!wl/NwlrIr rrt1rl (‘elllrlu~ Olrloc~r;,lolo~~. YY ( I YY4) 63-71

(: 1YY4 Else\,& Science Ireland. Ltd. O.iO3-7207/Y4/$07.OO

MCE 03142

Rapid modulation of ovarian 3P-hydroxysteroid dehydrogenase/A”-A” isomerase gene expression by prolactin and human chorionic gonadotropin

in the hypophysectomized rat

C&line Martel, Donald Gag& Jacques Couet, Yvan Labrie, Jacques Simard and Fernand Labrie M~iical R~ser~ch Cm~ncil Group m Moleculur Endocrirmlogy, C’HUL Revetwch Cmtrr oml Lrr/,ul Utlic~c,r.vity. 270.5 b’orrlec~tr~tl Luwicr,

Ste. k-c& Qlcdx~c GIV X2. Cullatlrr

(Received 2 June lYY3: accepted 20 September lYY3)

Key kvrds: Strroidogenesis: .ij%Hydroxysteroid dehydrogenase/A’-A’ isomerase: Ovary: In situ hybridization: Luteinizing hormone/human

chorionic gonadotropin; Prolactin

Summary

In order to better understand the role of prolactin (PRL) and luteinizing hormone (LH) on progesterone

biosynthesis in the ovary, we have investigated the time course (I-9 days) of the effect of PRL and human chorionic gonadotropin (hCG) on ovarian 3P-hydroxysteroid dehydrogenase/A”-A4 isomerase (3/3-HSD) expression in the hypophysectomized rat. As evaluated by quantitative in situ hybridization using a “S labelled type I 3/?-HSD cDNA probe, the administration of hCG for 2, 3 and 9 days induced increases of 63%, 145% and 146% above control,

respectively, in 3P-HSD mRNA levels in ovarian interstitial cells. The absence of apparent effect of the gonadotropin in other ovarian cell types could explain the small modulation of ovarian 3/3-HSD protein content and enzymatic activity observed in total ovarian tissue. On the other hand, treatment with PRL caused a rapid decrease

in 3j3-HSD mRNA levels in corpus luteum by 23%, 63%, 76% and 78% (P < 0.01) following 1, 2, 5 and 0 days of treatment, respectively. The short-term inhibitory effect of PRL was also observed on ovarian immunoreactive 3P-HSD protein, as measured by Western blot analysis, and on 3@-HSD activity measured by the conversion of [ “Cldehydroepiandrosterone into [ ‘“Clandrostenedione. The inhibitory effect of PRL on 3,!3-HSD expression and activity is correlated with a progressive decrease in serum progesterone concentration from a pretreatment value of 4.8 nM to levels below the limit of detection (< 0.13 nM) after 7-9 days of treatment with PRL, while serum pregnenolone levels were decreased by only approximately 55% after 9 days. The present findings indicate that the inhibition of 3/?-HSD gene expression and activity in corpora lutea occurrs early in the luteolytic process induced by

PRL and could well play an important role in this process.

Introduction

The production rates of various sex-steroid hor- mones change episodically during the ovarian cycle,

corresponding to the stage of development of follicles and corpus luteum and subsequent plasma concentra- tions of pituitary gonadotrophins. Whereas follicle- stimulating hormone (FSH) is the prime inducer of ovarian follicle maturation and is responsible for the development of granulosa cell responsiveness to luteinizing hormone (LH) and prolactin (PRL), LH stimulates the biosynthesis of androgens and is the most important hormone involved in the breakdown of the follicular wall, which results in ovum release. On

* Corresponding author. Tel.: (418) 654-2704; Fax: (418) 654-2735.

SSDI 0303-7207(93)E0252-P

the other hand, the PRL, which has been associated with some reproductive disorders such as anovulation, infertility, and amenorrhea (Schlechte et al., 1980), seems to be involved directly at the ovarian level for

the growth and development of the follicle and mainte- nance of luteal function (see for review McNeilly et al., 1982). However, the precise mechanisms by which PRL acts remain to be resolved.

To obtain a better understanding of sex-steroid hor- mone biosynthesis in the ovary, it is important to know the localization and hormonal regulation of steroido- genie enzymes, particularly in determining which cell types of the ovarian follicle and corpus luteum express these enzymes and how they are regulated. Among the steroidogenic enzymes implicated, 3/3-hydroxysteroid dehydrogenase/A’-A” isomerase, hereafter referred to as 3/3-HSD, plays a key role in ovarian steroidogenesis

bccausc it catalyzes the essential conversion of preg- ncnolonc and other S-ene-3@-hydroxysteroids into pro- gcstcronc as well as into the precursors of all andro- gcns and estrogens (Labrie et al., 1992).

