-day-old fetuses cultured for 48 h in the presence (C) or absence (D) of (Bu),cAMP, testes from 21.5-day-old fetuses: crude homogenate (E), mitochondrial fraction (F), cytosolic fraction (G). Arrow on the right indicates position of purified antigen; molecular weight (m wt) standards are indicated on the left. (b) From fetal and neonatal ovaries: autoradiograms of purified cytochrome P-450,,, (A), homogenates of ovaries from 10.5-day-old neonates cultured in the presence (B) or absence (C) of LH, homogenates of ovaries from 18.5-day-old fetuses cultured in the presence (D) or absence (E) of (Bu),cAMP. In order to enhance the sensitivity, immunoblotting was followed by treatment of the sheet with Amplify spray and exposure to X-Omat film.
117
hanced the level of ovarian pregnenolone content. An additive effect was observed when trilostane was added to medium supplemented with (Bu),cAMP and spironolactone. In these experimental conditions, no pregnenolone could be detected in media where control tissues (Mtillerian ducts and mesonephros) were cultured. Lastly, a series of six experiments was performed to study the effect of short-time exposure to (Bu),cAMP on the pregnenolone production. Ovaries from 18.5day-old fetuses were explanted for 24 h with spironolactone, trilostane and (Bu)~cAMP and the media were removed every 6 h and assayed. At t6 and t12, no pregnenolone could be detected in the media. At t18, pregnenolone production was 1.33 k 0.12 pmol/6 h/ovary and at t24 reached 1.63 + 0.30 pmol/6 h/ovary. Side-chain cleavage activity in isolated mitochondria (I) Side-chain cleavage activity in fetal testes. First. a series of six experiments was performed to study the time course of pregnenolone production. About 40 testes from 21.5day-old fetuses were pooled and the mitochondrial pellet was prepared as described in Materials and Methods. Aliquots of the mitochondrial fraction corresponding to 1.78 + 0.18 testes were incubated with radioactive and non-radioactive cholesterol and the production of pregnenolone was evaluated for 5, 10 and 20 min. The rate of production was linear from 0 to 20 min and declined thereafter (Fig. 2A). Second, the effect of increasing tissue concentration on the see activity was measured after 10 min of incubation in various dilutions of the mitochondrial pellet corresponding to 1.16-3.4 testes. The amount of pregnenolone produced was proportional to the tissue concentration (r = 0.999). In a third series of experiments, the activities per unit protein were compared in crude homogenate, mitochondrial pellet (7700 x g), microsomal pellet (100,000 X g) and cytosolic super-
natant (100,000 X g) (Table 1). The activity was primarily located in the mitochondrial fraction. Lastly, when mitochondria were incubated with aminoglutethimide, an inhibitor of see activity (Gower, 1974) conversion of [ 3H]cholesterol into [3H]pregnenolone was inhibited to a great extent (98.1%). (2) Side-chain cleavage activity in neonatal ovaries. Side-chain cleavage activity was studied in ovaries from 8.5 and 9.5-day-old neonates. Conversion of [ 3H]cholesterol to [ 3H]pregnenolone was observed at both stages. The production increased between 8.5 and 9.5 days (260 and 1610 dpm/lO mm/ovary, respectively). (3) Side-chain cleavage activity in fetal ovaries. First, experiments were performed on ovaries from 18.5- and 20.5-day-old fetuses. 48-60 ovaries were pooled and mitochondria isolated. Aliquots of the mitochondrial pellet containing about five ovaries were incubated for 5, 10 and 20 min. In all experiments no conversion of [3H]cholesterol to [ 3H]pregnenolone could be detected; the values observed were not different statistically from blank values obtained at time 0 of incubation (n = 4). Then experiments were performed on ovaries from 18.5-day-old fetuses cultured for 48 h in Ml99 in the presence or absence of (Bu),cAMP. For each experiment, 52-58 ovaries were pooled and mitochondria isolated. Aliquots of the mitochondrial pellet corresponding to about five ovaries were incubated for 5, 10 and 20 min. When the ovaries were cultured in the control medium no production of [ 3H]pregnenolone could be detected (n = 4). Therefore, after a pretreatment for 48 h with (Bu),cAMP, see activity was observed in the mitochondrial pellets (Fig. 2B). Immunoblot analysis and cytochemistry of the cytochrome P-450,,, The specific content of immunoreactive cytochrome, P-450,,, was studied in testes from 21.5day-old fetuses, ovaries from 10.5-day-old
Fig. 4. Immunolocalization of cytochrome P-450S, enzyme: testes from 14.5day-old fetus (A), 15.5day-old fetus (B), 18.5day-old fetus (C, G); ovary from 18.Sday-old fetus, precultured for 48 h in the presence of (Bu),cAMP (E); adrenal from 18.5day-old fetus (F); ovaries from 10.5-day-old neonates without preculture (D) and after treatment in vitro with LH for 48 h (H). Bars: 100 pm in A, B, C and E; 25pmin D, F,G and H.
