0013-7227/91/1296-2933$03.00/0 Endocrinology Copyright © 1991 by The Endocrine Society

Vol. 129, No. 6 Printed in U.S.A.

Tumor Necrosis Factor a Inhibits Gonadotropin Action in Cultured Porcine Leydig Cells: Site(s) of Action* C. MAUDUIT, D. J. HARTMANN, M. A. CHAUVIN, A. REVOL, A. M. MORERA, AND M. BENAHMED INSERM CJF No. 90-08, Groupe de Recherches sur les Communications Cellulaires, Laboratoire de Biochimie, Hopital Sainte Eugenie, Centre Hospitalier Lyon-Sud, 69310 Pierre-Benite and Centre de Radioanalyse, Institut Pasteur (DJH), 69007 Lyon, France

beyond cAMP formation. Furthermore, incubation of Leydig cells with 22R-hydroxycholesterol (5 Mg/ml, 2 h) reversed most of the inhibitory effect of TNF-a on androgen production. Indeed, the TNF-a (20 ng/ml, 72 h) inhibitory effect on testosterone production was limited to about 20% (P < 0.03) in Leydig cells supplied with 22R-hydroxycholesterol. Such a moderate effect of the cytokine in the presence of 22R-hydroxycholesterol compared with that observed when androgen secretion was stimulated with the gonadotropin (up to 90% inhibition) indicate that TNF-a acts by dramatically reducing cholesterol substrate availability in the mitochondria. Such an effect of TNF-a is directly exerted on Leydig cells since TNF-a receptors (dissociation constant «5.4 x 10"10 M) are present in primary cultures of purified porcine Leydig cells. Together, the present findings show that in Leydig cells TNF-a antagonizes the gonadotropin action on testosterone formation predominantly through a decrease in the availability of cholesterol substrate in the mitochondria. (Endocrinology 129: 2933-2940, 1991)

ABSTRACT. In the present study, we have tested the direct effects of tumor necrosis factor-a (TNF-a) on basal and human (h)CG-stimulated testosterone secretion by cultured purified Leydig cells isolated from immature porcine testes. TNF-a reduced (as much as 90% decrease) hCG-stimulated, but not basal testosterone secretion in a dose- and time-dependent manner. The maximal and half-maximal effects were, respectively, 3.75 ng/ml (2.2 x 10"10 M) and 0.66 ng/ml (3.9 x 10"11 M) of TNF-a after 48 h treatment. TNF-a antagonizes the gonadotropin hormonal action by affecting at least two types of biochemical steps. First, TNF-a reduced LH/hCG binding to a maximal decrease of 45% obtained with 2 ng/ml of TNF-a after 48 h of treatment. TNF-a also inhibited (44% decrease) hCG-stimulated cAMP production in optimal conditions (20 ng/ml, 72 h). Second, TNFa significantly (P < 0.001) reduced testosterone secretion stimulated with 8-bromo-cAMP (3 x 10"3 M) in a similar range (86% decrease) to that observed with the gonadotropin. Such an observation indicates that the antigonadotropic action of the cytokine is exerted in a predominant manner at a step(s) located

I

T IS NOW recognized that testicular Leydig cell steroidogenesis is not only under the predominant control exerted by luteinizing hormone (1), but also under a local control resulting from the multiple and complex interactions between different cell types (2-5). The factors potentially exerting such local control on the Leydig cell function include growth factors and cytokines. The potential source of testicular cytokines are the macrophages which represent about 20% of testicular interstitial tissue (6). Among the cytokines produced by the macrophages are interleukin l a (IL-la) and tumor necrosis factor a (TNF-a) (7). IL-la-like activity has been found at a high concentration in the human and the rat Received June 19,1991. Address all correspondence and requests for reprints to: Dr. M. Benahmed, Groupe de Recherches sur les Communications Cellulaires, Laboratoire de biochimie, Hopital Sainte Eugenie, Centre Hospitalier Lyon-Sud 69310 Pierre-Benite, France. * This work was supported by Institut No. National de la Sante et de la Recherche Medicale (INSERM, CJF No. 90-08), Ministere de l'Education Nationale, Fondation pour la Recherche en Hormonologie (FRH No. 699101).

testis (8, 9). In addition, different studies have reported an inhibitory (10, 11) or a stimulatory (12, 13) activity of IL-1 on testicular Leydig cell activities. TNF-a is known for its oncolytic activities but a large number of biological effects on cell metabolism have also been described (14). In the present study, we have studied whether TNF-a may play a role in the Leydig cell steroidogenic activity. By using a model of cultured immature porcine purified Leydig cells, this study investigates the effects of TNFa on immature Leydig cell steroidogenic activity and delineates further the potential biochemical mechanisms underlying the action of the cytokine. Additionally, the present study also aimed to determine whether TNF-a receptors are present on Leydig cells, supporting a direct interaction between the cytokine and the testicular cells.

Materials and Methods Materials Human (h)CG (hCG CR 121 13450 IU/mg) was a gift of Dr R. E. Canfield (Rockford, IL). Dulbecco's modified Eagle's

2933

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 December 2014. at 00:56 For personal use only. No other uses without permission. . All rights reserved.