Considering the important role of the enzyme 3p-

HSD in ovarian steroid biosynthesis, we studied the cffcct of short- and long-term treatment (l-9 days)

with PKL or hCG on 3/?I-HSD gene expression, protein content, and activity in the hypophysectomized rat ovary to assess the potential role of this enzyme in the lutcolytic process induced by PRL and in the lu- tcotrophic effect of hCG.

Materials and methods

nrzima/.s Hypophysectomized adult female Sprague-Dawley

rats [Crl:CD(SD)Br] (150-200 g> were purchased from Charles-River Canada (St-Constant, Quebec, Canada) and housed two per cage in a light- (14 h light/day; lights on at 0600 h) and temperature-(22 f l’C> con- trolled environment. The rats received Purina rat chow, oranges, apples, and 5% glucose in 0.9% NaCl ad lihitum.

Ovine PRL (NIDDK-oPRL-19; biopotency, 31 IU/ mg) was generously donated by the National Hormone and Pituitary Program (Baltimore, MD) and adminis-

tered in 0.1% (w/v) BSA, 10% (w/v> polyvinyl pyrroli- done-10 in 0.9% NaCl and 10 mM Tris-HCl, pH 9. Synthetic hCG (Profasi HP) was purchased from Phar- mascience (Montreal, Quebec, Canada), and adminis- tered in 1% gelatin in 0.9% NaCI.

Experimental procedures Adult female rats hypophysectomized 15 days be-

fore the beginning of treatment received twice daily

subcutaneous injections of oPRL (1 mg> or hCG (10 IU) for l-9 days. Prior the administration of oPRL or hCG for l-8 days, animals received the vehicle alone for 8-l days, respectively, in order to treat all animals for a total period of 9 days. The control group received the vehicle alone for 9 days. After killing of the rats by cervical dislocation on the morning of day 10, one ovary from each animal was promptly removed, freed from adhering tissue, weighed, frozen in liquid nitro- gen and stored at -80°C until determination of ovar- ian 3/3-HSD activity and protein content while the other ovary was submerged in 4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.2) for in situ hybridiza- tion.

Hormone measurements Serum pregnenolone and progesterone were mea-

sured by radioimmunoassays (RIA) in duplicate after

diethyl ether extraction followed by chromatography on LH-20 columns using antisera developed and char- acterized in our laboratory (BClanger et al., 1980).

Enzymatic assay of or%arian 3/3-HSD Frozen ovaries were homogenized with a polytron in

phosphate buffer (20 mM KH,PO,, 0.25 M sucrose, 1 mM EDTA, pH 7.5) containing protease inhibitors (1

mM phenylmethylsulfonylfluoride and 5 mg/ ml each of pepstatin A, antipain, and leupeptin) and cen- trifuged for 30 min at 1000 X g. Aliquots of the super-

natant were then incubated for 20 min at 37°C in 0.5 ml phosphate buffer (12.5 mM KH,PO,, 1 mM EDTA, pH 7.5) containing 1 PM [4-“CIDHEA (51 mCi/ mmol) (DuPont, Markham, Ontario, Canada) and 0.8 mM NAD+. The enzymatic reaction was stopped by

chilling the incubation mixture in an ice-water slurry, adding 3 ml of diethyl ether and mixing. The compo- nents were then frozen in a dry ice-ethanol bath. The liquid organic phase was kept while the aqueous phase was extracted again with diethyl ether. The two organic phases were then pooled and evaporated to dryness under a nitrogen stream.

“C-labeled 4-androstene-3,17-dione (A”-dione) pro- duced from [4-13C]DHEA by 3P-HSD activity was then separated by thin layer chromatography (TLC) on 60

F,,, silica gel plates (E. Merck, Darmstadt, Germany) in a toluene: acetone system (4: 1, v/v), and then au- toradiographed for 48 h. The TLC areas corresponding to A4-dione and DHEA were identified on the autora- diographs, scraped and transferred to scintillation vials containing 0.5 ml ethanol to which 10 ml scintillation fluid were added. The radioactivity was measured in a scintillation spectrometer and the rates of product for- mation were calculated and expressed as percentage of the control group. Ovarian protein content was mea- sured by the method of Bradford using BSA as stand- ard (Bradford, 1976).