neonates and ovaries from 18.5-day-old fetuses. Fetal ovaries were cultured for 48 h in the presence or absence of (Bu),cAMP and neonatal ovaries with and without LH (100 ng/ml). The experiment was repeated 3 times. The molecular weight of the cytochrome, as estimated from the position of standards was 50,000. An additional band present in all lanes with an apparent M, of less than the 45 kDa marker probably represents an artifact. Immunoreactive cytochrome P-450,,. was detectable both in the crude homogenate and in the mitochondrial pellet of testes, but absent in the cytosolic supernatant (Fig. 3~); autoradiographic band was undetectable in blots of control fetal ovaries (Fig. 3a and b) and fetal heart (Fig. 3a), but could be revealed in (Bu),cAMP-pretreated ovaries (Fig. 3a and 6). lmmunoreactive cytochrome P-450,,, was present in blots of neonatal ovaries and slightly enhanced by treatment with LH (Fig. 36). The immunohistochemical staining of P-450S,, was positive in the interstitial cells of 18.5-day-old fetal testes; immunoreactivity absent on day 14.5 could be detected on day 15.5 and the reaction was greatly enhanced on day 18.5 (Fig. 4A, B, C and G). The technique also showed positive staining of the cytoplasm in adrenocortical cells of l&5-day-old fetuses; adrenal medulla cells showed no such staining (Fig. 4F). Controls were performed with fetal testes included non-immune rabbit IgG, dilution of the antibody down to extinction of staining and presaturation of the antibody with purified bovine P-450,,, antigen or rat liver acetone powder (not shown). In postnatal ovaries, a faint reaction was detected in the cytoplasm of a few interstitial cells. The reaction was extended to more interstitial cells in ovaries cultured in the presence of LH (100 ng/ml) (Fig. 40 and H). However, no reaction was observed in sections of ovaries from 18.5-day-old fetuses cultured for 48 h in the presence or absence of (Bu),cAMP (Fig. 4E). Discussion
This study was aimed at evaluating the cholesterol side-chain cleavage activity in rat male and female gonads during fetal development. Our results indicate that the see activity was detectable
in fetal testes, but not in fetal ovaries. Production of pregnenolone was observed in media where testes were cultured in the presence of spironolactone and trilostane. Absence of pregnenolone in control media can be explained by an active metabolism of this steroid to testosterone since this latter one is extensively produced during fetal life (Picon, 1976). In contrast, pregnenolone was not detectable in female media, even in the presence of the inhibitors. This finding is consistent with the absence of conversion of radiolabelled cholesterol into pregnenolone. Indeed it should be noted that see activity was detected in mitochondria isolated from fetal testes and from ovaries of 8.5-day-old neonates. Using an immunoblotting technique, we detected cytochrome P-450,,, in lysates of these organs. Its absence in fetal ovaries could be explained by a lack or a very low expression of the enzyme. It might be suggested that the synthesis of cytochrome P-450S,c can be the ratelimiting step in ovarian steroidogenesis. In human fetuses, P-450,,, mRNA was not detected in Northern blot of ovarian mRNA unless a long exposure of the autoradiograph of dot-blots of the samples (Voutilainen and Miller, 1986). Our immunohistochemical study showed, for the first time, the localization of the enzyme in Leydig cells from 15.5-day-old male fetuses onwards, In female gonads, some interstitial cells were immunostained in ovaries as early as 10 days after birth. Progesterone production in vitro begins to be detectable in 4-day-old ovaries cultured for 48 h (Funkenstein et al., 1980) and it can be assumed that the interstitial immunoreactive cells represent the only source of de novo progesterone production. In previous studies performed on older female rats, the localization was extended to numerous interstitial cells and to some of the thecal internal cells (Bitzur and Orly, 1989; Le Goascogne et al., 1989). Our study shows that LH is effective in extending immunostaining to more interstitial cells in neonatal ovaries. This induction in cytochrome P-450,,, synthesis suggests that LH receptors are localized on interstitial cells at this stage of development. Treatment of the fetal ovaries (Bu),cAMP stimulated a synthesis of pregnenolone, but the values of secretion and content were found to be low. These results could be explained by a
119
metabolism of pregnenolone. The addition of spironolactone and trilostane increased both the production and the ovarian content of pregnenolone. In fetal rat ovaries, 3/?-hydroxysteroid dehydrogenase activity was detected and enhanced with the nucleotide (Picon et al., 1988). On the other hand, conversion of progesterone into androgens has led to contradictory results (Noumura et al., 1966; Weniger et al., 1984) and the presence of an 17a-hydroxylase/desmolase activity in unstimulated fetal ovaries cannot be stated. Therefore an inductive effect of the nucleotide on this enzyme in our experiments is not excluded. Recent data demonstrate that theCal-interstitial cells from immature rat ovaries lack the 17ol-hydroxylase enzyme which can be induced by LH in serum-free medium (Magoffin, 1989). The time course study of (Bu),cAMP on pregnenolone synthesis displayed a lag period of 18 h compatible with a chronic effect including transcriptional activation of the gene(s) for steroidogenesis rather than rapid mobilization of cholesterol to the inner mitochondrial membrane. The conversion of [ 3H]cholesterol into pregnenolone was also obtained in mitochondria from fetal ovaries pretreated with the nucleotide and cytochrome P-450,,, could be isolated in immunoblots as well. So far, in rats, the regulation of P-450,,, biosynthesis has been studied in immature ovarian cells in culture. Thus, the maintenance of the cytochrome can be obtained in the granulosa cells with follicle-stimulating hormone (FSH) (Trzeciak et al., 1986) or with et al., 1986; Zlotkin (Bu),cAMP (Trzeciak et al., 1986) and in the interstitial cells with LH (Goldring et al., 1986). No follicular structures can be observed in 18.5day-old fetal ovaries and unfortunately the immunostaining technique failed to localize the cytochrome P-450,, , which could be expressed at a low level in most somatic cells. The present study demonstrates that the nucleotide was able to induce pregnenolone synthesis in fetal ovaries, but not in other mesodermic non-steroidogenic tissues, which indicates the fetal female gonads are committed for expressing the first step of steroidogenesis. In male gonads, testosterone production (Picon, 1976) and differentiation of human chorionic gonadotropin (hCG)-LH receptors (Gangnerau et al., 1982) start at the same time. However, the
steroidogenesis in male indifferentiated gonads was shown to occur in serum-free medium without gonadotrophic hormone (Gangnerau and Picon, 1987). The appearance of actively steroidogenic cells in ovaries during postnatal life is induced by as yet unknown mechanisms. Differentiation of LH receptors has been noticed to begin at the same period of development and the question can be raised whether these two aspects of interstitial cell differentiation are independent. Acknowledgements We thank Dr. E. Chambaz (INSERM U244, Grenoble, France) for the anti-P-450,,, IgG and the P-450,,, antigen. This work was supported by a grant from the INSERM (grant No. 854019). References Anderson, C.M. and Mendelson, CR. (1985) Arch. B&hem. Biophys. 238, 378-387. Bitzur, S. and Orly, J. (1989) Endocrinology 124, 1471-1484. Bradford, M.M (1976) Anal. Biochem. 72, 248-254. Cigogarra, S.B., Dufau, M.L. and Catt, K.J. (1978) J. Biol. Chem. 253, 4297-4304.