MECHANISMS OF ACTION OF TNF-a ON LEYDIG CELLS

2934

medium (DMEM) and Ham's F-12 medium were obtained from GIBCO (Grand Island, NY). Collagenase/dispase (cat. 1097 113) and collagenase A (cat. 103 586) were obtained from Boehringer (Mannheim, F.R.G.), human recombinant tumor necrosis factor a (TNF-a) was obtained from Boehringer, and Promega (Promega Biotec, Madison, WI). Indomethacin, 22Rhydroxycholesterol (5-cholestene-3b,22(R)-diol: 22R-hydroxycholesterol), 8-bromo-cyclic AMP (8-bromo-cAMP), trypan blue stain (0.4% wt/vol), prostaglandin E2, forskolin(7/3-acetoxy-8,13-epoxy-la,6j3,9a-trihydroxy-labd-14-en-ll-one) cholera toxin, insulin, transferrin, vitamin E, HEPES (n-[2-hydroxyethyl]piperazine-A/"'-[2-ethanesulfonic acid]), deoxyribonuclease type I (DNase) were purchased from Sigma Chemical Co. (St. Louis, MO). Leydig cell preparation and culture Isolated Leydig cells were prepared from immature porcine testes (2-3 weeks old) by collagenase treatment as described by Mather and Philipps (15) and modified by Benahmed et al. (16). Briefly, decapsulated testes were minced and washed in DMEM/Ham's F-12 medium (1:1). After collagenase dissociation (0.5 mg/ml, 90 min at 32 C), cells were washed by centrifugation (200 X g for 10 min). The pellet was then resuspended and submitted to two successive sedimentations of 5 and 15 min. The crude interstitial cells were recovered from the supernatants and Leydig cells were prepared from this fraction by Percoll gradient centrifugation (16). The purity of Leydig cells was more than 90%, as determined by histochemical 3/3-hydroxysteroid dehydrogenase staining (16,17). Leydig cells were plated in Falcon (Los Angeles, CA) 24-multiwell plates (0.5 X 106 cells per dish) and cultured at 32 C in a humidified atmosphere of 5% CO2, 95% air in DMEM/Ham's F-12 medium (1:1) containing sodium bicarbonate (1.2 mg/ml), 15 mM 4-(2-hydroxyethyl)-l-piperazine ethanesulfonic acid (HEPES), and gentamicin (20 ^g/ml). This medium was supplemented with insulin (2 Mg/m0, transferrin (5 Mg/ml), and vitamin E (10 fig/ ml). At the end of experiments, the culture medium was collected and stored at —20 C until assayed for steroid hormone content. Determination of cell number and viability Culture medium of triplicate dishes of Leydig cells was removed and replaced by trypsin-EDTA. The trypsinized (10 min at 37 C) cell solution was centrifuged to obtain the cell pellet. Cellular viability was assessed using trypan blue stain. Total cell and viable cell number were determined in a hemacytometer. Cellular viability is expressed in terms of percentage of viable cells per dish. For determination of cell number, Leydig cells were detached from the culture dishes with trypsinEDTA and counted in a Coulter counter (Coulter Electronics, Margency, France). Lipoprotein preparation Low density lipoprotein (LDL) (density « 1.019-1.063) was isolated from human plasma by sequential ultracentrifugation (18). The lipoprotein fraction migrated as a homogeneous band on electrophoresis. The concentration of LDL is expressed in micrograms of protein per milliliter of medium (19).

Endo • 1991 Vol 129 • No 6

Leydig cell steroidogenic activity Cultured porcine Leydig cell steroidogenic activity was mainly evaluated through the secretion of testosterone. The main characteristics of this culture system is that the secretion of these steroids in response to LH/hCG remains high and stable for several days and particularly between day 2 and day 6 of culture (for references, see 15). Since in cultured porcine Leydig cells accumulation of unconjugated steroids was close to linear only during the first 4 h (20 and our unpublished data), the steroidogenic capacity of these cells was tested after a 3 h stimulation with hCG on day 6 of culture. For determination of steroidogenic enzyme activities, cultured Leydig cells were incubated with hydroxylated cholesterol derivative which readily passes through cell membranes and can be used as a substrate for the mitochondrial cholesterol side chain cleavage enzyme activity (cytochrome P450scc) (21), or incubated with different steroid substrates, i.e. pregnenolone, dehydroepiandrosterone (DHEA), and androstenedione for the other steroidogenic enzyme activities. Testosterone levels were measured in the culture medium by using a previously reported specific RIA (22). Determination of cAMP in the culture medium Leydig cells were stimulated with hCG for 30 min. At the end of the experiment, the culture media were collected into ethanol (80% vol/vol) at 4 C. After shaking, the supernatants were collected by centrifugation then evaporated to dryness under a stream of nitrogen. After reconstitution with 0.05 M phosphate buffer, pH 6.2, the samples were acetylated and assayed for cAMP using specific RIA (23) (cyclic AMP 126IRIA Pasteur kit, ERIA Diagnostics Pasteur, Marnes, France). 125

I-hCG binding study

Human CG was labeled with 125I by the Iodogen (Pierce Chem. Co.) method (24). Cultured Leydig cells were incubated at 32 C for 4 h with 125 I-hCG («4 X 10"10 M) in a total volume of 1 ml. The nonspecific binding was estimated in the presence of a 500-fold excess of unlabeled hCG (5IU Pregnyl, Organon International, West Orange, NJ). At the end of the incubation, the cells were washed twice with cold PBS/0.2% albumin (PBS/BSA) before solubilization in 0.5 N NaOH/0.4% deoxycholate. Radioactivity was measured in a 7-counter. 125