Ol,arian 3P-HSD protein immunoblotting Ovarian proteins were size-separated on a 5-15%

polyacrylamide gel (1.5 mm thick) and transferred to nitrocellulose filters. Purified human placental 3/3-HSD and BRL protein weight markers were used as positive control and for estimation of molecular size, respec- tively. The blots were treated with wash solution [5% fat-free milk (Carnation)/O.l% Nonidet P-40 in PBS] (3 X 30 min) before incubation for 18 h at 4°C with a 1: 2000 dilution of rabbit antiserum raised against puri- fied human placental 3/3-HSD. The blots were washed three times (30 min each) and incubated for an addi- tional 4 h at 4°C in a 1 : 1000 dilution of “‘I-labeled goat antirabbit immunoglobulin G. After three washes in the same solution, autoradiography was performed at -80°C with intensifying screens and XAR-5 films. The intensity of the signal emitted by [‘*‘I]antibodies

recognizing the 42 kilodalton protein was quantified using an Amersham RAS Image Analyzer System (Amershaml.

Preparation of cDNA probe A rat 3P-HSD type I (ro3PHSD56) EcoRI-Hind111

cDNA fragment (nucleotides + 136 to -t 1266) was used as probe (Zhao et al., 1991). The fragment of 1130 bp was then radiolabelled with [cu-“‘S]dCTPaS to high specific activity (1 X 10” cpm/pg) by the random- primed method (Feinberg and Vogelstein, 1983) and the probe was then purified by Sephadex G-50 column.

In situ hybridization Freshly excised rat ovaries were submerged 4 h in

4% paraformaldehyde in 0.1 M phosphate buffer (pH 7.2), followed by another 4 h immersion in 15% sucrose in 0.1 M phosphate buffer before embedding in O.C.T. compound. Ovaries were then cut into 10 pm-thick sections and mounted onto gelatin-coated glass slides. After a short-term treatment with 0.1% Triton X-100 in 2 :x SSC (20 X SSC: 3 M NaCl, 0.3 M sodium citrate), the sections of ovary were prehybridized at 37°C for 2 h in buffer consisting of 50% (vol/vol) formamide, 5 X SSPE (20 X SSPE: 0.18 M NaCI; 10 mM NaH,PO,, pH 7.4; 1 mM EDTA), 0.1% sodium dodecyl sulfate, 0.1% (wt/vol) BSA, 0.1% (wt/vol) ficoll, 0.1% (wt/ vol) polyvinylpyrrolidone, 200 pg/ml denatured salmon testis DNA. 20 kg/ml poly(A)+ RNA, 4% (wt/vol) dextran sulfate, and 10 mM dithiothreitol.

Thereafter, the heat-denatured 3P-HSD cDNA probe (2 X lOh cpm) was added to prehybridization buffer. After 16-17 h of hybridization at 37°C the sections were washed at 20°C for 120 min in 2 x SSC, 120 min in 1 X SSC, and 60 min in 0.5 X SSC, followed by a wash at 37°C for 60 min in 0.5 X SSC and finally, at 20°C for 60 min in 0.5 X SSC. The sections were then placed against a Kodak X-Omat AR film (East- man Kodak, Rochester, New York) for 3 days. After exposure, the sections were coated with a Kodak NTB-2 liquid photographic emulsion before staining with hematoxylin-eosin. As a negative control, sections of rat ovary were pretreated with 200 pg/ml ribonucle- ase A (Pharmacia) and 150 U/ml ribonuclease T (Pharmacia) in 2 X SSC before hybridization. The in- tensity of the signal emitted was quantified per unit area of corpus luteum or interstitial cells, using an Amersham RAS Image Analyzer System (Amersham) for interstitial cells and a Bioimage Analyser System (Millipore) for luteal cells.

Statistical analysis The results are expressed as means +_ SEM. Statisti-

cal significance was determined according to the multi- ple-range test of Duncan-Kramer (Kramer, 1956).

Results

Modulation of or%arian 3P-HSD mRNA, protein and actiL?ty lersels by PRL

As illustrated in Fig. lA, no significant effect on ovarian 3P-HSD activity was observed within one or two days of treatment with PRL in the hypophysec- tomized rat. However, after 3 days of treatment with PRL, 3P-HSD activity was decreased by 47% (P < 0.01) while the value of this parameter progressively dc- creased at later time intervals to reach maximal 70% to 78% (P < 0.01) inhibition at 7-9 days. As early as one day after PRL administration, 3/3-HSD protein content was decreased by 47% (P < 0.05) while this parameter was decreased by 77% to 88% (P < 0.01) following 3-X days of treatment with PRL (Fig. 1B).

