Search

The existence of a link between estrogen deprivation and osteoarthritis (OA) in postmenopausal women suggests that 17β-estradiol (17β-E2) may be a modulator of cartilage homeostasis. Here, we demonstrate that 17β-E2 stimulates, via its receptor human estrogen receptor α 66 (hERα66), type II collagen expression in differentiated and dedifferentiated (reflecting the OA phenotype) articular chondrocytes. Transactivation of type II collagen gene (COL2A1) by ligand-independent transactivation domain (AF-1) of hERα66 was mediated by "GC" binding sites of the -266/-63-bp promoter, through physical interactions between ERα, Sp1/Sp3, Sox9, and p300, as demonstrated in chromatin immunoprecipitation (ChIP) and Re-Chromatin Immuno-Precipitation (Re-ChIP) assays in primary and dedifferentiated cells. 17β-E2 and hERα66 increased the DNA-binding activities of Sp1/Sp3 and Sox-9 to both COL2A1 promoter and enhancer regions. Besides, Sp1, Sp3, and Sox-9 small interfering RNAs (siRNAs) prevented hERα66-induced transactivation of COL2A1, suggesting that these factors and their respective cis-regions are required for hERα66-mediated COL2A1 up-regulation. Our results highlight the genomic pathway by which 17β-E2 and hERα66 modulate Sp1/Sp3 heteromer binding activity and simultaneously participate in the recruitment of the essential factors Sox-9 and p300 involved respectively in the chondrocyte-differentiated status and COL2A1 transcriptional activation. These novel findings could therefore be attractive for tissue engineering of cartilage in OA, by the fact that 17β-E2 could promote chondrocyte redifferentiation.17β-E2 up-regulates type II collagen gene expression in articular chondrocytes. An ERα66/Sp1/Sp3/Sox-9/p300 protein complex mediates this stimulatory effect. This heteromeric complex interacts and binds to Col2a1 promoter and enhancer in vivo. Our findings highlight a new regulatory mechanism for 17β-E2 action in chondrocytes. 17β-E2 might be an attractive candidate for cartilage engineering applications.

International audience; The effects of some progestins on fish reproduction have been recently reported revealing the hazard of this class of steroidal pharmaceuticals. However, their effects at the central nervous system level have been poorly studied until now. Notwithstanding, progesterone, although still widely considered primarily a sex hormone, is an important agent affecting many central nervous system functions. Herein, we investigated the effects of a large set of synthetic ligands of the nuclear progesterone receptor on the glial-specific expression of the zebrafish brain aromatase (cyp19a1b) using zebrafish mechanism-based assays. Progesterone and 24 progestins were first screened on transgenic cyp19a1b-GFP zebrafish embryos. We showed that progesterone, dydrogesterone, drospirenone and all the progesterone-derived progestins had no effect on GFP expression. Conversely, all progestins derived from 19-nortesterone induced GFP in a concentration-dependent manner with EC50 ranging from the low nM range to hundreds nM. The 19-nortestosterone derived progestins levonorgestrel (LNG) and norethindrone (NET) were further tested in a radial glial cell context using U251-MG cells co-transfected with zebrafish ER subtypes (zfER alpha, zfER beta 1 or zfER beta 2) and cyp19a1b promoter linked to luciferase. Progesterone had no effect on luciferase activity while NET and LNG induced luciferase activity that was blocked by ICI 182,780. Zebrafish-ERs competition assays showed that NET and LNG were unable to bind to ERs, suggesting that the effects of these compounds on cyp19a1b require metabolic activation prior to elicit estrogenic activity. Overall, we demonstrate that 19-nortestosterone derived progestins elicit estrogenic activity by inducing cyp19a1b expression in radial glial cells. Given the crucial role of radial glial cells and neuro-estrogens in early development of brain, the consequences of exposure of fish to these compounds require further investigation.