I-TNF-a binding study

TNF-a was labeled with 125I by the Iodogen method (24) and had a specific activity of 1700 Ci/mmol. Cultured Leydig cells were incubated at 4 C for 4 h with 125I-TNF-a in a total volume of 0.2 ml. At the end of the incubation, cells were washed twice with cold PBS/BSA before solubilization in 0.5 N NaOH/0.4% deoxycholate. The specific binding is the difference between the total amount of radioactivity bound in the absence (total binding) and in the presence of excess (120 nM) unlabeled TNF-a (nonspecific binding). The nonspecific binding was less than 10% of the total binding. Binding data were analyzed by the method of Scatchard (25) and by using the EBDA computer program (26) which were converted to BIOSOFT by G. A. McPherson and distributed by Elsevier-BIOSOFT (Cambridge,

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 December 2014. at 00:56 For personal use only. No other uses without permission. . All rights reserved.

MECHANISMS OF ACTION OF TNF-a ON LEYDIG CELLS U.K.). All experimental data are presented as the mean ± SD of triplicate determinations of steroid production by three replicate cultures within each treatment group. All experiments reported here were repeated at least three times with independent cell preparations. A representative experiment of each series of experiments is presented. Statistical significance between groups was determined by Student's t test using the StatWorks (Hyden and son Ltd, London, U.K.) package on a Macintosh Plus computer. Differences are accepted as significant at P < 0.05. Results

Effects of TNF-a on Leydig cell testosterone secretion To assess the possibility that TNF-a may affect testosterone secretion, immature porcine Leydig cells were cultured in the absence or presence of increasing doses of TNF-a (0.01-15 ng/ml, 72 h) before being acutely (3 h) stimulated with a maximally efficient hCG dose (3 ng/ml) (Fig. 1A). Addition of TNF-a to primary cultures of Leydig cells had no significant effect on basal testosterone formation. By contrast, TNF-a inhibited hCGinduced testosterone production in a dose-dependent manner, with a half-maximal inhibitory dose (IC50) of 0.66 ng/ml (3.9 x 10~u M) and a maximal (82% decrease) effective dose of 3.75 ng/ml (2.2 x 10"10 M). Time course studies of TNF-a action disclosed inhi-

2935

bition of the gonadotropin effect for all time points studied (Fig. IB). As early as 12 h of treatment, TNF-a significantly (P < 0.001) inhibited (45% decrease) hCGstimulated androgen production and a maximal (78% decrease) effect of TNF-a was reached after 48 h of treatment. Addition of TNF-a (20 ng/ml, 72 h) to the Leydig cell culture medium resulted in consistent and significant (P < 0.001) inhibition (up to 90% decrease) of the gonadotropin effect for all doses tested (0.03-9 ng/ml), suggesting that the magnitude of the TNF-a effect may be independent of the concentration of hCG employed (Fig. 1C). Furthermore, the findings in Fig. ID indicate that the inhibitory effect of TNF-a on hCG-induced testosterone production was dependent upon Leydig cell densities (0.06-0.6 X 106 cells per well). While TNF-a was poorly active at low cell density (8 and 24% inhibition for 0.06 and 0.16 X 106 cells per well, respectively), it greatly inhibited hCG-induced androgen production at higher cell density (65 and 82% inhibition for 0.4 and 0.6 X 106 cells per well, respectively). The inhibitory effect of TNF-a on cultured Leydig cells was unaccounted for by a decrease in Leydig cell number and viability (Table 1). That TNF-a did not cause cellular damage was further confirmed by the data shown in Table 2. Indeed, the removal of the cytokine from Leydig cell culture medium resulted in a gradual and progressive recovery of most of the hCG-stimulated testosterone levels. The inhibition of hCG-induced anTABLE 1. Effects of TNF-a on cell number and viability Treatment

Cell number per well (X106)

None TNF-a hCG hCG + TNF-a

0.44 0.43 0.40 0.42

± ± ± ±

Viability (%) 87 ± 4

0.03 0.04 0.02 0.03

89 ±7 90 ± 4 90 ± 6

Leydig cells were cultured in the absence or the presence of TNF-a (20 ng/ml, 72 h) and in the absence or presence of hCG (3 ng/ml, 3 h). The results represent the mean ± SD of three separate determinations in three different dishes. 10'

hCG

1 0

J

1 0

J

( pg/ml)

I O

4

0.06

0.16

0.4

0.6

xio* Cells/well

FiG. 1. Effects of TNF-a on Leydig cell testosterone secretion. Leydig cells were cultured: A, for 72 h with or without increasing concentrations (0.01-15 ng/ml) of TNF-a, then stimulated for 3 h with a maximal efficient dose of hCG (3 ng/ml); B, for the duration indicated (12-72 h) with or without TNF-a (20 ng/ml) then simulated with hCG (3 ng/ ml, 3 h); C, for 72 h with or without TNF-a (20 ng/ml) then stimulated with increasing doses of hCG (0.01-9 ng/ml, 3 h); D, for 72 h at increasing cellular densities (0.06-0.6 x 106 cells per well) with or without TNF-« (20 ng/ml), then stimulated with hCG (3 ng/ml, 3 h). For Fig. 1, A, B, and C cell density was 0.5 x 106 cells per well. The results represent the mean ± SD of three separate determinations in three different dishes.