Defaye. G., Monnier, N., Guidicelli, C. and Chambaz, E.M. (1982) Mol. Cell. Endocrinol. 27, 157-168. Funkenstein, B., Nimrod, A. and Lindner, H.R. (1980) Endocrinology 106, 98-106. Gangnerau, M.N. and Picon, R. (1987) Arch. Androl. 18, 215-224. Gangnerau, M.N., Funkenstein, B. and Picon, R. (1982) Mol. Cell. Endocrinol. 28, 499-512. Goldring, N.B., Farkash, Y., Goldschmidt, D. and Orly, J. (1986) Endocrinology 119, 2821-2832. Gower, D.B. (1974) J. Steroid B&hem. 5, 501-523. Hall, P.F. (1966) Endocrinology 78, 690-698. Hall, P.F. and Koritz, S.B. (1965) Biochemistry 4, 1037-1043. Laemmli, U.K. (1970) Nature 227, 680-685. Lamprecht, !%A., Kohen, F., Ausher, J., Zor, U. and Lindner, H.R. (1976) J. Endocrinol. 68, 343-344. Le Goascogne, C., Sananes, N., Gouezou, M., Baulieu, E.E. and Robe], P. (1989) J. Reprod. Fertil. 85, 61-72. Magoffin, D.A. (1989) Endocrinology 125, 1464-1473. Mersmann, H.J., Goodman, J., Houk, J.M. and Anderson, S. (1972) J. Cell Biol. 53, 335-347. Morissey, M. (1981) Anal. B&hem. 117, 307-310. Noumura, T., Weisz, J. and Lloyd, C.W. (1966) Endocrinology 78, 245-253. Picon, R. (1976) J. Endocrinol. 71, 231-238. Picon, R., Pelloux, M.C., Benhaim, A. and Gloaguen, F. (1985) J. Steroid Biochem. 23, 995-1006.
120 Picon, R., Darmoul, D., Rouiller, V. and Duranteau, L. (1988) J. Steroid B&hem. 31, 839-843. Siebers, J.W., Peters, F., Zenzes, M.T., Schmidtke, J. and Engel, W. (1977) J. Endocrinol. 73, 491-496. Singh-Asa, P., Jenkin, G. and Thorbum, G.D. (1982) J. Endocrinol. 92, 2055212. Smeaton, T.C., Arcondoulis, D.E. and Steele, P.A. (1975) Steroids 26, 181-192. Smith-White, S. and Ojeda, S.R. (1981) Endocrinology 109, 152-161. Towbin, H., Staehelm, R. and Gordon, J. (1979) Proc. Natl. Acad. Sci. U.S.A. 76, 4350-4354.
Trzeciak, W.H., Waterman, M.R. and Simpson, E.R. (1986) Endocrinology 119, 323-330. Voutilainen, R. and Miller, W.L. (1986) J. Clin. Endocrinol. Metab. 63, 114551150. Warren, D.W., Haltmeyer, G.C. and Eik-Nes, K.B. (1973) Biol. Reprod. 8, 560-565. Waterman, M.R. and Simpson, E.R. (1985) Mol. Cell. Endocrinol. 39, 81-89. Weniger, J.P., Chouraqui, J. and Zeis, A. (1984) J. Steroid B&hem. 21, 347-349. Zlotkin, T., Farkash, Y. and Orly, J. (1986) Endocrinology, 119, 2809-2820.
MOLCEL
111- 120
111
02321
Cholesterol
side-chain
V. Rouiller
cleavage activity in rat fetal gonads: a limiting step for ovarian steroidogenesis
‘, M.N. Gangnerau
‘, J.L. Vayssiere
2 and R. Picon ’
’ Laboratoire de Physiologic du DCveloppement, CNRS, URA 307, Universitk Paris 7, 75251 Paris Cedex 0.5, France, and 2 Laboratoire de Biochimie CeNulaire, CoilGgede France, 75231 Paris Cedex 05, France (Received 9 April 1990; accepted 23 May 1990)
Key words: Fetus; Ovary; Testis; Pregnenolone; Cytochrome P-450,;
(Rat)
Summary The aim of this study was to examine the first step in steroidogenesis in male and female gonads of fetal rats. Pregnenolone production was measured by radioimmunoassay in organ culture, conversion of [ 3H]cholesterol to [ 3H]pregnenolone was evaluated in isolated mitochondria and cytochrome P-450,, was revealed by immunoblotting and immunocytochemical techniques. Our results clearly showed that in fetal testes (1) pregnenolone was produced in media where testes were cultured in the presence of trilostane and spironolactone, indicating an important metabolism of pregnenolone, (2) [ 3H]cholesterol was converted into [ 3H]pregnenolone in mitochondria, (3) cytochrome P-450,,, was revealed in immunoblots with a molecular weight of 50,000, (4) cytochrome P-450,, was localized in Leydig cells from 15.5-day-old fetal testes onwards. With respect to fetal ovaries, we were unable to detect any see activity, except after treatment with dibutyryl cyclic AMP. A lag period of 18 h was necessary to induce pregnenolone synthesis. However, the immunoperoxidase staining did not localize ovarian positive cells. Cytochrome P-450,,, could be revealed in postnatal ovaries by immunoblotting and some interstitial positive cells were observed with immunostaining; the reaction was enhanced in luteinizing hormone-pretreated ovaries. These data indicate that (a) the cholesterol see activity is present in fetal testes, (b) the conversion of cholesterol to pregnenolone is a limiting step for steroidogenesis in fetal ovaries. The inductive effect of the nucleotide on the enzyme suggests that the absence of gonadotrophic receptors in fetal female gonads could explain the lack of steroidogenesis before birth.