We next investigated the effect of PRL on 3P-HSD mRNA content, as evaluated by in situ hybridization,

OL --- 0 h-p-7-- ~ 1 3 5 7 9 1 3 5 7 -T

DAYS OF TREATMENT WITH PRL

Fig. 1. Modulation by PRL of rat ovarian 3P-HSD activity (A),

protein content (B) and mRNA levels in interstitial cells (C) and in

corpus luteum (D). Adult female rats hypophysectomized 15 days

previously received twice daily injections of PRL (I.0 mg) for l-9

days. Ovarian 3P-HSD activity was assayed hy measuring the rate of

formation of [4- lJC] androstenedione from [4- “C]dehydroepi-

androsterone while protein content was determined by immunoblot analysis. Ovarian 3P-HSD mRNA levels were measured by in situ

hybridization using the ‘“S-random primer-labeled EcoRI-Hirzdlll

cDNA fragment (nucleotides+ I36 to+ 1266) of rat ovary 3P-HSD.

Exposure time (panel B): 72 h. All data are expressed as means?

SEM tn = 3-6) and are represented as percentage of control group. In several case, the SEM is within the data point drawn. *, P < 0.05.

* *. P < 0.01 vs hypophysectomized control.

Fig. 2. X-Ray autoradiograph illustrating the effects of different time of treatment with PRL on ovarian type I 3/3-HSD mRNA levels evaluated by in situ hybridization. Fixed sections of ovaries from (k$f to r@hf, top to boftomf control hypophysectomizcd rats (Control) or ovaries pretreated

with ribonuclease fRNase) or ovaries from rats treated with PRL for 1 day, 2, 3, 5, 7 or 9 dzdys were hybridized to the %labeled rat ovary

3p-HSD cDNA fragment as described in Materials and methods. Magnification: X 15: Exposure time: 3 days.

67

in corpora lutea and in interstitial cells of the rat ovary.

AS illustrated in Fig. lC, no significant change in the

intensity of labelling in interstitial cells was observed

before 6 days of treatment with PRL, while a slight but significant decrease of 37% was observed when the rats

were treated for 6-9 days. On the other hand, it can be ycen in Fig. ID that a l-day treatment with PRL

already caused a 23% (P < 0.01) decrease in 3P-HSD mRNA levels, while after 2-4 days of treatment, 3/3- HSD mRNA levels were 60% (P < 0.01) below the control value and a maximal 87% (P < 0.01) inhibition was observed after 8 days of treatment. When the total

ovarian 3P-HSD mRNA content was measured, a 40% (P < 0.01) decrease was observed following 2-9 days of

treatment with PRL (data not shown). As illustrated in Fig. 2, PRL caused a reduction in ovarian volume mainly related to a decrease in the size and number of

the corpora lutea. It can also be seen that the intensity of labeling was already markedly reduced after 1 day of

PRL treatment while an almost complete disapearence of labeling was reached at 7 days.

Modulation of ocarian _I/%HSD mRNA, protein and actiuity 1eLlels by hCG

The effect of hCG on ovarian 3P-HSD expression and activity was also investigated under the same ex-

perimental conditions (Figs. 3 and 4). As can be seen in Fig. 3A, a small and non-significant 17% and 10% increases of 3P-HSD activity were seen after 3 and 4 days of treatment with hCG, respectively, while a maxi- mal 38% stimulation (P < 0.05) was observed after 5 days of treatment while no significant effect was ob-

served at later time intervals. As illustrated in Fig. 3B, total ovarian 3P-HSD protein content was not signifi- cantly affected by hCG treatment.

While total 3P-HSD activity and protein content

were poorly regulated by hCG, type I 3P-HSD mRNA levels in interstitial cells were markedly stimulated by such treatment (Figs. 3C and 4). As illustrated in Fig. 3C, 3P-HSD mRNA levels in interstitial cells were increased by 1.6-, 2.4-, and 2.6-fold (P < 0.01) after 2, 3

and 4 days of treatment with hCG, respectively, and a plateau of stimulation was observed between 4-9 days of hCG treatment. On the other hand, hCG adminis- tration caused a small but non-significant 24% increase in the intensity of the labeling of corpora lutea (Fig.