TABLE 2. Recovery of Leydig cell steroidogenesis after TNF-a removal Treatment hCG hCG + TNF-a

Testosterone (ng/106 cells) DayO

Dayl

Day 2

Day 3

31.2 ± 0.6 1.8 ± 0.2

40.5 ± 0.9 9.1 ± 0.5

38.0 ± 3.4 20.0 ± 0.4

33.3 ± 2 .7 22.5 ± 1.5

Leydig cells were initially cultured for 48 h in the absence or presence of TNF-a (20 ng/ml). The media were then removed, the cells washed and stimulated with hCG (3 ng/ml, 3 h) (day 0), or reincubated with TNF-a-free medium for an additional 24, 48, or 72 h (day 1, 2, or 3, respectively). At the end of each incubation, the cells were stimulated with hCG (3 ng/ml, 3 h). The results represent the mean ± SD of three separate determinations in three different dishes.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 December 2014. at 00:56 For personal use only. No other uses without permission. . All rights reserved.

MECHANISMS OF ACTION OF TNF-a ON LEYDIG CELLS

2936

drogen production was 94, 78, 47, and 32%, respectively, at the end of TNF-a treatment (day 0) and 24, 48, and 72 h after the removal of TNF-a from Leydig cell culture medium (Table 2). TNF-a site(s) of action To further delineate the mechanisms underlying the antagonistic interaction between the cytokine and the gonadotropin, we have examined the effects of TNF-a on several biochemical steps involved in the LH/hCG action. TNF-a reduced in dose- (Fig. 2A) and timedependent (Fig. 2B) manner LH/hCG binding to maximal decrease (45%) obtained with 2 ng/ml (1.2 x 10"10 M) after 48 h treatment. TNF-a also reduced (44% decrease) hCG-stimulated (extracellular) cAMP accumulation (without TNF-a: 13.9 ± 1.3 vs. with TNF-a: 7.8 ± 0.8 cAMP pmol/106 cells, P < 0.002). Furthermore, the inhibitory effects of the cytokine on androgen production were similarly observed whether Leydig cells were stimulated with the gonadotropin or with different pharmacological agents which enhance

cAMP levels including cholera toxin, forskolin, prostaglandin E (PGE2) (82, 92, and 85% inhibition for cholera toxin, forskolin, and PGE2, respectively, P < 0.001) (Fig. 3). These findings suggest that TNF-a probably interfered with cAMP production and/or action. Indeed, the significant (P < 0.001) and dramatic decline (86% decrease) in testosterone secretion was reproduced in TNFa-treated Leydig cells stimulated with 8-bromo-cAMP (3 X 10"3 M, 3 h) (Fig. 4). This indicates that the cytokine also affects Leydig cell steroidogenesis at a biochemical step(s) located beyond cAMP formation. Experiments were subsequently carried out to determine whether TNF-a may affect cholesterol substrate metabolism and/or the activity of the steroidogenic enzymes responsible for steroid biosynthesis. Because of the high dependency of Leydig cell steroidogenesis upon LDL cholesterol substrate, the effects of TNF-a were studied in this cell type incubated with saturating con-

hCG

2

4

TNFa

Cholera

Toxin F o r s k o l i n

PG E2

FIG. 3. Effects of TNF-a on Leydig cell steroidogenesis stimulated with cholera toxin, forskolin, PGE2. Leydig cells were cultured in the absence or the presence of TNF-a (20 ng/ml, 72 h) before being stimulated for 3 h with hCG (3 ng/ml) or cholera toxin (10 jig/ml) or forskolin (5 x 10"5 M) or PGE2 (10~6 M). The results represent the mean ± SD of three separate determinations in three different dishes.

6

(ng/ml)

48

Time

Endo • 1991 Vol 129 • No 6

72

(h)

FIG. 2. Effects of TNF-a on LH/hCG binding. Leydig cells were cultured: A, with TNF-a (0-6 ng/ml, 72 h) before measuring LH/hCG binding; B, for the duration indicated (12-72 h) with or without TNFa (20 ng/ml, 72 h) before evaluating out LH/hCG binding. The results represent the mean ± SD of three separate determinations in three different dishes.

hCG

8-bromo-cAMP

FIG. 4. TNF-a effect on gonadotropin- and 8-bromo-cAMP-stimulated testosterone secretion. Leydig cells were cultured in the absence or the presence of TNF-a (20 ng/ml, 72 h) before being stimulated for 3 h with hCG (3 ng/ml) or with 8-bromo-cAMP (3 X 10"3 M). The results represent the mean ± SD of three separate determinations in three different dishes.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 December 2014. at 00:56 For personal use only. No other uses without permission. . All rights reserved.