Calcitonin is a peptide which acts in the brain to modulate behavior and hormone release, possibly through an interaction with serotonergic systems. We investigated the effects of chronic systemic injections of salmon calcitonin on the [3H]-8-OHDPAT binding to 5-HT1A receptors in the frontal cortex and hippocampus in adrenalectomized and intact (non adrenalectomized) rats. The results show that salmon calcitonin increases the maximal density of 5-HT1A binding sites in both structures in adrenalectomized animals (and decreases the affinity in the frontal cortex only). Calcitonin does not alter this binding in intact rats. These results demonstrate the existence of interactions between calcitonin, serotonin and glucocorticoids, and raise the hypothesis of a neurotrophic effect of calcitonin on serotonergic neurons.

Background Testicular Leydig cells use either exogenous or de novo synthesized cholesterol as the substrate for the production of testosterone with hormone stimulation. Although the long-term effect of trophic hormones on Leydig cell cholesterol uptake, storage, and deesterification has been well documented, the early effects of the human choriogonadotropin (hCG) on cell cholesterol/lipid distribution are not yet known. Methods Sections of cells treated with hCG for 15 sec to 30 min were examined by electron microscopy (EM) for the surface density of lipid moieties in the cytoplasm. In addition, the time-dependent distribution of lipids within the cytoplasmic inclusions and the ultimate destination of this substrate were evaluated by EM. The results were analyzed with standard morphometric methods. Results and Conclusion The surface density of cytoplasmic lipid pools increased significantly within the 15 sec following the exposure of cells to hCG, and it tapered off to the control level in the subsequent 30 min. Such a fluctuation in the amount of cytoplasmic lipids may be due to (1) the quantity of released substrate from the reticular compartment or (2) the rate of its transport across the outer mitochondrial membrane to the inner mitochondrial cytochrome P450 side-chain cleavage, where steroidogenesis begins with the conversion of cholesterol to pregnenolone. These two processes were not quantitatively coordinated in the stimulated cell during the intitial 30 min, resulting in a surplus of cytoplasmic lipid pools. To compensate for such uneven metabolic balance, the cell apparently disposed of the excess substrate by a mechanism of molecular regrouping from a micellar configuration to a bilayer structure followed by exocytosis. Anat. Rec. 248:374-379, 1997. © 1997 Wiley-Liss, Inc.

We examined the topography of the MA-10 Leydig tumor cell mitochondrial peripheral-type benzodiazepine receptor (PBR). In previous studies, the 18 kDa PBR was found to be functionally associated with the voltage-dependent anion channel, located in the junctions between outer and inner membranes. Transmission electron (TEM) and atomic force microscopy (AFM) of immunogold labeled PBR on Leydig cell mitochondrial preparations showed that the 18 kDa PBR protein is organized in clusters of 4-6 molecules. Addition of hCG to Leydig cells induces a rapid, within 30 sec, increase in PBR ligand binding and morphological changes, namely redistribution of PBR molecules in large clusters (7 particles). These hCG-induced changes were inhibited by a cAMP-dependent protein kinase inhibitor and by the benzodiazepine flunitrazepam. AFM further demonstrated the rapid reorganization of the mitochondrial membrane, where the formation of contacts between the outer and the inner mitochondrial membrane may facilitate cholesterol transfer.

Native MA-10 mouse Leydig tumor cell mitochondrial preparations were examined by transmission electron (TEM) and atomic force (AFM) microscopic procedures in order to investigate the topography and organization of the peripheral-type benzodiazepine receptor (PBR). Mitochondria were immunolabeled with an anti-PBR antiserum coupled to gold-labeled secondary antibodies. Results obtained indicate that the 18,000 MW PBR protein is organized in clusters of 4-6 molecules. Moreover, on many occasions, the interrelationship among the PBR molecules was found to favor the formation of a single pore. Taking into account recent observations that the 18,000 MW PBR protein is functionally associated with the pore forming 34,000 MW voltage-dependent anion channel (VDAC) these results suggest that (i) the mitochondrial PBR complex could function as a pore, thus allowing the translocation of cholesterol and other molecules to the inner mitochondrial membrane, and (ii) the native receptor is a multimeric complex of an approximate 140,000 MW composed on an average of five 18,000 PBR subunits, one 34,000 VDAC subunit, and associated lipids.