Introduction In rats, production of steroids by male gonads has been demonstrated during fetal life from 15 days onwards (Warren et al., 1973; Picon, 1976)
Address for correspondence: R. Picon, Laboratoire de Physiologic du DCveloppement, CNRS, URA 307, UniversitC Paris 7, 75251 Paris Cedex 05, France. 0303-7207/~/$03.50
whereas ovarian steroidogenesis develops only after birth (Smeaton et al., 1975; Lamprecht et al., 1976; Funkenstein et al., 1980). However, some enzymes involved in steroidogenesis have been revealed to be active in the female gonads during fetal life. Thus aromatization of testosterone to estrogen could be observed in fetal ovaries as early as 14.5 days and this biochemical property allows to differentiate ovaries from testes at this stage (Picon et al., 1985). 3P-Hydroxysteroid dehydro-
Q 1990 Elsevier Scientific Publishers Ireland, Ltd.
112
genase was also detected in fetal female gonads (Picon et al., 1988). In adult tissues, the side-chain cleavage (see) of cholesterol to pregnenolone is considered to be the rate-limiting step in the synthesis of steroid hormones (Hall and Koritz, 1965; Hall, 1966; Anderson and Mendelson, 1985). Mediated by a single mitochondrial cytochrome P-450, this step is known to be an important site of regulation by trophic hormones (Waterman and Simpson, 1985). Luteinizing hormone (LH) receptors are present in testes from 15-day-old fetuses onwards (Gangnerau et al., 1982) whereas gonadotrophic receptors cannot be revealed in fetal ovaries (Siebers et al., 1977; Smith-White and Ojeda, 1981). Stimulation of estradiol-17P release from rat ovaries with LH was observed only from the beginning of week 2 after birth (Lamprecht et al., 1976). The aim of this study was to investigate the first step of steroidogenesis to give some insight into the mechanism controlling pregnenolone production in male and female gonads during fetal life. We therefore assessed the production of pregnenolone during culture of the fetal gonads; measured the conversion of radiolabelled cholesterol to pregnenolone in isolated mitochondria and examined the effect of dibutyryl cyclic AMP ((Bu),cAMP) on see activity in fetal ovaries. The presence of cytochrome P-450,,, was investigated with techniques of immunoblotting and immunohistochemistry. Materials and methods Animals Fetuses and newborn of the Wistar strain were used. Gestational age was based on the timing of ovulation as previously described (Gangnerau et al., 1982). Parturition occurred during the daytime of day 22 of gestation. Materials [4,7-3H]pregnenolone (spec. act. 19 Ci/mmol) and [1,2( n)- 3H]cholesterol (spec. act. 40 Ci/mmol) were obtained from the Radiochemical Centre (Amersham, U.K.). [4-‘4C]Pregnenolone (spec. act. 57 mCi/mmol) was purchased form NEN Research Products (Boston, MA, U.S.A.). Unlabelled steroids, (Bu),cAMP, spironolactone (3-[3-0x0-
7~-acetylthio-l7~-hydroxy-4-androstene-l7~-yl]propionic acid-y-lactone), bovine serum albumin (BSA), and 3,3’-diaminobenzidine tetrahydrochloride (DAB) were obtained from Sigma (St. Louis, MO, U.S.A.). Aminoglutethimide (Caminophenyl3-ethyl-2,6_piperidinedione) was obtained from Ciba-Geigy (Rueil-Malmaison, France). Acrylamide and bisacrylamide were purchased from Serva (Heidelberg, F.R.G.). Dithiothreitol (DTT) was obtained from Boeringher Mannheim (Meylan, France) and Tween 80 from Merck (Darmstadt, F.R.G.). Medium 199 with Earle’s salts (M199) was obtained from Eurobio (Paris, France). Tissue culture dishes were purchased from Falcon Plastics (Oxnard, CA, U.S.A.). Trilostane (4cY,S-epoxy-17/%hydroxy-3-keto-5cY-androstan-2acarbonitrile) was a gift from the Sterling-Winthrop Research Institute (Rensselaer, NY, U.S.A.) and ovine LH (NIH-LH-S19, 1.01 units/mg) was kindly supplied by Dr. P.G. Condliffe (NIAMDD. Bethesda, MD, U.S.A.). The antiserum obtained from a single bleed of rabbits immunized with pregnenolone-succinyl-bovine serum albumin was purchased from Steranti Research (St. Albano, U.K.). Rabbit antibody to bovine cytochrome P450,,, prepared by immunization with the protein purified from adrenocortical mitochondria of cattle (Defaye et al., 1982) was a generous gift from Dr. E. Chambaz (INSERM U244, Grenoble, France). Biotinylated IgGs raised against rabbit immunoglobulins and avidin-biotin-peroxidase complex (Vectastain reagents) were purchased from Vector Laboratories (Burlingame, CA, U.S.A.). Estimution of pregnenolone production by organ culture The pregnant rats were anesthetized by an intraperitoneal injection (90 pl/lOO g) of sodium pentobarbital (Lathevet, Paris, France) and fetuses were delivered aseptically by cesarean section. The fetal gonads were immersed in Medium 199 (100 pi/ovary and 500 pi/testis) supplemented with glutamine (0.35%), gentamicin (40 pg/ml), and penicillin (200 IU/ml), in the presence or absence of (Bu),cAMP (1 mM), spironolactone (10 PM) and trilostane (30 PM). The media were removed every 24 h and assayed for pregnenolone without extraction. Ovaries were combined (2-6) and ho-
113