MECHANISMS OF ACTION OF TNF-a ON LEYDIG CELLS

TABLE 3. Effects of TNF-a on Leydig cell steroidogenesis in the presence of LDL Testosterone (ng/106 cells) Treatment

-TNF-a 13.0 ± 0.9 23.2 ± 1

hCG hCG + LDL

-l-TNF-a 1.8 ± 0.2 3.3 ± 0.5

availability in the mitochondria rather than to the sterol conversion into androgens. TNF-a receptors in cultured Leydig cells To determine whether the effects of TNF-a reported above are directly exerted on Leydig cells, the following experiments were performed to identify TNF-a receptors on this cell type. Binding experiments were conducted using intact Leydig cells in near confluent monolayer cultures with various concentrations of 125I-TNF-a (0.03-1.94 nM) at 4 C for 4 h. As shown in Fig. 6, TNFa binding was concentration-dependent with a saturation observed at 1 nM. Scatchard plot analysis (25) of these data indicate that Leydig cells have a single class of high affinity binding sites (Fig. 6, inset). The calculated affinity (Kd value) was 5.4 x 10"10 M. Assuming that TNF-a binds as a trimer to its receptor (27), the calculated number of receptors per cell was 1.5 x 103. Discussion The present findings demonstrate that in cultured porcine Leydig cells TNF-a is a potent inhibitor of testosterone production under acute gonadotropin stimulation but not under basal conditions. TNF-a inhibited hCG-stimulated testosterone secretion to a maximal (up to 90%) decrease with an IC50 of 3.9 X 10"11 M after 48 h treatment. Among the other cytokines, namely IL-la, IL-1/3 (our unpublished data) and Interferon-7 (28) tested in this culture system, TNF-a appears to exert the most potent inhibitory action.

Leydig cells were cultured in the absence or the presence of TNF-a (20 ng/ml, 72 h) and incubated with or without LDL (100 tig/ml, 12 h) before being stimulated with hCG (3 ng/ml, 3 h). The results represent the mean ± SD of three separate determinations in three different dishes.

10

8 (0

200

150 0> O

• •

6 •

-TNFa +TNFa

C

50 -

22-R

•

>v •

C

10 Bound (pM)

20 ^

—

—

•

i

100 -

O r0)

Bound/Free (x 100)

centrations of LDL (100 /ug/ml). As shown in Table 3, both in the absence or presence of TNF-a, LDL cholesterol remained effective in terms of its ability to significantly (P < 0.001) increase (about 1.8-fold) hCG-stimulated testosterone formation. However, the addition to Leydig cells of the exogenous (LDL) cholesterol was still unable to prevent the inhibitory action (86% decrease, with or without LDL) of the cytokine on the gonadotropin action. By contrast, the data in Fig. 5 indicate that in Leydig cells incubated with 22R-hydroxycholesterol (a cholesterol substrate derivative which readily passes through cell membranes) the inhibitory effect of TNF-a on testosterone formation, although significant (P < 0.03) was moderate (about 20% of decrease). Furthermore, this slightly inhibitory action of the cytokine on the androgen formation was similarly observed whether cells were incubated with 22R-hydroxycholesterol, pregnenolone, DHEA, or androstenedione suggesting it might be exerted on 17j3-hydroxysteroid dehydrogenase activity. Together, these observations indicate that the potent inhibitory effect of the cytokine on the gonadotropin action occurs more at a step(s) related to cholesterol substrate

2937

P 5

DHEA

A 4

FIG. 5. Effects of TNF-a on steroidogenic enzyme activities. Leydig cells were cultured in the absence or the presence of TNF-a (20 ng/ml, 72 h) before being incubated with 22R-hydroxycholesterol (22-R, 5 Mg/ ml, 2 h) or pregnenolone (P5, 500 ng/ml, 2 h) or DHEA (500 ng/ml, 2 h) or androstenedione (A4, 500 ng/ml, 1 h). The results represent the mean ± SD of three separate determinations in three different dishes.

TNFa concentration

(nM)

FIG. 6. TNF-a receptors in cultured Leydig cells. Leydig cells were incubated with increasing concentrations (0.03-1.94 nM) of labeled TNF-a in the presence or absence of unlabeled TNF-a (120 nM) as described in Materials and Methods. The data were plotted as a saturation curve and treated by Scatchard analysis to calculate binding parameters (inset). B/F, Bound to free ratio.

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 December 2014. at 00:56 For personal use only. No other uses without permission. . All rights reserved.