International audience; OBJECTIVES: To investigate the mechanisms by which cytokines and 17beta-oestradiol (17beta-E2) modulate gene expression and activity of uridine diphosphoglucose dehydrogenase (UGDH), a key enzyme of GAG synthesis in articular chondrocytes. METHODS: Rabbit articular chondrocytes (RAC) from 3-week-old animals were incubated for 24 h with TGF-beta, insulin like growth factor-I (IGF-I), IL-1beta, IL-6 and 17beta-E2. GAG synthesis was measured by [35S]-sulphate labelling and the expression of the UGDH gene was estimated by both real-time polymerase chain reaction and western blotting, whereas the enzyme activity was assayed by a spectrophotometric procedure. In addition, the transcriptional activity of several UGDH gene promoter constructs was determined in RAC transiently transfected with wild-type or deleted human oestrogen receptor-alpha gene (hER alpha66 or hER alpha46, respectively). RESULTS: 17Beta-E2 and its receptor hER alpha66 enhanced GAG neosynthesis in rabbit articular chondrocytes, as did TGF-beta1 whereas IL-1beta decreased this synthesis. 17Beta-E2 was found to exert positive regulatory effects at mRNA, protein and UGDH activity levels. In addition, the receptor hER alpha66, but not hER alpha46, increased the transcriptional activity of the UGDH gene. In contrast, no clear correlation between transcription, translation and activity of the UGDH was found under the effects of the cytokines studied. However, TGF-beta enhanced the enzyme activity, whereas IL-1beta, IL-6 and IGF-I were without significant effect. CONCLUSIONS: 17Beta-E2 enhanced GAG synthesis in chondrocytes via up-regulation of the UGDH gene expression and enzyme activity. These data provide insights into the molecular mechanisms involved in the regulation of the UGDH gene and offer new approaches to investigate its potential alteration in joint diseases.

International audience; The estrogen receptor alpha (ER alpha) is key in regulating normal breast development and function and is closely involved in the onset and progress of cancers. ER alpha transcriptional activity is mediated through two activation functions, AF1 and AF2, whose activity is tightly regulated in a cell-specific manner through yet unknown processes. Here, we demonstrate that cell-cell junctions generate cell permissiveness to AF1 through an up-regulation of the activity of an AF1 sub-region termed box 1. Moreover, the loss of E-cadherin expression is shown to silence the AF1 activity of ER alpha, allowing the receptor to mainly act through its AF2. This switch from an AF1 to an AF2 cell permissiveness also consequently results in the attenuation of ER alpha activity. Therefore, a loss of cell-cell junctions, a key process that occurs during the epithelial-mesenchymal transition, should have a broad impact on ER alpha transcriptional functions.

In this study, we hypothesized that many of the reported effects of phthalate esters and other peroxisome proliferators (PPs) in the testis are mediated by members of the PPactivated receptor (PPAR) family of transcription factors through alterations in proteins involved in steroidogenesis. Exposure of Leydig cells to PPs prevented cholesterol transport into the mitochondria after hormonal stimulation and inhibited steroid synthesis, without altering total cell protein synthesis or mitochondrial and DNA integrity. PPs also reduced the levels of the cholesterol-binding protein peripheraltype benzodiazepine receptor (PBR) because of a direct transcriptional inhibition of PBR gene expression in MA-10 Leydig cells. MA-10 cells contain mRNAs for PPAR and PPAR/, but not for PPAR. In vivo treatment of mice with PPs resulted in the reduction of both testis PBR mRNA and circulating testosterone levels, in agreement with the proposed role of PBR in steroidogenesis. By contrast, liver PBR mRNA levels were increased, in agreement with the proposed role of PBR in cell growth/tumor formation in nonsteroidogenic tissues. However, PPs did not inhibit testosterone production and testis PBR expression in PPAR-null mice. These results suggest that the antiandrogenic effect of PPs is mediated by a PPARdependent inhibition of Leydig cell PBR gene expression. (Endocrinology 143: 2571–2583, 2002)