mogenized in ethyl acetate in a Kontes homogenizer at 4’ C. The homogenates were extracted twice with 500 1.11ethyl acetate with centrifugation and the combined extracts were evaporated under a stream of dry N,. The resulting residue was taken up in 350 ~1 of phosphate-buffered saline (pH 7) with 0.1% gelatin before the assay. The radioimmunoassay (RIA) of pregnenolone was performed by a procedure similar to that of the testosterone assay (Picon, 1976). The antiserum cross-reacted with progesterone (5.9%), less than 0.3% with 17a-hydroxypregnenolone, 17~hydroxyprogesterone and 20a-dihydroprogesterone and less than 0.01% with cholesterol. The limit of sensitivity was estimated to be 0.25 pmol in the radioimmunoassay. Conversion of [‘H]cholesterol to [3H]pregnenolone by mitochondria Mitochondria were prepared from cultured or fresh gonads. The gonads were homogenized in 3 ml of Mersmann medium (Mersmann et al., 1972). Homogenization and subsequent centrifugation steps were carried out at 4” C. The homogenate was centrifuged for 10 min at 480 x g and the supernatant was then centrifuged at 7700 X g for 10 min. The mitochondrial pellet was washed twice and resuspended in the isolation medium. The incubation was performed in a shaking water bath at 37 o C in the incubation medium composed of 9 mM MgCl,, 50 mM Tris, 49 mM KCl, 22 mM KH,PO,, 0.5% BSA and 10 mM succinate (pH 7.4). The reaction was initiated by the addition of 150 ~1 to 250 ~1 of mitochondrial pellet to the incubation medium (final volume 2 ml) supplemented with spironolactone (10 PM), trilostane (30 PM) and [ 3H]cholesterol (3.6 PCi). Unlabelled cholesterol (1 PM) was added to the medium in order to reduce possible differences between the endogenous pools present in the gonads. The reaction was stopped by the addition of 2 ml of chloroform-methanol (2/l). [i4C]Pregnenolone (400 dpm) was added to each tube to determine recovery and extraction was performed for 60 min. A two-phase system was established by addition of 2 ml of chloroform and 1 ml of water. The chloroform phase was reduced to a small volume. A preliminary separation of cholesterol, pregnenolone and progesterone was achieved by Sephadex
LH-20 partition chromatography in the solvent system hexane-benzene-methanol (85/5/4). Unlabelled pregnenolone (25 pg) was added as carrier, pregnenolone was further purified by silica gel G thin-layer chromatography in the system chloroform-ether (75/15). Pregnenolone was visualized with iodine vapor. The metabolites were scraped off, extracted with 1 ml of methanol overnight and counted in 8 ml of EP scintillation fluid (Beckmann, Glenrothes, U.K.) in an Intertechnique liquid scintillation spectrometer. Blank values obtained at time 0 of incubation were estimated and subtracted from the other values in each experiment. Protein immunoblotting Sodium dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis was performed essentially as described by Laemmli (1970). The separation gel contained 15% polyacrylamide (acrylamide-bisacrylamide, 75/l). The stacking gel was composed of 5 % polyacrylamide (acrylamide-bisacrylamide, 38/l). Samples were solubilized in 50 mM Tris,
izl Day 1 I Day2
a +
+
+
spironolactone
(l3u),cAMP
-
+
+
lrilostane
-
+
Fig. 1. Production and content of pregnenolone in rat female gonads from 185day-old fetuses in the presence of (Bu),cAMP (1 mM), spironolactone (10 PM) and trilostane (30 PM). Ovaries were cultured for 48 h. The amount of pregnenolone produced in the medium was estimated every 24 h and the pregnenolone content at the end of the culture by RIA. Each column represents the mean f SEM of the number of explantations indicated in parentheses. The dashed lines correspond to the limit of sensitivity (RIA sensitivity X dilution factor). (- - - -) for production; (__........) for content. **p < 0.01, * **p < 0.001 (Scheffe F-test).
114
5% glycerol, 50 mM dithiothreitol, 1% SDS, and boiled 5 min. Proteins were transferred from acrylamide gel onto nitrocellulose paper (Hybond C, Amersham) electrophoretically (Multiphor II/Novablot from LKB) as described by Towbin et al. (1979). Transfers were performed at 0.8 mA/cm2 of gel during 1 h. The efficiency of transfer was determined by silver staining of the remaining proteins on the gel, according to Morissey (1981). Retention of proteins in the blots was checked by brief Ponceau Red staining. The paper sheets were then incubated for 1 h at room temperature in a phosphate buffer saline (PBS) containing 5% dried skim milk (PBS-milk). After saturation, sheets were directly incubated overnight at room temperature with antibody against cytochrome P-450,,,, diluted (l/200) in PBS-milk. Sheets were then rinsed in PBS with 0.1% Tween 80 (PBS-T) (three 15 min washings) before incubation at room temperature for 60 min with 35Slabelled mouse anti-rabbit IgG (0.3 pCi/ml in PBS-T). After three 15 min washings in PBS-T, sheets were dried. The specific bands corresponding to the cytochrome P-450,,, were visualized by autoradiography. Immunoperoxidase
staining
Tissue blocks were fixed for 2 h in Carnoy’s fixative (absolute ethanol-chloroform-glacial acetic acid, 6/3/l). After dehydratation in absolute ethanol and 1-butanol, the tissue was embedded in paraffin wax and 7.5 pm sections were mounted on uncoated slides. Deparaffinized sections were
TABLE
510
20
TIME
510
(min)
TIME
Fig. 2. Time course of the conversion of [ 3H]cholesterol to [3H]pregnenolone by mitochondria. A: From 21.5-day-old fetal testes. B: From 18.5-day-old fetal ovaries precultured for 48 h in the presence of (Bu),cAMP (1 mM). In parentheses: number of experiments. Each assay was performed in duplicate.
rehydrated and rinsed with PBS and exposed for 30 min to 10% ovalbumin to block non-specific reactions. The sections were then incubated for 2 h in appropriate dilutions of anti-P-450,,,-IgG diluted in PBS, pH 7.4 containing 0.1% ovalbumin. After washing in PBS, the sections were incubated for 30 min in biotinylated goat secondary antibody, rinsed again in PBS and incubated for 30 min in avidin-biotin-peroxidase complex. Peroxidase was revealed in diaminobenzidine, 0.05% in PBS containing 0.035% hydrogen peroxide. Finally, the sections were rinsed in distilled water, dehydrated and mounted. Sections of fetal testes were included in each series to serve as positive control of the reaction. Negative controls were obtained with IgG from non-immune serum, the
1
CONVERSION
OF (‘HICHOLESTEROL
Each experiment
was performed
in duplicate
No. of testes
Protein content (pg/testis)
Expt. 1
40
134
Expt. 2
42
141
Expt. 3
38
160
TO [‘HJPREGNENOLONE
BY TESTES
FROM
21.5-DAY-OLD
FETUSES
(values in parentheses).