2938

MECHANISMS OF ACTION OF TNF-a ON LEYDIG CELLS

TNF-a was without significant effect on Leydig cell number or viability. Furthermore, the reversibility of TNF-a action was shown by the capacity of the Leydig cells to recover most of their steroidogenic activity in response to LH/hCG stimulation after removal of the cytokine. This indicates that, in the experimental conditions used in this study, TNF-a action may represent a specific regulatory (noncytotoxic) property of the cytokine. Furthermore, the inhibitory effect of TNF-a on gonadotropin-stimulated testicular androgen production may result from a direct interaction between the cytokine and Leydig cells. Indeed, the effect of the cytokine was studied using purified Leydig cells and recombinant TNF-a. This approach potentially excludes the possibility that the cytokine's inhibitory effect on steroidogenesis, reported here, might be related to a contaminating testicular cell type(s) and/or a contaminating factor(s) in the TNF-a preparation. That TNF-a may directly interact with Leydig cells was further supported by the identification of TNF-a receptors on the testicular cells. Leydig cells specifically bind 125I-TNF-a through a high affinity binding system represented by a single class of receptors with a Kd value of 5.4 X 10~10 M. This is the first report demonstrating the presence of the cytokine receptors on a steroidogenic cell type. Although the concentration of TNF-a present in the testis has not been determined, the IC50 of TNF-a was within the Kd range observed for TNF-a receptors present on porcine Leydig cells. This observation suggests that the effects of TNFa occur within a concentration range that might be expected under physiological conditions. The potential variations of TNF-a receptors during male gonadal development and in different physiological (and pathological) situations as well as their regulation by the major factors controlling Leydig cell functions will be interesting to investigate. However, in addition to the receptor mediated regulatory action of TNF-a on Leydig cells, the possibility exists that the cytokine may also regulate Leydig cell function via another cell type (macrophage, or other testicular cell types). With regard to the mechanisms involved in the inhibitory effect of the cytokine on hCG-induced androgen production, different biochemical steps appear to be affected but with different magnitudes. TNF-a induced changes in the structure and functions of the Leydig cell membrane as shown by its capacity to partially decrease LH/hCG receptors (maximal inhibition of 45%) and hCG-stimulated cAMP production (maximal inhibition of 44%). The similar, dramatic decrease in androgen production (in TNF-a-pretreated Leydig cells) was observed after an acute stimulation with the gonadotropin or with cAMP formation stimulators (e.g. cholera toxin, forskolin, PGE2) indicating that the TNF-a inhibitory action might be related to cAMP generation and/or

Endo • 1991 Voll29«No6

action. In this context, it is of interest to note that testosterone secretion was similarly and dramatically reduced (up to 90%) in TNF-a-treated Leydig cells stimulated either with the gonadotropin or with 8-bromocAMP, suggesting a predominant action of the cytokine at a site(s) located beyond cAMP generation. Furthermore, incubation of Leydig cells with 22R-hydroxycholesterol reversed most of the inhibitory effect of TNF-a (about 20% of inhibition in the presence of 22R-hydroxycholesterol vs. up to 90% of inhibition in the presence of the gonadotropin). These observations suggest that TNF-a antagonizes the gonadotropin hormonal action predominantly by decreasing cholesterol substrate transport and/or availability for cytochrome P450scc activity. Since cholesterol mobilization from the lipid droplets to the mitochondria involves cholesterol ester hydrolysis by cholesterol esterase (29), microfilaments (30), sterol carrier proteins (31), labile regulatory protein(s) (probably the steroidogenic activator polypeptide) (32), and the more recently identified GTP-regulatory protein(s) (33), it is tempting to speculate that these elements may be targets for the regulatory action of TNF-a in Leydig cells. Moreover, unlike some major growth factors, TNF-a appears to activate different intracellular signaling systems, explaining probably the extremely pleiotropic nature of its action. Among the TNF-a activated signaling pathways are protein kinase C, phospholipase A2 (resulting in an increase in arachidonic acid metabolites), and protein kinase A (for references, see 34). Therefore, an important step toward an understanding of the inhibitory effect of TNF-a on the gonadotropin action in Leydig cells would be to establish which of the TNF-a-activated intracellular signaling system(s) is involved, particularly in the decrease in cholesterol substrate availability in the mitochondria. Although we do not yet know which of these intracellular systems are activated by TNF-a in Leydig cells, the increase in arachidonic metabolites, particularly in prostaglandins, is probably not involved in TNF-a action. Indeed, combined treatment of cultured porcine Leydig cells with both TNF-a and indomethacin (10~6 M, 72 h) did not affect the antigonadotropin action of the cytokine (our unpublished data). Studies from other laboratories have also shown that cytokines may regulate testicular steroidogenesis. IL-1/3 has been shown to exert an inhibitory (10, 11) or a stimulatory (12, 13) action on LH/hCG-stimulated testicular androgen production. More recently, in cultured mature purified rat Leydig cells, TNF-a has been reported to exert no effect (35) or a stimulatory effect (13) on the gonadotropin hormonal action. The biochemical mechanisms involved in the stimulatory action on mature rat Leydig cells have not been further studied. The reasons for the differences observed between the data of

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 December 2014. at 00:56 For personal use only. No other uses without permission. . All rights reserved.

MECHANISMS OF ACTION OF TNF-a ON LEYDIG CELLS

Warren et al. (13) and our present findings are unknown. However, these discrepancies could be due to the types of cell cultures (porcine cells vs. rat cells) or the maturity of Leydig cells (immature porcine cells vs. mature rat cells) or the culture conditions, particularly the use of different media and the supplementation with different factors (insulin, transferrin, and vitamin E). Several reports have shown that cytokines modulate also the activity of the steroidogenic cells in the female gonad, namely granulosa and theca cells. TNF-a modulates steroid secretion both in the theca interna (36, 37) and granulosa cells (38-40). With regard to the mechanisms of action of TNF-a in the different steroidogenic cell types, it is of interest to note that the cytokine exerts its regulatory actions at distinct and crucial biochemical steps. Indeed, TNF-a affects the gonadotropin action at the putative second messenger (cAMP) production level (in rat granulosa cells, ref. 38), or cholesterol substrate availability in the mitochondria (in porcine Leydig cells, present study), or cholesterol conversion into pregnenolone (rat theca cells, ref. 36) or the steroidogenic enzymes, 17a-hydroxylase/17:20 lyase (rat theca cells, ref. 37). Whether the in vitro data presented here reflect a potential physiological (or physiopathological) role of TNF-a in the male gonad function requires further studies. In this context, it will be important to determine the site of TNF-a production. Although the presence of TNF-a or its mRNA in the testicular cells has not yet