The peroxisome proliferator perfluordecanoic acid (PFDA) has been shown to exert an antiandrogenic effect in vivo by acting directly on the interstitial Leydig cells of the testis. The objective of this study was to examine the in vitro effects of PFDA and identify its site of action in steroidogenesis using as model systems the mouse tumor MA-10 and isolated rat Leydig cells. PFDA inhibited in a time- and dose-dependent manner the hCG-stimulated Leydig cell steroidogenesis. This effect was localized at the level of cholesterol transport into the mitochondria. PFDA did not affect either the total cell protein synthesis or the mitochondrial integrity. Moreover, it did not induce any DNA damage. Morphological studies indicated that PFDA induced lipid accumulation in the cells, probably due to the fact that cholesterol mobilized by hCG did not enter the mitochondria to be used for steroidogenesis. In search of the target of PFDA, we examined its effect on key regulatory mechanisms of steroidogenesis. PFDA did not affect the hCG-induced steroidogenic acute regulatory protein (StAR) levels. However, it was found to inhibit the mitochondrial peripheral-type benzodiazepine receptor (PBR) ligand binding capacity, 18-kDa protein, and messenger RNA (mRNA) levels. Further studies indicated that PFDA did not affect PBR transcription, but it rather accelerated PBR mRNA decay. Taken together, these data suggest that PFDA inhibits the Leydig cell steroidogenesis by affecting PBR mRNA stability, thus inhibiting PBR expression, cholesterol transport into the mitochondria, and the subsequent steroid formation. Moreover, this action of PFDA on PBR mRNA stability indicates a new mechanism of action of peroxisome proliferators distinct from the classic transcription-mediated regulation of target genes.

Eukaryotic steroid hormones, derived from cholesterol, are involved in the maintenance of the organism’s homeostasis, adaptability to the environment, and developmental and reproductive functions. In addition to the well-defined actions in peripheral tissues, steroids have pleiotropic actions on the central nervous system (CNS), where they control a number of neuroendocrine and behavioral functions. Thus, comprehension of the molecular systems underlying the control of steroid hormone biosynthesis is essential for the study and treatment of a multitude of physiological disorders.

Gonadal steroidogenesis is regulated by pituitary gonadotropins and a locally produced, unidentified factor. A 70-kilodalton (kD) protein complex secreted from rat Sertoli cells was isolated. The complex, composed of 28- and 38-kD proteins, stimulated steroidogenesis by Leydig cells and ovarian granulosa cells in a dose-dependent and adenosine 3',5'-monophosphate-independent manner. The follicle-stimulating hormone-induced 28-kD protein appeared to be responsible for the bioactivity, but the 38-kD protein was indispensable for maximal activity. The 28- and 38-kD proteins were shown to be identical to the tissue inhibitor of metalloproteinase-1 (TIMP-1) and the proenzyme form of cathepsin L, respectively. Thus, a TIMP-1-procathepsin L complex is a potent activator of steroidogenesis and may regulate steroid concentrations and, thus, germ cell development in both males and females.