3D). Since the stimulatory effect of hCG is specific to

interstitial cells and these cells express only small amounts of 3/?-HSD comparatively to luteal cells, only a slight but non-significant increase of total ovarian 3,8-HSD mRNA content was obtained following hCG treatment (data not shown). In Fig. 4, it can be seen that hCG treatment increased the size and the inten- sity of labelling of the interstitial glands, while no significant change in 3@HSD mRNA was observed in

O+T-rT-i O+F-r-T- 9

DAYS OF TREATMENT WITH KG

Fig. 3. Modulation by hCG of rat ovarian 3P-HSD activity (A),

protein content (B) and mRNA levels in interstitial cells (C) and in

corpus luteum (D). Adult female rats hypophysectomized 15 days

previously received twice daily injections of hCG (10 IU) for l-9

days. Ovarian 3P-HSD activity and mRNA content was assayed like

before (see Fig. 1). Exposure time (panel B): 24 h. All data are

expressed as meansi:SEM (n = 3-6) and are represented as per-

centage of control group. In several case, the SEM is within the data

point drawn. *, P < 0.05, **, P < 0.01 vs hypophysectomized

control.

corpora lutea. Moreover, a progressive increase in ovarian size was noticed as the lenght of treatment with hCG increased.

Effect of PRL and hCG on ocarian weight and on serum pregnenolone and progesterone lecels

Figs. 2 and 4 show that treatment with PRL and hCG result in major changes in ovarian size. It can be

seen in Fig. 5 that ovarian weight was progressively decreased by treatment with PRL, a 25% (P < 0.05) decrease being observed after 4 days of treatment. The atrophy of the ovary was more important as the length of treatment increased and a maximal inhibition of 50% (P < 0.01) was seen after 9 days of treatment with PRL. On the other hand, hCG treatment caused a significant increase in ovarian weight from 1.6fold (P < 0.011 after 2 days of treatment to 2.1-fold (P < 0.01) after 9 days of treatment.

In order to obtain more direct and physiological information about the hormonal influence of the ob- served changes in ovarian 3/3-HSD expression and

Fig. 4. X-Ray autoradiograph illustrating the effects of different lime of treatment with hCG on ovarian type I 3p-HSD mRNA levels evaluated

by in situ hybridization. Fixed sections of ovaries from (k$ lo rigk, top to hottorn) control hypophysectomized rats (Control) or ovaries pretreated

with ribonuclease (RNasef or ovaries from rats treated with hCG for I day, 2, 3, 5, 7 or 9 days were hybridized to the 35S-lahrled rat ovary

3&HSD cDNA fragment as described in Materials and methods. Magnification: X 15; Exposure time: 3 days.

250 1 + OPRL **

01 0123456789

Days of treatment Fig. 5. Effects of l-9 days of treatment with hCG (w) or PRL (0) on

ovarian weight (milligrams / 100 g body weight) in hypophysec-

tomized adult rats. *, P < 0.05, **. P < 0.01 vs hypophysectomized

control.

activity, we next measured serum pregnenolone and progesterone levels. As can be seen in Table 1, the effects of PRL and hCG on ovarian weight and 3P-HSD expression and activity translated into qualitatively comparable effects on serum progesterone levels. Pro- gesterone levels were decreased by SO%, 85% and 88% (P < 0.01) after 2, 4 and 6 days of treatment with PRL, respectively, while the same treatment inhibited preg- nenolone levels by 42%, 42% and 58% (P < 0.011, respectively. Longer-term treatment with PRL (7-9 days) decreased serum progesterone concentration from a pretreatment value of 4.8 nM to levels below the limit of detection (< 0.13 nM) which represents

TABLE 1

EFFECTS OF TREATMENT WITH PRL OR HCG ON SERUM

PROGESTERONE AND PREGNENOLONE LEVELS IN HY-

POPHYSECTOMIZED ADULT RATS; SERUM STEROIDS ARE

EXPRESSED IN NMOL/L

Days Prolactin

of PREGN

:I:“,:: (nM)

FROG

(nM)

hCG

PREGN PROG

(nM) (nM)