been demonstrated, such a possibility is plausible in view of 1) the high percentage of macrophages present in the interstitial tissue (6); 2) the identification of immunoreactive TNF-a in granulosa cells (41), an observation which suggests that the male homologue (i.e. Sertoli cells) may be also a source of the cytokine. However, TNF-a may originate from the circulatory system and particularly in pathological conditions, i.e. septic shock, cachexia, cerebral malaria (14) where high plasma levels of the cytokine have been detected (42). In summary, by using a model of cultured purified immature porcine Leydig cells, this study has demonstrated that TNF-a antagonizes the gonadotropin action not only at the cell membrane level(s) but also in a more predominant manner, at the levels of cholesterol substrate availability in the mitochondria. The study has shown also the presence of TNF-a receptors on Leydig cells, supporting a direct interaction between the cytokine and the testicular cells.

Acknowledgments We are grateful to Dr. M. G. Forest for testosterone antisera, Mr. E. Villard and Mr. P. Bouteille for providing us with porcine testes. We thank Dr. D. Lawrence for reading the manuscript.

2939

References 1. Catt KJ, Harwood RN, Clayton RN, Davies TF, Chan V, Katikineni M, Nozu K, Dufau ML 1980 Regulation of peptide hormone receptors and gonadal steroidogenesis. In: Greep RO (ed) Recent Progress in Hormone Research. Academic Press, New York, vol 36:557-622 2. Benahmed M, Sordoillet C, Chauvin MA, Morera AM 1990 Local regulators of testicular function. In: Bouchard P, Haour F, Franchimont P, Schatz B (eds) Recent Progress on GnRH and Gonadal Peptides. Elsevier, Paris, pp 383-405 3. Bellve AR, Zheng W 1989 Growth factors as autocrine and paracrine modulators of male gonadal functions. J Reprod Fertil 85:771-793 4. Risbridger GP, de Kretser DM 1989 Paracrine regulation of the testis. In: Burger HG, de Kretser DM (eds) The Testis. Raven Press, New York, pp 225-268 5. Skinner MK 1991 Cell-cell interactions in the testis. Endocr Rev 12:45-77 6. Miller SC, Bowman BM, Rowland HG 1983 Structure, cytochemistry, endocytic activity, and immunoglobulin (Fc) receptors of rat testicular interstitial-tissue macrophages. Am J Anat 168:1-13 7. Le J, Vilcek J 1987 Tumor necrosis factor and interleukin 1: cytokines with multiple overlapping biological activities. Lab Invest 56:234-248 8. Khan SA, Soder 0, Syed V, Gustafsson K, Lindh M, Ritzen EM 1987 The rat testis produces large amounts of an interleukin-1like factor. Int J Androl 10:495-503 9. Khan SA, Schmidt K, Hallin P, DiPauli R, DeGeyter CH, Nieschlag E 1988 Human testis cytosol and ovarian follicular fluid contain high amounts of interleukin-1-like factor(s). Mol Cell Endocrinol 58:221-230 10. Calkins JH, Sigel MM, Nankin HR, Lin T 1988 Interleukin-1 inhibits Leydig cell steroidogenesis in primary culture. Endocrinology 123:1605-1610 11. Fauser BCJ, Galway AB, Hsueh AJW 1989 Inhibitory actions of interleukin-1/3 on steroidogenesis in primary cultures of neonatal rat testicular cells. Acta Endocrinol 120:401-408 12. Verhoeven G, Cailleau J, Damme JV, Billiau A 1988 Interleukin-1 stimulates steroidogenesis in cultured rat Leydig cells. Mol Cell Endocrinol 57:51-60 13. Warren DW, Pasupuleti V, Lu Y, Platler BW, Horton R 1990 Tumor necrosis factor and interleukin-1 stimulate testosterone secretion in adult male rat Leydig cells in-vitro. J Androl 11:353360 14. Beutler B, Cerami A 1989 Cachectin (tumor necrosis factor): a macrophage hormone governing cellular metabolism and inflammatory response. Endocr Rev 9:57-66 15. Mather JP, Phillips DM 1984 Primary culture of testicular somatic cells. In: Barnes DW, Sirbasku DA, Sato GH (eds) Methods for Serum-Free Culture of Cells of the Endocrine System. Alan R Liss, New York, pp 29-45 16. Benahmed M, Morera AM, Chauvin MA, De Peretti E 1987 Somatomedin C/insulin growth factor I as a possible intratesticular regulator of Leydig cell activity. Mol Cell Endocrinol 50:69-77 17. Mendelson C, Dufau M, Catt KJ 1975 Gonadotropin binding and stimulation of cyclic adenosine 3':5'-monophosphate and testosterone production in isolated Leydig cells. J Biol Chem 250:88188823 18. Havel RJ, Eder HA, Bragdon JH 1955 The distribution and chemical composition of ultracentrifugally separated lipoproteins in human serum. J Clin Invest 34:1345-1353 19. Lowry OH, Rosenbrough NJ, Farr AL, Randall RJ 1951 Protein measurement with the folin phenol reagents. J Biol Chem 193:265275 20. Orava M, Haour F, Leinonen P, Ruokonen A, Vihko R 1985 Relationships between unconjugated and sulphated steroids in porcine primary Leydig cell culture. J Steroid Biochem 22:507-512 21. Arthur JR, Blair HAF, Boyd GS, Mason JI, Suckling KE 1976 Oxidation of cholesterol and cholesterol analogues by mitochondrial preparations of steroid-hormone-producing tissue. Biochem J 158:47-51

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 December 2014. at 00:56 For personal use only. No other uses without permission. . All rights reserved.