In order to clarify the role of germ cells in the regulation of Sertoli cell secretions, three experimental models of germ cell depletion were used : hypodactyl rat mutation (testis weight [TW] : 55% less than controls), in utero busulfan treatment (TW : 88% less than controls), and neonatal experimental cryptorchidism (TW : 72% less than controls). The aim of this work was to compare the numbers of Leydig and Sertoli cells and the production of germ cells in each experimental model to the in vitro secretions of Leydig and Sertoli cells in conditioned media and to the hormonal serum profiles of the same animal in vivo. In the three models, serum levels of hypophyseal and testosterone hormones were significantly increased and decreased, respectively. In the absence of germ cells, the total length of seminiferous tubules, the total numbers of Sertoli and Leydig cells, and the daily production of germ cells were significantly diminished. The addition of both control and damaged seminiferous tubule culture media (STM : media conditioned by 10 cm of seminiferous tubules) to 10 5 control or damaged Leydig cells led to a further increase of testosterone production after ovine LH stimulation. However, expressed per Sertoli cell, testosterone production by control Leydig cells was reduced by addition of damaged STM as compared to addition of control STM, and, similarly, the addition of control STM to damaged Leydig cells enhanced testosterone production more than did the addition of damaged STM. Secretions of transferrin per Sertoli cell in STM were reduced as compared to controls by the absence of germ cells but to a lesser extent than was production of spermatocytes and of spermatids. In conclusion, secretions by Sertoli cells of the paracrine factor involved in the control of testosterone production by Leydig cells and of transferrin are modified by germ cells.

Summary In order to elucidate the respective effects of depletion of germ cells and of increase in testicular temperature, rats of the same Wistar strain were rendered experimentally bicryptorchid or sterilized by a busulfan injection in utero and compared to control animals. In both models, germ cells were depleted but numeric evolution and functions of somatic cells differed. The aim of that work was to compare the numeric evolutions of testicular somatic and germ cells to their respective functions in each model before puberty and in adult rats of the same strain. Serum concentrations of FSH, LH and testosterone were compared at 20, 40 and 110 days of age. Histological analyses of Sertoli and germ cells in the seminiferous tubules and of Leydig cells in the intertubular tissue were performed before puberty and at adulthood. Testosterone serum concentrations were depleted in both models starting at 40 days of age and more in busulfan-treated rats. Both FSH and LH levels were increased from 20 days onwards in experimental rats. Additional cryptorchidism in busulfan-treated rats depressed the serum testosterone concentration. At 17 days of age, the cryptorchidism do not modify somatic or germ cell populations while busulfan treatment has induced a decrease of both these populations. Conversely, the cross sectional area of the somatic testicular cells was not affected whatever the treatment. In adult testes of busulfan-treated and cryptorchid rats, the total numbers of Sertoli and Leydig cells and of germ cells per testis were decreased. The cellular size of the perivascular Leydig cells was not modified by any of the treatments whereas the size of the Sertoli cells was reduced. In conclusion, in both models the absence of germ cells induces a decrease in Sertoli cell function, while the increase in testicular temperature provokes degeneracies of Sertoli and germ cells in the seminiferous tubules of the rat.

The polypeptide diazepam-binding inhibitor (DBI) has been previously shown to stimulate testicular Leydig, adrenocortical, and glial-cell mitochondrial steroidogenesis in vitro. To assess the in situ role of DBI in trophic hormone-stimulated steroidogenesis, we suppressed DBI levels in the hormone-responsive MA-10 Leydig tumor cells, using a cholesterol-linked phosphorothioate oligodeoxynucleotide (Chol-odN) antisense to DBI. Treating MA-10 cells with Chol-odN antisense to DBI resulted in a dose-dependent reduction of DBI levels (ED50 = 1 microM). In contrast, Chol-odN sense to DBI did not affect its expression. Saturating amounts of human choriogonadotropin (hCG) increased MA-10 progesterone production by 150-fold. Addition of increased concentrations of Chol-odNs sense to DBI or of a nonrelated sequence did not reduce the MA-10 response to hCG. However, in the presence of Chol-odN antisense to DBI that could reduce DBI levels, MA-10 cells lost their ability to respond to hCG (ED50 = 1 microM). In these studies the hCG-stimulated cAMP levels and cytochrome P450 side-chain cleavage activity, as measured by metabolism of 22(R)-hydroxycholesterol, were not affected by the Chol-odNs used. These observations provide unequivocal evidence that DBI plays a vital role in the acute stimulation of steroidogenesis by trophic hormones.