[ 3Hlpregnenolone
20
(min)
(dpm/mg
protein
Crude homogenate
Mitochondrial fraction
3,350 (3,250-3,470) 2,330 (2,560-2,100) 5,227 (5,060-5,395)
32,580 (33,5Ot-31,660) 15,785 (14,930-16,640) 35,832 (30,150-41,515)
per min) Microsomal fraction 448 487 (540-435)
Cytosolic fraction ND ND
115
specific staining
antibody diluted down to extinction or exhausted by an excess of bovine antigen. All reactions were performed 4%C room temperature.
of Pat
spironolactone enhanced the production of pregnenolone and the production was found to increase with culture time. Spironolactone also enA
(a)
R
C
D
F
E
G
Protein determination Protein content of homogenates and of mitochondrial pellets was determined according to the technique described by Bradford (1976). Bovine serum albumin was used as standard.
70
451
c
Statistical analysis All values are means f SEM. Results were analyzed by the Student t-test or by one-way analysis of variance (ANOVA) (Scheffe F-test).
.
5
E
Results Pregnenolone detected by radioimmunoassay in vitro A first series of experiments was performed with testes from 16.5-day-old fetuses in vitro. The production of pregnenolone was undetectable in control media (n = 4). Therefore further metabolism of pregnenolone was inhibited with spironolactone, an inhibitor of 17ol-hydroxylase (Cigorraga et al., 1978) and trilostane, an inhibitor of 3fl-hydroxysteroid dehydrogenase (Sir@-Asa et al., 1982). For each fetus, one gonad was cultured in the presence of spironolactone and the contralateral testis in the presence of spironolactone and trilostane. The production of pregnenolone reached 15 + 2 pmol/gonad per 24 h (n = 3) in the presence of spironolactone and 50 + 3 pmol/gonad per 24 h (n = 3) in the presence of both inhibitors (p < 0.01, paired test). Then, ovaries from l&S-day-old fetuses were explanted for 48 h. Pregnenolone produced in the media every 24 h and pregnenolone content of the ovaries at the end of the culture were assayed. No production of pregnenolone was noted during the 2 days of culture in basal medium (M199) (n = 2) and in the media supplemented with spironolactone (n = 10) or with spironolactone and trilostane (n = 20). Pregnenolone was not detected in the corresponding organ extracts. When the fetal ovaries were cultured in the presence of (Bu),cAMP (Fig. l), a very low production of pregnenolone was detectable in the media and in the organ extracts. The addition of
(b)
A
0
C
D
E
Fig. 3. Immunoisolation of cytochrome P-454,,. For each sample, proteins (150 pg) were separated by SDS-polyacrylamide gel electrophoresis and electroblotted onto nitrocellulose. Cytochrome P-450,, was immunoisolated as described in Materials and Methods. (a) From fetal gonads: autoradiograms of purified cytochrome P-450,,, (A), homogenate of fetal heart (B), homogenates of ovaries from 18.5-day-old fetuses cultured for 48 h in the presence (C) or absence (D) of (Bu),cAMP, testes from 21.5-day-old fetuses: crude homogenate (E), mitochondrial fraction (F), cytosolic fraction (G). Arrow on the right indicates position of purified antigen; molecular weight (m wt) standards are indicated on the left. (b) From fetal and neonatal ovaries: autoradiograms of purified cytochrome P-450,,, (A), homogenates of ovaries from 10.5-day-old neonates cultured in the presence (B) or absence (C) of LH, homogenates of ovaries from 18.5-day-old fetuses cultured in the presence (D) or absence (E) of (Bu),cAMP. In order to enhance the sensitivity, immunoblotting was followed by treatment of the sheet with Amplify spray and exposure to X-Omat film.
117
hanced the level of ovarian pregnenolone content. An additive effect was observed when trilostane was added to medium supplemented with (Bu),cAMP and spironolactone. In these experimental conditions, no pregnenolone could be detected in media where control tissues (Mtillerian ducts and mesonephros) were cultured. Lastly, a series of six experiments was performed to study the effect of short-time exposure to (Bu),cAMP on the pregnenolone production. Ovaries from 18.5day-old fetuses were explanted for 24 h with spironolactone, trilostane and (Bu)~cAMP and the media were removed every 6 h and assayed. At t6 and t12, no pregnenolone could be detected in the media. At t18, pregnenolone production was 1.33 k 0.12 pmol/6 h/ovary and at t24 reached 1.63 + 0.30 pmol/6 h/ovary. Side-chain cleavage activity in isolated mitochondria (I) Side-chain cleavage activity in fetal testes. First. a series of six experiments was performed to study the time course of pregnenolone production. About 40 testes from 21.5day-old fetuses were pooled and the mitochondrial pellet was prepared as described in Materials and Methods. Aliquots of the mitochondrial fraction corresponding to 1.78 + 0.18 testes were incubated with radioactive and non-radioactive cholesterol and the production of pregnenolone was evaluated for 5, 10 and 20 min. The rate of production was linear from 0 to 20 min and declined thereafter (Fig. 2A). Second, the effect of increasing tissue concentration on the see activity was measured after 10 min of incubation in various dilutions of the mitochondrial pellet corresponding to 1.16-3.4 testes. The amount of pregnenolone produced was proportional to the tissue concentration (r = 0.999). In a third series of experiments, the activities per unit protein were compared in crude homogenate, mitochondrial pellet (7700 x g), microsomal pellet (100,000 X g) and cytosolic super-
natant (100,000 X g) (Table 1). The activity was primarily located in the mitochondrial fraction. Lastly, when mitochondria were incubated with aminoglutethimide, an inhibitor of see activity (Gower, 1974) conversion of [ 3H]cholesterol into [3H]pregnenolone was inhibited to a great extent (98.1%). (2) Side-chain cleavage activity in neonatal ovaries. Side-chain cleavage activity was studied in ovaries from 8.5 and 9.5-day-old neonates. Conversion of [ 3H]cholesterol to [ 3H]pregnenolone was observed at both stages. The production increased between 8.5 and 9.5 days (260 and 1610 dpm/lO mm/ovary, respectively). (3) Side-chain cleavage activity in fetal ovaries. First, experiments were performed on ovaries from 18.5- and 20.5-day-old fetuses. 48-60 ovaries were pooled and mitochondria isolated. Aliquots of the mitochondrial pellet containing about five ovaries were incubated for 5, 10 and 20 min. In all experiments no conversion of [3H]cholesterol to [ 3H]pregnenolone could be detected; the values observed were not different statistically from blank values obtained at time 0 of incubation (n = 4). Then experiments were performed on ovaries from 18.5-day-old fetuses cultured for 48 h in Ml99 in the presence or absence of (Bu),cAMP. For each experiment, 52-58 ovaries were pooled and mitochondria isolated. Aliquots of the mitochondrial pellet corresponding to about five ovaries were incubated for 5, 10 and 20 min. When the ovaries were cultured in the control medium no production of [ 3H]pregnenolone could be detected (n = 4). Therefore, after a pretreatment for 48 h with (Bu),cAMP, see activity was observed in the mitochondrial pellets (Fig. 2B). Immunoblot analysis and cytochemistry of the cytochrome P-450,,, The specific content of immunoreactive cytochrome, P-450,,, was studied in testes from 21.5day-old fetuses, ovaries from 10.5-day-old
Fig. 4. Immunolocalization of cytochrome P-450S, enzyme: testes from 14.5day-old fetus (A), 15.5day-old fetus (B), 18.5day-old fetus (C, G); ovary from 18.Sday-old fetus, precultured for 48 h in the presence of (Bu),cAMP (E); adrenal from 18.5day-old fetus (F); ovaries from 10.5-day-old neonates without preculture (D) and after treatment in vitro with LH for 48 h (H). Bars: 100 pm in A, B, C and E; 25pmin D, F,G and H.