2940

MECHANISMS OF ACTION OF TNF-a ON LEYDIG CELLS

22. Forest MG, Cathiard AM, Bertrand J 1973 Total and unbound testosterone levels in the newborns and normal hypogonadal children: use of a sensitive radio-immunoassay for testosterone. J Clin Endocrinol Metab 36:1132-1142 23. Harper JF, Brooker G 1975 Femtomole sensitive radioimmunoassay for cyclic AMP and cyclic GMP after 2' 0 acetylation by acetic anhydride in aqueous solution. J Cyclic Nucleotide Protein Phosphor Res 1:207-218 24. Tyszywski GP, Knight LC, Kornecki E, Srivastava S 1983 Labeling of platelet surface proteins with 125I by the iodogen method. Anal Biochem 130:166-170 25. Scatchard G 1949 The interaction of proteins with small molecules and ions. Ann NY Acad Sci 51:660-672 26. Munson PJ, Rodbard D 1980 LIGAND: a versatile computerized approch for characterization of ligand-binding systems. Anal Biochem 107:220-239 27. Jones EY, Stuart DI, Walker NP 1989 Structure of tumor necrosis factor. Nature 338:225-228 28. Orava M, Voutilainen R, Vihko R 1989 Interferon-7 inhibits steroidogenesis and accumulation of mRNA of the steroidogenic enzymes P450scc and P450cl7 in cultured porcine Leydig cells. Mol Endocrinol 3:887-894 29. Albert DH, Ascoli M, Puelt D, Coniglio JG 1980 Lipid composition and gonadotropin-mediated lipid metabolism of the M5480 murine Leydig cell tumor. J Lipid Res 21:862-867 30. Strott CA 1990 The search for the elusive adrenal steroidogenic "regulatory" protein. Trends Endocrinol Metab 1:312-314 31. Vahouny GV, Dennis P, Chanderbhan R, Fiskum G, Noland BJ, Scallen TJ 1984 Sterol carrier protein2 (SCP2)-mediated transfer of cholesterol to mitochondria inner membrane. Biochem Biophys Res Commun 122:509-517 32. Pedersen RC, Brownie AC 1987 Steroidogenesis-activator polypeptide isolated from a rat Leydig cell tumor. Science 236:188-190 33. Xu X, Xu T, Robertson DG, Lambeth JD 1989 GTP stimulates

34. 35. 36. 37.

38.

39.

40.

41. 42.

Endo• 1991 Voll29«No6

pregnenolone generation in isolated rat adrenal mitochondria. J Biol Chem 264:17674-17680 Vilcek J, Lee TH 1991 Tumor necrosis factor. New insights into the molecular mechanisms of its multiple actions. J Biol Chem 266:7313-7316 Calkins JH, Guo H, Sigel MM, Lin T 1990 Tumor necrosis factora enhances effects of interleukin-1/3 on Leydig cell steroidogenesis. Biochem Biophys Res Commun 166:1313-1318 Roby KF, Terranova PF 1990 Effects of tumor necrosis factor-a in vitro on steroidogenesis of healthy and atretic follicles of the rat: theca as a target. Endocrinology 126:2711-2718 Andreani CL, Payne DW, Packman JN, Resnick CE, Hurwitz A, Adashi EY 1991 Cytokine-mediated regulation of ovarian function. Tumor necrosis factor a inhibits gonadotropin-supported ovarian androgen biosynthesis. J Biol Chem 266:6761-6766 Adashi EY, Resnick CE, Croft CS, Payne DW 1989 Tumor necrosis factor a inhibits gonadotropin hormonal action in nontransformed ovarian granulosa cells. A modulatory noncytotoxic property. J Biol Chem 264:11591-11597 Darbon JM, Oury F, Laredo J, Bayard F 1989 Tumor necrosis factor-a inhibits follicle-stimulating hormone-induced differentiation in cultured rat granulosa cells. Biochem Biophys Res Commun 163:1038-1046 Emoto N, Baird A 1988 The effects of tumor necrosis factor/ cachectin on follicle-stimulating hormone-induced aromatase activity in cultured rat granulosa cells. Biochem Biophys Res Commun 153:792-798 Roby KF, Weed J, Lyles R, Terranova PF 1990 Immunological evidence for a human ovarian tumor necrosis factor-a. J Clin Endocrinol Metab 71:1096-1102 Calandra T, Baumgartner JD, Grau GE, Wum MM 1990 Kinetics and prognostic values of tumor necrosis factor/cachectin, Interleukin-1, a-interferon and 7-interferon in the serum of patients with septic shock. J Infect Dis 161:982-987

The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 15 December 2014. at 00:56 For personal use only. No other uses without permission. . All rights reserved.