The addition of seminiferous tubule (ST) culture medium (STM) prepared from testes of either busulfan-treated (Bus) or cryptorchid (Cryp) or genetically sterile (hd) rats, to Percoll purified Leydig cells leads to a further increase of LH-stimulated testosterone (T) output (26, 43 and 14%, respectively). Taking into account that the Sertoli cell number per cm of ST is 2.6, 1.8 and 1.4-fold greated in Bus, Cryp and hd rats than in controls, the above STM effects on T output, expressed per 10 6 Sertoli cells are in fact lower (63, 44 and 43%, respectively) that those of control STM. Similar results have been obtained for the STM transferrin levels which are decreased, 74, 67 and 45%, respectively in Bus, Cryp and hd animals. So, it is likely that the Sertoli cell secretion of both the paracrine factor involved on Leydig cell T production and the transferrin is influenced mainly by spermatids and to a lesser extent by spermatocytes of mature rat testis.

Intracellular cyclic AMP and testosterone productions by purified mature rat Leydig cells were stimulated by oLH (25 μg/l) 18- and 12-fold, respectively, after a 5-h incubation period. The replacement of the incubation medium by charcoal-treated testicular venous plasma (40%, v/v) from adult rams in the breeding season induced an inhibition of cyclic AMP and testosterone productions (82 and 66%, respectively, of oLH-stimulated values). Testicular arterial plasma is less effective than testicular venous plasma, even when it originates from non-breeding season rams; in that case, testicular venous and arterial plasma strongly inhibit testosterone productions (84 and 67%, respectively of oLH-stimulated values), which probably indicates that the inhibitory activity is higher in the non-breeding season. The addition of peripheral plasma leads to a testosterone production equal to 35 and 65% of the oLH-stimulated values, respectively, for ram blood collected in non-breeding and breeding seasons. The same concentration of ovine testicular lymph or rete testis fluid is without significant effect on cyclic AMP production; however, testosterone is slightly decreased by lymph but enhanced by rete testis fluid. Increasing amounts of venous or arterial testicular blood induce a dose-related decrease of the specific binding of labelled hCG in both rat and ram testicular membranes. This inhibiting factor is present in peripheral and testicular blood of either control or hypophysectomized or castrated rams, is a protein in nature, heat-sensitive, and has an apparent molecular weight higher than 10 000 daltons. These results suggest the existence of a control of LH-specific binding to its receptors and of Leydig cell cyclic AMP and testosterone outputs; these activities are not species-specific and are more concentrated in testicular venous than in arterial blood. The origin of this inhibiting factor remains to be determined, since it is not confined to the testis and not of pituitary origin.

A single i.p. injection of either cisplatin or bleomycin affects the testosterone production by the Leydig cells in mature rats. It is noteworthy that on day 30 after treatment, a complete recovery of Leydig cell function is found in bleomycin-treated rats whereas in cisplatin-injected animals, 25-48% decreases of testosterone synthesis are still observed, respectively, under basal and LH-stimulated conditions. Together with the germ cell destruction, the low testosterone levels probably contribute to the sterility in cisplatin-treated rats.