neonates and ovaries from 18.5-day-old fetuses. Fetal ovaries were cultured for 48 h in the presence or absence of (Bu),cAMP and neonatal ovaries with and without LH (100 ng/ml). The experiment was repeated 3 times. The molecular weight of the cytochrome, as estimated from the position of standards was 50,000. An additional band present in all lanes with an apparent M, of less than the 45 kDa marker probably represents an artifact. Immunoreactive cytochrome P-450,,. was detectable both in the crude homogenate and in the mitochondrial pellet of testes, but absent in the cytosolic supernatant (Fig. 3~); autoradiographic band was undetectable in blots of control fetal ovaries (Fig. 3a and b) and fetal heart (Fig. 3a), but could be revealed in (Bu),cAMP-pretreated ovaries (Fig. 3a and 6). lmmunoreactive cytochrome P-450,,, was present in blots of neonatal ovaries and slightly enhanced by treatment with LH (Fig. 36). The immunohistochemical staining of P-450S,, was positive in the interstitial cells of 18.5-day-old fetal testes; immunoreactivity absent on day 14.5 could be detected on day 15.5 and the reaction was greatly enhanced on day 18.5 (Fig. 4A, B, C and G). The technique also showed positive staining of the cytoplasm in adrenocortical cells of l&5-day-old fetuses; adrenal medulla cells showed no such staining (Fig. 4F). Controls were performed with fetal testes included non-immune rabbit IgG, dilution of the antibody down to extinction of staining and presaturation of the antibody with purified bovine P-450,,, antigen or rat liver acetone powder (not shown). In postnatal ovaries, a faint reaction was detected in the cytoplasm of a few interstitial cells. The reaction was extended to more interstitial cells in ovaries cultured in the presence of LH (100 ng/ml) (Fig. 40 and H). However, no reaction was observed in sections of ovaries from 18.5-day-old fetuses cultured for 48 h in the presence or absence of (Bu),cAMP (Fig. 4E). Discussion
This study was aimed at evaluating the cholesterol side-chain cleavage activity in rat male and female gonads during fetal development. Our results indicate that the see activity was detectable
in fetal testes, but not in fetal ovaries. Production of pregnenolone was observed in media where testes were cultured in the presence of spironolactone and trilostane. Absence of pregnenolone in control media can be explained by an active metabolism of this steroid to testosterone since this latter one is extensively produced during fetal life (Picon, 1976). In contrast, pregnenolone was not detectable in female media, even in the presence of the inhibitors. This finding is consistent with the absence of conversion of radiolabelled cholesterol into pregnenolone. Indeed it should be noted that see activity was detected in mitochondria isolated from fetal testes and from ovaries of 8.5-day-old neonates. Using an immunoblotting technique, we detected cytochrome P-450,,, in lysates of these organs. Its absence in fetal ovaries could be explained by a lack or a very low expression of the enzyme. It might be suggested that the synthesis of cytochrome P-450S,c can be the ratelimiting step in ovarian steroidogenesis. In human fetuses, P-450,,, mRNA was not detected in Northern blot of ovarian mRNA unless a long exposure of the autoradiograph of dot-blots of the samples (Voutilainen and Miller, 1986). Our immunohistochemical study showed, for the first time, the localization of the enzyme in Leydig cells from 15.5-day-old male fetuses onwards, In female gonads, some interstitial cells were immunostained in ovaries as early as 10 days after birth. Progesterone production in vitro begins to be detectable in 4-day-old ovaries cultured for 48 h (Funkenstein et al., 1980) and it can be assumed that the interstitial immunoreactive cells represent the only source of de novo progesterone production. In previous studies performed on older female rats, the localization was extended to numerous interstitial cells and to some of the thecal internal cells (Bitzur and Orly, 1989; Le Goascogne et al., 1989). Our study shows that LH is effective in extending immunostaining to more interstitial cells in neonatal ovaries. This induction in cytochrome P-450,,, synthesis suggests that LH receptors are localized on interstitial cells at this stage of development. Treatment of the fetal ovaries (Bu),cAMP stimulated a synthesis of pregnenolone, but the values of secretion and content were found to be low. These results could be explained by a
119