0 1.9 kO.3 4.x+_ 1.2 1.9+0.3 4.Kk 1.2

1 1.3 kO.2 * 1.7kO.4 ** 2.4 + 0.5 30.2k5.1 **

2 1.1 kO.1 ** 2.4kO.4 ** 2.5 * 0.4 14.7k2.4 **

3 1.0 *0.2 ** 1.5+0.3 ** 3.4* 1.4 10.4k2.6 **

4 1.1 kO.1 ** 0.7*0.2 ** 2.1+0.3 15.4+ 1.6 **

5 1.1 fO.l ** 0.8kO.l ** 2.1 kO.2 9.9k2.0 **

6 0.8 *0.1** 0.6kO.2 ** 2.3 + 0.3 11.0* 1.4 **

7 0.72+_0.01 ** LD ** 1.9+_0.2 9.1*1.8 **

8 0.70t0.02 ** LD ** 2.0+0.1 8.7? 1.1 *

9 0.85+0.04 ** LD ** 2.lkO.2 5.9 f 0.7

LD: Limit of detection: 0.13 nM.

*. P < 0.05, * *, P < 0.01 vs hypophysectomized control (day 0)

more than a 97% inhibition of this parameter, while pregnenolone levels were maximally inhibited by 63% after the g-day treatment with PRL.

On the other hand, as shown in Table 1, treatment with hCG for one day caused a 6.3-fold (V < 0.01) increase in serum progesterone levels while 2, 4 and 6 days of treatment with hCG caused respective 3.1-, 3.2-, and 2.3-fold (P < 0.01) increases of the value of this parameter. Longer-term treatment (7-9 days) had no further effect on progesterone concentrations. On the other hand, serum pregnenolone levels were not affected by treatment up to 9 days with hCG.

Discussion

The present data demonstrate the cell-specific, rapid and direct action of PRL and hCG/LH on 3/3-HSD gene expression in the rat ovary. In a previous experi- ment, we reported the inhibitory effect of a Y-day treatment with PRL on ovarian 3/3-HSD expression and activity as well as the decrease of ovarian weight and serum progesterone levels and the atresia of cor- pus luteum in hypophysectomized rats (Martel et al., 1990a). However, there was some interrogations con- cerning the active role of the PRL-induced inhibitory effect on 3P-HSD expression in the luteolytic process causing the atresia of the ovary. It was not excluded from our initial data (Martel et al., 1990a) that the observed inhibitory effect of PRL on 3P-HSD expres- sion and activity resulted from the important decrease of the size of corpora lutea which highly expressed 3/3-HSD. The present experiment clearly demonstrates that the inhibition of 3P-HSD gene expression and activity in rat corpora lutea occurred early in the luteolytic process induced by PRL even before gross morphologic changes could be detected. In fact, 38- HSD mRNA expression and the corresponding protein content were strongly inhibited in luteal cells after short-term treatment with PRL (l-3 days) while the size of corpora Iutea and ovary were apparently un- changed.

The decrease observed in type I 3@-HSD mRNA and protein content was reflected in 3P-HSD activity after a latency period of 2 days. Although three types of 3@HSD are expressed in the rat ovary (types I, II, and IV) (Simard et al., 1991, 1993; Zhao et al., 1990, 19911, all these isoenzymes cross-react with the 3/F&HSD antiserum used, and no obvious explanation can be offered for the 2-day delay between the fall in 3P-HSD protein levels and activity. Even if total ovarian enzy- matic activity apparently was not changed during the first two days of treatment with PRL, the observed 50-65% decreased in progesterone levels could be explained, at least in part, by the concomitant 32-42% decrease in pregnenolone levels. The PRL-induced decrease in pregnenolone levels could result from the

inhibition of the activity of the enzyme P45Oscc, which

is responsible for the conversion of cholesterol to preg- nenolone, or it might result from the inhibition of cholesterol movement to mitochondrial cytochrome P45Oscc (Cohn et al., 1992). As 3P-HSD activity was progressively inhibited by PRL (3 days of treatment or more), the decrease in progesterone levels becomes much more important than that observed in preg- nenolone levels, thus indicating the active role of the 3P-HSD in this process. The present findings showing the marked inhibitory action of PRL on 3P-HSD ex- pression and activity strongly support the involvement of this enzyme in the luteolysis of the corpora lutea. However, the molecular mechanisms whereby PRL in- duces luteolysis and inhibits 3/3-HSD expression re- main unknown.