The polypeptide diazepam binding inhibitor (DBI) and drug ligands for the mitochondrial peripheral-type benzodiazepine receptor (PBR) have been shown to regulate cholesterol transport, the rate-determining step in steroidogenesis, in hormone-responsive steroidogenic cells including the MA-10 Leydig tumor cells. The present study was designed to characterize the role of DBI and PBR in the R2C rat Leydig tumor constitutive steroid-producing cell model. Both DBI and PBR were present in R2C cells. R2C cell treatment with a cholesterol-linked phosphorothioate oligodeoxynucleotide antisense to DBI, but not sense, resulted in the reduction of DBI levels and a concomitant dramatic decrease of the amount of progesterone produced. These observations strongly suggested that DBI was important in maintaining constitutive steroidogenesis in R2C cells. Radioligand binding assays revealed the presence of a single class of PBR binding sites with an affinity 10 times higher (Kd approximately 0.5 nM) than that displayed by the MA-10 PBR (Kd approximately 5 nM). Photolabeling of R2C and MA-10 cell mitochondria with the photoactivatable PBR ligand [3H]1-(2-fluoro-5-nitrophenyl)-N-methyl-N-(1-methyl-propyl)-3- isoquinolinecarboxamide showed that the M(r) 18,000 PBR protein was specifically labeled. This indicates that the R2C cells express a PBR protein which has properties similar to the MA-10 PBR. Chemical crosslinking studies of purified metabolically radiolabeled DBI to mitochondria provided direct evidence that DBI specifically binds to the M(r) 18,000 PBR protein. Moreover, DBI and a PBR synthetic ligand were able to increase steroid production in isolated R2C cell mitochondria which express the 5 nM affinity receptor. However, mitochondrial PBR binding was increased by 6-fold upon addition of the post-mitochondrial fraction, suggesting that a cytosolic factor modulates the binding properties of PBR in R2C cells and is responsible for the 0.5 nM affinity receptor seen in intact cells. In conclusion, these data demonstrate that DBI plays a key role in maintaining R2C constitutive steroidogenesis by binding to the mitochondrial higher affinity PBR which promotes a continuous supply of cholesterol to the inner mitochondrial side chain cleavage cytochrome P450.

Previous studies demonstrated that the polypeptide diazepam binding inhibitor (DBI) and its receptor, the peripheral-type benzodiazepine receptor (PBR), are involved in the regulation of steroid biosynthesis and that one site of PBR action resides in mitochondria. In the present investigation, evidence is presented that a functional form of PBR is also present at the cell surface. First, PBR was immunolocalized in the rat testis using biotin-streptavidin peroxidase immunocytochemistry, and results revealed that PBR was present exclusively in the interstitial Leydig cells. Next, the distribution of PBR in MA-10 Leydig cells was further examined using confocal microscopy. MA-10 cells were either fixed and immunostained or fixed/permeabilized and immunostained for PBR, followed by generation of confocal microscope optical sections, three-dimensional reconstructions of these sections, and then generation of vertical confocal sections of the three-dimensional reconstruction. In the fixed/unpermeabilized cells, PBR immunostaining at the cell surface was clearly evident, whereas in the fixed/permeabilized cells, intracellular PBR distribution was more robust. These results suggest that the plasma membrane fraction of the receptor could mediate the action of extracellular PBR ligands on Leydig cell function. Next, we examined whether DBI, the naturally occurring PBR ligand, is secreted by testicular cells and whether it could activate the cell surface PBR. Immunolocalization of DBI demonstrated that it was present in both Leydig and Sertoli cells. Further, using an immunoblot assay, we demonstrated that DBI is present in rat testicular interstitial fluid. Metabolic labeling of cultured immature rat Sertoli cells and MA-10 mouse tumor Leydig cells, followed by immunoprecipitation of the secreted proteins with an anti-DBI antiserum, demonstrated that both Leydig and Sertoli cells secrete DBI and could serve as a cell source for the interstitial fluid DBI. Then, we partially purified the DBI present in conditioned medium and interstitial fluid by reverse phase chromatography and demonstrated it to be bioactive, based on displacement of a radiolabeled benzodiazepine (Ro5-4864)-specific ligand for PBR; pronase treatment of different preparations eliminated all bioactivity. We then examined the effects of DBI on Leydig cell function. DBI added to MA-10 cells affected DNA synthesis and cell growth in a biphasic manner; at low concentrations (1 nM), DBI was mitogenic, increasing [3H]thymidine incorporation and cell numbers by 30-40%, while at high concentrations (1 microM), DBI inhibited cell growth (30-40%). Similar effects on cell growth were obtained using the benzodiazepine Ro5-4864.