metabolism of pregnenolone. The addition of spironolactone and trilostane increased both the production and the ovarian content of pregnenolone. In fetal rat ovaries, 3/?-hydroxysteroid dehydrogenase activity was detected and enhanced with the nucleotide (Picon et al., 1988). On the other hand, conversion of progesterone into androgens has led to contradictory results (Noumura et al., 1966; Weniger et al., 1984) and the presence of an 17a-hydroxylase/desmolase activity in unstimulated fetal ovaries cannot be stated. Therefore an inductive effect of the nucleotide on this enzyme in our experiments is not excluded. Recent data demonstrate that theCal-interstitial cells from immature rat ovaries lack the 17ol-hydroxylase enzyme which can be induced by LH in serum-free medium (Magoffin, 1989). The time course study of (Bu),cAMP on pregnenolone synthesis displayed a lag period of 18 h compatible with a chronic effect including transcriptional activation of the gene(s) for steroidogenesis rather than rapid mobilization of cholesterol to the inner mitochondrial membrane. The conversion of [ 3H]cholesterol into pregnenolone was also obtained in mitochondria from fetal ovaries pretreated with the nucleotide and cytochrome P-450,,, could be isolated in immunoblots as well. So far, in rats, the regulation of P-450,,, biosynthesis has been studied in immature ovarian cells in culture. Thus, the maintenance of the cytochrome can be obtained in the granulosa cells with follicle-stimulating hormone (FSH) (Trzeciak et al., 1986) or with et al., 1986; Zlotkin (Bu),cAMP (Trzeciak et al., 1986) and in the interstitial cells with LH (Goldring et al., 1986). No follicular structures can be observed in 18.5day-old fetal ovaries and unfortunately the immunostaining technique failed to localize the cytochrome P-450,, , which could be expressed at a low level in most somatic cells. The present study demonstrates that the nucleotide was able to induce pregnenolone synthesis in fetal ovaries, but not in other mesodermic non-steroidogenic tissues, which indicates the fetal female gonads are committed for expressing the first step of steroidogenesis. In male gonads, testosterone production (Picon, 1976) and differentiation of human chorionic gonadotropin (hCG)-LH receptors (Gangnerau et al., 1982) start at the same time. However, the
steroidogenesis in male indifferentiated gonads was shown to occur in serum-free medium without gonadotrophic hormone (Gangnerau and Picon, 1987). The appearance of actively steroidogenic cells in ovaries during postnatal life is induced by as yet unknown mechanisms. Differentiation of LH receptors has been noticed to begin at the same period of development and the question can be raised whether these two aspects of interstitial cell differentiation are independent. Acknowledgements We thank Dr. E. Chambaz (INSERM U244, Grenoble, France) for the anti-P-450,,, IgG and the P-450,,, antigen. This work was supported by a grant from the INSERM (grant No. 854019). References Anderson, C.M. and Mendelson, CR. (1985) Arch. B&hem. Biophys. 238, 378-387. Bitzur, S. and Orly, J. (1989) Endocrinology 124, 1471-1484. Bradford, M.M (1976) Anal. Biochem. 72, 248-254. Cigogarra, S.B., Dufau, M.L. and Catt, K.J. (1978) J. Biol. Chem. 253, 4297-4304.
Defaye. G., Monnier, N., Guidicelli, C. and Chambaz, E.M. (1982) Mol. Cell. Endocrinol. 27, 157-168. Funkenstein, B., Nimrod, A. and Lindner, H.R. (1980) Endocrinology 106, 98-106. Gangnerau, M.N. and Picon, R. (1987) Arch. Androl. 18, 215-224. Gangnerau, M.N., Funkenstein, B. and Picon, R. (1982) Mol. Cell. Endocrinol. 28, 499-512. Goldring, N.B., Farkash, Y., Goldschmidt, D. and Orly, J. (1986) Endocrinology 119, 2821-2832. Gower, D.B. (1974) J. Steroid B&hem. 5, 501-523. Hall, P.F. (1966) Endocrinology 78, 690-698. Hall, P.F. and Koritz, S.B. (1965) Biochemistry 4, 1037-1043. Laemmli, U.K. (1970) Nature 227, 680-685. Lamprecht, !%A., Kohen, F., Ausher, J., Zor, U. and Lindner, H.R. (1976) J. Endocrinol. 68, 343-344. Le Goascogne, C., Sananes, N., Gouezou, M., Baulieu, E.E. and Robe], P. (1989) J. Reprod. Fertil. 85, 61-72. Magoffin, D.A. (1989) Endocrinology 125, 1464-1473. Mersmann, H.J., Goodman, J., Houk, J.M. and Anderson, S. (1972) J. Cell Biol. 53, 335-347. Morissey, M. (1981) Anal. B&hem. 117, 307-310. Noumura, T., Weisz, J. and Lloyd, C.W. (1966) Endocrinology 78, 245-253. Picon, R. (1976) J. Endocrinol. 71, 231-238. Picon, R., Pelloux, M.C., Benhaim, A. and Gloaguen, F. (1985) J. Steroid Biochem. 23, 995-1006.
120 Picon, R., Darmoul, D., Rouiller, V. and Duranteau, L. (1988) J. Steroid B&hem. 31, 839-843. Siebers, J.W., Peters, F., Zenzes, M.T., Schmidtke, J. and Engel, W. (1977) J. Endocrinol. 73, 491-496. Singh-Asa, P., Jenkin, G. and Thorbum, G.D. (1982) J. Endocrinol. 92, 2055212. Smeaton, T.C., Arcondoulis, D.E. and Steele, P.A. (1975) Steroids 26, 181-192. Smith-White, S. and Ojeda, S.R. (1981) Endocrinology 109, 152-161. Towbin, H., Staehelm, R. and Gordon, J. (1979) Proc. Natl. Acad. Sci. U.S.A. 76, 4350-4354.
Trzeciak, W.H., Waterman, M.R. and Simpson, E.R. (1986) Endocrinology 119, 323-330. Voutilainen, R. and Miller, W.L. (1986) J. Clin. Endocrinol. Metab. 63, 114551150. Warren, D.W., Haltmeyer, G.C. and Eik-Nes, K.B. (1973) Biol. Reprod. 8, 560-565. Waterman, M.R. and Simpson, E.R. (1985) Mol. Cell. Endocrinol. 39, 81-89. Weniger, J.P., Chouraqui, J. and Zeis, A. (1984) J. Steroid B&hem. 21, 347-349. Zlotkin, T., Farkash, Y. and Orly, J. (1986) Endocrinology, 119, 2809-2820.