Under our experimental conditions, PRL exerts a luteolytic effect and inhibits 3/3-HSD activity as well as progesterone formation. Inhibitory effects of PRL on progesterone secretion have also been described in some studies in small granulosa cells (Veldhuis et al., 1980), in developing follicles of mice (McNatty et al., 19761, in granulosa cells from women at various stages of cycles (McNatty et al., 1974) and in granulosa cells from hypophysectomized rats (Fortune and Vincent, 1986). However, it has been reported that under differ- ent experimental conditions, PRL may exerts lu- teotrophic effects and stimulates progesterone secre- tion or biosynthesis as well as 3P-HSD activity. PRL has been shown to stimulate progesterone production by increasing pregnenolone formation and 3P-HSD activity in rat FSH-primed granulosa cells (Jones et al.,

1983) and in rat granulosa cells as demonstrated by histochemistry (Madej, 1980). Stimulatory effects of

PRL on progesterone secretion have also been de- scribed in dispersed rat luteal cells during early preg- nancy (Wu et al., 1976) and in large porcine follicles (Veldhuis et al., 1981). Moreover, in vivo studies have shown that PRL can exert both luteotrophic and lute- olytic effects depending upon the time at which it was administered and the delay after hypophysectomy (Malven and Sawyer, 1966a,b). The effect of PRL on 3P-HSD expression and activity as well as on proges- terone secretion thus depend of the endocrine status of the animals studied as well as the dose used, the cell population, the stage of differentiation of the ovarian

cell and the time of treatment. The luteotrophic effect of LH/ hCG in ovary as well

as the stimulatory effect of LH/ hCG on 3/3-HSD activity have been shown in porcine thecal cells (Tonetta et al., 1987), in rat and porcine granulosa cells (Jones et al., 1983; Jones et Hsueh, 1982; Chedrese et al., 19901, in mice and rat testes (Hafiez et al., 1971), in luteal cells (Wu et al., 1976; Rothchild, 1981) and in ovary from immature (Yoshida, 1987) and adult (Martel et al., 1990b) female rat. In the present study, we

observed an important stimulation of the expression of 3P-HSD in interstitial cells while the activity and pro- tein content as well as the 3P-HSD mRNA levels in luteal cells were poorly stimulated. The absence of apparent effect of hCG in other ovarian cell types can explain the observed small modulation of total ovarian 3/3-HSD protein content and enzymatic activity. In fact, measurements of 3P-HSD mRNA levels in hy- pophysectomized rat ovary by in situ hybridization, which is a valuable tool in discrimining between cell- specific effect, clearly showed that basal expression of this enzyme in luteal cells is higher than in interstitial cells, this finding being in agreement with previous studies in rat and human ovaries (DuPont et al., 1990, 1992). These data suggest that high levels of basal 3/3-HSD expression and activity in corpora lutea masked the variations in 3/3-HSD mRNA occuring in interstitial cells such that measurements of 3P-HSD activity and protein content in whole ovaries did not show any regulation by hCG. On the other hand, short-term treatment with hCG stimulated proges- terone secretion while 3P-HSD activity and preg- nenolone levels were not changed. In porcine thecal cells, it has been demonstrated that hCG stimulated progesterone secretion while 3P-HSD activity was de- creased (Tonetta et al., 1987). These authors thus speculated about a possible ultra short-loop negative feedback within thecal cells by progesterone accumula- tion on 3/3-HSD activity.

The present data indicate that the inhibition of 3/3-HSD gene expression and activity in corpora lutea occurred early in the luteolytic process induced by PRL and could well play an active role in this process. In addition, this study confirms the opposite cell- specific effects of hCG and PRL on 3P-HSD expres- sion and activity as well as the time-dependent modula- tion of this enzyme by these two pituitary hormones. Moreover, these results could suggest that female in- fertility associated with hyperprolactinemia in women, could well be related, at least in part, to the potent inhibitory effect of PRL on ovarian 3/3-HSD activity and thus on progesterone levels, knowing that proges- terone, which is mainly produced by luteal cells (Csapo et al., 1972; MacDonald et al., 19911, is essential for nidation and maintenance of pregnancy.

References

BClanger, A., Caron, S. and Picard, V. C 19X0) J. Steroid Biochem. 13. 18S-IYO.

Bradford, M.M. (1976) Anal. Biochem. 72, 248-2.54.

Chedrese, P.J.. Luu-The, V., Labrie, F., Juorio, A.V. and Murphy, B.D. (1990) Endocrinology 126, 222X-2230.

Csapo, A.1.. Pulkkinen, M.O., Ruttner. B., Sauvage, J.P. and Wiest, W.G. (1972) Am. J. Obstet. Gynecol. 112, 1061-1067.

DuPont. E., Zhao, H.F.. RhCaume, E., Simard, J., Luu-The, V., Labrie, F. and Pelletier, G. (1990) Endocrinology 127. 1394-1403.

71

DuPont, E., Labrie, F., Luu-The, V. and Pelletier, G. (1992) J. Clin.

Endocrinol. Metab. 74, 994-99X.

Feinberg, A.P. and Vogelstein, B. (1983) Anal. Biochem. 132, 6-13.

Fortune, J.E. and Vincent, S.E. (1986) Biol. Reprod. 35, 84-91.

Hafiez, A.A.. Philpott, J.E. and Bartke. A. (1971) J. Endocr. SO, 619-623.

Jefcoate. C.R.. McNamara, B.C., Artemenko, 1. and Yamazaki, T.

(1992) J. Steroid Biochem. Mol. Biol. 43, 751 -767.

.Jones. P.B.C. and Hsueh. A.J.W. (1982) Endocrinology 1 IO, l663- 1671.

Jones. P.B.C.. Valk, C.A. and Hsueh, A.J.W. (19X3) Biol. Reprod. 2Y, 572-585.

Kramer, C.Y. (1956) Biometrics 12, 307-310.

Labrie, F.. Simard. J., Luu-The. V., Belanger, A. and Pelletier. G.

(1092) J. Steroid Biochem. Mol. Biol. 43. X05-826.

MacDonald, P.C., Dombroski, R.A. and Casey. M.L. (1991) Endocr.

Rev. 12, 372-401.

Madej, E. t 19X0) Acta Histochem. 67, 253-260.

Malven. P.V. and Sawyer, C.H. (1966a) Endocrinology 79, 268-274.

Malven, P.V. and Sawyer, C.H. tl966b) Endocrinology 78, 1259-1263.

Mattel. C.. Lahrie, C.. DuPont, E.. Couet, J.. Trudel. C., Rheaume.

E.. Simard, J.. Luu-The. V., Pelletier. G. and Labrie. F. (1990a)

Endocrinology 127, 2726-2737.

Martel. C., Lahrie, C.. CouEt, J., Du ont. E.. Trudel, C., Luu-The.

a V.. Takahashi, M., Pelletier. G. an Labrie, F. (1990b) Mol. Cell.

Endocrinol. 72, R7-R13.

McNatty, K.P.. Sawers, R.S. and McNeilly. A.S. (1974) Nature 250.

653-655.

McNatty, K.P., Neal, P. and Baker. T.G. (1976) J. Reprod. Fertil. 47. 155-156.

McNeilly, AS., Glasier, A., Jonassen. J. and Howie. P.W. (1082) J.

Reprod. Fertil. 6.5, 559-569.

Rothchild, I. (19x1) Recent Prog. Hormone Res. 37. 1X3-29X.

Schlechte, J.. Sherman. B.. Halmi, N., Van Gilder. J., Chapler, F..

Dolan. K.. Granner. D.. Duello. T. and Harris, C. (19X0) Endocr. Rev. I. 295-308

Simard, J., de Launoit, Y. and Labrie. F. (1991) J. Biol. Chem. 266. 14X42- 14845.

Simard, J., Couet, J., Durocher, F., Labrie, Y., Sanchez, R.. Breton.

N., Turgeon, C. and Labrie. F. (1903) J. Biol. Chem.. in press

Tonetta, S.A.. De Vinna. R.S. and DiZerega, G.S. (1987) J. Steroid

Biochem. 2X, 77-X2.

Veldhuis. J.D., Klase. P.A. and Hammond, J.M. (1980) Endocrino-

ogy 107. 42-46.

Veldhuis, J.D.. Klase. P.A. and Hammond, J.M. (1981) Endocrinol-

ogy 108, 1463-146X.

Wu, D.H., Wies!. W.G. and Enders, A.C. (1976) Endocrinology YX.

137X-1389.

Yoshida, Y. (1987) Nippon Sanka Fujinka Gakkai Zasshi 39, 1537-

1544.

Zhao, H.F., Rheaume, E.. Trudel, C.. Couet, J.. Labrie, F. and

Simard. J. (1990) Endocrinology 127, 3237-3239.

Zhao, H.F., Labrie, C., Simard, J., de Launoit, Y., Trudel, C..

Martel. C.. Rheaume, E., DuPont, E., Luu-The, V.. Pelletier, G.

and Labrie. F. (1991) J. Biol. Chem. 266, 583-593.